STR RANGE ASSIGNED TO --------- ----------------------- 1-96 -- 31 assigned to Matthias (DONE, check: 1, 18, 24, 45, 50, 76, 96) 97-265 -- 50 assigned to Matthias (DONE, check: 207, 214, 220, 260, 265) 270-448 -- 50 assigned to Albrecht (DONE, check: 294, 357, 419) 451-593 -- 50 assigned to Albrecht (DONE, nothing relevant) 599-763 -- 50 assigned to Greg (DONE, check: 614, 702) 785-1092 -- 50 assigned to Greg (DONE, check: 810, 849, 890, 1090) 1100-end -- 50 assigned to Albrecht (DONE, check: 1105, 1113, 1115, 1155, 1190, 1278, 1303, 1306, 1379)
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Child Windows and Double Windows |
Version: | 1.1.3 |
Created By: | mrsatan.users.sf |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
ciudadsatan.hotmail 15:42 May 20, 2003 | problem.png 5k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mrsatan.users.sf 22:27 Apr 04, 2003 | This is FLTK 1.1.2 on Cygwin, Win98/Win2k. There is a bug regarding child windows and double windows. Try this example: #include <FL/Fl_Double_Window.H> #include <FL/Fl.H> int main() { Fl_Double_Window* OUTER = new Fl_Double_Window (300, 300); Fl_Double_Window* INNER = new Fl_Double_Window (100, 100, 100, 100); INNER->box(FL_BORDER_BOX); OUTER->show(); return Fl::run(); } Before running this program, open some other program (explorer, for example), maximize it, then run the example. Now move the window so that the red box is partially out of screen. Now press alt-tab twice, fast, so that the maximized windows get the focus, then the example. The example window is updated, but not the red box, and performance decreases substantially (my computer even freezes for a second). This happens whenever an "outer" double_window has a child window. Try changing the tile example so that the main or child window is Double_Window; it does exhibit the same problem. A single window with a double_window as a child still gets redrawn, but the OS freezes for a second. | ||
mrsatan.users.sf 22:27 Apr 04, 2003 | Argh... the box isn't red, but anyway, you get the idea... But this bug is MUCH nastier than I thought... I was testing my app on WinXP, with a Double_Window inside a Single_Window, the bug happened and... WinXP crashed completely! I catched a brief glimpse of the BSoD, and read something about an "infinite bucle"... hope it helps. | ||
mike 22:28 Apr 04, 2003 | Can you run your app from a debugger (eg gdb) to see where the hangup is? Also, what graphics card are you using? Finally, can you try running things with one of the double windows changed to a single window? | ||
mrsatan.users.sf 22:28 Apr 04, 2003 | I've never used a debugger... The graphics card seems to be irrelevant, since this happens on every computer I've tried: the child isn't redrawn until the window is moved, and it is very difficult to move. Also, in XP, the system crashes. This happens everytime the main window is Double. If the main window is Single and the child is Double, the child is redrawn correctly, but it's still hard to move. If both are Single, no problem. | ||
mike 22:28 Apr 04, 2003 | I have just commited a possible fix for this in CVS; please try it and let me know if it works. | ||
mrsatan.users.sf 22:29 Apr 04, 2003 | It seems you misunderstood the problem... Ok, the patch fixes the "Double Window inside a Single Window" problem, but the bug is about any window inside any window, except if both of them are single. The problem remains when a Double Window has another window (Single or Double) embedded in it. Patch tested in Cygwin/Win98 only... | ||
mrsatan.users.sf 22:29 Apr 04, 2003 | Here probably the fix of a problem: in file fl_win32.cxx, procedure WndProc change the call of validating function in order to validate whole window, not the part, after redrawing. // ValidateRgn(hWnd,i->region); ValidateRgn(hWnd,NULL); Changes in flash method of Fl_Double_Window was not necessary and lead to actually disabling DoubleBuffering for childs. | ||
ciudadsatan.hotmail 23:53 May 16, 2003 | This last fix worked ok! Now only the previous patch needs to be undone, to reallow child windows to be double-buffered (I tested this patch without the previous; it worked ok in both mingw and cygwin) | ||
mike 12:10 May 20, 2003 | OK, can you try the current CVS and let me know if the change I put in (only validate against a region for the parent window) fixes your problems? I also removed the Fl_Double_Window hack. | ||
ciudadsatan.hotmail 13:37 May 20, 2003 | Same problem. However, if I change if (window->parent()) ValidateRgn(hWnd,0); else ValidateRgn(hWnd,i->region); to ValidateRgn(hWnd,0); it works. | ||
mike 14:23 May 20, 2003 | Well, we can't eliminate the region validation completely because it dorks up normal redraws. As I have not been able to reproduce the crash problem on any of our systems, and since the only "fix" for your system seems to be to break all others, I'm going to close this STR for now - you'll just need to keep your change in place for all future releases of FLTK. | ||
ciudadsatan.hotmail 15:44 May 20, 2003 | Just for the record, I've posted an image showing the problem using the tile example (after changing it to use two Fl_Double_Window. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | wrong libraries linked when cross compiling from linux to mingw |
Version: | 1.1.3 |
Created By: | mucki.gryps |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mucki.gryps 02:33 Apr 07, 2003 | when compiling a win32 hosted package with a linux to mingw cross compiler under linux the wrong libraries get linked into fluid.exe. The makefile tries to link with -lXext -lX11, but this should be -lgdi32 -lws2_32 -lole32 -luuid. | ||
mike 13:40 May 04, 2003 | Can you supply the necessary patches to configure.in in order to detect and use the GDI libs instead of X11? Thanks! | ||
mucki.gryps 04:08 May 06, 2003 | hmm ... I am not sure, but I will try :) | ||
mucki.gryps 02:27 May 14, 2003 | Hmm ... i cannot figure out how to test which platform the compile is for in the configure.in script. I think i am not really able to supply this patch, as i am a beginner with Autoconf :( | ||
mike 10:57 May 20, 2003 | OK, well if you can come up with a way to detect at configure time whether MingW is being used under Linux, please let us know and we'll try to incorporate it into the configure script. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | crashes on win2k prof |
Version: | 1.1.3 |
Created By: | unay.tcts.fpms.ac |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
unay.tcts.fpms.ac 04:25 Apr 10, 2003 | I downloaded extracted fltk version 1.1.3. Then I opened fltk.dsw and built demo project. after this step i checked fluid.exe and it works well. I tried to use the example you provided : cone3.cxx. I added the necessary libraries and compiled, 0 errors. Then I built cone3.cxx and nothing unusual. But the exe file does not work, it gives an error at initialization. I tried to debug adn saw that NTDLL.DLL is the problem. It is not initialized properly. I have windows2000 professional as OS. I don't know how to fix this problem. Can you help me? | ||
mike 05:53 Apr 10, 2003 | FLTK does not include a demo called "cone3"! | ||
unay.tcts.fpms.ac 06:02 Apr 10, 2003 | sorry, I am knocking the wrong door. I tried to use fltk within vtk and I had found an example of this ('cone3.cxx') from somewhere. But I cannot find where I took it:) sorry for inconveniant question | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Unassigned |
Summary: | fl_font doesn't work through exceed |
Version: | 1.1.0 |
Created By: | rodinw2000.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rodinw2000.yahoo 17:24 Apr 17, 2003 | The fonts example in test doens't work through Exceed. The label font doesn't change when I set to Curiur or Times. Running on Windows and native Unix terminal display are fine. But not if I using Exceed to display Unix window on Windows. | ||
mike 13:45 May 04, 2003 | What UNIX operating system are you using? If Linux, what distribution and version? Did you compile FLTK with Xft support (for antialiased fonts)? | ||
rodinw2000.yahoo 08:53 May 05, 2003 | I am using Linux. 2.4.9. I didn't compile FLTK with any option. Should I ? | ||
mike 12:47 May 15, 2003 | Do the other fonts work? I suspect this might be an issue of Exceed not mapping Times to Times New Roman and Courier to Courier New... | ||
rodinw2000.yahoo 11:46 May 20, 2003 | Only some work, like "clean bold", "fixed"... The most common ones don't, including "times", "helvetica"... | ||
mike 11:53 May 20, 2003 | OK, that makes sense - Windows doesn't normally have Times, Courier, or Helvetica. Instead it uses Times New Roman, Courier New, and Arial. I'm not sure there is any easy fix for this on the FLTK side of things, and I'm pretty sure you can configure Exceed to alias certain fonts to the available system fonts... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | in mingw , there is path accessing error by fluid for docuemntation help .html files |
Version: | 1.1.2 |
Created By: | ashishwave.yahoo |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ashishwave.yahoo 05:47 Apr 21, 2003 | i am reposting it here, as u said that bug will be only considered, when posted in STR. when building fltk1.1.2 and etc on Mingw and MSYS, it compiles well and fluid works very good, BUT when we try to open help manual from fluid's menu it gives file not found error, hence we have to goto that help html file in windows explorer and manually open it in internet explorer to get the fltk help. i think that this bug is just due to that when compiling you assume that it is posix shell(of msys), hence AFTER compilation and during actual running of fluid you will continue getting posix style paths. This is true in cygwin but not in mingw+msys. Hence remove this path access bug of help html file, so that even in mingw+msys compiled fltk, people can get integerated help. | ||
mike 15:05 May 04, 2003 | Well, the help works just fine (assuming you copy the files to the appropriate places) with VC++, so if MingW's libraries don't handle forward slashes then they are broken (since Microsoft's library *do* handle forward slashes...) Am I understanding the problem correctly? | ||
mike 11:03 May 20, 2003 | No response within 2 weeks, closing STR... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Unassigned |
Summary: | function fl_file_chooser returns null pointer on quick selects |
Version: | 1.1.3 |
Created By: | johnl.cs.wisc |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
johnl.cs.wisc 04:06 Apr 22, 2003 | When running the following in a button callback, I found that by selecting a file relatively quickly filename was sometimes null. filename = fl_file_chooser("Load sample", "*.wav", ""); I have only tried this on one machine, a duron 700 running Gentoo linux. | ||
mike 13:48 May 04, 2003 | Are you choosing files from a directory with a lot of files or files on an NFS-mounted filesystem? I can't duplicate the problem on any of our systems, so unless we can determine a cause I'll have to close this out without resolution. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Unassigned |
Summary: | sub-windows position child items relative to self, not top-window |
Version: | 1.1.3 |
Created By: | rsmith.ncbi.nlm.nih |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
rsmith.ncbi.nlm.nih 10:15 Apr 29, 2003 | fltk-subwindows.jpg 40k | ||
rsmith.ncbi.nlm.nih 10:15 Apr 29, 2003 | fltk-subwingroup.jpg 43k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rsmith.ncbi.nlm.nih 10:15 Apr 29, 2003 | Any items inside a sub-window will be positioned with its x and y values relative to the top right of its enclosing sub-window. This is different than how things are positioned inside Fl_Group. Fluid assumes all Widgets x and y refer to offsets within the top window. | ||
rsmith.ncbi.nlm.nih 10:18 Apr 29, 2003 | In the two enclosed graphic files, the left two windows are displayed by fluid showing the widget hierarchy and fluid's preview. The right window is from the running application. I have only tried it on Macintosh 10.1.5, w/Codewarrior 8.3, but due to the nature of the problem I guessed it probably effects all platforms (but I could be wrong!) | ||
mike 13:37 May 04, 2003 | This is as designed in FLTK 1.x; FLTK 2.0 is more consistent in this regard... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Copyright notices doesn't mention LGPL exceptions |
Version: | 1.1-current |
Created By: | nickasil |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nickasil 09:35 May 09, 2003 | The copyright notice of each source files doesn't mention that FLTK is distributed under the LGPL with exceptions. These exceptions can introduice trouble if somebody includes some pure LGPL code in FLTK without permission of the author. All copyright notices must be updated to show these license exceptions. | ||
mike 08:27 May 20, 2003 | Working on it... | ||
mike 21:23 Mar 10, 2004 | Determined to not be an issue... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #37 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Installatin Problem |
Version: | 1.1.3 |
Created By: | savvassg.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
savvassg.yahoo 03:24 May 20, 2003 | I have recently downloaded the FLTK 1.1.3 Source Code (.tar.gz) and I am trying to install it in a HP workstation running HP-UX B.11.00. Is there any problem installing or compiling the software under this operating system? The compiler I use is the default cc compiler of the operating system. The first step I am following is to run the configure command in order to change the installation directrory. Running this command the following message is appearing and the programm exits : configure: error: Configure could not find required X11 libraries Which are the X11 libraries that are required and which are the searching paths? Should I install some special X11 libraries on my workstation? | ||
mike 05:11 May 20, 2003 | Dupe of #37. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | seg-fault on fl_width() when called before Fl::run() is called |
Version: | 1.1.3 |
Created By: | dmoos.esigma-systems |
Assigned To: | spitzak |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dmoos.esigma-systems 23:49 May 21, 2003 | Calling fl_width(const char*) before Fl::run() is called results in a segmentation-fault. This is on X11 with fltk-1.3.1 Compilable code that demonstrates it: #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/fl_draw.H> int main(int argc, char **argv) { Fl_Window *window = new Fl_Window(100,100); window->end(); window->show(); double lala = fl_width("blurb"); Fl::run(); return 0; } Thanks for your time. Greetings, Darius. | ||
mike 05:08 May 25, 2003 | [From Bill Spitzak] You need to set the font with fl_font(...) before calling fl_width. The segfault is on purpose to make you do this. Some previous versions used a default font, but the problem there was that if any drawing had already been done it left the font some random value, and code that was measuring things would come out different sizes depending on what was already done. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | DND and FL_PASTE on OSX |
Version: | 1.1.4rc1 |
Created By: | jason.bryan |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jason.bryan 11:10 May 23, 2003 | i have noticed this during drag and drop on OSX: the FL_PASTE event is not sent to the widget currently under the mouse. instead, it always goes to whichever widget has the keyboard focus. the same code works under linux and windows, but not OSX. | ||
mike 07:02 May 26, 2003 | Hmm, OK, it looks like the handler doesn't focus() the widget under the mouse before it pastes; seeing what needs to be done... | ||
mike 07:24 May 26, 2003 | OK, I've done some more testing and am unable to see the problem you have with the "input" demo provided with FLTK. Can you provide additional information about how you are trying to use DND and under what conditions it fails. Also, can you provide a (small) sample program that demonstrates the problem? | ||
mike 21:42 Jun 14, 2003 | Closed after no response for 2 weeks. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Help_View does not handle tables with border properly (and more) |
Version: | 1.1.4rc1 |
Created By: | zaplia.mail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zaplia.mail 02:38 May 27, 2003 | 1.if table cell has </P> tag, it make additional border-lines in wrong place. 2.if table cell has <P> tag at the beginning of content it incorrectly makes new line. 4. There is a problem formatting tables with big width defined (with WIDTH property) 3.&ndash and some other symbols are not displayed. <TABLE BORDER=1 CELLPADDING=0 CELLSPACING=3> <TR> <TD> C1Ш </TD> <TD> C2Щ </TD> <TD> C3 </TD> <TD> C4 </TD> </TR> <TR> <TD><P> C1Ш </P> </TD> <TD><P> C2Щ </P> </TD> <TD><P> C3 </P> </TD> <TD><P> C4 </P> </TD> </TR> <TR> <TD><P> C1Ш </P> </TD> <TD><P> C2Щ </P> </TD> <TD><P> C3 </P> </TD> <TD><P> C4 </P> </TD> </TR> </TABLE> | ||
mike 07:06 May 27, 2003 | Fl_Help_View is not a true HTML widget; it only supports a limited subset of HTML and most certainly doesn't support Unicode character entities (since FLTK doesn't support them, yet) You can re-file as an RFE against 2.0cvs, but this will not be corrected in 1.x. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Help_View make extra newline in list item if it starts from <P> |
Version: | 1.1.4rc1 |
Created By: | zaplia.mail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zaplia.mail 03:28 May 27, 2003 | Improper handling of <P> tag inside another tag is shown. It also valid for table cells. try the next text: <UL> <LI> i1 </LI> <LI> <P> i2 </P></LI> <LI> <P> i3 </P></LI> <LI> <P> i4 </P></LI> <LI> <P> i5 </P></LI> </UL> | ||
mike 07:09 May 27, 2003 | Fl_Help_View is not a true HTML widget; it only implements a subset of HTML and P elements nested inside LI elements are not supported. You can file this as an RFE against 2.0cvs, however this will not be changed in 1.x. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Modal window should not prevent from operating with help window |
Version: | 1.1.4rc1 |
Created By: | zaplia.mail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zaplia.mail 03:39 May 27, 2003 | Now there are next kinds of window: normal,modal,nonmodal. Help window should stay operational if it was displayed once, regardless which new window shown (modal or not). At present, if modal window shown, help window is blocked from events receiving. So: it would be good to have special kind of window : help_window, which will coexist with modal windows. | ||
mike 07:11 May 27, 2003 | We will not be adding a new type of window. Assuming that we could somehow simulate this on all OS's (and I don't think we could unless the help window was provided by a separate process), it would mean a substantial change to FLTK for little benefit. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | fluid: selecting an object in a tab does not bring tab to foreground |
Version: | 1.1.4rc1 |
Created By: | burgers.ecn |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
burgers.ecn 06:15 May 27, 2003 | If you select in the fluid program an object in a tab, the object is selected (red contour shown), but the tab to which it belongs, is not brought to the foreground which I would have expected. ONly the red box is visible, but not the object itself. | ||
mike 07:07 May 27, 2003 | This behavior is as designed; if you wish to re-file this as an RFE against 2.0cvs, please feel free, however it will not be changed in 1.x. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Config Files |
Summary: | fltk.config --prefix=DIR not working |
Version: | 1.1.4rc1 |
Created By: | ciudadsatan.hotmail |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ciudadsatan.hotmail 09:07 May 28, 2003 | Given this command: fltk-config --prefix=/home/fltk114cvs --compile dd.cxx It executes: gcc -I/usr/local/include [...] shouldn't it include /home/fltk114cvs/ ? A trailing "/" in the prefix doesn't help, either. The "--prefix=DIR" option works only with the "--prefix" and "--exec-prefix" options, which isn't... er... useful. | ||
mike 11:45 May 29, 2003 | The --prefix functionality has never worked, and it will be removed in a future release. The correct method is to use the --prefix option at configure time and then use the corresponding version of fltk-config. | ||
greg.ercolano 17:47 Aug 29, 2003 | Hmm, it still doesn't seem to work, even if you build FLTK (1.1.4rc2) with 'config --prefix /some/path' first: ---- snip % cd /usr/local/src/fltk-1.1.4rc2 % ./configure --prefix /usr/local/src/fltk-1.1.4rc2 [no errors] % make [no errors] % cd /var/tmp % /usr/local/src/fltk-1.1.4rc2/fltk-config --compile foo.cxx gcc -I/usr/local/src/fltk-1.1.4rc2/include [..] --- snip ^^^^^^^ The problem with the above is the "/include" appended to the -I prefix path; the fltk base directory does not have an 'include' subdir; "FL" is at the top level. I can work around if I go into the fltk-1.1.4rc2 directory and create an 'include' symlink to ".", then it will work: ---- % cd /usr/local/src/fltk-1.1.4rc2 % ln -s . include % cd /var/tmp % /usr/local/src/fltk-1.1.4rc2/fltk-config --compile foo.cxx gcc -I/usr/local/src/fltk-1.1.4rc2/include [..] [compiles OK] ---- | ||
greg.ercolano 17:54 Aug 29, 2003 | PS. Hmm, somehow the form posting wordwrapped my carats (^^^^^^^) to the next line during submit, even though the WYSIWYG text viewer didn't show it :/ Might be my Mozilla client, or it might be the php. | ||
mike 18:10 Aug 29, 2003 | Try using the --includedir=foo option when you configure. The fltk-config script isn't designed for use with a non-installed source build, and we don't support it as such. | ||
greg.ercolano 19:21 Aug 29, 2003 | Thanks; Re-resolved. I found if I make the prefix same as the includedir, it worked. Thought they'd be the same by default, but I can see why by default they're not. So this worked for me: ./configure --prefix /usr/local/src/fltk-1.1.4rc2 --includedir=/usr/local/src/fltk-1.1.4rc2 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | OS X CodeWarrior project file is incomplete |
Version: | 1.1-current |
Created By: | matt |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
matt 09:24 May 28, 2003 | The OS X version of the CodeWarrior project file does not generate an FLTK library. It is merely a rough layout to generate "hello". | ||
mike 12:59 Mar 25, 2005 | |||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | need "FL/Fl_xyz_Button.H" inorder to compile project |
Version: | 1.1.0 |
Created By: | ggutierrez.marketscan |
Assigned To: | matthiaswm |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ggutierrez.marketscan 07:22 May 30, 2003 | I need "FL/Fl_xyz_Button.H" in order to compile project but I can find any where to get it. | ||
ciudadsatan.hotmail 16:20 May 30, 2003 | Fl_xyz_Button? That class doesn't exist, unless you created it. Are you asking about the example in the documentation (common.html) ? Add FL/Fl_Button.H, FL/Fl_Light_Button.H, FL/Fl_Round_Button.H or whatever you are using. Ask this kind of things in the newsgroup ;) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Can not start FLUID. |
Version: | 1.1.0 |
Created By: | shiqizheng147.sina |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shiqizheng147.sina 05:22 Jun 02, 2003 | I am using Red Hat Linux 6.0. After compile FLUID,I start it with ./fluid, but it says:"Can not get font.",then exit. Please tell me how to fix this problem??? Thanks very much. | ||
mike 10:45 Jun 05, 2003 | Please post general help/usage questions to the fltk.general group/mailing list. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | probelm using FLTK with opencv |
Version: | 1.1.4rc1 |
Created By: | sarfira_jat.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sarfira_jat.yahoo 19:25 Jun 04, 2003 | hello all, i was trying to use fltk woth open cv when i getthe following error. I tried serching for the files realted to the error but have been unlucky. can someone please help me out with this thank you Linking... fltkd.lib(Fl_x.obj) : error LNK2001: unresolved external symbol __imp___TrackMouseEvent@4 Release/FaceDetect.exe : fatal error LNK1120: 1 unresolved externals Error executing link.exe. FaceDetect.exe - 2 error(s), 0 warning(s) | ||
mike 16:25 Jun 11, 2003 | Please read the FAQ - you are missing a library on your link line. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | X11 Libraries |
Version: | 1.1.3 |
Created By: | savvassg.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
savvassg.yahoo 01:31 Jun 05, 2003 | My problem is that when I run the configure file the following message appears interrupting the configuration procedure : checking for X... no configure: error: Configure could not find required X11 libraries So I would like to ask which are the required X11 libraries? Thank you. | ||
mike 16:29 Jun 11, 2003 | Please post this usage question to the fltk.general newsgroup. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Fl_Menu_ item select with mouse and keyboard |
Version: | 1.1.4rc1 |
Created By: | ciudadsatan.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ciudadsatan.hotmail 04:08 Jun 08, 2003 | This has been tested in mingw/win32 only, although I think it applies to all systems: When selecting a subitem from a Fl_Menu_*, whenever the mouse is moved, the selection is screwed up or something. It doesn't matter where the mouse moves, as long as is inside the window. As a test, in the "menubar" demo, highlight some option of the MenuBar, then try navigating with the keyboard while moving the mouse elsewhere: you can't. It seems like moving the mouse outside the menu resets the selection to none. This causes some tricky tricks with pop-up menus, for example. In the same demo, open the pop-up menu and suppose it is very long, so that to select the last item you need to scroll a long way. There is a shortcut: open the pop-up menu, move the mouse outside the menu, then press the "up" key. Voila! Last item selected. | ||
mike 16:24 Jun 11, 2003 | This does not happen on Linux or OSX, so it might be specific to Windows. What version of Windows are you using? | ||
ciudadsatan.hotmail 17:10 Jun 11, 2003 | Windows 98, Second Edition | ||
mike 18:12 Jun 11, 2003 | This is a known side-effect under Windows 9x - they don't support tracking entering/leaving windows properly, which causes the problem. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | tile test shutting down twm |
Version: | 1.1.4rc1 |
Created By: | mnicolet.satlink |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mnicolet.satlink 09:09 Jun 08, 2003 | Running all test programs, when playing with tile, twm stopped running ( don´t know if orderly or sigsegv´d ). Of course, the visible effect was all decorations vanished. The strange thing was that a previous program window appeared ( specifically, navigation ). But I could type ´exit´ on my ´login´ xterm ( the one on the foreground ) to orderly terminate the X session, that is, the server was up and healthy. Later, I tried to reproduce the thing, this time redirecting twm stdout and stderr to files, so catching any messages. But all went well playing some minutes with tile. I am using XFree86 4.2, on QNX 4.25 compiling with Watcom 10.6, and the official ( QSSL ) TCP/IP stack and toolkit. Moreover, because my version of the TCP stack lacks support for Unix sockets, all the X suite is compiled to use only INET sockets. I am a newbie to X. I builded the X suite some two weeks ago, and the first ´extensive´ usage was this: looking at fltk widgets ... so I cannot say if twm is intrinsically stable in my build. Thank you in advance | ||
mike 16:22 Jun 11, 2003 | Unable to duplicate on other platforms; does this happen with other apps? | ||
mnicolet.satlink 09:23 Jun 19, 2003 | Thank you. As said, I am starting to use X, I am not using a graphical environment all the time ( because I don´t had the time to get a decent tools kit - I have a very good editor, the fltk one, but I lacks some other things by now ) so I cannot tell if some other app could bring down twm. Nevetheless, some people has told me that twm is fragile or buggy, so I downloaded and compiled fvwm. May be the question is twm ... but if you could not reproduce the condition, may be it´s _my_ twm build which fails. And that is my trouble: being X a so complex software, I am not absolutely sure my build of the fundamental pieces are OK. In fact ... how could I test it extensively ? Is there some test suite ? Thank you again Marcelo Nicolet | ||
mike 20:15 Jul 13, 2003 | I don't know of any specific test suites you can use, sorry. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Image Support |
Summary: | Specify image data from a buffer than a file |
Version: | 1.1.4 |
Created By: | caruccio.operamail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
caruccio.operamail 10:18 Jun 10, 2003 | Hi, I'm using Fl_PNG_Image to show a PNG image downloaded from a site, but this image can't be save to a file (no disk write permission). There are any FL_PNG_Image's public method (or trick) to define a image's data using a *char buffer ? If I just derive Fl_PNG_Image class and use data() ? Any inconvenient ? tnx, Mateus Caruccio | ||
mike 16:35 Jun 11, 2003 | Your best bet is to copy the current Fl_PNG_Image code and add the appropriate hooks to provide the PNG data in a memory buffer to the PNG library. FLTK 2.0 will probably support C++ streams, so you'll be able to load from a memory stream using the normal C++ mechanisms. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | linker error |
Version: | 1.1.4rc1 |
Created By: | lorenz.giefing.utanet |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
lorenz.giefing.utanet 12:09 Jun 11, 2003 | using vc6.0 on w2k: fltkd.lib(Fl_x.obj) : error LNK2001: Nichtaufgeloestes externes Symbol __imp___TrackMouseEvent@4 USE_TRACK_MOUSE defined! thanks! | ||
mike 16:30 Jun 11, 2003 | Please read the FAQ. You are missing a library on your link line. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #77 |
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | FL_GL_Window in FL_Tabs does not hide properly |
Version: | 1.1.3 |
Created By: | salah.gris.uni-tuebingen |
Assigned To: | matthiaswm |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
salah.gris.uni-tuebingen 01:48 Jun 13, 2003 | This problem also exist in Windows. I am using fltk with windows and MS visual C++. I have an FL_Tabs with three groups, one of them contains an FL_GL_Window. After this FL_GL_Window is shown for the fisrt time, it will not hide when you choose other groups | ||
mike 13:54 Jun 14, 2003 | Duplicate of 77. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #77 |
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | FL_GL_Window in FL_Tabs does not hide properly |
Version: | 1.1.3 |
Created By: | salah.gris.uni-tuebingen |
Assigned To: | matthiaswm |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
salah.gris.uni-tuebingen 01:49 Jun 13, 2003 | This problem also exist in Windows. I am using fltk with windows and MS visual C++. I have an FL_Tabs with three groups, one of them contains an FL_GL_Window. After this FL_GL_Window is shown for the fisrt time, it will not hide when you choose other groups | ||
mike 13:54 Jun 14, 2003 | Duplicate of 77. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | menu - scroll widget crash - windows only |
Version: | 1.1.4rc1 |
Created By: | pdevries.ebw-electronics |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
pdevries.ebw-electronics 06:35 Jun 18, 2003 | test.cxx 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
pdevries.ebw-electronics 06:35 Jun 18, 2003 | I have a little test program which crashes under windows98, but not under linux. The program involves Fl_menu, Fl_tabs, Fl_scroll, and Fl_value_output. I can't find a place on this form to post the test program, and so If I can't figure out how to post it, please tell me where to e-mail it. | ||
mike 20:13 Jul 13, 2003 | I am unable to reproduce the problem with your test program; I'm guessing that you are running into a compiler problem. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Building fltk with VisualStudio.NET on Windows XP |
Version: | 1.1.4rc1 |
Created By: | joettaa.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joettaa.hotmail 20:48 Jun 18, 2003 | Hi, In attempting to build fltk.sln with Visual C++ in VisualStudio.NET(academic version) on Windows XP operating system I got the following error ------ Build started: Project: threads, Configuration: Debug Win32 ------ Compiling... threads.cxx e:\Registration_Segmentation\fltk-1.1.4rc1\test\threads.cxx(64) : error C2668: 'sqrt' : ambiguous call to overloaded function c:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(626): could be 'long double sqrt(long double)' c:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(578): or 'float sqrt(float)' c:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(200): or 'double sqrt(double)' while trying to match the argument list '(int)' Build log was saved at "file://e:\Registration_Segmentation\fltk-1.1.4rc1\visualc\threads_\BuildLog.htm " threads - 1 error(s), 0 warning(s) Of note, I have previously successfully built fltk using Visual C++ in VisualStudio.NET(academic version) on Windows 2000 operating system without any errors. I tried doing a search regarding this problem in the archives but didn't find anything with regard to this problem. Do you know how to fix this problem - is it an fltk problem, a Microsoft problem, or something else? Your help/suggestions are greatly appreciated. Joetta | ||
mike 20:30 Jul 13, 2003 | Make sure you have the latest service pack... Unfortunately, none of the developers use VC.NET so there isn't much we can do... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl::wait() not Fl::flush()ing |
Version: | 1.1.4rc1 |
Created By: | wilson.afn |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wilson.afn 10:51 Jun 20, 2003 | From the documentation for Fl::wait(): "What this really does is call all idle callbacks, all elapsed timeouts, call Fl::flush() to get the screen to update, and then wait ... for any events from the user ..." I have an application where MyWidget::handle() delegates to MyWidget::draw() by setting specific damage bits and relying on FLTK to call draw() *before* handle() is presented the next user event. This does not appear to be happening. Example code: http://www.afn.org/~wilson/foo.fl | ||
mike 20:21 Jul 13, 2003 | This is as designed; if the argument passed to Fl::wait() is 0.0 or less, then the flush is done after; otherwise it is done before. Use Fl::flush() before calling Fl::wait()... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | FLUID |
Summary: | FLUID generated code is missing initializers |
Version: | 1.1.3 |
Created By: | dan.photon |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
dan.photon 10:50 Jul 29, 2003 | test.fl 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dan.photon 14:03 Jun 24, 2003 | Fluid generated code is missing initializers for FL_Menu_Item. Produces a significant number of warning messages. This occurs when using GCC 3.x on RedHat Linux 8.0 (probably same on all platforms). | ||
mike 20:40 Jul 13, 2003 | Please post a sample .fl file that demonstrates the problem. | ||
mike 19:09 Jul 28, 2003 | Closed with no response in 2 weeks. | ||
dan.photon 10:51 Jul 29, 2003 | The warnings only show up when compiled with the -W option. This occurs for any simple menu as shown in the example fluid file. | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | OpenGL |
Summary: | FLTK on multiple monitors |
Version: | 1.1.4rc1 |
Created By: | yao.mit |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yao.mit 12:04 Jun 26, 2003 | I was using Fl_Gl_Window to draw on 2 monitors. Although I initialized the window's position as (1280, 0) which is the upper-left corner of the 2nd monitors, the window just aligned with the right side of the 1nd monitor and didn't go to the 2nd monitor at all. I am using VC.NET in WinXP and Nvidia GeForce4 Ti 4200 card. My 1st monitor uses the digital output and 2nd uses the analog. Thanks, Yao | ||
mike 20:25 Jul 13, 2003 | OpenGL rendering across multiple monitors may or may not work depending on the OpenGL drivers that come with the graphics card. This isn't a FLTK issue. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Config Files |
Summary: | missing Fl_Help_View on libfltk_images.a |
Version: | 1.1.4rc1 |
Created By: | caruccio.operamail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
caruccio.operamail 12:42 Jun 26, 2003 | We are unable to link against libfltk_images.a cause Fl_Help_View.cxx is missing on IMGCPPFILES var in src/Makefile. I'm using Mingw32, but think this should happen on any system. Mateus Caruccio | ||
mike 20:08 Jul 13, 2003 | Fl_Help_View is part of libfltk.a | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | FL_GL_Window |
Version: | 1.1.3 |
Created By: | salah.gris.uni-tuebingen |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
salah.gris.uni-tuebingen 01:53 Jul 08, 2003 | I am using an image viewer that inherits from an FL_GL_WINDOW. Problem: The first time the widget is shown, it does not display the image data! Instead, it displays either a clean grey window (on Windows 2000) or a garbage-filled window (on Windows XP). It displays the image data only after I interact with the Widget, a mouse click inside is enough. | ||
mike 20:23 Jul 13, 2003 | We'll need sample code to reproduce the problem. | ||
mike 19:07 Jul 28, 2003 | Closed with no response in 3 weeks. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | fltk-config |
Version: | 1.0.11 |
Created By: | mchittur.cs.nmsu |
Assigned To: | matthiaswm |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mchittur.cs.nmsu 10:59 Jul 14, 2003 | I'm trying to compile a application in NetBSD, which requires FLTK. I installed the package fltk-1.0.11nb2. There is no fltk-config executable present. Thankx for the help in advance. | ||
matt 20:41 Jul 17, 2003 | The correct way to configure nd compile FLTK is to call ./configure and then make from within the fltk directory. I'd also recommend that you download fltk 1.1.3 . It is 99% source code compatible to 1.0.11 and fixes quite a lot of bugs. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #102 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | config.h |
Version: | 1.1.3 |
Created By: | even68.aol |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
even68.aol 15:24 Jul 22, 2003 | Re: STR 102. Thank you for replying, but I didn't get the information I need. I have copied all of the appropriate files to the main diectory, including config.mingw. Is config.mingw the file you say I didn't copy/move? I did. Would you please direct me to the location of the config.h file make is looking for and also where I should copy it to? Guess I need more specific instructions. Again, thanks, Nikki | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Memory leak in Fl_Menu_Item |
Version: | 1.1.4rc1 |
Created By: | dicuccio.ncbi.nlm.nih.gov |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dicuccio.ncbi.nlm.nih.gov 05:27 Jul 25, 2003 | If I create a menu in FLUID, the code generated for the menu uses the Fl_Menu_::menu() function to generate the menu widgets. If I then dynamically add to the menu using Fl_Menu_::add(), new Fl_Menu_Item objects are created. The text labels for these items are allocated via strdup(), and are not freed unless the whole menu was created with Fl_Menu_::copy(). | ||
mike 13:42 Jul 25, 2003 | Unless you can supply a fix, this won't be corrected; FLTK 2.0 replaces this code, and you can use the Fl_Menu_::copy() method to create a copy of the menu prior to manipulating it... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | test/shiny broken |
Version: | 1.1.4rc2 |
Created By: | wilson.afn |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wilson.afn 12:48 Jul 29, 2003 | Shiny's buttons appear momentarily when sliders are moved, but are soon obscured by black veil. RH9/GCC 3.2.2/XF86 4.3.0 but NVIDIA souped-up (proprietary) driver. (Rc1's shiny works; other OGL demos work). | ||
mike 05:42 Aug 02, 2003 | First, the shiny demo is deprecated. Second, I just tried it with the standard XFree86 OpenGL drivers with an ATI Radeon card and did not see the black boxes; it is probably a bug in the NVIDIA driver... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | adds FL_UPPERCASE_INPUT type |
Version: | 1.1.3 |
Created By: | sandersd.aries.tucson.saic |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sandersd.aries.tucson.saic 14:27 Jul 29, 2003 | I am developing an fltk application for which valid alphabetical input must be uppercase. To make this easier on the user I have been patching each fltk version to provide an uppercase input type. When this type is set a-z are translated to A-Z automatically and all other characters are passed unmodified. In other words it is impossible to enter lowercase letters. This patching is kind of a pain so I was hoping that if the developers think this would be a useful addition to fltk that they would make it a permanent feature... Here are my patches (to fltk-1.1.3) via diff -u Fl_Input_.H: --- /c/net/fltk/fltk-1.1.3/FL/Fl_Input_.H 2003-01-30 14:39:23.000000000 -0700 +++ Fl_Input_.H 2003-07-29 09:30:29.000000000 -0700 @@ -1,5 +1,5 @@ // -// "$Id: Fl_Input_.H,v 1.6.2.4.2.8 2003/01/30 21:39:23 easysw Exp $" +// "$Id: Fl_Input_.H,v 1.1 2003/05/01 15:21:20 sandersd Exp sandersd $" // // Input base class header file for the Fast Light Tool Kit (FLTK). // @@ -43,6 +43,7 @@ #define FL_INPUT_WRAP 16 #define FL_MULTILINE_INPUT_WRAP (FL_MULTILINE_INPUT | FL_INPUT_WRAP) #define FL_MULTILINE_OUTPUT_WRAP (FL_MULTILINE_INPUT | FL_INPUT_READONLY | FL_INPUT_WRAP) +#define FL_UPPERCASE_INPUT 6 class FL_EXPORT Fl_Input_ : public Fl_Widget { @@ -138,5 +139,5 @@ #endif // -// End of "$Id: Fl_Input_.H,v 1.6.2.4.2.8 2003/01/30 21:39:23 easysw Exp $". +// End of "$Id: Fl_Input_.H,v 1.1 2003/05/01 15:21:20 sandersd Exp sandersd $". // Fl_Input.cxx: --- /c/net/fltk/fltk-1.1.3/src/Fl_Input.cxx 2003-01-30 14:41:57.000000000 -0700 +++ Fl_Input.cxx 2003-07-29 09:30:28.000000000 -0700 @@ -1,5 +1,5 @@ // -// "$Id: Fl_Input.cxx,v 1.10.2.15.2.17 2003/01/30 21:41:57 easysw Exp $" +// "$Id: Fl_Input.cxx,v 1.1 2003/05/01 15:21:11 sandersd Exp sandersd $" // // Input widget for the Fast Light Tool Kit (FLTK). // @@ -34,6 +34,7 @@ #include <FL/fl_draw.H> #include <FL/fl_ask.H> #include "flstring.h" +#include <stdlib.h> void Fl_Input::draw() { if (input_type() == FL_HIDDEN_INPUT) return; @@ -87,7 +88,15 @@ if (del || Fl::event_length()) { if (readonly()) fl_beep(); - else replace(position(), del ? position()-del : mark(), + else if (type() == FL_UPPERCASE_INPUT){ + char *upper = strdup(Fl::event_text()); + for(char *up = upper; *up; ++up) + if(islower(*up)) *up = toupper(*up); + replace(position(), del ? position()-del : mark(), + upper, Fl::event_length()); + free(upper); + } + else replace(position(), del ? position()-del : mark(), Fl::event_text(), Fl::event_length()); } return 1; @@ -407,5 +416,5 @@ } // -// End of "$Id: Fl_Input.cxx,v 1.10.2.15.2.17 2003/01/30 21:41:57 easysw Exp $". +// End of "$Id: Fl_Input.cxx,v 1.1 2003/05/01 15:21:11 sandersd Exp sandersd $". // | ||
mike 06:51 Aug 02, 2003 | OK, based upon the discussions on the FLTK developer newsgroup, I think it makes more sense to use a callback or your own subclass than to make this part of the FLTK core. At the very least we'd want to extend this for arbitrary filtering... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | OpenGL |
Summary: | Fl_Double_Window not swapping buffers on initialisation |
Version: | 1.1.4rc2 |
Created By: | ariock.bluebottle |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ariock.bluebottle 02:55 Aug 12, 2003 | I have a Fl_Gl_Window inside a Fl_Double_Window, and when my app is started I tend to have my GL context not rendered until I do some action that make my app to ask for a render explicitely. This seems to happen only afther a launched my app once. The version of Fltk i'm using is 1.1.4rc2, and I compile under MVS 6.0 | ||
mike 08:24 Aug 18, 2003 | Can you provide a short example program that demonstrates this problem? Also, what version of Windows are you using, and what graphics card do you have? | ||
ariock.bluebottle 01:40 Aug 20, 2003 | Sorry but I cannot provide a sample program right now my windows version is: Win2000 pro sp2 openGl version is: Intel Brookdale-G 1.3.0 - Build 4.13.01.3069 The trouble seem to occur only on a few machine, and it's not dependent of my program becaus on other computers with different implementation of openGl, it does not occur... | ||
mike 10:41 Aug 20, 2003 | OK, then the problem is likely with the graphics driver... | ||
ariock.bluebottle 23:01 Aug 20, 2003 | well I guess not because with the Fltk 1.1.4.rc1 the problem did not occur at all... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | makeinclude sets CXX=gcc, should be "g++" |
Version: | 1.0.11 |
Created By: | larry.siden.myself |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
larry.siden.myself 14:00 Aug 12, 2003 | In order to get fltk 1.0.11 to compile, I had to manually modify "makeinclude" after running configure. I had to change the line "CXX=gcc" to "CXX=g++". Otherwise, the sources would compile, but they would not link, because gcc (the C compiler) invokes the linker with a different set of standard libraries than g++ (the C++ compiler). I have GCC 3.2.3 and I'm running on Linux 2.4.21. | ||
mike 16:26 Aug 12, 2003 | First, 1.0.11 is no longer maintained. Second, the correct fix is to detect and use the supc++ library when using GCC 3.x; otherwise you introduce unnecessary dependencies upon the standard C++ library, of which there are literally hundreds of versions and none of them standard on Linux and other platforms. If you want a fix, grab the changes from the current 1.1.x release. | ||
larry.siden.myself 19:20 Aug 12, 2003 | I had to use 1.0.11 in order to build the Palm OS Emulator, which will not build with later versions b/c of name changes. I do not know how to detect libsupc++ or even where to find it. But I got fltk to build by changing CXX=gcc to CXX=g++. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | Shiny-Demo does not look as it should |
Version: | 1.1.4rc2 |
Created By: | a-goebel.gmx |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
a-goebel.gmx 11:17 Aug 14, 2003 | I just compiled rc2 on a recent cygwin installation. The shiny-demo looks almost black and flickers when adjusting the sliders. The other gl demos work fine. | ||
mike 11:26 Aug 14, 2003 | The shiny demo is obsolete and we are removing it from the final 1.1.4 release (it doesn't work, can't work, and won't work...) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | GLUT Emulation |
Summary: | glutBitmapCharacter() and glutBitmapWidth() declarations are not compatible with GLUT |
Version: | 1.1.4rc2 |
Created By: | yazhuk.ncbi.nlm.nih |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yazhuk.ncbi.nlm.nih 13:22 Aug 14, 2003 | We have a suite of application written using FLTK. We developed a new command line application that uses OpenGL to generate image files and found that calls to functions working with bitmap fonts initiate X connection (what is undesirable) and what is worse - text is not printed in the output. We tried to link to MESA GLUT instead of FLTK library, but found that glutBitmapCharacter() and glutBitmapWidth() are declared as C++ functions while in GLUT their C functions. This makes FLTK and MESA not interchangeable on the linkage stage, what is inconvenient. We resolved this problem by adding "extern "C"" to declarations of the functions in "glut.h". | ||
mike 08:27 Aug 18, 2003 | Please post a question to the fltk.development group about this; I'm hesitant to change the language binding since that will seriously break binary compatibility between FLTK releases. Also, we don't guarantee that the GLUT functions in FLTK will work without a recompile, anyways... | ||
mike 12:19 Sep 03, 2003 | Closed after two weeks of inactivity... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Problems with fl_filename_list |
Version: | 1.1.4rc2 |
Created By: | a-goebel.gmx |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
a-goebel.gmx 11:35 Aug 15, 2003 | The following code makes my application crash: char* dirname = fl_dir_chooser("Wählen Sie ein Verzeichnis", 0); int dateizahl; dirent*** liste; dateizahl = fl_filename_list(dirname, liste); //here it crashes for (int i = dateizahl; i > 0;) { free((void*)(liste[--i])); } free((void*)liste); With 1.1.3 id doesn´t crash, but it freezes. Best regards, Andreas | ||
mike 12:24 Aug 15, 2003 | What operating system and compiler are you using? | ||
a-goebel.gmx 01:30 Aug 16, 2003 | Win32 with recent cygwin installation, compiled with -mno-cygwin. | ||
a-goebel.gmx 01:31 Aug 16, 2003 | Oh, and gcc version 3.2. | ||
a-goebel.gmx 09:04 Aug 16, 2003 | I have made a few more tests. Unfortunately I am not very experienced with debugging. Something very strange happened. The following minimal example works: //Test fuer filename #include <FL/fl.h> #include <FL/filename.H> #include <iostream> int main() { int dateizahl; dirent*** liste; dateizahl = fl_filename_list("C://cygwin//", liste); std::cout << "Die Zahl der Dateien ist: " << dateizahl << "\n"; for (int i=0; i<dateizahl; i++) std::cout << (*liste)[i]->d_name << "\n"; for (int i = dateizahl; i > 0;) { free((void*)(liste[--i])); } free((void*)liste); return 0; } The very same code inside my program doesn´t work! I really do not understand this. I´d rather have thought that not beeing in the main event-loop would result in not letting this code work. Maybe this is all my fault, but unfortunately there is no example for anyone using the fl_filename_list in the demo-programs. Best regards, Andreas | ||
mike 08:34 Aug 18, 2003 | This sounds very much like you are overwriting the end of an array or buffer in your program; there really isn't much that can go wrong with fl_filename_list()... If you were using Linux I would suggest Valgrind; on WIN32 your only options are commercial memory checkers like Purify, as far as I know... | ||
a-goebel.gmx 06:21 Aug 19, 2003 | Thanks for your reply. I made sure that there is no problem with my own code. The following code - just hello world with fl_filename_list query inserted - works with mingw (recent installation), but it crashes with a recent cygwin installation (cygwin disabled). So this either is a cygwin-bug or a bug in your cygwin-branch. This is not so improbable, as the code for fl_filename_list is highly system-dependent. Strange though, as cygwin uses mingw anyway... Best regards Andreas //Test fuer filename #include <FL/Fl.h> #include <FL/filename.H> #include <iostream> #include <FL/Fl_Window.H> #include <FL/Fl_Box.H> int main(int argc, char **argv) { Fl_Window *window = new Fl_Window(300,180); Fl_Box *box = new Fl_Box(20,40,260,100,"Hello, World!"); box->box(FL_UP_BOX); box->labelsize(36); box->labelfont(FL_BOLD+FL_ITALIC); box->labeltype(FL_SHADOW_LABEL); window->end(); int dateizahl; dirent*** liste; dateizahl = fl_filename_list("C://cygwin//", liste); std::cout << "Die Zahl der Dateien ist: " << dateizahl << "\n"; for (int i=0; i<dateizahl; i++) std::cout << (*liste)[i]->d_name << "\n"; for (int i = dateizahl; i > 0;) { free((void*)(liste[--i])); } free((void*)liste); window->show(argc, argv); return Fl::run(); } | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Unable to hear sounds when using the "fl_beep" function |
Version: | 1.1.3 |
Created By: | sumac.oregon.usa |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sumac.oregon.usa 10:02 Aug 16, 2003 | OS: Linux Mandrake 9.1 Window Manager: Black Box, WindowMaker, KDE G++: 3.2.2 The "fl_beep" functions compiles fine and the app runs but no sounds are made. Sound system does work. -Daniel | ||
mike 08:22 Aug 18, 2003 | Open the audio mixer and make sure that the PC speaker volume is turned up (the beep is from the PC speaker output, not from the normal PCM output...) | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #133 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | GLUT Emulation |
Summary: | extern "C" for all GLUT symbols |
Version: | 1.1.4rc2 |
Created By: | dicuccio.ncbi.nlm.nih.gov |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dicuccio.ncbi.nlm.nih.gov 17:18 Aug 16, 2003 | FLTK provides implementations of many GLUT functions, but several GLUT API's are not extern "C". I support a large codebase that needs to use both GLUT and FLTK. This makes it impossible to use FLTK's OpenGL headers reliably in such circumstances, without patching glut.H. Can this be changed? A patch is below: diff -urN fltk-1.1.3/FL/glut.H fltk-1.1.3.new/FL/glut.H --- fltk-1.1.3.old/FL/glut.H 2003-01-30 16:40:50.000000000 -0500 +++ fltk-1.1.3/FL/glut.H 2003-06-14 15:45:20.000000000 -0400 @@ -92,6 +92,8 @@ //# define GLUT_API_VERSION This does not match any version of GLUT exactly... +extern "C" { + FL_EXPORT void glutInit(int *argcp, char **argv); // creates first window FL_EXPORT void glutInitDisplayMode(unsigned int mode); @@ -426,8 +428,6 @@ // files in GLUT remain as they currently are so that there are // not symbol conflicts with the above. -extern "C" { - extern int APIENTRY glutExtensionSupported(char *name); /* Stroke font constants (use these in GLUT program). */ | ||
mike 08:29 Aug 18, 2003 | This is a duplicate of STR #133; in summary, if we change this then we break binary compatibility with prior FLTK releases. Also, we don't guarantee binary compatility with GLUT, and specifically require C++ compiles for both GLUT and XForms apps... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | mingw fltk linking problems |
Version: | 1.1.4rc2 |
Created By: | ashishftp1.rediffmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ashishftp1.rediffmail 10:55 Aug 31, 2003 | I am having problems compiling FLTK using mingw .I tried almost every trick known to me,but failed .Or Maybe i dont know enough.These are things i tried out 1.Installed Mingw and Mingsys [gcc version 3.2 (mingw special 20020817-1)] 2.Copied sh.exe and make.exe from Minsys to Mingw\bin 3.In the fltk directory made the following .bat file ///////contents of bat file/////////////////////// set path=%path%;C:\msys\1.0\bin;C:\mingw\bin;C:\mingw\ copy makefiles\config.mingw configure copy makefiles\makefile.mingw makefile make /////////////////////////////////// 4.Now make starts ,libfltk.a and others ARE BUILT ...BUT the procedures stops when LINKING fluid .Other files also dont compile when i tried manually .I also tried using a IDE MinGW Developer Studio.I thing there is a problem while linking to libraries .I dont where the problem lies I have tried with most recent version of fltk .All other things just dont work .I am desperate to get FLTK work with mingw .Please tell me the excat options that i should use with gcc and/or linker to get it done. (Your web site does not look goot with msie) | ||
mike 12:47 Sep 03, 2003 | Please post usage questions to the FLTK newsgroups/mailing lists. You need to copy the files to their correct names (config.h, makeinclude, Makefile) for them to be used! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Changing fonts for dynamic menus does not work OK on some OS/Compilers |
Version: | 1.1.4rc2 |
Created By: | mpaliwoda.interia |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mpaliwoda.interia 06:22 Sep 01, 2003 | Hi, When building menu dynamically by using Fl_Menu_:add() method, after the menu is built, setting fonts for menu items has no effect. For example on Win98 / Bcc551 / Fltk1.1.4rc2, new fonts are used in drawing menu item shortcut, not menu item text itself. Greg Ercolano reported that under some other systems (RH9.0) fonts are not used too. When the menu is statically defined, everything works ok. Attached is a small program demonstrating the problem. ---------- cut --------------- #include <FL/Fl.H> #include <FL/Fl_Menu_.H> #include <FL/Fl_Window.H> #include <FL/Fl_Menu_Bar.H> #include <FL/Fl_Menu_Item.H> /* statically defined menu */ Fl_Menu_Item menutable[] = { {"&Edit", FL_F+2, 0, 0, FL_SUBMENU}, {"Undo", FL_ALT+'z', 0}, {"Redo", FL_ALT+'r', 0, 0, FL_MENU_DIVIDER}, {"Cut", FL_ALT+'x', 0}, {"Copy", FL_ALT+'c', 0}, {"Paste", FL_ALT+'v', 0}, {"Inactive", FL_ALT+'d', 0, 0, FL_MENU_INACTIVE}, {"Clear", 0, 0, 0, FL_MENU_DIVIDER}, {"Invisible", FL_ALT+'e', 0, 0, FL_MENU_INVISIBLE}, {"Preferences", 0, 0}, {"Size",0,0}, {0}, {"&Font", 0, 0, 0, FL_SUBMENU }, {"Normal", 0, 0, 0, 0, 0, 0, 14}, {"Bold", FL_ALT+'1', 0, 0, 0, 0, FL_BOLD, 14}, {"Italic", FL_ALT+'2', 0, 0, 0, 0, FL_ITALIC, 14}, {"BoldItalic",FL_ALT+'3', 0, 0, 0, 0, FL_BOLD+FL_ITALIC, 14}, {"Small", FL_ALT+'4', 0, 0, 0, 0, FL_BOLD+FL_ITALIC, 10}, {"Emboss", FL_ALT+'5', 0, 0, 0, FL_EMBOSSED_LABEL}, {"Engrave", FL_ALT+'6', 0, 0, 0, FL_ENGRAVED_LABEL}, {"Shadow", FL_ALT+'7', 0, 0, 0, FL_SHADOW_LABEL}, {"@->", FL_ALT+'8', 0, 0, 0, FL_SYMBOL_LABEL}, {0}, {0} }; int main(int argc, char **argv) { int i = 65; Fl_Menu_Item *mi; Fl_Window w(300,100,400,200); Fl_Menu_Bar *m = new Fl_Menu_Bar(0, 0,400,20,"Dynamic"); Fl_Menu_Bar *n = new Fl_Menu_Bar(0,20,400,20,"Static"); w.end(); /* setting dynamically defined menu */ m->add((const char *)"File", 0, (Fl_Callback *)0, (void *)0, FL_SUBMENU+0); m->add((const char *)"Font", 0, (Fl_Callback *)0, (void *)0, FL_SUBMENU+0); m->add((const char *)"Font/Normal", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Bold", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Italic", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/BoldItalic",FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Small", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Engrav", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Emboss", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/Shadow", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); m->add((const char *)"Font/@->", FL_ALT+i++, (Fl_Callback *)0, (void *)0, 0); /* setting statically defined menu */ n->menu( menutable ); /* setting general attributes */ m->textsize(12); n->textsize(12); /* setting properties of dynamically added Fl_Menu_Items */ mi = ( (Fl_Menu_Item *)m->menu() ) + 4; /* pointer to Font/Bold Fl_Menu_Item */ ( mi++ )->labelfont( FL_BOLD ); ( mi++ )->labelfont( FL_ITALIC ); ( mi++ )->labelfont( FL_BOLD+FL_ITALIC ); ( mi )->labelfont( FL_BOLD+FL_ITALIC ); ( mi++ )->labelsize( 10 ); ( mi++ )->labeltype( FL_ENGRAVED_LABEL ); ( mi++ )->labeltype( FL_EMBOSSED_LABEL ); ( mi++ )->labeltype( FL_SHADOW_LABEL ); ( mi++ )->labeltype( FL_SYMBOL_LABEL ); w.show(argc, argv); return Fl::run(); } ---------- cut --------------- | ||
mike 12:42 Sep 03, 2003 | I just tried with VC++ 6.0 on Win2k and with GCC 2.96 on RedHat 7.3 and was unable to see a problem - both font menus showed different fonts for each menu item, as expected. I will test on a RH9 system later, but so far I am unable to duplicate your problem. | ||
mike 07:29 Sep 14, 2003 | Unable to duplicate on any platform, closing... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #116 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | shiny.exe show the black screen |
Version: | 1.1.4rc2 |
Created By: | amnikitin.eimb |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
amnikitin.eimb 02:52 Sep 05, 2003 | shiny.exe demo from most of FLTK versions has different problems on WindowsNT (MSVC-6.0). It can be screen flicker or buttons disappearance. shiny.exe from 1.1.4rc2 show the black screen on the start. During window resizing screen flicker show traces of buttons. | ||
mike 05:22 Sep 05, 2003 | Duplicate of 116. Short summary: shiny can't work and is history. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | X11 |
Summary: | Regarding Overlay Window |
Version: | 1.1.4 |
Created By: | contacthaseeb.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
contacthaseeb.yahoo 03:01 Sep 09, 2003 | i want to use Fl_Overlay_Window, now problem is that, when i use Xi Accelerated X Server in 16M Overlay Mode, i can use Overlay but Can't see other controls, but when switch to 16M (non Overlay) mode i can see other widgets but then i am using Overlay emulation not actual hardware overlays. (i am using Matrox G450 with Hardware Overlay) can i use XFree86 for Hardware Overlay? what i want is: One Fl_Window One Fl_overlay_window (on Top of it) and a Control Pannel Besides these two windows. Can i achieve this ? (using Hardware Overlay) | ||
mike 07:17 Sep 14, 2003 | This is a driver issue; consult the documentation that comes with Xi to determine a combination of settings that will allow both overlay and normal visuals. It could also be an Xi bug, but since I don't have Xi on any system it is impossible for me to help you. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Text_Buffer can't save to a file properly |
Version: | 1.1.4 |
Created By: | alexeyp.m7.tts-sf |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
alexeyp.m7.tts-sf 14:32 Sep 09, 2003 | Fl_Text_Buffer::outputfile() method creates an empty read-only file and errs, when trying to save to a not existing file. The solution is to open file in "w+" mode instead of "w" mode. | ||
mike 15:13 Sep 09, 2003 | OK, that doesn't make any sense - "w+" opens a file for read/write access and will create the file with the same permissions as "w"... What OS and compiler are you using? This doesn't appear to be a problem on RH Linux 9, at least. | ||
mike 02:29 Sep 27, 2003 | Closed after two weeks of inactivity... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | crash with cygwin: Error while dumping state |
Version: | 1.1.4 |
Created By: | emmanuel.debanne.free |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
emmanuel.debanne.free 15:02 Sep 15, 2003 | Under Cygwin, my application always crash after one or two minutes. I send a test case in order to reproduce the bug. Build Main.cpp with these commands: $ g++ -I/usr/local/include -mwindows -DWIN32 -g -c Main.cpp $ g++ -g -o main.exe Main.o -L/usr/local/lib -lfltk_images -lpng -ljpeg -lz -lfltk -mwindows -lole32 -luuid -lcomctl32 -lwsock32 And to launch it: $ ./main.exe When "i" reaches about 2600 (by moving the mouse for example), the appplication crashes: Segmentation fault (core dumped) $ cat main.exe.stackdump 76110298 [main] main 714113 handle_exceptions: Error while dumping state (probably corrupted stack) $ fltk-config --version 1.1.4 ////////////// Main.cpp /////////////////////////////// #include <cstdio> #include <string> #include <FL/Fl.H> #include <FL/fl_draw.H> #include <FL/Fl_Output.H> #include <FL/Fl_Window.H> std::string translate( int val ) { char toto[48]; sprintf( toto, "%d", val); return std::string(toto); } Fl_Output* t; class Foo : public Fl_Window { public: Foo() : Fl_Window(0,0,200,200), i(0) { } int i; int handle( int event ) { if ( event == FL_MOVE || event == FL_KEYDOWN ) { ++i; t->value(translate(i).c_str()); redraw(); while(true) { if ( Fl::ready() ) { Fl::check(); } } } return 1; } }; int main( int argc, char** argv ) { Foo* w = new Foo(); t = new Fl_Output(10,10,100,30); w->end(); w->show(argc,argv); return Fl::run(); } | ||
mike 16:56 Sep 15, 2003 | Your handle() method should not recursively call itself; you are overflowing the stack and causing the crash that way... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | MacOS |
Summary: | More case-insensitive filesystem issues (from STR#164) |
Version: | 1.1.4 |
Created By: | dmacks.netspace |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
dmacks.netspace 00:57 Sep 18, 2003 | case-insensitive-flag.patch 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dmacks.netspace 00:00 Sep 16, 2003 | Okay, that worked. But now I see that HLINKS is not the only place where the same problem (using OS to choose determine FS case-sensitivity) occurs. I see it in fluid/fluid.cxx update_history() and src/filename_absolute.cxx fl_filename_relative(), haven't finished browsing the rest of the code yet. Maybe the ./configure flag should be more general (--{with,without}-case-insensitive-FS) that sets a flag in config.h (with default based on $uname) and base fluid.cxx (etc.) behavior on that? | ||
dmacks.netspace 00:41 Sep 16, 2003 | Those two files are the only places __APPLE__ is used to mean "case-insensitive filesystem". | ||
mike 07:36 Sep 16, 2003 | At this point we will not be changing this code; the default filesystem on OSX is case-insensitive, and if we disable this code then other problems WILL occur on HFS filesystems. | ||
dmacks.netspace 12:51 Sep 16, 2003 | That's exactly the opposite of the problem I'm trying to solve (sorry if my description was unclear). I'm on an OS X (and so have __APPLE__ defined) but am using X11 graphics. The easiest way to get that it to undefine __APPLE__, but I still need HFS. I'll write a patch tonight... | ||
dmacks.netspace 01:01 Sep 18, 2003 | 'case-insensitive-flag.patch' (really a cvs diff) makes FS case-insensitivity independent of __APPLE__. Instead it piggybacks on the setting of HLINKS. The --with-links configure flag is now a bit poorly named, but the functionality I need is there in a generic way without breaking anything (I think:). | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fluid crashes on jpeg and png images om win32 (build on mingw or cygwin - rezults are the same) |
Version: | 1.1.4 |
Created By: | freebsd_my.ua |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
freebsd_my.ua 23:21 Sep 21, 2003 | camera.png 7k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
freebsd_my.ua 08:40 Sep 18, 2003 | fluid.exe caused an Access Violation at location 0022f698 Writing to location 6de0ac60. Registers: eax=6de0ac60 ebx=0022f698 ecx=6de024d4 edx=77c29473 esi=00440b42 edi=0022faec eip=0022f698 esp=0022f5a0 ebp=0022f6f8 iopl=0 nv up ei pl nz na po nc cs=001b ss=0023 ds=0023 es=0023 fs=0038 gs=0000 efl=00000206 Call stack: 0022F698 004836AA fluid.exe:004836AA fl_check_images(char const*, unsigned char*, int) fl_images_core.cxx:84 ... #ifdef HAVE_LIBPNG if (memcmp(header, "\211PNG", 4) == 0)// PNG file > return new Fl_PNG_Image(name); #endif // HAVE_LIBPNG ... 0D000000 | ||
mike 11:37 Sep 20, 2003 | We'll need a whole lot more information to provide you with any support. What version of Mingw or Cygwin are you using? What version of PNG, JPEG, and ZLIB are you using? Can you attach one or more test images that exibit this problem? Does the same thing happen with the file_chooser demo? | ||
freebsd_my.ua 23:26 Sep 21, 2003 | mingw version (this is the latest(but on older versions start from 1.0.8 is the same error)) MINGW32_NT-5.1 ANDRIYP 1.0.10(0.46/3/2) 2003-07-29 09:21 i686 file_chooser from tests(without debug, but error is in the same location) file_chooser.exe caused an Access Violation at location 0022f2e8 Writing to location 6de0ac60. Registers: eax=6de0ac60 ebx=0022f2e8 ecx=6de024d4 edx=77c29473 esi=004307c2 edi=0022faec eip=0022f2e8 esp=0022f1f0 ebp=0022f348 iopl=0 nv up ei pl nz na po nc cs=001b ss=0023 ds=0023 es=0023 fs=0038 gs=0000 efl=00000206 Call stack: AddrPC AddrReturn AddrFrame AddrStack Params 0022F2E8 004061B3 0022F348 0022F1F0 003F7C88 003F93B8 00000004 77C23EB0 0022F2E8 004061B3 0D000000 0022F368 0022F1F0 003F93B8 0022F390 00000040 77C4ACE0 004061B3 file_chooser.exe:004061B3 fl_check_images(char const*, unsigned char*, int) fl_images_core.cxx:84 0D000000 00000000 0A1A0A0D 0022F1F0 00000000 00000000 00000000 00000000 0D000000 | ||
freebsd_my.ua 23:32 Sep 21, 2003 | libpng and zlib - latest versions if I compile with VC++ 6 - in the same location fluid crashes! With bmp images no such error, and for now i do png2bmp and work so, but this is not normal 8( | ||
mike 17:18 Oct 31, 2003 | I am unable to reproduce this on any platform using the current ZLIB and PNG libraries. I would suspect that you are running into a PNG library bug and not a problem in FLTK... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | win32 font patch |
Version: | 1.1.4 |
Created By: | freebsd_my.ua |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
freebsd_my.ua 08:46 Sep 18, 2003 | diff -ru3p fltk-1.1.4rc2-my/src/Fl_Widget.cxx fltk-1.1.4rc2/src/Fl_Widget.cxx --- fltk-1.1.4rc2-my/src/Fl_Widget.cxx Mon Sep 8 10:44:58 2003 +++ fltk-1.1.4rc2/src/Fl_Widget.cxx Thu Jan 30 21:42:58 2003 @@ -72,7 +72,7 @@ int Fl_Widget::handle(int) { return 0; } -int FL_NORMAL_SIZE = 11; +int FL_NORMAL_SIZE = 14; Fl_Widget::Fl_Widget(int X, int Y, int W, int H, const char* L) { | ||
mike 11:46 Sep 20, 2003 | We will not change the default font size. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fl_filename_list and fl_dir_chooser malfunction |
Version: | 1.1.4 |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 13:50 Sep 26, 2003 | Contrary to the documentation fl_filename_list expects a trailing '/' (from the manual: "d is the name of a directory; it does not matter if it has a trailing slash or not."). <code> d = "./test/"; dir = fl_filename_list(d, list); </code> gives me a list of files in directory 'test', while <code> d = "./test"; dir = fl_filename_list(d, list); </code> results in a list with the single entry 'test'. fl_dir_chooser cuts off the trailing slash in a directory string so the result looks like "./test" although the dialog box shows me "./test/" (this behaviour led me to the fl_filename_list problem). My configuration: PC PIII, Windows2000, Gnu compiler. | ||
mike 02:29 Sep 27, 2003 | Hmm, can't seem to duplicate this with VC++. Are you using MingW or Cygwin or ??? | ||
kontakt.andreas-schoemann 13:03 Sep 27, 2003 | I use MingW 3.2.3. | ||
mike 17:37 Oct 31, 2003 | Well, I'm unable to duplicate the problem... :( | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Win32/RegisterWindowClass, two FLTKstatic-linked DLL loaded to one app cause crash |
Version: | 1.1.4 |
Created By: | gyohng.netscape |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
gyohng.netscape 11:31 Oct 01, 2003 | A crash happens, if side application is loading 2 different dll, and both of dlls have FLTK linked statically. 1. App loads DLL1 2. App shows DLL1 window 3. DLL1 registers window class 'FLTK' 4. DLL1 uses class 'FLTK' to show window 5. App loads DLL2 6. App shows DLL2 window 7. DLL2 registers class 'FLTK' (ignored by windows - class exists) 8. DLL2 uses class 'FLTK' from DLL1 9. DLL1's WndProc is fooled with Fl_Window, belonging to DLL2 10.Crash! Fix, file Fl_win32.cxx: const char* class_name = /*w->xclass(); if (!class_name) class_name =*/ "FLTK"; // create a "FLTK" WNDCLASS const char* message_name = "FLTK::ThreadWakeup"; + UnregisterClass(class_name,fl_display); WNDCLASSEX wc; // Documentation states a device context consumes about 800 bytes No side effects found. Please, apply. | ||
gyohng.netscape 18:23 Oct 13, 2003 | Side effect of patch found. Though application doesn't crash, it doesn't show valid interface on the concurrently opened window from the second DLL (shows blank). This happens only when ONE application loads TWO dll files, and both DLL files have FLTK linked statically. | ||
gyohng.netscape 19:20 Oct 21, 2003 | Problem is solved by assigning fl_display to hInstance, which is passed to DllMain (#ifdef WIN32) GetModuleHandle indeed returns instance of a parent executable, and not hInstance of a DLL module (which has FLTK linked statically). Wonder if there could be any intelligent workaround for this? | ||
mike 17:39 Oct 31, 2003 | All I can recommend is to link both DLLs and the app against the FLTKDLL library... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Fl_Window::size_range malfunction? |
Version: | 1.1.4 |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 07:07 Oct 02, 2003 | I want to restrict window size while my application is running but found that the command Fl_Window::size_range does nothing if I have opened a window at that time. Only if I apply the command _prior_ to opening the first window it works. PC PIII, Windows2000, MingW 3.2.3 | ||
mike 08:02 Oct 02, 2003 | This is a windows-specific issue that cannot be resolved due to the way the Microsoft did things. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | unresolved symbol main when building fltk on solaris |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 19:54 Oct 24, 2003 | I'm trying to build FLTK on Sun solaris 2.7 with gcc 2.95.2. I was already able to build successfully using the same archive on windows xp and redhat linux 7.1. The first problem I had was that, when running "make", it complained that makeinclude could not be found. I fixed this by manually running "./config.status --recheck;./config.status" and then executing main. The "src" and "fluid" built built without trouble, however when the makefile went on to build "test", it gives the following link error. I tried removing colbrowser from the build, but then the next link (demo) gives the same error, so I'm presuming this problem applies to all the links within the test directory. === making test === Compiling adjuster.cxx... Compiling arc.cxx... Compiling ask.cxx... Compiling bitmap.cxx... Compiling boxtype.cxx... Compiling browser.cxx... Compiling button.cxx... Compiling buttons.cxx... Compiling checkers.cxx... Compiling clock.cxx... Compiling colbrowser.cxx... Linking colbrowser... Undefined first referenced symbol in file main /usr/local/encap/gcc-2.95.2/lib/gcc-lib/sparc-sun-solaris2.7/2.95.2/crt1.o ld: fatal: Symbol referencing errors. No output written to colbrowser collect2: ld returned 1 exit status *** Error code 1 The following command caused the error: gcc -I.. -O2 -Wall -Wunused -fno-exceptions -I/usr/local/X11R6/include -fpermissive -o colbrowser -L../lib -lfltk_forms -lfltk -L/usr/local/X11R6/lib -R/usr/local/X11R6/lib -ldl -lm -lXext -lX11 -lsocket -lnsl make: Fatal error: Command failed for target `colbrowser' Thanks, Ash. | ||
jedimasterthrash 19:57 Oct 24, 2003 | Sorry, this should have been priority 3 I guess. I was too quick to think that not being able to build the toolkit constituted key functionality. I went back and changed the priority to 3 and submitted again, but it turns out that there isn't a confirmation screen before it is actually posted, so now there are two copies of my question. Sorry about that. Ash. | ||
mike 17:23 Oct 31, 2003 | I think the problem might be caused by the standard Sun make program; try using GNU make and let me know if the problem goes away... | ||
mike 21:22 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #188 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | unresolved symbol main when building fltk on solaris |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 19:55 Oct 24, 2003 | I'm trying to build FLTK on Sun solaris 2.7 with gcc 2.95.2. I was already able to build successfully using the same archive on windows xp and redhat linux 7.1. The first problem I had was that, when running "make", it complained that makeinclude could not be found. I fixed this by manually running "./config.status --recheck;./config.status" and then executing main. The "src" and "fluid" built built without trouble, however when the makefile went on to build "test", it gives the following link error. I tried removing colbrowser from the build, but then the next link (demo) gives the same error, so I'm presuming this problem applies to all the links within the test directory. === making test === Compiling adjuster.cxx... Compiling arc.cxx... Compiling ask.cxx... Compiling bitmap.cxx... Compiling boxtype.cxx... Compiling browser.cxx... Compiling button.cxx... Compiling buttons.cxx... Compiling checkers.cxx... Compiling clock.cxx... Compiling colbrowser.cxx... Linking colbrowser... Undefined first referenced symbol in file main /usr/local/encap/gcc-2.95.2/lib/gcc-lib/sparc-sun-solaris2.7/2.95.2/crt1.o ld: fatal: Symbol referencing errors. No output written to colbrowser collect2: ld returned 1 exit status *** Error code 1 The following command caused the error: gcc -I.. -O2 -Wall -Wunused -fno-exceptions -I/usr/local/X11R6/include -fpermissive -o colbrowser -L../lib -lfltk_forms -lfltk -L/usr/local/X11R6/lib -R/usr/local/X11R6/lib -ldl -lm -lXext -lX11 -lsocket -lnsl make: Fatal error: Command failed for target `colbrowser' Thanks, Ash. | ||
mike 17:20 Oct 31, 2003 | Dupe... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | gcc only inlines when see inline definition before use |
Version: | 1.1.4 |
Created By: | tmwg-fltk.inxservices |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
tmwg-fltk.inxservices 05:15 Oct 27, 2003 | inline.patch 4k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
tmwg-fltk.inxservices 05:15 Oct 27, 2003 | Patch rearranges order of some inline functions, so gcc will inline them. Suggest adding -Winline to gcc flags. | ||
mike 17:57 Oct 31, 2003 | I'm not sure what effect the changes will have, however we definitely can't change the forms emulation functions to use ints instead of shorts - that will break binary compatibility when the functions are not inlined. What effect does this specifically have? And what version of GCC are you using? | ||
tmwg.inxservices 18:39 Oct 31, 2003 | That was a mistake. The patch to Fl/forms.H should be removed from the rest of the patch. It should not have been included in the first place. The purpose of the patch is to make sure that gcc REALLY inlines the functions that have been moved. Each of those functions is called by other inline functions in the headers. But the call to the moved function comes BEFORE the definition of the moved function. gcc doesn't deal with that properly: it only inlines a function if it has ALREADY seen its definition. For instance, in Fl_Widget.H, draw_focus() call x(), y(), etc. But x() is defined AFTER draw_focus(). So gcc doesn't actually inline x(), etc. It makes local copies of the inline functions in every place that draw_focus() is called. This is not what the authors expected, and doesn't happen with many other compilers. But it does with every version of gcc I've ever used. All the patch was supposed to do was reorder some inline functions so that gcc would really inline them. As I said, use -Winline in the gcc flags, and you'll see what I mean quite clearly. Sorry for the confusion with the forms.H patch getting included and muddying the waters. I know that you can't change the shorts to ints. Obviously, I've done it in my local copy, and let the patch leak to you. | ||
mike 19:30 Oct 31, 2003 | What version of GCC are you using? I added -Winline and recompiled using GCC 2.96 and 3.3 and got no warnings or errors... | ||
tmwg-fltk.inxservices 20:25 Oct 31, 2003 | Nor do I. I'm using 3.3.2. I made this patch a long time ago, and no longer remember what code I was compiling at the time where I saw the symptom. But I can't reproduce it now. | ||
mike 16:52 Nov 02, 2003 | OK, please file a new STR if you are able to reproduce this problem... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Why the FLTK toolkit is sow slow in drawing |
Version: | 1.1.4 |
Created By: | mama_mari.hotmail |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mama_mari.hotmail 02:19 Oct 28, 2003 | Dear Sir, I would like to ask you about the FLTK toolkit libraries. When I make some programmes, for ex. Editor or even some of the ex. provided with the FLTK. The programs Components blinks like a flasher ... ! I think it is because of the FLTK drawing is not too fast. (may be!!) I use the Dev-cpp updated with the latest files. Is there any way to solve that? Thanks. Mariwan Sweden | ||
mike 17:59 Oct 31, 2003 | What kind of computer and video card do you have? We've never had any drawing speed complaints, although it might be worthwhile to use Fl_Double_Window for your application to get rid of any flickering. | ||
mama_mari.hotmail 05:12 Nov 03, 2003 | Dear Michael; I do not think it is computer's problem. Description of my computer: New Dell computer Celerion 2GHz 256 RAM Intel (R) 82845G/GL/GE/PE/GV Graphics controller The problem that I have mentioned is appearing on many computers. Just One more Info. I have tried even Gtk. It has the same problem but not so clear like FLTK. I think the redraw is not so fast. It is the problem. If you compare the file editor which comes with Dev-cpp which is written in C, It is much faster, smaller. Why? When you just make a small window. The size is 660KByte. It is too big. I wish to make objects to the toolkit, as it is the easiest toolkit I have seen! (I do not know if there are others!). Could you send me some example which you have maybe and works perfectly without blinking?!. Thanks Mariwan J. | ||
mama_mari.hotmail 00:35 Jan 23, 2004 | As I mentioned before that problem, It is the fl_window which is not using double buffering technique, because of that the window blinks. I suggest that FLTK and all the other widgets should be with the double windows technique to avoid the blinking. I used a Fl_Double_Window with the file chooser sample comes with FLTK 1.1.4. It blinks even with the Fl_Double_Window. Because the fl_chooser itself is not built with Fl_Double_Window. But when I used the Fl_Double_Window there was no more blinking … There is only one way to solve that. It is by using the Fl_Double_Window to the fl_chooser itself. What is the difference between the fl_Window and Fl_Double_Window? Can I re-compile the FLTK with Fl_Double_Window for all widgets available with the toolkit? Does this make FLTK slower or bigger as code after compilation? I would appreciate answering my question. Thanks .. As I said I used the toolkit with Windows XP and with Windows 98 and even with Linux... Redhat.. with different computer/graphic cards..etc. Mariwan | ||
ariock.bluebottle 23:29 Feb 19, 2004 | To solve this blinking problem try adding this line before any other call to fltk (as mentionned in the doc) Fl::visual(FL_DOUBLE|FL_INDEX); | ||
mama_mari.hotmail 22:52 Feb 22, 2004 | HI, I do not think it will help. I have tried that. The only thing which could help is the Fl_Double_Window. But it will be slow in drawing. There should be a better solution for eliminating the problem. I am thinking about that.... Thank about complex program that has about 1000 widget? It is impossible with flicker widget to build such a program. It should be the first what we think about when we design a toolkit for cpp. As I mentioned, I have tried many toolkits. Most of them have the same problem. There should be a way...! But how? I do not know really. thanks for your reply | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | Incorrect behaviour of FL_ENTER and FL_LEAVE events |
Version: | 1.1.4 |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
kontakt.andreas-schoemann 07:37 Nov 06, 2003 | main.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 07:37 Nov 06, 2003 | Please perform the attached program to verify the following steps: 1) When I enter the subclassed output widget it says "Hello Mouse". 2) If I now move the mouse slowly upwards and leave the widget it says "Bye Mouse". I get the same result if I slowly move to the left and leave the widget. 3) If I move the mouse _fast_ upwards or to the left and leave the widget nothing happens. The widget obviously does not get a FL_LEAVE event. 3) If I move to the right or downwards and leave the widget nothing happens, no matter how slow or fast I move. The widget obviously does not get a FL_LEAVE event. Configuration: PC PIII, Windows 2000 (German), MingW 3.2.3 | ||
mike 07:51 Nov 06, 2003 | This is a Windows bug and there is absolutely nothing we can do about it. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Config Files |
Summary: | fltk-config incorrect output |
Version: | 1.1.4 |
Created By: | shaiay.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shaiay.yahoo 20:45 Nov 08, 2003 | fltk-config --ldflags outputs libraries as well as flags. This is a problem because while flgs should probably come before the object file in the linker command line, libraries should definitely come after. the canonical way to invoke the linker is: $(LDD) `fltk-config --ldflags` -o $(EXECUTABLE) $(OBJECT_FILES) `fltk-config --libs` on my system: >> fltk-config --ldflags -L/usr/local/lib -Wl,-rpath,/usr/local/lib -L/usr/X11R6/lib -lfltk -lXft -lm -lXext -lX11 -lsupc++ >> fltk-config --libs /usr/local/lib/libfltk.a so the resulting command line is $(LDD) -L/usr/local/lib -Wl,-rpath,/usr/local/lib -L/usr/X11R6/lib -lfltk -lXft -lm -lXext -lX11 -lsupc++ -o $(EXECUTABLE) $(OBJECT_FILES) /usr/local/lib/libfltk.a while on RH9 Linux the program links just fine, cygwin complains about not finding all the fltk functions unless I put all the libraries at the end, i.e. $(LDD) -L/usr/local/lib -mwindows -o $(EXECUTABLE) $(OBJECT_FILES) -lfltk -lole32 -luuid -lcomctl32 -lwsock32 -lsupc++ I think that all libs, e.g. everything with a "-l" options should be moved to fltk-config --libs | ||
mike 07:40 Nov 11, 2003 | The "--ldflags" option is meant to provide all of the link-time options, including libraries, and can be grouped with the libraries normally placed in the "LIBS" variable in a makefile. The "--libs" option writes the library filename(s) for use in build dependencies. This is covered in the man page... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | OpenGL |
Summary: | Setting color for FL_GL_Window |
Version: | 1.1.4 |
Created By: | ypatodia.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ypatodia.hotmail 03:58 Nov 09, 2003 | I am using a demo program "cube", that came with the fltk toolkit. I wanted to know how to change the background color for the cube. It appears that the color is set to black, and after trying a lot, I couldnt get to change it. The cube class uses FL_GL_Window, so I guess that if I can change the color in the FL_GL_Window, then I will be able to change it for the cube too. I tried Fl::background(), but that just changes the background outside the cube. Thanks. | ||
mike 07:43 Nov 11, 2003 | You must change the glClearColor() values, since OpenGL windows are rendered using OpenGL calls, not FLTK drawing calls. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | rotated text drawing |
Version: | 1.1-current |
Created By: | zaplia.mail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zaplia.mail 22:34 Nov 11, 2003 | It would be nice to have ability to draw rotated text and also it will be very nice to apply matrix transformation to text. At least, drawing of text in 90,180,270 directions should be possible without lot of work. -- It is not a problem in Win32, but on X it requires some brain. -- Have You any suggestions, how to make it? | ||
mike 21:14 Mar 10, 2004 | This isn't something we plan to add anytime soon, sorry. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | undefined references |
Version: | 1.1.4 |
Created By: | dbernick.soe.ucsc |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dbernick.soe.ucsc 22:33 Nov 14, 2003 | I must be missing a library reference..any ideas? thanks, Dave ProteinGui.o(.text+0x47): In function `makeWindow()': : undefined reference to `windowCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x13b): In function `makeWindow()': : undefined reference to `connectDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x27a): In function `makeWindow()': : undefined reference to `getConfigDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x37b): In function `makeWindow()': : undefined reference to `getConfigByIdDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x478): In function `makeWindow()': : undefined reference to `windowCB(Fl_Double_Window*, void*)' | ||
mike 08:39 Nov 17, 2003 | You are missing those functions in your own code; they aren't part of FLTK... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | function objects for callbacks. |
Version: | 1.1.4 |
Created By: | rainer.koecher.ok-automation |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rainer.koecher.ok-automation 01:53 Nov 19, 2003 | hello, I did a little work for testing with fltk. The toolkit looks very professional. But I'm missing one feature: I'd like to use function objects for callbacks instead of static member functions of classes. I think if you want to implement a object oriented system it's not good to have static member functions. Here is a little example of what i'm meaning. now callbacks work in that way: class Example{ void callback_function_i(FL_Button *bt, void *v) { // do something useful. } //v would be a pointer to "this" static void callback_function(FL_Button *bt, void *v) { v->callback_function_i(bt, v); } public: Example(){ //initialize the Window } } i wold like to do this in that way: class callback{ public: void operator()(Fl_Button *bt, void *v) { // do something useful. } } class Example{ //define a callback object which operator() will be called. callback button_callback; public: Example(){ //initialize the Window } } | ||
mike 21:16 Mar 10, 2004 | There are some enhancements along these lines available on the bazaar/links page. Currently they require "advanced" C++ features which are not universally available, but they may be added in the future. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | FL_MOVE events not passed through Fl_Menu_Button POPUP3 |
Version: | 1.1.4 |
Created By: | chris.mobilemascots |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
chris.mobilemascots 12:26 Nov 25, 2003 | In making an app from the CubeView example I am experiencing strange behavior. I added a POPUP3 Menu to the Gl_Window and when the Menu is not visible it will pass through FL_PUSH and FL_DRAG events to the Gl_Window below. But it will not pass through FL_MOVE events. This is on Linux. I have not tested on OSX or WIN32. | ||
chris.mobilemascots 17:53 Nov 25, 2003 | I checked my project on OSX, and the behavior is the same. | ||
chris.mobilemascots 20:00 Nov 25, 2003 | Since I started trying to find where I could patch the 1.1.4 src's, I decided to move the Project over the 2.0.x I have not observed the problem! And things are generally nicer in 2.0.x Thanks. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | FLUID |
Summary: | Fl_Double_Window --- used to be FL_Window |
Version: | 1.1.4 |
Created By: | dbernick.soe.ucsc |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dbernick.soe.ucsc 16:00 Dec 02, 2003 | When trying to rebuild a product that generated correctly on 1.1.3, I am now gettin F1_Double_Window inserted as parameters where previously Fl_Window was generated. The .fl file does not have any calls (that I can see) that refer to Fl_Double_Window. This is causing link errors due to the extern being unresolved for the new calls. Running Linux RedHat 9 Here is an example: ProteinGui.o(.text+0x47): In function `makeWindow()': : undefined reference to `windowCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x13b): In function `makeWindow()': : undefined reference to `connectDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x27a): In function `makeWindow()': : undefined reference to `getConfigDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x37b): In function `makeWindow()': : undefined reference to `getConfigByIdDialogCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x478): In function `makeWindow()': : undefined reference to `windowCB(Fl_Double_Window*, void*)' ProteinGui.o(.text+0x713): In function `makeWindow()': : undefined reference to `windowCB(Fl_Double_Window*, void*)' | ||
mike 18:59 Feb 25, 2004 | FLUID prior to 1.1.4 did not honor the single/double buffer state in a .fl file; edit the .fl with FLUID and select the proper mode for the window widget. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Problem with CGAL 3.0 and FLTK |
Version: | 1.1.0 |
Created By: | cfranco.exatas.unisinos |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cfranco.exatas.unisinos 07:29 Dec 08, 2003 | I'm having problems trying to compile an application based on FLTK, using CGAL 3.0, in VC++ .NET. I got a lot of compile errors, in VC header files and in CGAL header files. When I compile the application without use CGAL it compiles fine. The same when I compile only the CGAL based class that I created. The problem is when I integrate both. | ||
mike 14:32 Feb 29, 2004 | Without a lot more information, it is impossible to help you. Try posting your problems to the FLTK newsgroups/mailing lists. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | fluid-generated code names all widgets "o" |
Version: | 1.1.4 |
Created By: | ucko.ncbi.nlm.nih |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ucko.ncbi.nlm.nih 11:28 Dec 09, 2003 | The code FLUID generates for creating nested widgets names all of the widgets o (except for the outer window, which is w), using nested C++ scopes to delimit their use. Although this technique is perfectly legal, it causes warnings on at least Sun's WorkShop compiler, which proceeds to spew messages of the form ".../foo_.cpp", line N: Warning: o hides the same name in an outer scope. I would suggest giving them unique names (o1, o2, etc.), resetting the numbering for each function. | ||
mike 14:43 Feb 29, 2004 | Feel free to submit a patch! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | Fl_Menu_, Fl_Menu_Item, Fl_Choice - FLTK crash |
Version: | 1.1.4 |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
kontakt.andreas-schoemann 13:38 Dec 13, 2003 | test.zip 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 13:38 Dec 13, 2003 | Please compile the attached program and perform the following actions: 1. Press 'add_1'. This adds a menu item to the choice menu to the right of 'add_1'. This menu is then assigned to the 'choice_1' menu. 2. Press 'add_2'. This adds a menu item to the choice menu to the right of 'add_2'. This menu is then assigned to the 'choice_2' menu. 3. Try to look at the 'choice_1' menu. Step 3 causes FLTK to crash with an 'unhandled exception'. (This example is deduced from an application I'm currently developping. I've tried to simplify it as much as possible hence it may look somewhat artificial.) Configuration: PC PIII, Windows XP, Visual C++ Regards, Andreas Schömann | ||
mike 17:32 Feb 25, 2004 | Unable to duplicate - make sure you have all of the current VC++ service packs installed and are not using the "optimize for speed" setting since VC++ doesn't optimize things safely. (compile with debug to test...) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fd handler not working properly in XP |
Version: | 1.1.4 |
Created By: | shaia.biocontrol.co |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shaia.biocontrol.co 01:25 Dec 24, 2003 | First, thank you for a great product. I have fltk 1.1.4 compiled with cygwin support on win XP pro. I add an fd handler for a socket which I create with the socketpair system call. when I send things to the socket, the handler does get called until the mouse is over one of the fltk windows of the application. using the same setup in linux (same code, same fltk version) it works just fine and the handler gets called once there is data in the socket. I have a fairly up to date cygwin system, and I compiled fltk 1.1.4 from source using the --enable-cygwin flag | ||
shaiay.yahoo 02:13 Feb 13, 2004 | As a follow up to my previous report: there was a typo in the previous mail: fd_handler does NOT get called until the mouse is over one of the fltk's application windows regardless of whether ther is data in the socket. I have worked around this using a timeout callback which uses select on the socket and calles the fd_callback if necesary. This is a code snippet showing the workaround, where void fd_handler(int,void*) is the original fd_handler which did not work for me (in win XP). this code works in both win XP and linux (rh9, fltk 1.1.4 compiled from source) #define TIMEOUT 0.1 fd_set socks; int _fd; void timeout_callback(void *_n) { struct timeval timeout; timeout.tv_sec=0; timeout.tv_usec=0; FD_ZERO(&socks); FD_SET(_fd,&socks); select(_fd+1, &socks, 0, 0, &timeout); if(FD_ISSET(_fd,&socks)) fd_callback(_fd,0); Fl::repeat_timeout(TIMEOUT,timeout_callback); } | ||
mike 05:08 Feb 29, 2004 | This is probably a side-effect of the hack we have to do to support add_fd() on Windows. You might have better luck with the asynchronous selection code (by default commented out right now, since we've had reports of problems on some versions of Windows) - try uncommenting the "#define USE_ASYNC_SELECT" line in src/Fl_win32.cxx and let me know if that solves the problem. | ||
shaiay.yahoo 10:55 Mar 02, 2004 | Thanks for the reply. As I posted before on Feb 13, I have a fix for this which works under XP and Linux. This fix is probably cygwin dependant for XP since it relies on the "select" function. Since my application is now working, I do not need to try the fix you suggest, especially since it means modifing the fltk source, which means my code will not work with pre-compiled fltk versions found on many systems. However, if you think that me trying what you suggest will help the fltk project, I will be happy to try it and report the results to you. Thanks again for fltk Shai p.s. I seem to used two seperate email adresses, however, I assure you I am the same person. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | set_non_modal note/issue |
Version: | 1.1.4 |
Created By: | jelmerc.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jelmerc.hotmail 08:57 Jan 07, 2004 | After a lot of hours puzzeling with Fl_Window's I found out that the Fl_Window::set_non_modal function has to be called AFTER you show () it. This wasn't mentioned in the docs so I think it would be nice if it was. BTW, this only applies to Win32, I didn't test on X so it might be not an issue there. hope this helps someone - Jelmer Cnossen | ||
jelmerc.hotmail 09:01 Jan 07, 2004 | I forgot to mention the actual issue here :D When you call set_non_modal () before show (), it initially works fine for some situations, but it is then somehow linked to the focus window or the currently set window (set Fl::current (w)) so that when you hide that window the non modal window is hidden too. Also it will mess up the widgets when you add the non modal window to another group. | ||
jelmerc.hotmail 09:10 Jan 07, 2004 | Oh great, sorry for this spamming, but the 'fix' described above doesn't actually work, it just makes it a modal dialog instead of non modal. So bottomline is this is a bug and I can't fix it. Plz help - Jelmer Cnossen | ||
mike 17:56 Feb 29, 2004 | non-modal windows are connected to the parent window on purpose; essentially they are toolbar windows that are not modal. Not a bug, a feature! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Library incompatibilty on Redhat 9 |
Version: | 1.1.4 |
Created By: | echlan.cs.umbc |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
echlan.cs.umbc 13:35 Jan 12, 2004 | I am getting the folowing warning and error messages and and am not sure what should be done about it. It is not clear if the undefined references are related to the warning or not. /usr/local/bin/ld: warning: libstdc++.so.5, needed by /usr/X11R6/lib/libGLU.so, may conflict with libstdc++.so.3 /usr/lib/crt1.o: In function `_start': ../sysdeps/i386/elf/start.S:77: undefined reference to `main' /cs/lib/libfltk.a(Fl.o): In function `Fl::handle(int, Fl_Window*)': Fl.o(.text+0xe19): undefined reference to `__ctype_b' Fl.o(.text+0xe56): undefined reference to `__ctype_tolower' Fl.o(.text+0xf5c): undefined reference to `__ctype_toupper' /cs/lib/libfltk.a(Fl_x.o): In function `Fl_X::make_xid(Fl_Window*, XVisualInfo*, unsigned long)': Fl_x.o(.text+0x1ea7): undefined reference to `__ctype_b' Fl_x.o(.text+0x1f0f): undefined reference to `__ctype_toupper' Fl_x.o(.text+0x20ea): undefined reference to `__ctype_toupper' /cs/lib/libfltk.a(fl_draw.o): In function `fl_draw(char const*, int, int, int, int, Fl_Align, void (*)(char const*, int, int, int), Fl_Image*, int)': fl_draw.o(.text+0xa97): undefined reference to `__ctype_b' /cs/lib/libfltk.a(fl_draw.o): In function `fl_measure(char const*, int&, int&, int)': fl_draw.o(.text+0xe6a): undefined reference to `__ctype_b' /cs/lib/libfltk.a(fl_font.o): In function `fl_find_fontsize(char*)': fl_font.o(.text+0x1ae): undefined reference to `__ctype_b' fl_font.o(.text+0x215): undefined reference to `__ctype_b' /cs/lib/libfltk.a(Fl_Pixmap.o): In function `Fl_Pixmap::color_average(Fl_Color, float)': Fl_Pixmap.o(.text+0xc77): undefined reference to `__ctype_b' /cs/lib/libfltk.a(Fl_Pixmap.o)(.text+0xcbd): more undefined references to `__ctype_b' follow collect2: ld returned 1 exit status Thanks | ||
mike 14:38 Feb 29, 2004 | Recompile FLTK with your current compiler. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | X11 |
Summary: | xft fonts dont seem to work |
Version: | 1.1.4 |
Created By: | mangosoft.comhem |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
mangosoft.comhem 04:17 Jan 15, 2004 | fluid.png 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mangosoft.comhem 04:17 Jan 15, 2004 | Fonts look like this (submitted screenshot) in all fltk programs if i install with --enable-xft, i dont have a clue what's going on, other programs use xft just fine, and fltk works fine without xft | ||
mike 14:40 Feb 29, 2004 | What OS, and if Linux, what Linux distro? I've never seen this before - Xft generally "just works". | ||
mangosoft.comhem 15:22 Feb 29, 2004 | running gentoo linux | ||
mike 17:00 Feb 29, 2004 | What versions please??? We need the versions of at least the following components to help you at all: Kernel XFree86 Xft Freetype GCC | ||
mangosoft.comhem 22:14 Feb 29, 2004 | Kernel - 2.4.21 XFree86 - 4.3 Xft - the one included in XFree Freetype - 2.1.5 GCC - 3.3.2 | ||
mangosoft.comhem 22:22 Feb 29, 2004 | ah, a bit progress, it seems related to locales more specifically, it doesnt work if LC_ALL is exported as an empty variable | ||
mangosoft.comhem 22:26 Feb 29, 2004 | I'm not sure why i put "export LC_ALL=", i must have meant "unset LC_ALL"... Works fine with that (it works fine with LC_ALL actually set to something, even something invalid) | ||
mike 21:12 Mar 10, 2004 | OK, this isn't a FLTK bug, please report this to the XFree86 folks... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Problem compiling flcdsim 1.1 |
Version: | 1.1.0 |
Created By: | pierre42d.9online |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
pierre42d.9online 19:09 Jan 18, 2004 | # gmake install Installing flcdsim program in /usr/local/bin Installing flcdsim manpage in /usr/local/share/man /usr/local/bin/cp: cannot stat `./doc/flcdsim.1': No such file or directory gmake: *** [install] Error 1 | ||
mike 14:36 Feb 29, 2004 | Remove the "doc/" from the makefile where it installs the file; will look into fixing this, although I am no longer developing/using flcdsim right now... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Default font... |
Version: | 1.1.4 |
Created By: | robysimoni.inwind |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
robysimoni.inwind 02:06 Jan 19, 2004 | Hi, why isn't there the possibility to set a default font for all widgets. It would be useful... Thanks Bye | ||
mike 18:12 Feb 29, 2004 | Use: Fl::set_font(FL_HELVETICA, fontindex) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Box application crash |
Version: | 1.1.4 |
Created By: | robysimoni.inwind |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
robysimoni.inwind 02:11 Jan 20, 2004 | Hi. I'm trying to create a StatusBar widget. So I start to insert some Fl_Box to view "panels". I declare panels like: -> StatusBar.H private: Fl_Box * pBox0_; -> StatusBar.cxx in the constructor I set se pBox0_ to NULL, then I create the Fl_Box with 'new Fl_Box(..., ..., etc...) When I close the application I have a segmentation fault. If I change all Fl_Box with Fl_Button all run correctly. I think there are something wrong in this widget. Probably the missing desctructor? My compiler is GNU GCC of cygwin distribution. I compile on Windows 2000. Thanks Bye | ||
mike 17:34 Feb 25, 2004 | We need a LOT more information to help you; the class definition for your widget would help! | ||
mike 21:16 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | OpenGL |
Summary: | gl_start() only works for single GL window |
Version: | 1.1.4 |
Created By: | david.morgan |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
david.morgan 16:52 Jan 25, 2004 | gl_start.cxx 6k | ||
david.morgan 20:06 Jan 29, 2004 | gl_start.cxx 6k | ||
david.morgan 12:24 Feb 17, 2004 | gl_start.cxx 6k | ||
david.morgan 16:18 Feb 29, 2004 | gl_start_diff 5k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
david.morgan 16:52 Jan 25, 2004 | Because gl_start.cxx only maintains a single context as a static variable, it is impossible to use the gl_start/gl_end mechanism with multiple windows. Attached is a fixed gl_start.cxx that supports N contexts. | ||
david.morgan 20:07 Jan 29, 2004 | Second attached gl_start.cxx has additional code to properly manage contexts when a window is hidden and then shown again, which requires creation of a new context. | ||
david.morgan 12:25 Feb 17, 2004 | Third update fixes a bug exposed by showing/hiding windows without changing their position. Now all static variables are tracked. | ||
mike 14:31 Feb 29, 2004 | Can you provide a patch file (diff -uw oldfile newfile)? Also, I'm not keen on your current implementation - it looks like a structure might be in order rather than all of those parallel arrays. Also, I'm not sure what the purpose is for this - if you have multiple OpenGL windows, gl_start/end is not the way to do it - subclassing Fl_Gl_Window is... | ||
david.morgan 16:43 Feb 29, 2004 | I merely expanded the original statics into arrays, trying to limit the scope of the change. I agree it is ugly. I ran into the gl_start problem because I was trying to display multiple windows, each with a mixture of FLTK widgets and GL rendering. I was unable to get Fl_Gl_Window to draw child objects. I now see that the official way to do this is to make the Fl_Gl_Window a child of the main window containing the other widgets. Anyway, I guess it can't hurt to make gl_start more general. | ||
mike 21:00 Mar 10, 2004 | Holding this change pending some other potential OpenGL changes. | ||
mike 13:50 Jul 26, 2004 | OK, upon further consideration we won't be using this patch; future releases of FLTK will likely address things. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Fl_Scroll inside Fl_Scroll doesn't scroll correctly |
Version: | 1.1.4 |
Created By: | isaque.uol.com |
Assigned To: | mike |
Fix Version: | 1.1.5rc1 |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
isaque.uol.com 09:07 Jan 29, 2004 | scroll.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
isaque.uol.com 09:07 Jan 29, 2004 | Hi guys! When I put a Fl_Scroll (let's name it "in") inside a Fl_Scroll ("ex") and "in" is bigger than "ex", if I try to scroll "in", it shows me other parts of my screen (win32). I don't know if this happens to other os. Thanks. Isaque | ||
mike 19:10 Feb 25, 2004 | Unable to duplicate, probably a video card driver bug. | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | Flipping images with fl_draw_image |
Version: | 1.1.4 |
Created By: | didac.iiia.csic |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
didac.iiia.csic 02:52 Feb 03, 2004 | I've tried to vertically flip an image using fl_draw_image, but the application crashes when I try it. According to the documentation, I should do the following: fl_draw_image(myBitmap,x,y,width,height,3,-width*3); to flip it. But it doesn't work. It simply crashes. However, flipping horizontally works perfect. Anyone with the same problem? Any solution to it? Thanks! | ||
mike 17:50 Feb 25, 2004 | The "mybitmap" pointer needs to point to the start of the last line. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Text_Editor can't scroll with keys when contains very long wrapped lines. |
Version: | 1.1.4 |
Created By: | alexey.parshin |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
alexey.parshin 08:54 Feb 03, 2004 | 1) I'm creating an Fl_Text_Editor widget. 2) assigning a text with a very long lines. Wrapping is set to 1 character. Text is displayed correctly, occupies 10-15 screens. 3) scrolling it back and forth with mouse - just fine 4) put cursor into the first line of text and start going down with the 'Down' keyboard button. After reaching the lowest visible line it scrolls fine for some time then stops scrolling. FYI, the features as 'goto(line,position)' and 'current_line()' would be very useful, as well as the ability to display line numbers | ||
mike 17:58 Feb 25, 2004 | Can you provide a small example program that demonstrates the problem? | ||
mike 21:17 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Fl_Gl_Window cannot be moved from one parent Fl_Window to another |
Version: | 1.1.4rc2 |
Created By: | yazhuk.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
yazhuk.yahoo 08:18 Feb 04, 2004 | FL_Window_bug.zip 4k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yazhuk.yahoo 08:18 Feb 04, 2004 | This appears on Windows and most probably is different on other platforms. If Fl_Gl_Window is removed from its parent container (with Fl_Window as a root) and inserted in the container belonging to another Fl_Window its corresponding native window still remains inserted into native window associated with original Fl_Window. I am attaching a demo project derived from "gl_overlays" FLTK demo. Press "Move to New Frame" button to see the problem then try to resize new window. void new_frame(Fl_Widget *o, void *p) function contains code demonstrating the problem. Try various values with "int scenario" variable to see 3 different scenarios showing various sides of the problem. As I can see from the FLTK source code the native window have to be re-created by call to CreateWindowEx, but this does not happen automatically. Trying to force it by calling hide() and show() makes it even worse. Another issue is - calling show()/hide() functions via pointers of different types gives completely different results what is very confusing (scenario "0" vs. scenario "1"). show() and hide() are declared as non-virtual in Fl_Widget and as virtual in Fl_Window. If you need more information I will be glad to help. Andrey Yazhuk. | ||
mike 05:10 Feb 29, 2004 | We don't support moving widgets to other windows/parents. In particular, moving a Fl_Gl_Window or any other subclass of Fl_Window will be very unpredictable since the Windows resources are tied to the original window (as you've seen). Since all OpenGL windows in FLTK share a common OpenGL context, try adding a new Fl_Gl_Window to the new window and then delete the old one. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Fix for the problem "Fl_GL_Window initialization takes a long time on Windows" |
Version: | 1.1.4rc2 |
Created By: | yazhuk.yahoo |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
yazhuk.yahoo 08:44 Feb 04, 2004 | Fl_Gl_Choice.cxx 10k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yazhuk.yahoo 08:44 Feb 04, 2004 | Creation of every instance of Fl_Gl_Window takes a long (2-3) seconds time. I provide a possible workaround for this problem. With this fix the delay is experienced only for the first window being created, all other windows are created almost instantly. Fix is located in lines 166 - 203 in the file attached. The first delay is still a problem that need to be resolved. Andrey Yazhuk | ||
mike 20:36 Mar 10, 2004 | It isn't clear what this actually fixes, and current CVS already has changes to the pixel format code which may do what you want. As for the speed, this is likely an issue with your graphics card driver and not with FLTK. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | Fl_Gl_Windows broken under Fedora |
Version: | 1.1.4 |
Created By: | nalkar.earthlink |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nalkar.earthlink 13:08 Feb 09, 2004 | Classes Drived form Fl_Gl_Window no longer function with Fedora. Where the opengl content should be displayed the window instead shows the background of the X11 Desktop. This occurs even if the application is run from another machine and displayed though ssh. The affected applications work fine on other computers. | ||
mike 17:46 Feb 25, 2004 | Need more info - things work great with my laptop with ATI OpenGL drivers and FC1. | ||
mike 21:20 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_File_Chooser doesn't support environment variables |
Version: | 1.1.3 |
Created By: | dan.photon |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dan.photon 03:45 Feb 11, 2004 | Environment variables cannot be passed to the directory method of the Fl_File_Chooser widget. (e.g. chooser->directory("$HOME")). This will result in the fl application hanging indefinately. | ||
mike 14:48 Feb 29, 2004 | Environment variables don't work, and are not documented to work, for Fl_File_Chooser. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | AIX 4.3 - fltk-1.1.4 compiled ok, but on run: Illegal instruction(coredump) |
Version: | 1.1.4 |
Created By: | rogx.libero |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rogx.libero 08:01 Feb 12, 2004 | Anagni, Italy 2004/02/12 Hi! my machine is an Ibm rs6000 with aix 4.3, the compiler is gnu gcc 3.2.3 I have fltk 1.1.2 that is ok, and I am trying to compile fltk 1.1.4 What I have done: ./configure (everythings ok) make (everythings ok) make install (everythings ok) but when i try to run fluid or any of the tests i get only: Illegal instruction(coredump) Can somebody help me? rogx@libero.it | ||
mike 17:47 Feb 25, 2004 | Please post your request to the FLTK newsgroups. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | unresolved external symbol |
Version: | 1.1.4 |
Created By: | popatr.yahoo |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
popatr.yahoo 18:30 Feb 14, 2004 | I downloaded the source and opened demo.dsw found in the visualc directory. (I have visual c++ 6.0) When I built the project this error came up: demo.obj : error LNK2001: unresolved external symbol "public: void __thiscall Fl_Group::forms_end(void)" (?forms_end@Fl_Group@@QAEXXZ) ../test/demod.exe : fatal error LNK1120: 1 unresolved externals What is supposed to be linked in to fix this? (And why wasn't it linked in?) | ||
mike 14:42 Feb 29, 2004 | Do a "rebuild all"; it sounds like the fltkforms.lib library isn't getting linked (check that the demo project depends on fltkforms) | ||
mike 21:22 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | make crashes on fedora |
Version: | 1.1.4 |
Created By: | mlhughes.rogers |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
mlhughes.rogers 22:12 Feb 14, 2004 | logs.tgz 9k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mlhughes.rogers 22:12 Feb 14, 2004 | make crashes when trying to compile the fl_draw_image.cxx file. complains about U32 not being defined. It ran fine on winXP attached is a makerun and a config log file. | ||
mike 17:48 Feb 25, 2004 | What OS and compiler is it not working with??? | ||
mlhughes.rogers 20:19 Feb 25, 2004 | Originally it was fedora core 1 that it first crashed on I loaded redhat 8.0 on the machine and tried again and it crashed in the same place. On the redhat 8.0 system : gcc -- version reports " gcc (GCC) 3.2 20020903 (redhat linux 8.0 3.2-7) I don't have the fedora up at this time. the log files should be useful in seeing the error. | ||
mike 21:02 Mar 10, 2004 | I am not able to reproduce this on any RedHat or Fedora system in my lab, sorry... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | Resizable widget attribute doesn't work after first "show" call. |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 16:54 Feb 22, 2004 | The small piece of code below demonstrates the problem. When the initial "show" call is done, the double_window remains unresizable regardless of the call to resizable(pTD). When you comment out the initial "show" call, resizability works as expected. I came across this becuase I want to, at a later time, add a resizable widget to an existing (already shown) window. Calling "hide" after "show" has no effect. Fl_Double_Window *pDW = new Fl_Double_Window(400, 400, ""); pDW->show(); //comment this line out pDW->hide(); //this line has no effect pDW->begin(); Fl_Text_Display *pTD = new Fl_Text_Display(100, 100, 200, 200); pDW->resizable(pTD); pDW->end(); pDW->show(); return Fl::run(); Thanks, Ash. | ||
mike 18:24 Feb 25, 2004 | This is a side-effect of the window manager - there is nothing you can do about it, unfortunately... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Unassigned |
Summary: | fl_cursor() not working under windows |
Version: | 1.1.4 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 00:58 Feb 23, 2004 | Attempts to change the cursor with fl_cursor() seem to yield temporary results; cursor quickly changes back to normal. Appears to be Windows specific; seems OK under IRIX/OSX/LINUX. The problem can be demonstrated with the test\cursor.exe program; the cursor does not change. Tested with 1.1.4 on Windows 2000. | ||
greg.ercolano 01:14 Feb 23, 2004 | Correction: the problem is in fltk-utf8-1.1.4 (emphasis on utf8). Regular 1.1.4 is OK. Will follow up to Jean-Marc, current maintainer of the utf8 version. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | Fluid and #extern "C" |
Version: | 1.1.4 |
Created By: | eric.fraxinous.cisi.cnes |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
eric.fraxinous.cisi.cnes 06:18 Feb 23, 2004 | In fluid, I can't enter the declaration line : #extern "C" { char *test; } The error message is : "missing }" | ||
mike 18:29 Feb 25, 2004 | Which dialog are you entering this declaration in? | ||
eric.fraxinous.cisi.cnes 09:49 Feb 27, 2004 | It's in declaration dialog. | ||
mike 21:10 Mar 10, 2004 | Unable to reproduce, sorry! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Problem with dynamic menu and file chooser |
Version: | 1.1-current |
Created By: | mpaliwoda.interia |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
mpaliwoda.interia 12:09 Feb 23, 2004 | menu.cxx 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mpaliwoda.interia 12:09 Feb 23, 2004 | Platform:Windows98, compiler:Borland Bcc 5.5, fltk:last CVS. Attached is a small program which demonstrates inapropiate dynamic menu behavior after file chooser was opened and then closed. To demonstrate the problem you need to run the program, then switch "font" menu atributes using "SwitchMenu" button, then choose any file thru "ChooseFile" button, and finaly try to swith "font" menu once again. After closing file chooser the menu cannot be manipulated anymore... By contrast staticly defined menu works ok in the same situation. | ||
mike 06:08 Feb 29, 2004 | The menubar is taking focus away from the button. Click the button a second time to see what I mean. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Header file fails |
Version: | 1.1.2 |
Created By: | pjwjr.ccioh |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
pjwjr.ccioh 04:32 Feb 24, 2004 | I can not compile This is what I get : BUILD PRODUCT In file included from /home/perry/src/hdr/FL/Fl_Widget.H:29, from /home/perry/src/hdr/FL/Fl_Group.H:30, from /home/perry/src/hdr/FL/Fl_Window.H:29, from c/hello.c:2: /home/perry/src/hdr/FL/Enumerations.H:179: error: syntax error before "fl_define_FL_ROUND_UP_BOX" /home/perry/src/hdr/FL/Enumerations.H:179: warning: data definition has no type or storage class /home/perry/src/hdr/FL/Enumerations.H:182: error: syntax error before "fl_define_FL_SHADOW_BOX" /home/perry/src/hdr/FL/Enumerations.H:182: warning: data definition has no type or storage class | ||
mike 14:25 Feb 29, 2004 | What operating system? What compiler? It looks like you are compiling a C file and not C++ - make sure you use a valid C++ source file extension (.cxx, .c++, .cpp, .cc) | ||
mike 21:20 Mar 10, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Window::label() function does not store title unless shown() |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 14:34 Feb 24, 2004 | Fl_Window::label() does not store the window title text. If you call label(sTitle) before show(), the title is lost and you get a window with no title. You have to call label() after show() to have effect. The FLTK documentation page 300 does not say anything about this feature. "Gets or sets window title bar label." Also, as future enhancement, label() might save the name is !shown(), and then when the window gets created it could automatically apply the saved title. Thanks. | ||
mike 14:48 Feb 29, 2004 | The label() method does not copy the contents of the string. This applies to ALL widgets. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Re-activated menu doesn't get shown until damage() called |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 17:54 Feb 24, 2004 | If you have a first-level (one of the main menu's) sub-menu, and you deactivate it: { "&Messages", 0, 0, 0, FL_MENU_INVISIBLE | FL_SUBMENU }, And then reactivate it pMainMenu->mode(12, pMainMenu->mode(12) & ~FL_MENU_INVISIBLE); The menu does not initially reappear in the main menu bar. However, if you click on the location where the menu should be, the title and menu show up, and the menu does work, but as soon as you unclick the menu, it goes back to being invisible. So it is correctly activated. It just is not correctly re-drawn when not clicked. Following the re-activation with pMainMenu->damage(FL_DAMAGE_ALL); or otherwise damaging it via a minimize/maximize on the window results in the menu text getting drawn correctly from then on. Thanks. | ||
mike 17:54 Feb 29, 2004 | Calling redraw() is required for most widgets after changing the state of the widget (this is on purpose) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | X11 |
Summary: | Error in loading 4-bit bitmap. |
Version: | 1.1.4 |
Created By: | nilesh_89.yahoo |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nilesh_89.yahoo 03:38 Mar 04, 2004 | i have ported pixmap_browser sample app. to ARM Linux Based Hand-Held Device. This device has 4-bit display (320*240) resolution. this is the error message i get on my hendheld device. Can't do 4 bits_per_pixel colormap What could the prob? Thanks, Nilesh | ||
mike 04:43 Mar 04, 2004 | FLTK currently requires at least an 8-bit pseudo-color visual on X11 to operate. I don't know how difficult it will be to support a 4-bit display, but I will do some research and get back to you... | ||
mike 17:57 Apr 09, 2004 | It looks like supporting 4-bit displays is not going to be possible, sorry. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Text Display doesn't highlight selctions |
Version: | 1.1.4 |
Created By: | vimal.parikh |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
vimal.parikh 10:59 Mar 04, 2004 | I have text Display widget which is displaying multiple lines of text. I want to highlight one of the lines. I tried using the text buffer's highlight method but the text is not highlighted. Here are the calls I am making for the Text_Display and Text_Buffer widgets: textDisplay->show_cursor(1); textDisplay->cursor_style(Fl_Text_Display::HEAVY_CURSOR ); textDisplayinsert_position(line_start); textDisplay->buffer()->select(line_start, line_end); textDisplay->redraw(); I tried using 'highlight' instead of 'select' but it didn't work either. I have verified that the line_start and line_end are valid positions. | ||
mike 19:59 Mar 10, 2004 | Hmm, I just tried to duplicate this with the editor demo and wasn't able to duplicate the problem. Can you try adding the following at line 480 of test/editor.cxx: textbuf->select(0, 80); and then try the editor demo to see if it highlights the first line and part of the second? It might also be the selection color is set to the same value as the background color... | ||
mike 17:55 Apr 09, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Unassigned |
Summary: | help for a student |
Version: | 1.2-current |
Created By: | rchokhani.yahoo |
Assigned To: | rokan |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rchokhani.yahoo 19:56 Mar 04, 2004 | we have solaris machines in our labs at our university and we have the fltk toolkit for glut on it. i am trying to make opengl with glut using fltk toolkit on it. but i am having certain problems could you kindly suggest the steps to solve this problem: i get the following error: denali: ~] > g++ -c fltk2.cpp fltk2.cpp:1:19: Fl/Fl.H: No such file or directory fltk2.cpp:2:27: Fl/Fl_Window.H: No such file or directory fltk2.cpp:3:24: Fl/Fl_Box.H: No such file or directory fltk2.cpp: In function `int main(int, char**)': fltk2.cpp:7: `Fl_Window' undeclared (first use this function) fltk2.cpp:7: (Each undeclared identifier is reported only once for each function it appears in.) fltk2.cpp:7: `window' undeclared (first use this function) fltk2.cpp:7: syntax error before `(' token fltk2.cpp:8: `Fl_Box' undeclared (first use this function) fltk2.cpp:8: `box' undeclared (first use this function) fltk2.cpp:8: syntax error before `(' token fltk2.cpp:9: `FL_UP_BOX' undeclared (first use this function) fltk2.cpp:11: `FL_BOLD' undeclared (first use this function) fltk2.cpp:11: `FL_ITALIC' undeclared (first use this function) fltk2.cpp:12: `FL_SHADOW_LABEL' undeclared (first use this function) fltk2.cpp:15: `Fl' undeclared (first use this function) fltk2.cpp:15: syntax error before `::' token fltk2.cpp is a simple hello world program from fltk.org the program is: here is my program: #include <Fl/Fl.H> #include <Fl/Fl_Window.H> #include <Fl/Fl_Box.H> int main(int argc, char **argv) { Fl_Window *window = new Fl_Window(300,180); Fl_Box *box = new Fl_Box(20,40,260,100,"Hello, World!"); box->box(FL_UP_BOX); box->labelsize(36); box->labelfont(FL_BOLD+FL_ITALIC); box->labeltype(FL_SHADOW_LABEL); window->end(); window->show(argc, argv); return Fl::run(); } somebody kindly do the needful as soon as possible regards rohit | ||
youdontneedmy.emailaddress 11:07 May 01, 2004 | You need to set the directory for the header files, I believe it is done by the -I option. Like this I think: gcc text.cpp -I /usr/local/fltk/ or maybe gcc text.ccp -I/usr/local/fltk/ I'm not sure which, try both. | ||
mike 19:42 Jun 03, 2004 | If this is still an issue, please use the FLTK newsgroups to ask your question; this form is primarily for reporting bugs and asking for new features... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | osx file browser subwindow misses mouse down event |
Version: | 1.1.4 |
Created By: | choffman3.nyc.rr |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
choffman3.nyc.rr 08:42 Mar 07, 2004 | flt.C 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
choffman3.nyc.rr 08:42 Mar 07, 2004 | osx NavGetFile browser subwindows don't get mouse down event and are unresponsive. browser hangs until mouseclick on another process when the unresponsive subwindow disappears. This happens if NavGetFile is called from a normal or menubar button. If NavGetFile is called from a menubar subwindow it works ok! Once the file browser window comes up, click on the top center bar. It brings up a menu of root locations which you can't get out of without clicking to switch to another task. The example program has both working and not working menu bar items. Thanks---Bob Hoffman | ||
mike 19:47 Mar 10, 2004 | Not sure what is going wrong, but we don't support using the Carbon file chooser directly from FLTK. If you come up with a cause or a fix, we'll consider adding it to FLTK, but we don't have the resources to track these kinds of issues down ourselves. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | DeleteObject fails in function clear_xmap in fl_color_win32.cxx. Resource leakage |
Version: | 1.1.4 |
Created By: | gboado.nahuelsat.com |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
gboado.nahuelsat.com 03:46 Mar 10, 2004 | fl_color_win32.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
gboado.nahuelsat.com 03:46 Mar 10, 2004 | Hi. I have an application using fltk 1.1.4 for the GUI under windows 98. The app works fine but it looks as it has a GDI leak or so because the GDI resources are depleted in about 10 minutes. If I close the app, the resoureces return to normal. The app doesn't use any windows specific code (no explicit use of the GDI resources, just the code within fltk). I tracked the bug using DevPartner to the function clear_xmap: Argument 1 in DeleteObject (the object HGDIOBJ__ hObject = 0x000034EE=0x000034EE) is still selected in to hDC 0x000034F2. Current Call Stack clear_xmap H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 73 fl_color H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 116 fl_color H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 90 fl_rectf ..\FL/fl_draw.H 77 draw_box H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_boxtype.cxx 295 draw_box H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_boxtype.cxx 285 draw h:\users\gboado\boado\develop\myprojects\tigris2\tigris_fltk\custom_widgets.cpp 375 update_child H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 547 draw_children H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 531 draw H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 540 draw H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Window.cxx 99 flush H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 984 flush ..\FL/win32.H 84 flush H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 363 wait H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 278 run H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 286 main h:\users\gboado\boado\develop\myprojects\tigris2\tigris_fltk\tigris6.cxx 6254 mainCRTStartup crtexe.c 338 C:\WINDOWS\SYSTEM\KERNEL32.DLL!0x0001b532 C:\WINDOWS\SYSTEM\KERNEL32.DLL!0x0001b3e4 | ||
mike 19:34 Mar 10, 2004 | Hmm, I'm looking at the code for clear_xmap() and it is doing exactly what it is supposed to - it selects a new pen and brush and deletes the old ones... Can you confirm this problem on another system with a different graphics card? | ||
mike 17:57 Apr 09, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Unassigned |
Summary: | Trouble compiling example under visual 6.0 |
Version: | 1.2-current |
Created By: | dlippman.coe.neu |
Assigned To: | rokan |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dlippman.coe.neu 13:19 Mar 22, 2004 | I created a project in Visual Studio 6.0 and pasted your example program (I'll include it at the bottom). I then added the FLTK headers and lib files to the appropriate VSC directories and added the fltkd.lib and wsock32.lib libraries to the project. When I try to build the project I get errors: --------------------Configuration: fltkTest - Win32 Debug-------------------- Linking... MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _toupper already defined in LIBCD.lib(toupper.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _tolower already defined in LIBCD.lib(tolower.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _free already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _realloc already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _malloc already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _sprintf already defined in LIBCD.lib(sprintf.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _exit already defined in LIBCD.lib(crt0dat.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _memmove already defined in LIBCD.lib(memmove.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _strtol already defined in LIBCD.lib(strtol.obj) LINK : warning LNK4098: defaultlib "MSVCRTD" conflicts with use of other libs; use /NODEFAULTLIB:library fltkd.lib(Fl_x.obj) : error LNK2001: unresolved external symbol __imp___TrackMouseEvent@4 Debug/fltkTest.exe : fatal error LNK1120: 1 unresolved externals Error executing link.exe. fltkTest.exe - 11 error(s), 1 warning(s) here's the Example Code: #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Box.H> int main(int argc, char **argv) { Fl_Window *window = new Fl_Window(300,180); Fl_Box *box = new Fl_Box(20,40,260,100,"Hello, World!"); box->box(FL_UP_BOX); box->labelsize(36); box->labelfont(FL_BOLD+FL_ITALIC); box->labeltype(FL_SHADOW_LABEL); window->end(); window->show(argc, argv); return Fl::run(); } | ||
rokan 14:16 Apr 09, 2004 | These are linking/conflict errors between your application and fltk: Please ensure that you are linking against multithreaded dll runtime libs (project/options/c++/code generation) and also project/options/link/input/ignore libraries: libcd does not do any harm. Also link your project to comctl32.lib as well (to get rid of TrackMouseEvent link error). Roman | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | show() window does not maintain focus on Unix/Linux, also text selection feature |
Version: | 1.1.4 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 20:43 Mar 31, 2004 | First off, I'll say that everything works on win32 (xp). When I call show() on a window, the window comes to the front and takes the focus. Then I put it on both Unix: Machine hardware: sun4u OS version: 5.8 Processor type: sparc Hardware: SUNW,Sun-Fire-280R gcc version 2.95.2 19991024 (release) And Linux: RedHat Linux 7.3 gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-112) Both OSes exhibit the following problem identically. When you call show() on window-A as a result of clicking on window-B (for instance, clicking a button or handling a mouse event in window-B), window-A will raise to the front, and sometimes will get the focus for a split second, but then the focus always goes back to window-B (with window-A still remaining forward). This happens reliably every time, assumming both window-A and window-B were already open. If window-A is not yet open, and a click on window-B causes window-A to show for the first time, then window-A becomes the raised window and correctly maintains focus. I have also tried following the show() of window-A with a take_focus() of window-A (both before and after completing the handle() in window-B). Neither has any effect. window-A still relinquishes the focus back to window-B without my consent. A related problem is that the graphical highlight of a text selection is not displayed when the test_display window does not have the focus. I believe it would be a better feature if the text selection or highlight was graphically displayed even when the window did not have the focus. Two examples: 1. I have a window with two child-windows. One displays text, the other displays status. For some reason, the status window always takes the focus when the parent window gets shown. This means that selections in the text_display do not show up. I have to override the FL_FOCUS message in handle of the parent window to force the text window to take the focus instead of the status window. 2. I have the problem that this STR is about. I have a find dialog and a text window. When I click find in the dialog, it selects the text in the text window and brings the text window forward. However, since the find dialog (in unix/linux) steals the focus back, the user never sees the selected found text in the text_display. Thanks, Ash. | ||
mike 07:15 Apr 01, 2004 | This is entirely an issue with your window manager; some window managers behave like Windows, and most offer "focus policy" options to allow for pointer, click, and other various policies that do not allow you to take focus away from a window the user has intentially focused on. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Unassigned |
Summary: | Fl_Preferences::size() returns inconsistent value for binary ("void *") data |
Version: | 1.1.4 |
Created By: | cbiegl.bellsouth |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cbiegl.bellsouth 11:37 Apr 05, 2004 | "Fl_Preferences::size()" always returns the length of the string stored in the preferences data base. This ends up being twice the size of the binary data added with "Fl_Preferences::set(const char *, void *, int)", as such data is converted to a string with two characters per original data byte. | ||
mike 08:28 Apr 06, 2004 | This is how the size() method is meant to operate since all values are stored as text in the preferences file. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Text_Display: horizontal scrollbar appears unnecessarily |
Version: | 1.1.4 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 14:40 Apr 11, 2004 | Even if there is no text going beyond the right edge of the text window, the horizontal scrollbar can appear. It seems if the vertical scrollbar appears, the horizontal scrollbar will also automatically appear. Replication: 1) Run the editor.cxx application (no scrollbars) 2) Enter many short lines of text (no wider than screen) until text scrolls off bottom; vert AND horiz scrollbar appear Horiz scroll bar should not appear unless there's somewhere to scroll to. | ||
mike 17:42 Apr 11, 2004 | Looking into this; I suspect this was done to avoid an infinite loop, but I'll check. | ||
mike 15:20 May 15, 2004 | Comments from the Fl_Text_Display::resize() code: Decide if the horizontal scroll bar needs to be visible. If there is a vertical scrollbar, a horizontal is always created too. This is because the alternatives are unatractive: * Dynamically creating a horizontal scrollbar based on the currently visible lines is what the original nedit does, but it always wastes space for the scrollbar even when it's not used. Since the FLTK widget dynamically allocates the space for the scrollbar and rearranges the widget to make room for it, this would create a very visually displeasing "bounce" effect when the vertical scrollbar is dragged. Trust me, I tried it and it looks really bad. * The other alternative would be to keep track of what the longest line in the entire buffer is and base the scrollbar on that. I didn't do this because I didn't see any easy way to do that using the nedit code and this could involve a lengthy calculation for large buffers. If an efficient and non-costly way of doing this can be found, this might be a way to go. So, unless you can think of a good algorithm for quickly calculating the maximum line width, the code will stay as-is. | ||
greg.ercolano 10:07 May 18, 2004 | > * Dynamically creating a horizontal scrollbar based on the currently > visible lines is what the original nedit does, but it always wastes > space for the scrollbar even when it's not used. Since the FLTK > widget dynamically allocates the space for the scrollbar and > rearranges the widget to make room for it, this would create a very > visually displeasing "bounce" effect when the vertical scrollbar is > dragged. Trust me, I tried it and it looks really bad. Maybe the bounce effect can be avoided if the calculation for detecting the hscroll isn't run during scrolling, ie. if the mouse is in the scrollbar, and one of the buttons is down, don't do the calculation.. just schedule the calcuation to occur when the mouse is released. > * The other alternative would be to keep track of what the longest > line in the entire buffer is and base the scrollbar on that. I > didn't do this because I didn't see any easy way to do that using > the nedit code and this could involve a lengthy calculation for > large buffers. If an efficient and non-costly way of doing this > can be found, this might be a way to go. This sounds like the right solution, as it seems like what most editors do; I'm used to seeing the hscroll's tabsize indicate the widest line in the file. It's also consistent with how the vscroll tab works; it shows the longest line count in the file. eg. the Mozilla 'Edit|View Source' window, and 'kedit' to name a few examples. I can see where whenever a large block of text is added, a complete char-by-char scan would be needed to determine maximum line width. For optimization, a public method that allows the caller to pass in the max line length could help that, eg: add(const char *text, int maxlen = -1 ); In my app, I have to parse through the file anyway before adding the lines to the widget, so I can include the maxlen calculation as an optimization, so the widget doesn't have to do it. Then it's just a matter of making sure the widget detects max line widths while the user types new lines of text. | ||
mike 05:54 May 21, 2004 | We're open to patches... :) | ||
mike 08:04 Jul 27, 2004 | Please open a new STR if you come up with a solution (patch)... Thanks! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | DeleteObject fails in function clear_xmap in fl_color_win32.cxx |
Version: | 1.1.4 |
Created By: | gboado.nahuelsat.com |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
gboado.nahuelsat.com 10:34 Apr 12, 2004 | Sorry, I missed the follow-up mail and the previous STR was closed, so here it is again. Hi. I have an application using fltk 1.1.4 for the GUI under windows 98. The app works fine but it looks as it has a GDI leak or so because the GDI resources are depleted in about 10 minutes. If I close the app, the resoureces return to normal. The app doesn't use any windows specific code (no explicit use of the GDI resources, just the code within fltk). I tracked the bug using DevPartner to the function clear_xmap: Argument 1 in DeleteObject (the object HGDIOBJ__ hObject = 0x000034EE=0x000034EE) is still selected in to hDC 0x000034F2. Current Call Stack clear_xmap H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 73 fl_color H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 116 fl_color H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_color_win32.cxx 90 fl_rectf ..\FL/fl_draw.H 77 draw_box H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_boxtype.cxx 295 draw_box H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\fl_boxtype.cxx 285 draw h:\users\gboado\boado\develop\myprojects\tigris2\tigris_fltk\custom_widgets.cpp 375 update_child H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 547 draw_children H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 531 draw H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Group.cxx 540 draw H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl_Window.cxx 99 flush H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 984 flush ..\FL/win32.H 84 flush H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 363 wait H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 278 run H:\Users\gboado\boado\Develop\MyProjects\fltk\fltk-1.1.4\src\Fl.cxx 286 main h:\users\gboado\boado\develop\myprojects\tigris2\tigris_fltk\tigris6.cxx 6254 mainCRTStartup crtexe.c 338 C:\WINDOWS\SYSTEM\KERNEL32.DLL!0x0001b532 C:\WINDOWS\SYSTEM\KERNEL32.DLL!0x0001b3e4 ----------------------------------------------------------- To answer Michael question about the video card, I verified that the problem happens in the following configurations: 1) Pentium 166, Windows 95 OSR2, 64MB RAM, S3 Trio 64V+ video card with 2 MB DRAM, standard S3 windows driver. 2) Pentium 3 700 Mhz, 128 MB RAM, Windows 98 first ed., TNT2 M64 video card with 32 MB SDRAM. 3) Athlon 2000XP, 512MB SDRAM, Windows XP Pro, TNT2 Ultra 32 MB SDRAM. In the third configuration the system didn't deplete the resources (because of the higher O.S. limit) but in both conf.2 and conf.3 the problem was clearly detected through DevPartner. Exactly the same error. I agree. The code looks fine, but it fails. | ||
mike 14:58 May 15, 2004 | Well, since the code is written specifically to Microsoft's guidelines, and it *is* deleting things as it is supposed to, I don't know what we can do here. Perhaps if you can post a *short* example program that demonstrates the problem, we may be able to figure out what is going on, but at the moment I am inclined to point the finger at Windows and close this out. Please post an example program within the next two weeks. | ||
mike 08:05 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Different "bugs" in 1.1.5rc1 |
Version: | 1.1.5rc1 |
Created By: | ariock.skynet |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ariock.skynet 00:16 Apr 13, 2004 | I, while compiling the latest 1.1.5rc1 relaease I comme accros those littles anoying bugs: My config: WinXP + vc6 + intel compiler As stated before: in the mailing list, the configure.h of vc6 should not have the HAVE_LIBPNG symbol defined (...) - When building cubemain.cxx it misses cubeviewUI.h as well as cubeviewUI.cxx (and I cannot find them in the test directory) - The fast_slow.cxx file is missing - When building the fltkdll files there a huge number of *.h ( example string.h, sys/dir.h and so on ) file missing that are foud in static link part of the build - In Fl_File_Icon2.cxx there is still a warning for an integer conversion at line: 273 - inactive.cxx file is missing - keyboard_ui.h is missing and keyboard_ui.cxx as well - mandelbrot_ui.h and mandelbrot_ui.cxx are missing - preferences.cxx is missing - radio.cxx is missing - resize.cxx is missing - tabs.cxx is missing - valuators.cxx is missing Just so those can be tracked down. | ||
mike 07:45 Apr 13, 2004 | We don't have nor have we tested with the Intel compiler. This is likely the cause of most of your problems. As for the changes to config.h, we are inclined to leave them in since a lot of users want to have JPEG and PNG support but don't know to change config.h to enable that; a download page will be provided with pre-compiled copies of the image libraries that can be used with VC++... The demo programs depend on scripting support in your copy of VC++; if you have the student version it won't work... They are generated automatically using FLUID (perhaps if FLUID did not compile then the files were not created for that reason...) | ||
ariock.skynet 23:14 Apr 25, 2004 | The issiues are not from the intel compiler for most of them as previous version of fltk compile perfectly. Mainly (not talking about advantages over VC) the intel compiler is just a littlbe bit more strict than the VC compiler. The scripting possibility is there as I use the pro version of VC... Now I know some warning are from my configuration, but the missing files are still missing... | ||
mike 19:53 Apr 28, 2004 | Does fluid compile successfully? Without FLUID, none of the example programs built using FLUID will be able to be compiled. The fltkdll warnings are a side-effect of how VC++ builds the export list; it is a known bug in VC++ and is not a unique problem to FLTK. The warnings can be safely ignored. | ||
ariock.skynet 23:54 Apr 28, 2004 | Yes fluid build succesfully althought I don't use it much... | ||
mike 15:14 May 15, 2004 | Well, we are unable to reproduce this problem, so I will close this STR. Please post a new STR if you come up with a fix (other than adding the generated source files...) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | X11 |
Summary: | Optern AMD64 segfault if fl_draw_pixmaps |
Version: | 1.1.5rc1 |
Created By: | d.brust.dbrust.cs.kent |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
d.brust.dbrust.cs.kent 15:37 Apr 13, 2004 | ./configure gives the variable U64 the value UNSIGNED LONG on opteron machines during the configure stage, which seems to cause some troubles, namely a segfault, in the fl_draw_pixmap class. Changing this to UNSIGNED stops the segfaults and allows for normal execution, but the pixmaps are no longer colored correctly. Whether changing this value directly affects the coloring of the pixmaps or not, I have no idea. This happens in both the 1.14 and 1.2 versions. This happens with GCC versions 3.2.2 and 3.3.2 on various Linux Distro's | ||
d.brust.dbrust.cs.kent 15:56 Apr 13, 2004 | Changing to config.h to /*#undef U64 */ solves the problem! | ||
mike 11:30 Apr 15, 2004 | Thanks, Erco's set me up with access to his AMD64 machine, so I'll be figuring out a proper fix for both 1.1.5 and 1.2 real soon now... | ||
mike 07:41 Apr 16, 2004 | OK, I've done some basic testing on Erco's AMD64 test system and don't see any segfaults when displaying images. Did you see the problem with a specific demo program, or can you provide a small test program that segfaults? Thanks! | ||
d.brust.dbrust.cs.kent 11:56 Apr 16, 2004 | I ran into the problem when using the FLU library. When I pressed the 'Browse...' button on the file chooser, it would segfault while trying to display. After a few hours of good 'ol GDB and alot of debuging printf's, I narrowed it down into the fl_pixmaps, although, when I went a couple layers deeper and found that it was the U64 ultimatly responsible. It certainly could be something that pixmaps was calling, or something associated. Once I found the culprit, I stopped narrowing it down, because I was back into developer mode! I can throw something together small that will cause the segfault if you like. Just let me know, and I can have it by sunday or so. Darren | ||
mike 18:32 Apr 24, 2004 | Yes, we'll need a small example that does not use FLU (just the core FLTK stuff) to determine that FLTK is at fault. | ||
mike 17:39 May 21, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fl_filename_absolute() and Windows drive letter case. |
Version: | 1.1.4 |
Created By: | shane.hill.dsto.defence.gov |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
shane.hill.dsto.defence.gov 20:54 Jun 15, 2004 | filepathTest.cpp 1k | ||
shane.hill.dsto.defence.gov 20:55 Jun 15, 2004 | Makefile 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shane.hill.dsto.defence.gov 16:39 Apr 14, 2004 | I have a program that uses fl_filename_absolute() and has been compiled under Cygwin using the -mwindows and -mno-cygwin switches to make it a pure Windows application not dependent on Cygwin. However, when I run the program from a Cygwin bash shell a lower case drive letter is always returned like "c:\MyDir". Whereas when the same program binary is run via double clicking its icon or through CMD.EXE or COMMAND.EXE the drive letter returned is always uppercase like "C:\MyDir". Is this change in drive letter case intentional or is it something deep within Cygwin maybe? The reason I'm asking about this is because I write my fltk code to be able to use both fl_file_chooser() and FLU's flu_file_chooser(). The FLU code always returns an uppercase drive letter no matter how the program is initiated. However, if I want to take FLU's flu_file_chooser() result and pass it through fl_filename_relative(), the relative path will not be correct unless the drive letter case is the same as that returned by fl_filename_absolute(). I have had a look at the filename_absolute.cxx file ane can't see where there is a problem, unless the resultant drive letter is comming from the getcwd() command. Have Fun !! Shane Hill Aircraft Vehicles Division, Platforms Sciences Laboratory, Defence Science & Technology Organisation, 506 Lorimer St, Fishermens Bend, Vic. 3207, AUSTRALIA. | ||
mike 14:59 May 15, 2004 | It is probably a cygwin issue, however all of the FLTK code is written to treat filenames as case-insensitive on Windows, so it shouldn't matter one bit. Can you post a *short* example program along with instructions to duplicate the problem? | ||
shane.hill.dsto.defence.gov 19:08 May 17, 2004 | G'day Michael, I'll try and get you a small example ASAP. Thanks | ||
shane.hill.dsto.defence.gov 20:54 Jun 15, 2004 | G'day Michael, Sorry for taking so long, but I have been very busy. The next two articles are my test code project. They include filepathTest.cpp and Makefile. Just type "make" under a cygwin shell and the Makefile will compile the code using the gcc MINGW32 version of the compiler under Cygwin. The resulting executable will be non-Cygwin dependent. This assumes that Fltk has also been compile with the MINGW32 style, which is the default for Cygwin. When run under a Cygwin shell the filepathTest.exe code will print a lower case drive letter like: c:/Documents and Settings/........... Run the same EXE by double clicking under the Windows explorer window or from a command prompt will give: C:/Documents and Settings/........... Using fl_filename_absolute() and fl_filename_relative() combinations work fine under both environments because they both have the same case change. However, if I get my absolute file path using the FLU library flu_file_chooser(), it will always return the same case no matter how invoked. This then means fl_filename_relative() will work okay under say the Windows execution method, but not under the Cygwin method. Did you want me to provide you with an FLU example program to show this? Anyway, test routine follows. My suggestion is to get Fltk to ensure that only one case type is returned for drive letters or to make fl_filename_relative() case independent under Windows. Have Fun !! Shane Hill | ||
shane.hill.dsto.defence.gov 20:59 Jun 15, 2004 | G'day Michael, I forgot to mention that the program should also print: WIN32 defined. __MINGW32__ defined. If correctly compiled. I can post the filepathTest.exe file if you would like. Have Fun !! Shane Hill | ||
mike 13:42 Jul 26, 2004 | OK, the problem is pretty simple - when you build FLTK with Cygwin it makes certain assumptions about filenames being case sensitive. I don't know why this is the case (all Windows filesystems are case insensitive), however that is the way the people who ported FLTK to Cygwin wanted things. It isn't clear what the best solution is here - please post this to the fltk.development group and we'll see if there are any Cygwin users that prefer one way over another. | ||
mike 19:21 Sep 12, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | about fl_pack children |
Version: | 1.1.4 |
Created By: | jxbking.263 |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jxbking.263 01:32 Apr 16, 2004 | Hi, I found a bug in Fl_Pack widget. When children of pl_pack more then 1310, fl_pack can't show it's children. example: ./test/pack.cxx Replace all lines from line66 to line89( new Fl_Button(35, 35, 25, 25, "b1");......new Fl_Button(265, 265, 25, 25, "b24"); )with line for(int i =0 ; i < 1320; i++ ) new Fl_Button(35, 35, 25, 25, "b1"); Fl_Pack will show nothing about those button, but if you change i < 1310, it's show good. any number large than 1310, Fl_Pack doesn't work. Regards Arcane | ||
mike 17:55 Apr 17, 2004 | This isn't a bug in Fl_Pack - you are running into the 16-bit coordinate limits for 1.1.x, which limit the maximum X and Y position to 32767. This limit is also present in X11 and WIN32. FLTK 1.2.x and 2.0.x use 32-bit coordinates with clipping (to avoid problems with sending 32-bit coords to the underlying graphics system) and do not have this issue. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Config Files |
Summary: | fltk-config --ldflags --use-glut doesn't include -lglut in the output |
Version: | 1.1.5rc1 |
Created By: | marwan.badawi.irisa |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
marwan.badawi.irisa 08:36 Apr 19, 2004 | When generating linking info with fltk-config --ldflags --use-glut the output is exactly the same as when --use-gl is used. The -lglut does not appear in the library list that is generated. I am using FLTK 1.1.5rc2 under Linux but I think this problem is not limited to this version or this operating system. | ||
mike 13:57 Apr 24, 2004 | The --use-glut option is provided for compatibility with FLTK 2.x which separates the GLUT emulation code. You need to use the GLUT-provided hooks to get the GLUT library included. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | I've installed, with checkinstall, the fltk-1.1.5rc1 package |
Version: | 1.1.5rc1 |
Created By: | gene.heskett.verizon |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
gene.heskett.verizon 07:07 Apr 22, 2004 | I'm trying to find a decent mixer gui for use with alsa. I've built and installed that soft in version 1.0.4. The OS is linux, currently running a 2.6.5-mm5 kernel. A friend recommended the alsamixergui, so I found that and dl'd it. It in turn exits during the ./configure phase, saying that fltk is not installed. But it is, and "fluid" seems to be ok, but I don't get any output for fltk libraries from a "ldconfig -v |grep fltk" It seems like I should. Ideas anyone? | ||
gene.heskett.verizon 18:48 Apr 22, 2004 | I've since found that I should have made it with the option --enable-shared, which I've now done. No change other than ldconfig shows the libs now. alsamixergui still can't find it at all, and anything gtk built won't make. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Unassigned |
Summary: | Fluid crashes on MacOS X when switching applications |
Version: | 1.1.5rc1 |
Created By: | shelbydavis.mac |
Assigned To: | matthiaswm |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
shelbydavis.mac 17:04 Apr 22, 2004 | fltkDPP 6k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shelbydavis.mac 17:04 Apr 22, 2004 | opened a file from 1.1.4 and switched applications and it crashed. MacOS X 10.3.3, freshly compiled source stack trace: #0 0x927d8a7c in HIObject::GetEventTarget() () #1 0x9281c6c0 in _AXUIElementCopyParentHIObject () #2 0x9281c59c in PostUpdateWindowVisibility(VisChangeState*) () #3 0x9281ec00 in _ShowHideWindows () #4 0x9284aedc in ShowHide () #5 0x928832f8 in ShowHideWindows () #6 0x928652b0 in HIApplication::HandleActivated(unsigned char, OpaqueWindowPtr*) () #7 0x927e1ca4 in HIApplication::EventObserver(unsigned long, OpaqueEventRef*, void*) () #8 0x927d221c in _NotifyEventLoopObservers () #9 0x927dec34 in AcquireEventFromQueue () #10 0x927dcc0c in ReceiveNextEventCommon () #11 0x927fe7cc in AcquireNextEventInMode () #12 0x927fbc9c in RunApplicationEventLoop () #13 0x00034f74 in do_queued_events(double) () #14 0x00034fd8 in fl_wait(double) () #15 0x000253d4 in Fl::wait(double) () #16 0x00025450 in Fl::run() () #17 0x00018ff0 in main () #18 0x000023a8 in _start (argc=0, argv=0x8b2f8, envp=0x95224) at /SourceCache/Csu/Csu-46/crt.c:267 #19 0x0000221c in start () | ||
mike 18:34 Apr 24, 2004 | Will investigate on Monday, thanks! | ||
matt 21:04 Jun 05, 2004 | I can't replicate this crash (Mac OS X 10.3.3, FLTK 1.1.5.cvs). Is there are reliable way to repeat this? I loaded your .fl file and switched applications many times with various fluid windows opened and closed. Any other ideas? Matthias | ||
mike 10:56 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | Special character sequences in strings and comments (fluid) output bad code |
Version: | 1.1.5rc2 |
Created By: | mkajdas.mkprod |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mkajdas.mkprod 11:49 Apr 28, 2004 | In 'fluid' code (FLTK-1.1.5), special sequences (i.e. \0,\r,\n) in strings and comments confuse program generator and result in bad output code. Partial fluid code example: Function {Temp()} {open private return_type void } { code { ... memset(str, '\0', sizeof(str)); // '\0' causes bad output ... memcpy(str, ( "example with \0" "more with \0"), sizeof(str)); ... // example comment with "\r\n", rest of comment ... } {} } The '\0' removed the rest of the line. The "\r\n" caused newline, with the rest of the comment on the new line. This problem is probably in other versions of FLTK/fluid also. | ||
mkajdas.mkprod 16:02 Apr 28, 2004 | The same problem happens with fluid code like this: memcpy(str, "inches \", ", sizeof(str)); // the slash in \" gets removed Similar problem: // unmatched brace { in a comment The commented out brace is treated as part of code and fluid parser gets confused. | ||
mkajdas.mkprod 09:16 Apr 30, 2004 | After some more testing I remembered that special characters need to have the backslash doubled. I believe this is documented somewhere. \\r\\n\\0 instead of: \r\n\0 This also applies to comments. There are more problems with the comments though: { and # confuse the parser, i.e. // comment with #1 causes error // { in this line causes error if { is not matched in another comment The parser should ignore all comments in the code sections of the fluid file. | ||
mike 14:35 May 15, 2004 | Um, I just tried this and was unable to duplicate the problem - the .fl file contains \\ for each backslash as expected. Can you attach a sample .fl file that contains the wrong data? | ||
mkajdas.mkprod 11:30 May 18, 2004 | When I tried to duplicate the problem in Fluid, the problem did not occur. After some experimenting I duplicated the problem. What I often do is to write/modify the .fl file in an external text editor and then run FLUID on that file. That is when the problem occurs. When the file is generated in FLUID, FLUID puts backslashes in all the strategic places and everything works. This limitation/requirement forces users to use FLUID for code editing which many times is not convenient and complicated. I thought that CODE sections of the .fl files are straight 'C' code but this is not exactly true and this is what I stumbled upon. I do not know if this is a real bug anymore, but it is a confusing limitation that should be clearly explained. In summary: FLUID modifies the 'C' code to be FLUID acceptable or it may not run! | ||
mike 05:51 May 21, 2004 | We don't support or advocate modifying .fl files by hand. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Debug build works, Release build doesn't! |
Version: | 1.1.3 |
Created By: | joshuaf.vnet |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joshuaf.vnet 10:55 May 11, 2004 | Trying to write a simple FLTK program under Visual Studio .NET version 7.0.9466. When I compile in Debug mode, it works as planned; when the button is pushed the callback is called and the program ends. However, in Release mode the callback gets called, but the widget that called it has a different address than closeBtn. Am I doing something wrong? Is this a problem with this version of compiler? My three source files follow. Thanks, Joshua Foster /////////////////////////////////////// // Main.cpp /////////////////////////////////////// #include <FL/Fl.h> #include "MainWindow.h" int main() { MainWindow(); return Fl::run(); } /////////////////////////////////////// // MainWindow.h /////////////////////////////////////// #ifndef MAINWINDOW_H #define MAINWINDOW_H #include <FL/Fl_Window.h> #include <FL/Fl_Button.h> class MainWindow { private: Fl_Window *win; Fl_Button *closeBtn; public: MainWindow(); ~MainWindow(); static void HandleButtonPress( Fl_Widget* w , void* ); void HandleButtonPress( Fl_Button* ); }; #endif // MAINWINDOW_H /////////////////////////////////////// // MainWindow.cpp /////////////////////////////////////// #include <stdlib.h> #include <FL/fl_ask.h> #include "MainWindow.h" MainWindow::MainWindow() { win = new Fl_Window( 500, 500, 200, 100 ); closeBtn = new Fl_Button( 150, 50, 32, 32, "X" ); closeBtn->callback( HandleButtonPress, this ); win->end(); win->clear_border(); win->show(); } MainWindow::~MainWindow() { } void MainWindow::HandleButtonPress( Fl_Widget *w, void* mw ) { ((MainWindow*)mw)->HandleButtonPress( (Fl_Button*)w ); } void MainWindow::HandleButtonPress( Fl_Button* b ) { if( b == closeBtn ) { exit( 0 ); } } | ||
mike 14:55 May 15, 2004 | Please post to the fltk.general newsgroup for assistance; this is a C++ language issue (namely, your code is not valid C++) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | mouse and kbd events go to wrong subwindows |
Version: | 1.1.5rc1 |
Created By: | choffman3.nyc.rr |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
choffman3.nyc.rr 15:59 May 18, 2004 | f2.C 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
choffman3.nyc.rr 15:59 May 18, 2004 | create main window create 2 subwindows end main window move mouse over subwindows and events go to wrong subwindow. kbd events also go to wrong subwindow but different... clicking on a subwindow doesn't raise it to the front (probably because the mouse events never get to it?) fl_mac.cxx code seems to indicate that the window the mouse event is passed in the event handler is not correct or has a problem thus some other (wrong) method for identifying the window is used... | ||
mike 05:50 May 21, 2004 | Looking into it, will get back to you... | ||
mike 12:50 Sep 24, 2004 | Since MacOS X doesn't support true subwindows, the behavior you are seeing is correct. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | unable to get visual C++ to accept the include <FL/Fl.H>. |
Version: | 1.1.5rc1 |
Created By: | mentgel.eou |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mentgel.eou 19:06 May 23, 2004 | I followed the installation process found on your web page by double clicking on the fltk.dsw file under the visualc folder. This opened the workspace in Visual C++ so I built the program. This still did not allow the include statement to be recognized. Furter assistance will be greatly appriciated. | ||
mike 19:09 May 25, 2004 | Please ask this question on the fltk.general newsgroup. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | Unable to build FLTK_1.1.5rc1 on Windows |
Version: | 1.1.5rc1 |
Created By: | ydu1.nd |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ydu1.nd 20:30 May 27, 2004 | Hi, I am having trouble building FLTK_1.1.5rc1 for the ITK InsightApplications. FLTK_1.1.5rc1 does not build on my Windows machine (I use MS VC++6.0). The linking error says "cannot open file jpeg.lib", and it also has linking error with "png.lib". | ||
mike 17:30 May 31, 2004 | Go to the PNG, JPEG, and ZLIB home pages, download the source packages, and install them on your system. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Fl_Input can't get focus via mouse , before press any key (Under Linux) |
Version: | 1.1.3 |
Created By: | pat0116.iris.seed.net |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
pat0116.iris.seed.net 08:28 May 31, 2004 | I create a window for login one normal fl_input, another is secret fl_input fl_input can't get focus (can't see cursor) if I press any key via keyboard, fl_input can get focus and input my data. ps.win32 is ok | ||
mike 17:31 May 31, 2004 | What X server are you using? What window manager are you using? | ||
mike 08:05 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #401 |
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | MinGW compilation error |
Version: | 1.1.5rc1 |
Created By: | sanel.z |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sanel.z 01:54 Jun 01, 2004 | In makefiles/makeinclude.mingw this should be added: LDLIBS = -mwindows -lole32 -luuid -lcomctl32 -lwsock32 -lsupc++ GLDLIBS = -mwindows -lglu32 -lopengl32 -lole32 -luuid -lcomctl32 -lwsock32 -lsupc++ (libuuid.a and libole32.a are missing). | ||
mike 07:43 Jun 01, 2004 | Dupe | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | Trouble with winXP Heap |
Version: | 1.1.4 |
Created By: | nacholarrabide.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nacholarrabide.yahoo 16:44 Jun 02, 2004 | Hi, I developed a software (www.skullydoo.com.ar) using fltk 1.1.0 and VTK and now I'm having problems to adapt it to use fltk 1.1.4. The mayor problems have arrised with the windows heap. It apears that when I try to open a new windows, it gives a assertion problem (tries to free some memory at Fl_Menu::clear() ) and crushes down. Please help. ThanX Nacho | ||
mike 19:40 Jun 03, 2004 | We'll need a small sample program that reproduces the problem and a makefile/project file we can use to test it. Also, please let us know what operating system and compiler you are using. Finally, have you tried things with the current 1.1.5rc1 release candidate? | ||
nacholarrabide.yahoo 04:52 Jun 04, 2004 | I'm working over windows XP and compiling with Visual C++ 7 .Net. Y tried with release 1.1.5rc but I couldn't even compile my aplication with it (it keept giving some linking problems with the fltkimages.lib) ThanX again Nacho | ||
mike 19:10 Jun 07, 2004 | Edit vcnet/config.h and comment out the "HAVE_LIBxyz" definitions so that you can compile without the image libraries. Also, we still need a small example program which demonstrates the same problem, preferably some code that does not depend on VTK. | ||
mike 07:56 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #407 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Fl_Input_::handletext() docs |
Version: | 1.1.5rc1 |
Created By: | wilson.afn |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wilson.afn 06:55 Jun 03, 2004 | The docs for Fl_Input_::handletext(int e,int,int,int,int) claim it returns a "void", but perusing the source shows it returns an int. Indeed, this becomes important for a subclass which needs a reliable value to return from it's own handle() method. Furthermore, I find this somewhat confusing: "You must pass it the same bounding box as passed to draw()." Presumably, this should read "draw_text()", the method taking four integers, and not "Fl_Input_::draw()", the protected method which takes no parameters. Also, the sentence seems to imply the mysterious draw() is somehow called *before* handletext(i,i,i,i,i). Perhaps the intention is: "... the same bounding box *you want* passed to draw_text()." or similar mumbo jumbo explaining the parameter's affect on FLTK | ||
mike 19:43 Jun 03, 2004 | Dupe of 407 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | Program crashes when linked along with Microsoft Agent |
Version: | 1.1.4 |
Created By: | carthick84.hotmail |
Assigned To: | matthiaswm |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
carthick84.hotmail 09:54 Jun 04, 2004 | Any Program which links both fltk and any one of the following libraries crashes in WIN32 platform. 1. SAPI 4 SDK (spchwrap.lib) 2. Microsoft Agent 3. Digital Persona Fingerprint SDK The programs run fine when build as Debug versions, but crashes when build as release versions. Please help! | ||
matt 10:07 Jun 07, 2004 | Since I don't have access to any of these files, I am unable to test this. Have you checked if those libraries are compiled the same way as you FLTK libraries (if linking ens in some "library conflict" style error, you may be using conflicting clibs with different heap implementations). You could also compile and link as a release version , but with symbols enabled, and check, where your app fails. As a last resort, and if those libraries have a copyright that allows it, you may send either one or all of them and I will give it a test run. | ||
mike 10:56 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | Patch to fltk-1.1.5rc1 to allow the MacOS/X -- X11 combination |
Version: | 1.1.5rc1 |
Created By: | hugues.talbot |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
hugues.talbot 23:59 Jun 06, 2004 | fltk_1.1.5r1_ht.patch.bz2 22k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
hugues.talbot 23:59 Jun 06, 2004 | Sorry it took me so long to finally make good on this, I lost the first version I had after a laptop disappeared. This is a patch for FLTK-1.1.5rc1 to make it use the X11 back end rather than the Carbon one. This was very easy to do, and did not require major changes to the code, pretty much the only things are configure.in changes and changing the #ifdef __APPLE__ macros everywhere to something more flexible. The result is a GUI toolkit for Darwin/X11 combination which might be useful for servers. I'm currently doing the same stuff for Win32, i.e compile for Cygwin and/or Interix and X11. The target audience is the same. I'll post it here if anyone is interested (contact me) Cheers | ||
mike 19:07 Jun 07, 2004 | Thanks, will apply for 1.1.5rc2 tomorrow. | ||
mike 13:47 Jul 26, 2004 | Sorry, I just got a chance to look at the patch; unfortunately, it won't work as-is since the definitions used in various places to control X11/Carbon/GDI are in config.h, which is *not* installed in the FL include dir! I don't know what I want to do with this right now, but I can't apply it as-is since it will seriously break things... :( | ||
hugues.talbot.csiro 16:44 Jul 26, 2004 | Hi Mike, I'm not sure I understand the problem. Some stuff has to be in config.in because on both MacOS and Win32, deciding to compile for X11 or the native toolkit is a config decision! Then, when people are compiling with FLTK, the correct configuration is given when invoking the fltk-config script, e.g: On my linux box: % /usr/local9/bin/fltk-config --cflags -I/usr/local9/include -I/usr/X11R6/include On my MacOS box: % /usr/local/bin/fltk-config --cflags -I/usr/local/include -DMACOSX_CARBON If you don't like that mechanism, I can add a .H file in the FL directory that will have the correct define, this is trivial to do. | ||
mike 07:56 Jul 27, 2004 | Well, not everyone uses fltk-config, and the #define names you are using could easily conflict with other apps. I'm not sure what the correct fix is for this; creating a header file is fine for configure script users, but complicates installation and usage on Windows via VC++ (i.e. we can create a FL subdir under visualc and vcnet, but then you have to remember to copy that file if you install it in a system-wide location...) How about the following: 1. Dump the MACOSX_CARBON and WINDOWS_GDI constants. 2. Add "FL_USE_X11" which specifies the X11 should be used. 3. Change any conditional code which checks for __APPLE__ or WIN32 to also check for FL_USE_X11. a. We could automate this somewhat by providing a private header in the src directory which defines FL_USE_CARBON and FL_USE_GDI; if all of the system-specific files include x.H, then it could go there... b. There are still some files (Enumerations.H, Fl.H) which need the definitions as well; including x.H is not appropriate there but maybe we add a common stub header which does the definitions and the other headers include it? 4. fltk-config updated to define FL_USE_X11 in the CFLAGS output if --with-x is used to configure FLTK. Thoughts? | ||
hugues.talbot.csiro 08:24 Jul 27, 2004 | Hi, OK, thanks for the suggestions. If we are going to use the current WIN32 and __APPLE__ definitions, do we need the FL_USE_CARBON and FL_USE_WIN32 defines as well? Moreover we can probably ignore the problem of compiling with VC++ for the X11 backend, how likely is this to happen? What we need to do is come up with a solution that requires no copying if the default GUI backend is used (GDI on win32 and Carbon on MacOS). Using X11 on those platforms is for true unix diehards that will use configure. Let me rephrase: 1- Create a fl_gui_backend.H file (say) in include which by default has #undef FL_USE_X11 2- Check for FL_USE_X11 as you suggest 3- The configure script can overwrite what is in include/fl_gui_backend.H 4- All the include files that need it include fl_gui_backend.H, especially x.H 5- fltk-config defines FL_USE_X11 when needed. That ought to work for both ./configure and VC++ config files as the fl_gui_backend.H file will always be present. Have I forgotten something? Other than that I'm going on holiday tomorrow, can this wait for another week? I'll be back Wednesday next week (04/Aug). | ||
mike 08:57 Jul 27, 2004 | Hugues Talbot wrote: > Hi, > > OK, thanks for the suggestions. If we are going to use the current > WIN32 and __APPLE__ definitions, do we need the FL_USE_CARBON and FL_USE_WIN32 defines as well? No, I'm just thinking that it might be easier (read: lazy) to use a single constant rather than: #if defined(__APPLE__) && !defined(FL_USE_X11) #if defined(WIN32) && !defined(FL_USE_X11) everywhere we want to switch the code... > Moreover we can probably ignore the problem of compiling with VC++ > for the X11 backend, how likely is this to happen? What we need to do > is come up with a solution that requires no copying if the default > GUI backend is used (GDI on win32 and Carbon on MacOS). Using X11 on > those platforms is for true unix diehards that will use configure. > > Let me rephrase: > > 1- Create a fl_gui_backend.H file (say) in include which by default > has #undef FL_USE_X11 > 2- Check for FL_USE_X11 as you suggest > 3- The configure script can overwrite what is in > include/fl_gui_backend.H > 4- All the include files that need it include fl_gui_backend.H, > especially x.H > 5- fltk-config defines FL_USE_X11 when needed. If we rely on having FL_USE_X11 defined to override the default, then we don't need to generate the header file, just add it to CFLAGS. That would also allow for the hypothetical VC++ build for X11, since then all you'd need to do is add FL_USE_X11 to the preprocessor definitions (as we do today for FL_DLL) If we add a header file, it can be static and look for FL_USE_X11 to define the other constants. > Other than that I'm going on holiday tomorrow, can this wait for > another week? I'll be back Wednesday next week (04/Aug). I'd like to get the next release candidate out this week, so I might take a crack at it if you don't get me anything before tomorrow. Thanks! | ||
mike 12:07 Oct 18, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #413 |
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | patch for support of win32 + X11 |
Version: | 1.1.5rc1 |
Created By: | hugues.talbot |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
hugues.talbot 01:00 Jun 10, 2004 | fltk-1.1.5rc1_ht3.patch.bz2 73k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
hugues.talbot 01:00 Jun 10, 2004 | This is the patch against 1.1.5rc1 that augments support to X11 to both macs and PCs. It also allows compilation from arbitrary directories. Let me know what you think. | ||
mike 09:56 Jun 12, 2004 | Thanks, working on integrating into 1.1.x and 1.2.x. | ||
mike 13:48 Jul 26, 2004 | Closing this one out; please track on STR #413... Thanks! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Image Support |
Summary: | Fl_XXX_Image classes need a status function |
Version: | 1.1.4 |
Created By: | anonymous |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
anonymous 11:31 Jun 10, 2004 | The Fl_XXX_Image classes, such as Fl_PNG_Image, Fl_JPEG_Image, etc. take a filename as the required argument to the constructor. There is no mechanism to tell if an image file loaded correctly, and probably more importantly, to find why a file failed to load. Possible failure reasons include at least: file does not exist file not readable file format incorrect premature end-of-file (FYI: I have wasted two full days attempting to load a BMP, PNG, or JPEG image using these classes and it failed on every attempt. I finally wrote my own BMP file reader.) Mark | ||
mike 09:48 Jun 12, 2004 | IIRC, the width, height, and depth [w(), h(), and d()] will all be 0 if the image cannot be loaded. Unfortunately, C++ doesn't offer much to handle this outside of exceptions, and those aren't (yet) portable. We can't add extra fields to the image classes without breaking binary compatibility, however it should be possible to add an inline method which checks for invalid image dimensions and reports whether there is an error in the image or not. Is a separate method important to you, or is checking for w() and h() == 0 sufficient? | ||
mike 08:12 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Link error in Solaris |
Version: | 1.1.5rc1 |
Created By: | cougers.126 |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cougers.126 06:13 Jun 12, 2004 | I'v just began to use fltk to work on this platform, and I write a simple sample. But after I compiled it with "g++ -I/usr/local/include -c test.cxx" and linked it with "g++ -o testGui -L/usr/local/lib test.o", it droped error like that: undefine symbol in file main /usr/local/lib/gcc-lib/sparc-sun-solaris2.8/2.95.1/crt1.o My gcc version is 2.95.1, what wrong with it? | ||
mike 09:41 Jun 12, 2004 | You probably have the wrong build of GCC for your version of Solaris. Either that, or you don't have a "main" function in your test program. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | building FLTK 1.4 or 1.5 --enable-threads |
Version: | 1.1.5rc1 |
Created By: | bigswift.ufl |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
bigswift.ufl 05:33 Jun 15, 2004 | when attempting to compile 1.5 with the --enable-threads option on OS X 10.2.8 i get Compiling threads.cxx... Linking threads... ld: Undefined symbols: Fl::lock() Fl::awake(void*) Fl::unlock() i need threads to work, is there a fix for this? | ||
matt 23:21 Jun 18, 2004 | Threads are not supported at this point on the OS X version of FLTK 1.1. Any contribution is greatly appreciated. | ||
mike 11:31 Jul 27, 2004 | Um, I just built FLTK with threading support and the threads demo worked fine out-of-the-box. What version of MacOS X are you using? | ||
mike 19:15 Sep 12, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Config Files |
Summary: | missing zlib.h and jpeglib.h |
Version: | 1.1.5rc1 |
Created By: | dlippman.coe.neu |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dlippman.coe.neu 07:10 Jun 17, 2004 | Compiling Fluid fails due to: C:\fltk-1.1.5rc1\src\Fl_JPEG_Image.cxx(53) : fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory C:\fltk-1.1.5rc1\src\Fl_PNG_Image.cxx(44) : fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory I'm having trouble locating a source for these files, are they something that can be posted at FLTK.org, or do you have suggestions on where to find these? | ||
mike 07:48 Jun 17, 2004 | You'll find the necessary headers and project files at: JPEG: http://www.ijg.org/ ZLIB: http://www.zlib.org/ PNG: http://www.libpng.org/ If you don't care about PNG or JPEG image support, comment out the HAVE_LIBPNG/JPEG/Z lines in visualc/config.h and remove the references to jpeg.lib, png.lib. and z.lib from the project files. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | missing dirent.h need substitude |
Version: | 1.1.5rc1 |
Created By: | dlippman.coe.neu |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dlippman.coe.neu 09:43 Jun 18, 2004 | When trying to compile my application under VS 6.0 using nmake from CLI on a WINXP machine I get C:\fltk-1.1.5rc1\FL/filename.H(84) : fatal error C1083: Cannot open include file: 'dirent.h': No such file or directory reading filename.H it suggests alternate locations for a similar file however none of those exists. Is there another file for VS 6.0? thanks, Devan | ||
matt 23:37 Jun 18, 2004 | Please load teh VisualC 6 project file into VC, select 'demo' in the list of sub projects, then click on 'compile and run' in the menu. This should compile fltk, fluid, and all test programs for you. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Help_Dialog improper rendering of < in text |
Version: | 1.1.5rc1 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 19:44 Jun 21, 2004 | If a webpage contains the following: R7 <- R1 + R2 The Fl_Help_View will not display anything inbetween the "<" and the next html command (regardless of the "-". This may also be by design, I'm not sure what the html spec says about rendering non-matched invalid "<" tags. (IE actually prints the "<", which is what I'd ideally want, but it's probably non-standard behavoir). But regardless, a second problem still exists. To get around the above issue, you use the "<" in place of the "<" character. This allows the page to render a less-than sign without trying to intepret it as a bogus HTML command. However, the Fl_Help_View will display it as the text "<" rather than rendering it as a less-than sign. I also checked, and it appears that it isn't rendering any of the &xx's. If this is just a case of < not falling under the "most html 2.0 elements", then I'd like to request that it be added in as soon as possible. I need the Help_View for on-line help for my program, which provides lots of sample code and descriptions, including many cases of R7 <- X, etc. Thanks, JMT. | ||
mike 02:38 Jun 22, 2004 | First, "-" is actually a valid (if undefined) element name in SGML and XML, and so there isn't much we can do to automatically detect the situation (all we can look for is whitespace following the "<") Second, character entities are &...;, you are missing the semicolon at the end. < is supported, not < (which is invalid). | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Revisit of Fl_Return_Button not executing when Enter hit. |
Version: | 1.1.5rc1 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 18:08 Jun 26, 2004 | This is a follow up to STR #324, which, although it says "Closed w/Resolution", it was actually closed w/o resolution. Almost every single dialog and window in my program that has an Fl_Return_Button in it does not execute the Return_Button when Enter is pressed. Even if you tab over to the Return_Button, it still doesn't execute when you hit Enter. Once you have forced the Return_Button to execute by either clicking on it with the mouse, or by Tabbing over to it and using the Space bar to click it, or if you click in one of the Input fields before hitting Enter, from then on that Return_Button will work normally in that window while it is still visible. I have a find dialog, a find/replace dialog, and add define dialog, as well as toolbars in windows, and they all don't work consistently. Nearly every time I open up the dialog, hitting Enter doesn't work. (Randomly, it will work for a couple of times in a row, and then just stop working again.) I don't know what to do. I would suggest maybe a review of the related code just to see if there's something that could allow this to happen. Or maybe I can e-mail someone my whole workspace or have them VNC in or something. I'd submitted sample code before, but the sample code did not do it for you. The funny thing is it never happens with the find/replace dialog in the editor demo code that comes with FLTK. To make things even more funny, I tried making a change to make my code look more like the editor demo, and all of a sudden Return_Button was working consistently. And then I made a small change, and then it stopped working again. I removed thechange, so it was back to the editor-demo way, and blimey it no longer works. And so I reverted my code all the way back to my original way, which was *not* working consistently 10 minutes ago, and now it is working consistently. (pause inbetween writing this STR)... and not it is not working again. There's no rhyme or reason, I swear. It also didn't work on Unix and Linux last time I tried. There must be something in the code that's preventing the KEYBOARD message for Enter from getting through the Fl_Window sub-widgets to the Return_Button unless some condition is met. And it just happens that my code isn't meeting that condition. ::sigh:: JMT. | ||
mike 13:20 Jul 23, 2004 | Can you repost an example here, along with a detailed description of your environment (OS, window manager, compiler, focus policy, etc.) so that we can try to duplicate things here? Thanks! | ||
mike 08:02 Jul 27, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Wrong window raises on show() command |
Version: | 1.1.5rc1 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 18:43 Jun 26, 2004 | I have several text editor windows which have find dialogs (yeah, same dialogs as before, but different problem this time). When the dialogs are closed, I do a show() on the text editor window which called the dialog. The reason I do that call is because, without it, the windowing system seems fit to randomly raise one of the other windows when the dialog window is closed, rather than to consistently raise the window which launched the dialog. First, I believe this may be a bug or lack of feature in the FLTK code. I think it should be able to remember which window was the previous topmost window, and when the topmost window is closed, it should re-raise this window. Well, to get around that problem, in the CloseCB() callback for when the dialog is closed, I force a show() on the text editor which created this dialog. This should gaurantee that the text editor which launched this dialog always gets re-raised when the dialog exits. However, there is one scenario when the window that I call show() on actually *doesn't* get raised. Instead, the 2nd from the top window (in my testing, the text editor which created the dialog is always the 1st from the top window) always gets raised instead. The scenario happens when the dialog causes an fl_alert() message. If no message is raised by the dialog, then the correct text editor window gets re-raised (since I am forcing a show() on it). If the fl_alert() message happens, the 2nd from the top window gets raised. I've verified in debug that I am calling show() on the correct window. Even so, as soon as I resume simulation, the 2nd from the top window is the only one that gets raised, not the correct text_editor. And for the love of kami, I actually have code that reproduces this behavoir! #include <FL/Fl_Double_Window.H> #include <FL/Fl_Button.H> class SpecialEditor : public Fl_Text_Editor { public: Fl_Window *pDlg; SpecialEditor(unsigned int x, unsigned int y, unsigned int w, unsigned int h) : Fl_Text_Editor(x, y, w, h) { pDlg = new Fl_Window(200, 200); Fl_Button *pB = new Fl_Button(50,50,100,50); pB->callback(ButtonCB, this); pDlg->end(); pDlg->callback(CloseDlgCB, this); pDlg->set_non_modal(); end(); } void ShowDlg() { pDlg->show(); } static void ButtonCB(Fl_Widget *pW, void *pV) { fl_alert(""); } static void CloseDlgCB(Fl_Widget *pW, void *pV) { ((SpecialEditor *)pV)->pDlg->hide(); ((SpecialEditor *)pV)->parent()->show(); } }; class SpecialWindow : public Fl_Double_Window { public: Fl_Button *pB; SpecialEditor *pTE; SpecialWindow(unsigned int w, unsigned int h) : Fl_Double_Window(w, h) { pB = new Fl_Button(50,50,200,50); pB->callback(ButtonCB, this); pTE = new SpecialEditor(50, 150, 200, 50); end(); callback(CloseCB, this); show(); } int handle(int Event) { int RetVal = Fl_Double_Window::handle(Event); switch(Event) { case FL_FOCUS: pB->take_focus(); } return RetVal; } static void CloseCB(Fl_Widget *pW, void *pV) { //delete ((SpecialWindow *)pW)->pDWNew; ((SpecialWindow *)pW)->hide(); delete (SpecialWindow *)pW; } static void ButtonCB(Fl_Widget *pW, void *pV) { ((SpecialWindow *)pV)->pTE->ShowDlg(); } }; int main(int argc, char **argv) { SpecialWindow *p1 = new SpecialWindow(300, 300), *p2 = new SpecialWindow(300, 300); return Fl::run(); } To reproduce the behavoir, do the following: You'll notice two windows open up. re-arrange them as necessary so that you can see them both. Both windows have a button in the middle. Clicking the button raises a dialog (actually just another window with a button). If you hit ESC to close the dialog, the window which launched the dialog (the one you clicked the button on) will correctly raise as the topmost window. Now, click a button in one of the two windows to open up the dialog again. This time, click on the button inside of the dialog. This button causes an fl_alert() to display. Now click ESC twice, the first to close the fl_alert(), and the second to close the dialog. This time, you'll see that the wrong window gets raised. Even though I am explicitly calling show() on the right window. You'll notice that it very consistently always raises the other window in this scenario. Thanks, JMT. | ||
mike 13:21 Jul 23, 2004 | What OS, compiler, window manager, focus policy, etc.? | ||
jedimasterthrash 13:29 Jul 24, 2004 | It happens on Windows XP pro SP1. VC6 Ent SP5 No special window managers or focus policies. The behavoir also occurs on: RedHat Linux with the following GCC data: Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --enable-shared --enable-threads=posix --disable-checking --with-system-zlib --enable-__cxa_atexit --host=i386-redhat-linux Thread model: posix gcc version 3.2.2 20030222 (Red Hat Linux 3.2.2-5) But I'm using Exceed 6.2 as the window manager, so it may just be my windows or exceed and not the linux manager that's doing it. Thanks, JMT. | ||
mike 12:46 Jul 26, 2004 | This is almost certainly the window manager setting the focus to the last window which had focus. I tested on Linux with KDE and was unable to duplicate your problem. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Unable to compile the library |
Version: | 1.1.4 |
Created By: | raitech.bol.com |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
raitech.bol.com 19:56 Jul 06, 2004 | README.win32 8k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
raitech.bol.com 19:56 Jul 06, 2004 | My machine is very slow, then trying to compile with MSYS or Cygwin is very very slow. So I try to compile under mingw, which is needed sh.exe and the make.exe. I put this progs. on bin dir. of mingw, but its not work, do not compile!!! In the error, an process try to be created, but because an 'touch' its do not work. Then I found a touch.exe (like my sh.exe, it do not need a msys nor a cygwin dll) and put it in bin dir. When I typing this, the library is compiling. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | c:/fltk/lib/libfltk.a |
Version: | 1.1.5rc1 |
Created By: | campus-joe.web |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
campus-joe.web 04:59 Jul 19, 2004 | Hi, i wanna compile code-file wich should show me the string "Hello World" in a new window...but when i try to compile in cygwin-shell it returns the massage: gcc: c:/fltk/lib/libfltk.a: No such file or directory ..and there is really no file with such name..what shell i do?? Thx for answering...on next friday i will have exame *g* Olli | ||
mike 14:09 Jul 23, 2004 | Please use the fltk.general newsgroup for this question. Thanks! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Need to call Fl::run? |
Version: | 1.1.4 |
Created By: | ahlborn.cs.ucdavis |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ahlborn.cs.ucdavis 13:04 Jul 21, 2004 | Is it necessry to call Fl::run to display Fl_Windows/Fl_Gl_Windows, or subclasses of them? I am having trouble doing so, but it may just be me. Is it necessary for event handling? Thanks, -Ben | ||
mike 14:07 Jul 23, 2004 | You need to at least call Fl::check() to handle any pending window events. Please continue this discussion on the fltk.general group. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Undefined Fl_Help_View symbols when linking with fltk_images |
Version: | 1.1-current |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 15:59 Jul 21, 2004 | This happens on both Solaris Unix and RedHat Linux, using both gcc 2.95 and 3.2. I did not have any building or linking problems before I added usage of the Fl_Help_Dialog, which forced me to include the fltk_images library. I did have to manually undefine the HAVE_LIBPNG etc. in config.h because I didn't need to link with them and I can't find out where they are on the system. I'm using all the latest files from the 1.1.5 CVS. g++ -O3 -w -fpermissive ...assorted .cpp and object files... -Lfltk-1.1.5/lib/Solaris_Sparc -lfltk -lfltk_images -L/usr/X11R6/lib -lX11 -lsocket -o AshIDE.out Undefined first referenced symbol in file Fl_Help_View::topline(int) fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) Fl_Help_View::format(void) fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) Fl_Help_View::Fl_Help_View(int, int, int, int, char const *)fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) Fl_Help_View::find(char const *, int)fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) Fl_Help_View::load(char const *) fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) Fl_Help_View::topline(char const *) fltk-1.1.5/lib/Solaris_Sparc/libfltk_images.a(Fl_Help_Dialog.o) ld: fatal: Symbol referencing errors. No output written to AshIDE.out collect2: ld returned 1 exit status *** Error code 1 Thanks, Ash. | ||
jedimasterthrash 08:55 Jul 22, 2004 | You can delete this STR. I was too quick to the gun because I was anxious to get this built and done since all I had left to do was build unix binaries for my app. I later realized it's probably something I can fix with the makefiles, so I posted in the Newsgroups forum and got a quick reply. JMT. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Choice doens't set the changed flag |
Version: | 1.1.5rc1 |
Created By: | ariock.skynet |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ariock.skynet 06:22 Jul 23, 2004 | Fl_Choice appears not to set at all the changed flag as shown in the test program below... I tried many of the when possibilities but it doesn't change anything, the changed() flag is never set. #include <FL/FL.h> #include <FL/FL_Window.h> #include <FL/FL_Choice.h> #include <FL/FL_ask.h> void cbCh(Fl_Widget *w,void *v) { if(w->changed()) fl_alert("changed"); } int main() { Fl_Window w(100,100,"test choice"); Fl_Choice ch(0,0,100,20); ch.add("choice 1"); ch.add("choice 2"); ch.add("choice 3"); ch.add("choice 4"); ch.value(1); ch.callback(cbCh); ch.when(FL_WHEN_RELEASE_ALWAYS); w.end(); w.show(); return Fl::run(); } | ||
mike 13:08 Jul 23, 2004 | changed() is not set when you have a callback defined. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | projects declared multiple times in rc2 VC workspace |
Version: | 1.1.5rc2 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 18:12 Jul 28, 2004 | When I opened up the visualc\fltk.dsw from the new 1.1.5rc2, I'm getting about a hundred message boxes saying: " This workspace file is corrupted: it contains more than one project called 'cubeview'. Those extra projects will be removed from the workspace. " The above happend for about every single project in the fltk workspace. Then I'm also getting message boxes saying: " F:\program files\...\visualc\jpeg.dsp This makefile was not generated by Developer Studio. Continuing will create a new Developer Studio project to wrap this makefile. You will be prompted to save after the new project has been created. Do you want to continue? " The above happened for about every single library project in teh fltk workspace. I clicked yes on everything, but then got some message box about a sharing violation for fltk.dsw, and then MSVC "encountered a problem an needs to close". I'm using MSVC 6 SP5 on WinXP pro SP1. I never had any problems opening the workspaces from other FLTK releases. Thanks, JMT. | ||
jedimasterthrash 18:35 Jul 28, 2004 | The next two times I opened up the rc2 workspace, MSVC crashed. But the third time is a charm! It all builds and works now. Still no problems with the rc1 workspace. Thanks, JMT. | ||
mike 13:37 Sep 07, 2004 | Dupe of #476... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | mingw gcc 3.4 build of color_chooser.exe doesn't work |
Version: | 1.1.5rc2 |
Created By: | baraclese |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
baraclese 11:10 Jul 31, 2004 | color_chooser.exe 0.3M | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
baraclese 11:10 Jul 31, 2004 | Hi, I build fltk using mingw gcc 3.4 (WinXP). All tests run fine except for color_chooser. No window shows up, no message box indicating an error, nothing remains in the task manager either. Not sure if attaching executables is common but here it is for those who want to try it. | ||
mike 13:25 Sep 07, 2004 | Sorry, not much we can do here to help you; more than likely it is a MingW problem... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Re-Revisit of Fl_Return_Button not executing when Enter hit. |
Version: | 1.1.5rc2 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 09:44 Aug 01, 2004 | This is a follow up of STR #324 and # 450. I don't know why you closed #450 so fast. It was only 3 days and 20 hours since you posted asking if I can repost an example. It took me 3 days to have time to come up with another example. And then the STR was closed because it wasn't "updated by the submitter for two or more weeks". Yeah, it had been two weeks since I last updated it, but it was only 3 days since I had gotten a response asking me to do something. Anyway, I recreated another example. This is just the text editor stripped out of my full project. My full project demonstrates the problem consistently every time, but this stripped-down one is more sporadic. http://wisegirl.dyndns.org/fltk/GUITest.zip The first time I built and ran it, it exhibited the behavoir (return button not executed when you hit return). The next several times I ran it, it did not. Then I did a build clean a couple times, and after the 3rd build clean, it started exhibiting it again. I know you weren't able to get my previous demo code to work, but it may just have been the sporadic nature of it. If it doesn't work right away, just keep re-building it or re-running it. Eventually it will. Sometimes it helps if you make an insignificant change in the code on a re-build. If it happens to be in a good mood and the bug is being exhibited, do the following: Run it. It prompts you to open a file, and then displays it in the text editor. Do a ctrl-f to open the find dialog box. Don't hit any other keys or mouse buttons. Then type your search word (it is automatically entered into the find input). I suggest choosing a funky word that is not actually in the document, that way you'll get an error box when it's not found, which is the easiest way to tell that the return-button did in fact work. Then hit Enter. If you don't get the message box, then the return-button didn't work. You can keep hitting enter all you want, it still won't work. It will begin working if you click with the mouse anywhere in the find dialog. This leads me to believe it may be a focus issue of some kind. But that's wierd because the keyboard text is getting correctly routed to the find input field, and you can TAB around the buttons. The environment I've tested on is WinXP SP1 VC6 SP5. It also exhibits the same behavoir in Solaris and RedHat with GCC, but again, I'm using the GUI through Exceed. The fltk texteditor demo never exhibits this behavoir. I also modified my code to behave exactly the same as the texteditor demo, and mine still does exhibit the behavoir. But there is one main difference. The texteditor demo uses the fltk DLL, while my code is using the fltk LIB. Thanks, JMT. | ||
mike 19:18 Sep 12, 2004 | I'm still not able to duplicate your problems, and since all of your work is through Windows (even for the X11 stuff on non-Windows platforms), I'm guessing that there is a focus issue specific to your Windows system that we are unable to duplicate... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Heap error, debug crash |
Version: | 1.1.5rc2 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 09:56 Aug 01, 2004 | I actually tripped across this bug with the code I made for the return-button STR. http://wisegirl.dyndns.org/fltk/GUITest.zip This bug is irrespective of the return-button bug. It just happens to show up with this sample code. Here's how to reproduce it: 1. Run the program in debug mode in MSVC (use F5). It prompts you to open a file. It opens it in a texteditor window. 2. Do ctrl-f to open the find dialog. This time, enter a word that *does* exist in the text file. Then click enter to find it. It will highlight the found word. 3. Now, hit just the ESCape key to close the find dialog (don't hit anything else or click anywhere else). The find dialog closes, the texteditor window raises to be the current window, and the found word is still highlighted in the text display. 4. Now, without clicking on anything else, just click the X in the upper right corner to close the editor window. You'll notice that the editor does not close. I have verified in debug that the CloseCB callback is never called. Click it a few times and it still doesn't close. 5. Click somewhere else in the text editor window, either in the text display or on a menu. Now the focus seems to correctly get re-registered with the window. Now you can click the X in the upper-right corner again, and this time it will close. 6. This is why you need to run in debug mode in MSVC. After the application returns from main, MSVC's debugger does its post-execution checks, including running through the malloc Heap to look for memory leaks. MSVC reports an error while executing the heap_dbg checking code. I've debugged these before with my own code. It's usually a result of code writing off the edge of its allocated buffer, or writing to heap space that it has previously free'd/deleted. This error only shows up when the bug of "clicking the upper-right X doesn't close the window" occurs. So I assume they are related. I'm suspecting that FLTK code somewhere is doing something to handle the upper-right X, but since it's never actually calling the CloseCB command, maybe it's running off some bad pointer or something, which then is later detected as this debug-heap error. The environment I've tested on is WinXP SP1 VC6 SP5. Thanks, JMT. | ||
mike 19:38 Aug 07, 2004 | First, you need to use "friend class File_Text_Editor" in FileWindow.h, not just "friend File_Text_Editor". I tried it, but was unable to duplicate your problem - as soon as I clicked on the "X" it closed the window. I tried this on Linux with Valgrind (which actually tells you where memory problems are caused) and it shows that the problem originated on line 295 of FileWindow.cpp, so it looks like doing "delete this" inside the callback interferes with the new code that clears the "changed" bit after a callback. Will look into a fix for that... | ||
jedimasterthrash 12:15 Aug 10, 2004 | Yeah, I do a "delete this" in the close methods of all my main windows, because I'm one of those rare individuals who makes sure I clean up all my memory. Otherwise just closing a window by hide()ing it doesn't remove the class from memory. I read that the "class" specifier is only needed if the class is not yet declared. In my example, the File_Text_Editor is declared directly above the friend statement, so it's then not necessary. Though my source for this info is MSDN (copied below), which probably doesn't hold much weight around here. I searched the web hoping to find the standard somewhere, but all I found was a couple posts for the GCC compiler about how it was going to change its behavoir to flag an error if the "class" keyword is missing, but it appears it's just their interpretation of the standard, and there wasn't a reference for the change besides a post by someone saying it was so. And I wouldn't trust the GCC compiler to be 100% accurate in its interpretation of the C++ standard anymore than I trust MSVC. I've found several bugs in GCC 3.2, including a really nasty one, where the compiler implements a rotate shifter for left-shifting 64-bit numbers by a variable amount instead of a regular left-shifter. I could go on to list a dozen more but it's off topic. To be fair, the list of bugs I've had to work around in MSVC6 is larger than the bugs I've had to work around to support either GCC 2.9x or GCC 3.2. That's what I love about FLTK though. When I find a bug, it can get fixed! What a concept! I wish I could do the same thing with Windows. " When you declare a friend class in a member list, you can omit the member-declarator-list. For more information on friends, see friend Specifier in Chapter 6, and Friends in Chapter 10. Even if a class name has not been introduced, it can be used in a friend declaration. This friend declaration introduces the name. However, in member declarations for such classes, the elaborated-type-specifier syntax must be used, as shown in the following example: class HasFriends { public: friend class NotDeclaredYet; }; In the preceding example, there is no member-declarator-list after the class declaration. Because the declaration for NotDeclaredYet has not yet been processed, the elaborated-type-specifier form is used: class NotDeclaredYet. A type that has been declared can be specified in a friend member declaration using a normal type specifier: class AlreadyDeclared { ... }; class HasFriends { public: friend AlreadyDeclared; }; " | ||
mike 18:23 Sep 12, 2004 | Unfortunately, it doesn't look like we can correct this without eliminating the changed() change, and we don't want to do that. Currently the only workaround I can think of is to hide the widget in the callback and delete the widget outside the callback. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Picking "double Buffered" in the 2 screen fullscreen demo makes the menu window vanish |
Version: | 1.1.5rc2 |
Created By: | wayne.r.jones.boeing |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wayne.r.jones.boeing 11:29 Aug 02, 2004 | Run the fullscreen demo in 2-window mode (ie: fullscreen -2) on a Win32 PC. Then pick the "Double Buffered" button. This causes the menu window to completely and irretrievably vanish. This has been broken since fltk-1.1.3. I would very much like to be able to run in this 2 window mode so I hope you can fix this. | ||
mike 17:35 Aug 18, 2004 | I am unable to duplicate this on my WinXP system. Can you test on another PC? It could very well be a graphics driver bug. | ||
wayne.r.jones.boeing 18:55 Aug 21, 2004 | Michael, I tried this on my home PC and got the same result. More detail: I have a Dell Dimension 8300, XP professional, a 128MB DDR ATI RADEON 9800 PRO graphics card, system acquired just before Xmas 2003. Before starting, I went to the Dell website to see if there was an updated driver. There was, dated Jan, 2004 and I downloaded and installed it. Next, I downloaded a copy of fltk-1.1.5rc2 to install (I had not previously installed 1.1.5 at home). As before, I changed the end of line characters in jpeg.dsp and libpng.dsp and fixed my settings to see jpeglib.h, etc. and did a batch build of everything sucessfully. To test the problem, I copied over my .bat file from 1.1.4 containing the single line: fullscreend -2. I executed the .bat file bringing up the two window version, I did nothing other than push the "Double Buffered" button and the menu window vanished. When I try to follow execution in the debugger, I see it getting to Line 382 in Fl_Gl_window.cxx where it proceeds to hide the menu window. I have not tried to make sense out of exactly how it gets to that point. If you can't reproduce the problem, would it make sense to send you my executable? Is there anything else I can do to help you identify and fix this problem? -Wayne | ||
mike 08:53 Sep 24, 2004 | No, we don't run user executables under any circumstances here, so that won't help. Do both PCs have ATI cards? | ||
wayne.r.jones.boeing 09:54 Sep 24, 2004 | Yes, both machines have ATI cards. The one at home is the one I listed above. The one I am running at work is a RADEON 9700. | ||
mike 10:46 Sep 24, 2004 | Well, you probably are using the same driver or a driver with the same bugs; I can't duplicate this on my (admittedly old) TNT2-based Windows test system. | ||
wayne.r.jones.boeing 14:25 Sep 28, 2004 | 1) It also fails on a PC with a garden variety Intel 82815 Graphics Controler. 2) It also fails when built with a different delivery disk of Visual C++. 3) It also fails on on my SGI O2 running IRIX 6.5.14. Are you sure you are running the 2-screen fullscreen demo to check for this problem? The single screen (default) version seems ok. Can I loan you a more up to date graphics card to test on? | ||
mike 07:53 Sep 29, 2004 | I am testing with "./fullscreen -2" on Linux and running with the "-2" option in VC++ on Windows. I have both SGI Indigo, Indigo2, O2, and Octane machines in our lab and will test on one of them. I'll see about testing on a dual-head Win2k system we have with a NVidia Quadro card in it, but it doesn't have any developer tools on it... | ||
mike 06:34 Nov 21, 2004 | Unable to duplicate, unable to fix. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Config Files |
Summary: | CPPFLAGS is not used |
Version: | 1.1.5rc2 |
Created By: | prj.po.cwru |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
prj.po.cwru 13:26 Aug 02, 2004 | The "./configure --help" output claims that $CPPFLAGS is influential. It is used by the configure script, but it doesn't find its way into the Makefiles, so my -I flags disappear, and the jpeg and other headers can't be found: Compiling Fl_JPEG_Image.cxx... Fl_JPEG_Image.cxx:53:23: jpeglib.h: No such file or directory Also, make still exits 0, so it's difficult for a script to detect that the build failed. | ||
mike 17:40 Aug 18, 2004 | CPPFLAGS is used for testing include files, but not for compiling. Set CFLAGS and CXXFLAGS before running ./configure. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | A really strange Fl_Widget behavior |
Version: | 1.1.5rc2 |
Created By: | ybourbeau.mrgtech |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ybourbeau.mrgtech 13:45 Aug 02, 2004 | If i create a Fl_Widget directed class like a Fl_Pack like this way: Fl_Pack* pack1 = new Fl_Pack(0,0,0,0,""); And then add buttons to it, with pack1->add, the callbacks attached to this buttons are not active at runtime (ie: clicking on the button do nothing, the pushed effect not showing up, nothing, nada), if i do a resize (I have to do a resize to actually view the buttons, yes), nothing seem to change, but if i init this way: Fl_Pack* pack1 = new Fl_Pack(0,0,1000,1000,""); And then do a resize(), the callbacks are preserved... then my question: WHY? I got a really hard time with this strange behavior because im using non-fixed height widgets to pass on a Fl_Pack widget | ||
mike 17:34 Aug 18, 2004 | Please attach a short example program which demonstrates this problem. | ||
mike 19:14 Sep 12, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Help_Dialog.cxx missing in src/Makefile |
Version: | 1.1.5rc2 |
Created By: | lists.rumori |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
lists.rumori 17:04 Aug 02, 2004 | Makefile 17k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
lists.rumori 17:04 Aug 02, 2004 | the file Fl_Help_Dialog.cxx is missing in src/Makefile in CPPFILES. therefore it's not linked and the symbols are missing when linking applications using Fl::Help_Dialog. bests, martin | ||
mike 17:41 Aug 18, 2004 | It is not missing, you just aren't linking against the fltk_images library. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Summary of VC6 build problems and fixes |
Version: | 1.1.5rc2 |
Created By: | nospam.fakemailaddress12345 |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nospam.fakemailaddress12345 09:01 Aug 03, 2004 | Some of these issues don't surface until you try to build fluid.exe. Problem 1: jpeg.dsp and libpng.dsp projects appear invalid to VC6. Fix: convert CR to CRLF in these files. Problem 2: invalid #include path for jpeg.h, png.h and zlib.h Fix: insert their directory names. #include "jpeg.h" to "jpeg/jpeg.h" #include "png.h" to "png/png.h" #include "zlib.h" to "zlib/zlib.h" Problem 3: jpeg.lib, png.lib and z.lib not found after successful compile Fix: modify their .dsp file to output to ../lib directory. they are being output to the wrong directory. | ||
ktwn.dellmail 08:48 Aug 13, 2004 | Excuse me sir, i am a learner in VC6. you said, change CR to CRLF in jpeg.dsp and libpng.dsp. how to fix and where to? pls, reply me to ktwn at dellmail dot com. thanks | ||
mike 13:33 Sep 07, 2004 | The proposed changes to the includes won't work for normal builds against pre-installed versions of those libraries. The real problem is the missing include paths in the project files. Dupe of STR #476. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Some strange things under MacOsX |
Version: | 1.1.5rc2 |
Created By: | jeromevernet.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jeromevernet.hotmail 01:58 Aug 04, 2004 | I have some problem with the 1.1.5rc2 release when I compil on MacOsX 10.3.4: - the overall performance of the programs compiled with 1.1.5 is about 10 time *slower* than with 1.1.4, but a key mapping problems in the 1.1.4 version prevent to use 1.1.4. - I cannot use Fl_Help_Dialog, something is missing and I get errors messages when linking, reporting missing _jpg, etc... J. Vernet | ||
matt 13:41 Aug 09, 2004 | You have to compile and link with the jpeg (and png) libraries if you are going to use images in you application. Make sure the the 'configure' script is able to find all required libraries by checking the configure log. How did you see that 1.1.5 is that much slower? Please give us the application you used to measure the fltk performance, and explaint what you saw. This is the only way to find and fix the problem. | ||
jeromevernet.hotmail 22:53 Aug 09, 2004 | The application is an emulator for an old Computer. The project will have soon a Sourceforge repository. The programs call fl_wait(0) often to check for any event. It seems that this function is much slower (on MacOsX) than with the previous version of fltk -which is unusuable on Macintosh, as there is key mapping problems-, as the emulator report only a max speed of the emulated CPU at '7 Mhz' with 1,1,5, and report more than '70 Mhz' whith 1,1,4. Jerome | ||
mike 17:36 Aug 18, 2004 | fl_wait() is a private function; does "Fl::check()" work for you? | ||
jeromevernet.hotmail 06:07 Aug 23, 2004 | >fl_wait() is a private function; does "Fl::check()" work for you? Fl:check() works, yes. But it's slower than with previous version of fltk on my Mac (10.3.4/XCode 1.2/gcc 3 ?) with wait(0) calls. I 've changed all the wait(0) calls by check(), but no changes. Jerome | ||
jeromevernet.hotmail 09:07 Sep 07, 2004 | Hi, No news for this problem ? Does anybody have encountered this ? Jerome | ||
mike 18:53 Sep 12, 2004 | First, ALL SUPPORT IS PROVIDED BY VOLUNTEERS. We don't guarantee any kind of response time. Second, no one else has reported this problem. Unless you can post a short program which demonstrates the problem there isn't much we can do to help you. I would recommend that you try the current CVS or wait a few days for 1.1.5rc3 to see if our recent MacOS X changes improve the performance for you. | ||
jeromevernet.hotmail 01:21 Sep 13, 2004 | Hi, You can try VirtualT, from here: http://sourceforge.net/projects/virtualt/ It's an Tandy Model 100 Emulator. To compile it under MacOsX, you should have to make a very few changes in the sources and Makefile to make it running). Under W2k on a 1.8Ghz PIV, you will get around 100 Mhz for the emulated Tandy. On my 2x1.25 Ghz Power Macintosh, you have only a 6/8 Mhz emulated Tandy. And 9 Mhz on a 2x2Ghz G5. It seems that the Fl_check() function is far far slower on Mac than on PC, using the same release of FLTK -the latest 1.1.5 version you can get thru CVS-. Jerome Vernet | ||
mike 16:38 Oct 18, 2004 | We can't test against large programs. If you can reproduce this problem with a small test program, please attach the source and we'll look into this further. | ||
rollos.starband 08:55 Nov 18, 2004 | This is not a trouble report, but a request for info; I don't know where else to post it: Where can I download a binary of latest FLTK for Mac OS X 10.2 (Jaguar)/Darwin? | ||
mike 17:18 Nov 19, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | swap() for Fl_Browser only swaps first two items |
Version: | 1.1.5rc2 |
Created By: | johnrrdrake.netscape |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
johnrrdrake.netscape 08:21 Aug 06, 2004 | The newly added swap function to the Fl_Browser only works for the first two selected rows, regardless of how many rows are selected and in which order they were selected. This behaviour is not intuitive. (Of course, I can't imagine how you could possibly make the behaviour intuitive for all persons). I would expect (in no particular order, and not necessarily compatible): (1) swap() must be enabled as an attribute for the browser and permits selecting only 2 rows (optionally, it would be nice to be able to swap regions as well - so it could be generalized to 2 regions). (2) swap() would rotate selected items. So if 3 rows were selected, they rows would either rotate up or down. Up or down rotation could be user selectable behaviour. (3) swap() swaps all selected rows (if 6 rows were selected, then 1&6, 2&5, 3&4 would swap with one another) Of course, (2) and (3) could be combined with the behaviour being either rotate or swap all selected items. Of course, this should all be simple, not bulk up the code, and satisfy everyone. :-) rr | ||
mike 13:41 Sep 07, 2004 | Can you attach a short example program that shows the problem? swap() swaps the two rows that are specified in the call, not the current selection... | ||
johnrrdrake.netscape 14:09 Sep 07, 2004 | I observed the problem using the browser example in the \test directory. I have not actually got around to coding a test example myself. | ||
mike 13:28 Sep 12, 2004 | Please post a new STR if you manage to come up with a problem in your own code using the swap() method. The browser demo does not handle more than 2 selected lines as that would require multiple Fl_Browser::swap() calls. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | problem compiling fltk - incomplete build |
Version: | 1.1.5rc2 |
Created By: | ggutierrez.marketscan |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ggutierrez.marketscan 13:54 Aug 06, 2004 | I am receiving the following errors and warnings when building fltk115rc2 on msys for dev-cpp4990 w/mingw3.3 (it will finish but wont build the whole library): Here is the configure section: checking for gcc... gcc checking for C compiler default output... a.exe checking whether the C compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... .exe checking for suffix of object files... o checking whether we are using the GNU C compiler... yes checking whether gcc accepts -g... yes checking for gcc option to accept ANSI C... none needed checking for g++... g++ checking whether we are using the GNU C++ compiler... yes checking whether g++ accepts -g... yes checking for nroff... no checking for htmldoc... no checking for ranlib... ranlib checking for ar... /mingw/bin/ar checking whether byte ordering is bigendian... no checking how to run the C preprocessor... gcc -E checking for egrep... grep -E checking for ANSI C header files... yes checking for sys/types.h... yes checking for sys/stat.h... yes checking for stdlib.h... yes checking for string.h... yes checking for memory.h... yes checking for strings.h... yes checking for inttypes.h... yes checking for stdint.h... yes checking for unistd.h... yes checking for short... yes checking size of short... 2 checking for int... yes checking size of int... 4 checking for long... yes checking size of long... 4 checking whether the compiler recognizes bool as a built-in type... yes checking for dirent.h that defines DIR... yes checking for library containing opendir... none required checking sys/select.h usability... no checking sys/select.h presence... no checking for sys/select.h... no checking sys/stdtypes.h usability... no checking sys/stdtypes.h presence... no checking for sys/stdtypes.h... no checking for scandir... no checking for vsnprintf... yes checking for snprintf... yes checking for strings.h... (cached) yes checking for strcasecmp... yes checking for strlcat... no checking for strlcpy... no checking for library containing pow... none required checking for jpeg_CreateCompress in -ljpeg... yes checking for gzgets in -lz... yes checking for png_set_tRNS_to_alpha in -lpng... yes checking png.h usability... yes checking png.h presence... yes checking for png.h... yes checking GL/gl.h usability... yes checking GL/gl.h presence... yes checking for GL/gl.h... yes checking GL/glu.h usability... yes checking GL/glu.h presence... yes checking for GL/glu.h... yes checking if libsupc++ is required... yes checking if GCC supports -fno-exceptions... yes configure: creating ./config.status config.status: creating makeinclude config.status: creating fltk.list config.status: creating fltk-config config.status: creating FL/Makefile config.status: creating config.h Here are the error messages from the make: Compiling Fl_Color_Chooser.cxx... Fl_Color_Chooser.cxx: In member function `virtual Fl_Color_Chooser::~Fl_Color_Chooser()': Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here Fl_Color_Chooser.cxx: In member function `virtual Fl_Color_Chooser::~Fl_Color_Chooser()': Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here Fl_Color_Chooser.cxx:423: warning: inlining failed in call to `virtual Flcc_Value_Input::~Flcc_Value_Input()' Fl_Color_Chooser.cxx:480: warning: called from here File chooser gives these warnings: Fl_File_Chooser.cxx: In constructor `Fl_File_Chooser::Fl_File_Chooser(const char*, const char*, int, const char*)': Fl_File_Chooser.cxx:138: warning: declaration of `o' shadows a previous local Fl_File_Chooser.cxx:135: warning: shadowed declaration is here Compiling Fl_Text_Buffer.cxx... Fl_Text_Buffer.cxx: In static member function `static int Fl_Text_Buffer::expand_character(char, int, char*, int, char)': Fl_Text_Buffer.cxx:945: warning: array subscript has type `char' Fl_Text_Buffer.cxx: In static member function `static int Fl_Text_Buffer::character_width(char, int, int, char)': Fl_Text_Buffer.cxx:972: warning: array subscript has type `char' Compiling Fl_x.cxx... In file included from Fl_x.cxx:27: Fl_win32.cxx: In static member function `static void Fl::add_fd(int, int, void (*)(int, void*), void*)': Fl_win32.cxx:160: warning: comparison between signed and unsigned integer expressions Fl_win32.cxx:161: warning: comparison between signed and unsigned integer expressions Fl_win32.cxx:162: warning: comparison between signed and unsigned integer expressions Fl_win32.cxx: In static member function `static void Fl::add_fd(int, void (*)(int, void*), void*)': Fl_win32.cxx:167: warning: declaration of `fd' shadows a global declaration Fl_win32.cxx:139: warning: shadowed declaration is here In file included from Fl_x.cxx:27: Fl_win32.cxx: In member function `virtual void Fl_Window::resize(int, int, int, int)': Fl_win32.cxx:891: warning: declaration of `flags' shadows a member of `this' In file included from Fl_x.cxx:27: Fl_win32.cxx: In static member function `static Fl_X* Fl_X::make(Fl_Window*)': Fl_win32.cxx:1023: warning: declaration of `w' shadows a parameter Fl_win32.cxx:934: warning: shadowed declaration is here Compiling fl_draw_pixmap.cxx... fl_draw_pixmap.cxx: In function `int fl_draw_pixmap(const char* const*, int, int, Fl_Color)': fl_draw_pixmap.cxx:150: warning: array subscript has type `char' Compiling scandir.c... In file included from scandir.c:20: scandir_win32.c:35: warning: declaration of `select' shadows a global declaration W:/Dev-Cpp/include/winsock2.h:565: warning: shadowed declaration is here scandir_win32.c: In function `fl_scandir': scandir_win32.c:46: warning: passing arg 1 of `malloc' as unsigned due to prototype scandir_win32.c:75: warning: passing arg 2 of `calloc' as unsigned due to prototype scandir_win32.c:98: warning: passing arg 2 of `qsort' as unsigned due to prototype Compiling Fl_Gl_Choice.cxx... Compiling Fl_Gl_Overlay.cxx... Compiling Fl_Gl_Window.cxx... Fl_Gl_Window.cxx: In member function `int Fl_Gl_Window::mode(int, const int*)': Fl_Gl_Window.cxx:98: warning: unused variable `Fl_Gl_Choice*oldg' Compiling Fl_Menu_Type.cxx... Fl_Menu_Type.cxx: In function `void __static_initialization_and_destruction_0(int, int)': Fl_Menu_Type.cxx:90: warning: inlining failed in call to `virtual Fl_Submenu_Type::~Fl_Submenu_Type()' Fl_Menu_Type.cxx:90: warning: called from here Compiling Fl_Group_Type.cxx... Fl_Group_Type.cxx: In function `void __static_initialization_and_destruction_0(int, int)': Fl_Group_Type.cxx:151: warning: inlining failed in call to `virtual Fl_Tabs_Type::~Fl_Tabs_Type()' Fl_Group_Type.cxx:151: warning: called from here Compiling factory.cxx... factory.cxx: In function `void __static_initialization_and_destruction_0(int, int)': factory.cxx:665: warning: inlining failed in call to `virtual Fl_Value_Slider_Type::~Fl_Value_Slider_Type()' factory.cxx:665: warning: called from here Compiling checkers.cxx... checkers.cxx: In function `void evaluateboard(node*, int)': checkers.cxx:224: warning: array subscript has type `char' checkers.cxx: In function `void movepiece(node*, int, node*)': checkers.cxx:416: warning: array subscript has type `char' Compiling demo.cxx... demo.cxx: In function `void push_menu(const char*)': demo.cxx:179: warning: array subscript has type `char' demo.cxx: In function `void pop_menu()': demo.cxx:188: warning: array subscript has type `char' Linking demo.exe... editor.cxx: In constructor `EditorWindow::EditorWindow(int, int, const char*)': editor.cxx:425: warning: declaration of `h' shadows a member of `this' editor.cxx:425: warning: declaration of `w' shadows a member of `this' Linking editor.exe... Generating fast_slow.fl... make[1]: *** [fast_slow.cxx] Error 128 gast_slow.cxx and fast_slow.h are in the /test directory | ||
mike 17:43 Aug 18, 2004 | fast_slow is a demo; the error 128 sounds like fluid crashed - no core/dump file? Without more information there isn't much we can do - we don't use MingW here... | ||
mike 19:15 Sep 12, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | test/fonts.cxx doesn't display fonts correctly |
Version: | 1.1-current |
Created By: | Albrecht.Schlosser |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Albrecht.Schlosser 04:26 Aug 10, 2004 | fonts_diff.txt 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Albrecht.Schlosser 04:26 Aug 10, 2004 | Display of fonts is truncated at character '@' and 0xA0. Workaround: the attached modification replaces these characters by spaces. This seems not to be the correct way, but shows the full fonts without truncation. Maybe the problem has to do with "@ symbol" drawing. Additional modification: other sizes of the widgets and another resizable() (the full window) allows display of bigger fonts in the display area. See attached file for modifications. Albrecht Schlosser | ||
Albrecht.Schlosser 09:59 Aug 10, 2004 | Correction/adddition: Doubling the '@' sign works (as it should), maybe I made a mistake with my previous tests :-( 0xA0 works, too, don't know what the problem was ... Albrecht | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | have error in compiling fltk 1.1.5rc2 |
Version: | 1.1.5rc2 |
Created By: | paradise-dtp.mail4u.com |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
paradise-dtp.mail4u.com 08:36 Aug 12, 2004 | When i open fltk.dsw using ms visual c++ 6.0, two error found. No.1 fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory Fl_PNG_Image.cxx No.2 fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory Fl_PNM_Image.cxx Error executing cl.exe. pls | ||
mike 13:32 Sep 07, 2004 | Dupe of #476 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Memory leaks in Hello World example |
Version: | 1.1.5rc2 |
Created By: | nospam.fakemailaddress12345 |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
nospam.fakemailaddress12345 10:02 Aug 14, 2004 | //Listing 1 - "hello.cxx" (MODIFIED TO DETECT MEMORY LEAKS IN VC6) //Upon program exit, 3 memory leaks are noted: 1100 bytes, 68 bytes, 128 bytes. Details shown at the end of program listing. Recommend using DbgView (freeware) to see the debug messages. #define _CRTDBG_MAP_ALLOC #include <cstdlib> #include <crtdbg.h> #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Box.H> int main(int argc, char **argv) { _CrtSetDbgFlag( _CRTDBG_ALLOC_MEM_DF | _CRTDBG_LEAK_CHECK_DF); Fl_Window *window = new Fl_Window(300,180); Fl_Box *box = new Fl_Box(20,40,260,100,"Hello, World!"); box->box(FL_UP_BOX); box->labelsize(36); box->labelfont(FL_BOLD+FL_ITALIC); box->labeltype(FL_SHADOW_LABEL); window->end(); window->show(argc, argv); return Fl::run(); } /* [2176] Detected memory leaks! [2176] Dumping objects -> [2176] d:\apps\microsoft visual studio\vc98\include\crtdbg.h(552) : [2176] {99} [2176] normal block at 0x00995708, 1100 bytes long. [2176] Data: < \ > 00 00 00 00 5C 03 0A CF 1B 00 00 00 1B 00 00 00 [2176] d:\apps\microsoft visual studio\vc98\include\crtdbg.h(552) : [2176] {52} [2176] normal block at 0x00992DB8, 68 bytes long. [2176] Data: <lPF - ^@ > 6C 50 46 00 00 2D 99 00 D0 5E 40 00 00 00 00 00 [2176] d:\apps\microsoft visual studio\vc98\include\crtdbg.h(552) : [2176] {51} [2176] normal block at 0x00992D00, 128 bytes long. [2176] Data: < PF ]@ > A0 50 46 00 00 00 00 00 CB 5D 40 00 00 00 00 00 [2176] Object dump complete. */ | ||
mike 13:15 Sep 07, 2004 | We make the assumption that the OS will reclaim any memory when a program exits, so certain global state data (such as the list of available fonts on the system) is cached for the duration of the process. If you have show a general memory leak that occurs while the program is running, please submit a new STR with the corresponding example program. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | Dead keys in Linux doesn't work |
Version: | 1.1.5rc2 |
Created By: | tux.inmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
tux.inmail 08:09 Aug 19, 2004 | If I use FLTK in Linux, deadkeys (like ì¹èø¾ýáÈ©Ø...) doesn't works. But, another deadkeys like éáú works fine. In windows version (compiled in Linux by mingw-cross tools) all works fine... | ||
cyrpier7.hotmail 17:10 Aug 26, 2004 | I have kind of same problem, but with WIndows 98 and I compiled fltk 1.1.5rc2 with mingw and msys. But me, the "é" doesn't work neither... | ||
mike 11:09 Sep 07, 2004 | I'm not able to reproduce the problem here; what character set are you using? | ||
tux.inmail 03:27 Sep 08, 2004 | It's iso-8859-2. I'm from Czech Republic. I can change coding in fltk components to iso-8859-2, but can't write this characters... For example è is in czech written by this two keys. +c (It means Shift+key where is + and than c)... Fltk ignore first character, write · and than c. I have this problem only in Linux, windows (and WINE TOO!!!!) has no problem. | ||
mike 18:55 Sep 12, 2004 | OK, that would explain it. The current FLTK I18N support is limited to ISO-8859-1 for input; you'll need 2.0 to support the full range of Unicode characters... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #449 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Menus don't drag with window |
Version: | 1.1.5rc2 |
Created By: | derek.signalmarketing |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
derek.signalmarketing 11:13 Aug 23, 2004 | On Mac OSX if you click to drop down a menu, then drag the window by its titlebar, the menu stays in place as the window moves. The Favorites drop down menu in the file chooser exhibits the same behaviour. | ||
mike 13:53 Sep 07, 2004 | Dupe of #449. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | Strange segfaults when using callback |
Version: | 1.1.4 |
Created By: | sirver.gmx |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
sirver.gmx 16:17 Sep 05, 2004 | scoiattolo.tgz 5k | ||
anonymous 13:27 Sep 07, 2004 | problem.cc 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sirver.gmx 16:17 Sep 05, 2004 | Hi, maybe i'm just plain stupid, but i seem to have this problem which won't go away no matter which os (linux 2.4, win98, fltk1.1, gcc-3.4 on both os'es). The problem occurs following: i have a main window as my os window, clicking a button there opens the 'add-word' window (it's gonna be an article trainer for some language students). when the ok button there is pressed, all fields in the dialog should be set to empty again and the new word should be included in the wordlist. All the time, the programm crashes. following source snippets: *****************+ snip **********************+ Fl_Double_Window* o = new Fl_Double_Window(690, 175); o->user_data(static_cast<void*>(this)); ******* some lines missing ***************** m_artikel = new Fl_Input(60, 60, 105, 25, "Artikel:"); m_german_w = new Fl_Input(290, 60, 395, 25, "Deutsches Wort:"); m_italien_w = new Fl_Input(290, 90, 395, 25, "Italienisches Wort:"); std::cerr << "german: " << m_german_w << std::endl; std::cerr << "this: " << this << std::endl; // Return button Fl_Return_Button* r = new Fl_Return_Button(245, 135, 130, 25, "Hinzuf\374gen"); r->callback(ms_hinzufuegen_callback); *********************** snap*********************+ ************************ snip ********************** void Add_Word::ms_hinzufuegen_callback(Fl_Widget* o, void* data) { Add_Word* v=static_cast<Add_Word*>(o->parent()->user_data()); std::cerr << "user data: " << v << std::endl; std::cerr << "german1: " << v->m_german_w << std::endl; v->m_german_w->value(""); std::cerr << "ALIVE--" << std::endl; std::cerr << "ALIVE" << std::endl; v->m_artikel->value(""); ******************** snap ********************** The ouput is the following: german: 0x8090810 this: 0xbfffe250 user data: 0xbfffe250 german1: 0x93 Segmentation fault (core dumped) or similar. There are no obvious bugs (to me), i've attached the whole project (just started <1000 lines). Just hack make and click on "Wörter hinzufügen" and return. I'd really appreciate some help here. Problem occurs with 1.1.x and 1.2.x. Thanks. Holger | ||
mike 12:57 Sep 07, 2004 | Can you put together a smaller, single-file example that demonstrates this problem? I'm having trouble following your code... | ||
anonymous 13:21 Sep 07, 2004 | Hi, the problem doesn't occure when i leave the first message box out, so a single file example doesn't crash. I'm sorry, but i can't change much to make it easier: the problem dissappears. Holger | ||
anonymous 13:30 Sep 07, 2004 | Ok, i've but a everything in a single file. This version still crashes here. compiled with: g++ `fltk-config --cxxflags` `fltk-config --ldflags` -o problem problem.cc $ ./problem german: 0x8066ab8 this: 0xbfffe7f0 user data: 0xbfffe7f0 german1: 0x804ab58 ALIVE-- ALIVE Segmentation fault (core dumped) $ g++ --versiong++ (GCC) 3.3.4 (Debian 1:3.3.4-3) It won't compile at all with g++-2.95 (lotsa linker failures) just run programm, click button in 1 window, then click button in second Hope this helps. Holger | ||
mike 17:53 Sep 12, 2004 | Ok, the problem is in your code; in MainFialog::cb_W_i() you have a local variable of type "Add_Word", which is deleted after the win->show() (when the function returns and the local variable is deleted...) I recommend that you have an Add_Word * in your Main_Dialog class and initialize things there... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Can't build fltk 1.1.5rc2 on windows 2000 using visual studio.net |
Version: | 1.1.5rc2 |
Created By: | r2matthews.students.latrobe.edu |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
r2matthews.students.latrobe.edu 01:15 Sep 07, 2004 | Hello, I've tried to build fltk on my windows 2000 professional system. Some of the programs don't build. I build from fltk.sln. I'm using visual studio.net (microsoft .net framework 1.1.4322 SP1)) Here is the build log.. it's a bit long. ------ Build started: Project: libpng, Configuration: Release Win32 ------ Compiling... pngwutil.c pngwtran.c pngwrite.c pngwio.c pngtrans.c pngset.c pngrutil.c pngrtran.c pngrio.c pngread.c pngpread.c pngmem.c pngget.c pngerror.c png.c Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" libpng - 0 error(s), 0 warning(s) ------ Build started: Project: zlib, Configuration: Release Win32 ------ Compiling... zutil.c uncompr.c trees.c infutil.c inftrees.c inflate.c inffast.c infcodes.c infblock.c gzio.c deflate.c crc32.c compress.c adler32.c Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" zlib - 0 error(s), 0 warning(s) ------ Build started: Project: jpeg, Configuration: Release Win32 ------ Compiling... jutils.c jquant2.c jquant1.c jmemmgr.c jidctred.c jidctint.c jidctfst.c jidctflt.c jfdctint.c jfdctfst.c jfdctflt.c jerror.c jdtrans.c jdsample.c jdpostct.c jdphuff.c jdmerge.c jdmaster.c jdmarker.c jdmainct.c jdinput.c jdhuff.c jddctmgr.c jdcolor.c jdcoefct.c jdatasrc.c jdatadst.c jdapistd.c jdapimin.c jctrans.c jcsample.c jcprepct.c jcphuff.c jcparam.c jcomapi.c jcmaster.c jcmarker.c jcmainct.c jcinit.c jchuff.c jcdctmgr.c jccolor.c jccoefct.c jcapistd.c jcapimin.c Compiling... jmemnobs.c Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" jpeg - 0 error(s), 0 warning(s) ------ Build started: Project: fltkimages, Configuration: Release Win32 ------ Compiling... Fl_PNM_Image.cxx Fl_PNG_Image.cxx \fltk-1.1.5rc2\src\Fl_PNG_Image.cxx(44) : fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory Fl_JPEG_Image.cxx \fltk-1.1.5rc2\src\Fl_JPEG_Image.cxx(53) : fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory fl_images_core.cxx Fl_Help_Dialog.cxx Fl_GIF_Image.cxx Fl_File_Icon2.cxx Fl_BMP_Image.cxx Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fltkimages - 2 error(s), 0 warning(s) ------ Build started: Project: fltk, Configuration: Release Win32 ------ Compiling... vsnprintf.c scandir.c numericsort.c flstring.c Fl_XPM_Image.cxx Fl_XBM_Image.cxx Fl_x.cxx Fl_Wizard.cxx Fl_Window_iconize.cxx Fl_Window_hotspot.cxx Fl_Window_fullscreen.cxx Fl_Window.cxx Fl_Widget.cxx Fl_visual.cxx fl_vertex.cxx Fl_Value_Slider.cxx Fl_Value_Output.cxx Fl_Value_Input.cxx Fl_Valuator.cxx Fl_Tooltip.cxx Fl_Tiled_Image.cxx Fl_Tile.cxx Fl_Text_Editor.cxx Fl_Text_Display.cxx Fl_Text_Buffer.cxx Fl_Tabs.cxx fl_symbols.cxx Fl_Slider.cxx Fl_Single_Window.cxx fl_show_colormap.cxx fl_shortcut.cxx Fl_Shared_Image.cxx fl_shadow_box.cxx fl_set_fonts.cxx fl_set_font.cxx Fl_Scrollbar.cxx fl_scroll_area.cxx Fl_Scroll.cxx fl_rounded_box.cxx Fl_Round_Button.cxx fl_round_box.cxx Fl_Roller.cxx Fl_Return_Button.cxx Fl_Repeat_Button.cxx fl_rect.cxx Fl_Progress.cxx Fl_Preferences.cxx Fl_Positioner.cxx fl_plastic.cxx Fl_Pixmap.cxx Fl_Pack.cxx Fl_own_colormap.cxx Fl_Overlay_Window.cxx fl_overlay_visual.cxx fl_overlay.cxx fl_oval_box.cxx Fl_Multi_Label.cxx Fl_Menu_Window.cxx Fl_Menu_global.cxx Fl_Menu_Button.cxx Fl_Menu_Bar.cxx Fl_Menu_add.cxx Fl_Menu_.cxx Fl_Menu.cxx Fl_lock.cxx fl_line_style.cxx Fl_Light_Button.cxx fl_labeltype.cxx Fl_Input_.cxx Fl_Input.cxx Fl_Image.cxx Fl_Help_View.cxx Fl_Group.cxx Fl_grab.cxx Fl_get_system_colors.cxx Fl_get_key.cxx fl_font.cxx Fl_File_Input.cxx Fl_File_Icon.cxx fl_file_dir.cxx Fl_File_Chooser2.cxx Fl_File_Chooser.cxx Fl_File_Browser.cxx fl_engraved_label.cxx fl_draw_pixmap.cxx fl_draw_image.cxx fl_draw.cxx Fl_Double_Window.cxx fl_dnd.cxx Fl_display.cxx fl_diamond_box.cxx Fl_Dial.cxx fl_curve.cxx fl_cursor.cxx Fl_Counter.cxx Fl_compose.cxx Fl_Color_Chooser.cxx fl_color.cxx Fl_Clock.cxx Fl_Choice.cxx Fl_Check_Button.cxx Fl_Check_Browser.cxx Fl_Chart.cxx fl_call_main.c Fl_Button.cxx Fl_Browser_load.cxx Fl_Browser_.cxx Fl_Browser.cxx fl_boxtype.cxx Fl_Box.cxx Fl_Bitmap.cxx fl_ask.cxx Fl_arg.cxx fl_arci.cxx fl_arc.cxx Fl_Adjuster.cxx Fl_add_idle.cxx Fl_abort.cxx Fl.cxx filename_setext.cxx filename_match.cxx filename_list.cxx filename_isdir.cxx filename_ext.cxx filename_expand.cxx filename_absolute.cxx Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fltk - 0 error(s), 0 warning(s) ------ Build started: Project: pixmap_browser, Configuration: Release Win32 ------ Compiling... pixmap_browser.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\lib\fltkimages.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" pixmap_browser - 1 error(s), 0 warning(s) ------ Build started: Project: fltkforms, Configuration: Release Win32 ------ Compiling... forms_timer.cxx forms_pixmap.cxx forms_fselect.cxx forms_free.cxx forms_compatability.cxx forms_bitmap.cxx Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fltkforms - 0 error(s), 0 warning(s) ------ Build started: Project: fltkdll, Configuration: Release Win32 ------ Compiling... vsnprintf.c scandir.c numericsort.c glut_font.cxx glut_compatability.cxx gl_start.cxx gl_draw.cxx forms_timer.cxx forms_pixmap.cxx forms_fselect.cxx forms_free.cxx forms_compatability.cxx forms_bitmap.cxx flstring.c Fl_XPM_Image.cxx Fl_XBM_Image.cxx Fl_x.cxx Fl_Wizard.cxx Fl_Window_iconize.cxx Fl_Window_hotspot.cxx Fl_Window_fullscreen.cxx Fl_Window.cxx Fl_Widget.cxx Fl_visual.cxx fl_vertex.cxx Fl_Value_Slider.cxx Fl_Value_Output.cxx Fl_Value_Input.cxx Fl_Valuator.cxx Fl_Tooltip.cxx Fl_Tiled_Image.cxx Fl_Tile.cxx Fl_Text_Editor.cxx Fl_Text_Display.cxx Fl_Text_Buffer.cxx Fl_Tabs.cxx fl_symbols.cxx Fl_Slider.cxx Fl_Single_Window.cxx fl_show_colormap.cxx fl_shortcut.cxx Fl_Shared_Image.cxx fl_shadow_box.cxx fl_set_fonts.cxx fl_set_font.cxx Fl_Scrollbar.cxx fl_scroll_area.cxx Fl_Scroll.cxx fl_rounded_box.cxx Fl_Round_Button.cxx fl_round_box.cxx Fl_Roller.cxx Fl_Return_Button.cxx Fl_Repeat_Button.cxx fl_rect.cxx Fl_Progress.cxx Fl_Preferences.cxx Fl_Positioner.cxx Fl_PNM_Image.cxx Fl_PNG_Image.cxx \fltk-1.1.5rc2\src\Fl_PNG_Image.cxx(44) : fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory fl_plastic.cxx Fl_Pixmap.cxx Fl_Pack.cxx Fl_own_colormap.cxx Fl_Overlay_Window.cxx fl_overlay_visual.cxx fl_overlay.cxx fl_oval_box.cxx Fl_Multi_Label.cxx Fl_Menu_Window.cxx Fl_Menu_global.cxx Fl_Menu_Button.cxx Fl_Menu_Bar.cxx Fl_Menu_add.cxx Fl_Menu_.cxx Fl_Menu.cxx Fl_lock.cxx fl_line_style.cxx Fl_Light_Button.cxx fl_labeltype.cxx Fl_JPEG_Image.cxx \fltk-1.1.5rc2\src\Fl_JPEG_Image.cxx(53) : fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory Fl_Input_.cxx Fl_Input.cxx fl_images_core.cxx Fl_Image.cxx Fl_Help_View.cxx Fl_Help_Dialog.cxx Fl_Group.cxx Fl_grab.cxx Fl_Gl_Window.cxx Fl_Gl_Overlay.cxx Fl_Gl_Choice.cxx Fl_GIF_Image.cxx Fl_get_system_colors.cxx Fl_get_key.cxx fl_font.cxx Fl_File_Input.cxx Fl_File_Icon2.cxx Fl_File_Icon.cxx fl_file_dir.cxx Fl_File_Chooser2.cxx Fl_File_Chooser.cxx Fl_File_Browser.cxx fl_engraved_label.cxx fl_draw_pixmap.cxx fl_draw_image.cxx fl_draw.cxx Fl_Double_Window.cxx fl_dnd.cxx Fl_display.cxx fl_diamond_box.cxx Fl_Dial.cxx fl_curve.cxx fl_cursor.cxx Fl_Counter.cxx Fl_compose.cxx Fl_Color_Chooser.cxx fl_color.cxx Fl_Clock.cxx Fl_Choice.cxx Fl_Check_Button.cxx Fl_Check_Browser.cxx Fl_Chart.cxx Fl_Button.cxx Fl_Browser_load.cxx Fl_Browser_.cxx Fl_Browser.cxx fl_boxtype.cxx Fl_Box.cxx Fl_BMP_Image.cxx Fl_Bitmap.cxx fl_ask.cxx Fl_arg.cxx fl_arci.cxx fl_arc.cxx Fl_Adjuster.cxx Fl_add_idle.cxx Fl_abort.cxx Fl.cxx filename_setext.cxx filename_match.cxx filename_list.cxx filename_isdir.cxx filename_ext.cxx filename_expand.cxx filename_absolute.cxx Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\fltkdll\BuildLog.htm" fltkdll - 2 error(s), 0 warning(s) ------ Build started: Project: fluid, Configuration: Release Win32 ------ Compiling... widget_panel.cxx function_panel.cxx Fluid_Image.cxx fluid.cxx Fl_Window_Type.cxx Fl_Widget_Type.cxx Fl_Type.cxx Fl_Menu_Type.cxx Fl_Group_Type.cxx Fl_Function_Type.cxx file.cxx factory.cxx code.cxx alignment_panel.cxx align_widget.cxx about_panel.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\lib\fltkimages.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\fluid___\BuildLog.htm" fluid - 1 error(s), 0 warning(s) ------ Build started: Project: fltkgl, Configuration: Release Win32 ------ Compiling... glut_font.cxx glut_compatability.cxx gl_start.cxx gl_draw.cxx Fl_Gl_Window.cxx Fl_Gl_Overlay.cxx Fl_Gl_Choice.cxx Creating library... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fltkgl - 0 error(s), 0 warning(s) ------ Build started: Project: shape, Configuration: Release Win32 ------ Compiling... shape.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" shape - 0 error(s), 0 warning(s) ------ Build started: Project: scroll, Configuration: Release Win32 ------ Compiling... scroll.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" scroll - 0 error(s), 0 warning(s) ------ Build started: Project: resizebox, Configuration: Release Win32 ------ Compiling... resizebox.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" resizebox - 0 error(s), 0 warning(s) ------ Build started: Project: resize, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... resize.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\resize.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" resize - 1 error(s), 0 warning(s) ------ Build started: Project: radio, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... radio.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\radio.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" radio - 1 error(s), 0 warning(s) ------ Build started: Project: preferences, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... preferences.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\preferences.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" preferences - 1 error(s), 0 warning(s) ------ Build started: Project: subwindow, Configuration: Release Win32 ------ Compiling... subwindow.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" subwindow - 0 error(s), 0 warning(s) ------ Build started: Project: pixmap, Configuration: Release Win32 ------ Compiling... pixmap.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" pixmap - 0 error(s), 0 warning(s) ------ Build started: Project: pack, Configuration: Release Win32 ------ Compiling... pack.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" pack - 0 error(s), 0 warning(s) ------ Build started: Project: overlay, Configuration: Release Win32 ------ Compiling... overlay.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" overlay - 0 error(s), 0 warning(s) ------ Build started: Project: output, Configuration: Release Win32 ------ Compiling... output.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" output - 0 error(s), 0 warning(s) ------ Build started: Project: navigation, Configuration: Release Win32 ------ Compiling... navigation.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" navigation - 0 error(s), 0 warning(s) ------ Build started: Project: minimum, Configuration: Release Win32 ------ Compiling... minimum.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" minimum - 0 error(s), 0 warning(s) ------ Build started: Project: tiled_image, Configuration: Release Win32 ------ Compiling... tiled_image.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" tiled_image - 0 error(s), 0 warning(s) ------ Build started: Project: menubar, Configuration: Release Win32 ------ Compiling... menubar.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" menubar - 0 error(s), 0 warning(s) ------ Build started: Project: symbols, Configuration: Release Win32 ------ Compiling... symbols.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" symbols - 0 error(s), 0 warning(s) ------ Build started: Project: line_style, Configuration: Release Win32 ------ Compiling... line_style.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" line_style - 0 error(s), 0 warning(s) ------ Build started: Project: label, Configuration: Release Win32 ------ Compiling... label.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" label - 0 error(s), 0 warning(s) ------ Build started: Project: keyboard, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... keyboard_ui.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\keyboard_ui.cxx': No such file or directory keyboard.cxx \fltk-1.1.5rc2\test\keyboard.cxx(41) : fatal error C1083: Cannot open include file: 'keyboard_ui.h': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" keyboard - 2 error(s), 0 warning(s) ------ Build started: Project: input, Configuration: Release Win32 ------ Compiling... input.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" input - 0 error(s), 0 warning(s) ------ Build started: Project: inactive, Configuration: Release Win32 ------ Create .cxx and .h file with fluidd '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... inactive.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\inactive.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" inactive - 1 error(s), 0 warning(s) ------ Build started: Project: image, Configuration: Release Win32 ------ Compiling... image.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" image - 0 error(s), 0 warning(s) ------ Build started: Project: iconize, Configuration: Release Win32 ------ Compiling... iconize.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" iconize - 0 error(s), 0 warning(s) ------ Build started: Project: help, Configuration: Release Win32 ------ Compiling... help.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\lib\fltkimages.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" help - 1 error(s), 0 warning(s) ------ Build started: Project: hello, Configuration: Release Win32 ------ Compiling... hello.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" hello - 0 error(s), 0 warning(s) ------ Build started: Project: glpuzzle, Configuration: Release Win32 ------ Compiling... glpuzzle.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" glpuzzle - 0 error(s), 0 warning(s) ------ Build started: Project: gl_overlay, Configuration: Release Win32 ------ Compiling... gl_overlay.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" gl_overlay - 0 error(s), 0 warning(s) ------ Build started: Project: fullscreen, Configuration: Release Win32 ------ Compiling... fullscreen.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fullscreen - 0 error(s), 0 warning(s) ------ Build started: Project: fractals, Configuration: Release Win32 ------ Compiling... fracviewer.cxx fractals.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fractals - 0 error(s), 0 warning(s) ------ Build started: Project: forms, Configuration: Release Win32 ------ Compiling... forms.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" forms - 0 error(s), 0 warning(s) ------ Build started: Project: fonts, Configuration: Release Win32 ------ Compiling... fonts.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fonts - 0 error(s), 0 warning(s) ------ Build started: Project: tabs, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... tabs.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\tabs.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" tabs - 1 error(s), 0 warning(s) ------ Build started: Project: valuators, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... valuators.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\valuators.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" valuators - 1 error(s), 0 warning(s) ------ Build started: Project: threads, Configuration: Release Win32 ------ Compiling... threads.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" threads - 0 error(s), 0 warning(s) ------ Build started: Project: mandelbrot, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... mandelbrot_ui.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\mandelbrot_ui.cxx': No such file or directory mandelbrot.cxx \fltk-1.1.5rc2\test\mandelbrot.cxx(26) : fatal error C1083: Cannot open include file: 'mandelbrot_ui.h': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" mandelbrot - 2 error(s), 0 warning(s) ------ Build started: Project: tile, Configuration: Release Win32 ------ Compiling... tile.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" tile - 0 error(s), 0 warning(s) ------ Build started: Project: message, Configuration: Release Win32 ------ Compiling... message.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" message - 0 error(s), 0 warning(s) ------ Build started: Project: file_chooser, Configuration: Release Win32 ------ Compiling... file_chooser.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\lib\fltkimages.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" file_chooser - 1 error(s), 0 warning(s) ------ Build started: Project: fast_slow, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... fast_slow.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\fast_slow.cxx': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" fast_slow - 1 error(s), 0 warning(s) ------ Build started: Project: editor, Configuration: Release Win32 ------ Compiling... editor.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\test\fltkdll.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" editor - 1 error(s), 0 warning(s) ------ Build started: Project: doublebuffer, Configuration: Release Win32 ------ Compiling... doublebuffer.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" doublebuffer - 0 error(s), 0 warning(s) ------ Build started: Project: adjuster, Configuration: Release Win32 ------ Compiling... adjuster.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" adjuster - 0 error(s), 0 warning(s) ------ Build started: Project: curve, Configuration: Release Win32 ------ Compiling... curve.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" curve - 0 error(s), 0 warning(s) ------ Build started: Project: cursor, Configuration: Release Win32 ------ Compiling... cursor.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" cursor - 0 error(s), 0 warning(s) ------ Build started: Project: cube, Configuration: Release Win32 ------ Compiling... cube.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" cube - 0 error(s), 0 warning(s) ------ Build started: Project: color_chooser, Configuration: Release Win32 ------ Compiling... color_chooser.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" color_chooser - 0 error(s), 0 warning(s) ------ Build started: Project: colbrowser, Configuration: Release Win32 ------ Compiling... colbrowser.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" colbrowser - 0 error(s), 0 warning(s) ------ Build started: Project: clock, Configuration: Release Win32 ------ Compiling... clock.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" clock - 0 error(s), 0 warning(s) ------ Build started: Project: checkers, Configuration: Release Win32 ------ Compiling... checkers.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" checkers - 0 error(s), 0 warning(s) ------ Build started: Project: buttons, Configuration: Release Win32 ------ Compiling... buttons.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" buttons - 0 error(s), 0 warning(s) ------ Build started: Project: button, Configuration: Release Win32 ------ Compiling... button.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" button - 0 error(s), 0 warning(s) ------ Build started: Project: browser, Configuration: Release Win32 ------ Compiling... browser.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" browser - 0 error(s), 0 warning(s) ------ Build started: Project: boxtype, Configuration: Release Win32 ------ Compiling... boxtype.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" boxtype - 0 error(s), 0 warning(s) ------ Build started: Project: bitmap, Configuration: Release Win32 ------ Compiling... bitmap.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" bitmap - 0 error(s), 0 warning(s) ------ Build started: Project: ask, Configuration: Release Win32 ------ Compiling... ask.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" ask - 0 error(s), 0 warning(s) ------ Build started: Project: arc, Configuration: Release Win32 ------ Compiling... arc.cxx Linking... Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" arc - 0 error(s), 0 warning(s) ------ Build started: Project: CubeView, Configuration: Release Win32 ------ Create .cxx and .h file with fluid '..\fluid\fluid' is not recognized as an internal or external command, operable program or batch file. Compiling... CubeViewUI.cxx c1xx : fatal error C1083: Cannot open source file: '\fltk-1.1.5rc2\test\CubeViewUI.cxx': No such file or directory CubeView.cxx CubeMain.cxx \fltk-1.1.5rc2\test\CubeMain.cxx(28) : fatal error C1083: Cannot open include file: 'CubeViewUI.h': No such file or directory Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\CubeView___\BuildLog.htm" CubeView - 2 error(s), 0 warning(s) ------ Build started: Project: demo, Configuration: Release Win32 ------ Compiling... demo.cxx Linking... LINK : fatal error LNK1181: cannot open input file '\fltk-1.1.5rc2\test\fltkdll.lib' Build log was saved at "file://c:\fltk-1.1.5rc2\vcnet\Release\BuildLog.htm" demo - 1 error(s), 0 warning(s) ---------------------- Done ---------------------- Build: 50 succeeded, 18 failed, 0 skipped | ||
mike 13:37 Sep 07, 2004 | Dupe of #476. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | OpenGL drawing in a widget is broken under OSX |
Version: | 1.1.5rc2 |
Created By: | derek.signalmarketing |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
derek.signalmarketing 15:00 Sep 07, 2004 | glwid.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
derek.signalmarketing 15:00 Sep 07, 2004 | using gl_start() and gl_finish() in an Fl_Widget()'s draw routine completely corrupts the contents of the window. Calling fl_push_clip before gl_start() does not limit the corruption, nor does it clip subsequent gl drawing calls. | ||
mike 16:09 Sep 09, 2004 | gl_start/end are deprecated, and don't work well on OSX. Instead, subclass Fl_Gl_Window to use OpenGL in your widget. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Re: Str#517 isprint in .net causes assert with debug fltkd.lib |
Version: | 1.1.5rc2 |
Created By: | eamon.sonyoxford.co |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
eamon.sonyoxford.co 02:38 Sep 08, 2004 | Hi Michael. > Actually, the correct fix is to add &255 to all of the isfoo() calls. I thought of that, however, it is a significant performance hit. It adds a costly "and ecx,0x000000FF" (6 bytes) to each isfoo() call. "Actually" the "correct" name is "Fast" light toolkit. Eamon. (Grin) | ||
mike 11:15 Sep 11, 2004 | First, we're not talking about something that is executed that often, only when updating the preview image (which can take a lot longer if you are previewing a large image, for example) Second, using unsigned char does not guarantee the 0-255 range - the result is compiler-dependent. We need to mask the bits to be portable. Finally, smart compilers can reduce "& 255" to a byte load or cast if the underlying processor supports 8-bit ops. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #485 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Some variables referenced after their malloc block is freed |
Version: | 1.1-current |
Created By: | larry.finger.lwfinger |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
larry.finger.lwfinger 20:56 Sep 08, 2004 | In using valgrind to check my program for memory leaks, the following errors were reported. Each of them is due to reading a variable in a widget after the associated memory block has been freed. These errors occurred with FLTK-1.1.x CVS, downloaded at 17:00 CDT, Sep. 8. Error #1: Invalid write of size 4 at 0x807AC51: Fl_Widget::clear_changed() (Fl_Widget.H:180) by 0x8099F42: Fl_Button::handle(int) (Fl_Button.cxx:101) by 0x8094305: send(int, Fl_Widget*, Fl_Window*) (Fl.cxx:564) by 0x80945F0: Fl::handle(int, Fl_Window*) (Fl.cxx:644) Address 0x41DAD03C is 44 bytes inside a block of size 76 free'd at 0x4002AFF3: __builtin_delete (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x4002B011: operator delete(void*) (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x809A4BB: Fl_Button::~Fl_Button() (Fl_Button.cxx:188) by 0x809B7C2: Fl_Group::clear() (Fl_Group.cxx:337) Error #2: Invalid read of size 4 at 0x807AC76: Fl_Widget::do_callback() (Fl_Widget.H:188) by 0x8099F42: Fl_Button::handle(int) (Fl_Button.cxx:101) by 0x8094305: send(int, Fl_Widget*, Fl_Window*) (Fl.cxx:564) by 0x80945F0: Fl::handle(int, Fl_Window*) (Fl.cxx:644) Address 0x41DAD018 is 8 bytes inside a block of size 76 free'd at 0x4002AFF3: __builtin_delete (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x4002B011: operator delete(void*) (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x809A4BB: Fl_Button::~Fl_Button() (Fl_Button.cxx:188) by 0x809B7C2: Fl_Group::clear() (Fl_Group.cxx:337) Error #3: Invalid read of size 4 at 0x8095404: Fl_Widget::parent() const (Fl_Widget.H:109) by 0x809B7F7: Fl_Group::~Fl_Group() (Fl_Group.cxx:343) by 0x8094AFF: Fl_Window::~Fl_Window() (Fl.cxx:832) by 0x809A985: Fl_Double_Window::~Fl_Double_Window() (Fl_Double_Window.cxx:351) Address 0x41D97C24 is 116 bytes inside a block of size 280 free'd at 0x4002AFF3: __builtin_delete (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x4002B011: operator delete(void*) (in /usr/lib/valgrind/vgskin_memcheck.so) by 0x807C04A: Flu_Spinner::~Flu_Spinner() (Flu_Spinner.cxx:130) by 0x809B7C2: Fl_Group::clear() (Fl_Group.cxx:337) | ||
mike 15:39 Sep 09, 2004 | The problem is caused by deleting a widget that has focus. Use the take_focus() method on a widget that you are not deleting before you delete... | ||
larry.finger.lwfinger 21:07 Sep 10, 2004 | I added statements that called take_focus() for an input widget on the main page before each of the child windows destructor is called, but the error still occurs. | ||
mike 18:50 Sep 12, 2004 | This is a duplicate of #485. Currently there is no way to fix this, as it is a side effect of supporting the changed() flag for callbacks that was added in 1.1.5. Since having a working changed() implementation is more important than supporting widget deletion from the widget's callback function, we won't be fixing this. The workaround is to hide() the widget in the callback and delete it outside of the callback. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | WIN32 |
Summary: | Build errors/warnings is MSVS 7 .net |
Version: | 1.1.5rc2 |
Created By: | adskewes.uh |
Assigned To: | mike |
Fix Version: | 1.1.5rc3 |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
adskewes.uh 07:46 Sep 11, 2004 | Receive the following build errors for CubeView in C++ .net 7.0. Any Ideas? c:\Program Files\Insight\fltk-1.1.5rc2\test\CubeMain.cxx(28): fatal error C1083: Cannot open include file: 'CubeViewUI.h': No such file or directory C:\Program Files\Insight\fltk-1.1.5rc2\src\Fl_JPEG_Image.cxx(53): fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory c:\Program Files\Insight\fltk-1.1.5rc2\src\Fl_JPEG_Image.cxx(53): fatal error C1083: Cannot open include file: 'jpeglib.h': No such file or directory c:\Program Files\Insight\fltk-1.1.5rc2\test\keyboard.cxx(41): fatal error C1083: Cannot open include file: 'keyboard_ui.h': No such file or directory c:\Program Files\Insight\fltk-1.1.5rc2\test\mandelbrot.cxx(26): fatal error C1083: Cannot open include file: 'mandelbrot_ui.h': No such file or directory C:\Program Files\Insight\fltk-1.1.5rc2\src\Fl_PNG_Image.cxx(44): fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory c:\Program Files\Insight\fltk-1.1.5rc2\src\Fl_PNG_Image.cxx(44): fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory CubeView fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\CubeViewUI.cxx': No such file or directory fast_slow fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\fast_slow.cxx': No such file or directory inactive fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\inactive.cxx': No such file or directory keyboard fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\keyboard_ui.cxx': No such file or directory mandelbrot fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\mandelbrot_ui.cxx': No such file or directory preferences fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\preferences.cxx': No such file or directory radio fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\radio.cxx': No such file or directory resize fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\resize.cxx': No such file or directory tabs fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\tabs.cxx': No such file or directory valuators fatal error C1083: Cannot open source file: '\Program Files\Insight\fltk-1.1.5rc2\test\valuators.cxx': No such file or directory editor fatal error LNK1181: cannot open input file '\Program Files\Insight\fltk-1.1.5rc2\test\fltkdlld.lib' demo fatal error LNK1181: cannot open input file '\Program Files\Insight\fltk-1.1.5rc2\test\fltkdlld.lib' pixmap_browser fatal error LNK1181: cannot open input file 'png.lib' fluid fatal error LNK1181: cannot open input file 'png.lib' help fatal error LNK1181: cannot open input file 'png.lib' file_chooser fatal error LNK1181: cannot open input file 'png.lib' shape warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification scroll warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification resizebox warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification subwindow warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification pixmap warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification pack warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification overlay warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification output warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification navigation warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification minimum warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification tiled_image warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification menubar warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification symbols warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification line_style warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification label warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification input warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification image warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification iconize warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification hello warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification glpuzzle warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification gl_overlay warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification fullscreen warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification fractals warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification forms warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification fonts warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification threads warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification tile warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification message warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification doublebuffer warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification adjuster warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification curve warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification cursor warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification cube warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification color_chooser warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification colbrowser warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification clock warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification checkers warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification buttons warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification button warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification browser warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification boxtype warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification bitmap warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification ask warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification arc warning LNK4075: ignoring '/EDITANDCONTINUE' due to '/INCREMENTAL:NO' specification | ||
adskewes.uh 07:49 Sep 11, 2004 | Tried to redo entire installation, but still same errors. | ||
mike 11:07 Sep 11, 2004 | Dupe of #476. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #476 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | cannot open file "jpeg.lib" during linking |
Version: | 1.1.5rc2 |
Created By: | tanche21.comcast |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
tanche21.comcast 16:54 Sep 11, 2004 | I am new to this FLTK. I just download the source code and try to complile it. I passes all the compliling barriers, now it can not link due to missing out the "jpeg.lib". I have spent some effort to look for this library on the internet, but got no luck. Could someone please point me to the right direction. Thank you in advanced for your help. Tan | ||
mike 20:18 Sep 11, 2004 | Dupe of #476; this is a known problem and will be fixed in rc3. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Re: Str#523 and Str#517 isprint in .net causes assert with debug fltkd.lib |
Version: | 1.1.5rc2 |
Created By: | eamon.sonyoxford.co |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
eamon.sonyoxford.co 09:57 Sep 13, 2004 | First if something is inefficient then it's inefficient, doing it in a loop makes it worse. Second, using unsigned char will guarantee the 0-255 range; it is called an "integral promotion". However if you KNOW that there are compilers broken in this way out there, then I bow to your superior knowledge. I’ve checked a few, GNU 3.X, Visual Studio 6 and .net, DEC OSF cxx, on both Alpha and I386 processors and they all behaved as expected. Before you say there MIGHT be one, there MIGHT be a compiler that assigns 2 to i in the expression i=1; > Finally, smart compilers can reduce "& 255" to a byte load or cast if the underlying processor supports 8-bit ops. Third, there is no 32 processor that I know off, and I have programmed most of them at one time or another, that will get "and x, 0xFF" down to less than 4 bytes (680X0), An i386 will always take 6 bytes and the only way for the compiler to get it less than that involves setting up registers which take longer and more bytes than the and instruction itself. I don’t understand the bit about "cast if the underlying processor supports 8-bit ops". Finally I’ve had enough of this, in future if I find a bug I will just report it and let you find the solution, it’s bound to be better than anything I can come up with. Have fun. Eamon. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | osx fltk - wheres fluid? |
Version: | 1.1.2 |
Created By: | bennyp.rogers |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
bennyp.rogers 16:41 Sep 13, 2004 | I compiled fltk on my iBook using gentoo macos now i'm trying to compile zynaddsubfx but it doesn't work because fluid is not in /usr/local/bin and locate fluid yields no good results see http://www.kvr-vst.com/forum/viewtopic.php?t=55892 for more info | ||
mike 08:48 Sep 24, 2004 | Please ask this question on fltk.general or ask the application developer... Also, the current stable release is 1.1.4... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fl_draw_image_mono() does not draw image under windows |
Version: | 1.1.5rc2 |
Created By: | innersmile.gmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
innersmile.gmail 18:30 Sep 16, 2004 | drawTest.cpp 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
innersmile.gmail 18:30 Sep 16, 2004 | Hi folks, Can anybody show me an example for drawing images with full source code? I am totally confused by the FLTK. I made a simple program using Visual C++ 6.0 last year. It is supposed to open a file_chooser, and pick an image file (bmp or tif) and then display the image. At that time, it can display the image only at debug mode, and just gives me a blank window at release mode. But now when I return to that program, it can not display image in either debug mode or release mode!! I wrote a simple file to test it and got the same behavior. Please help point out the problems, the code is given below. Many thanks in advance! // ======== beginning ================== #include <Fl/fl_file_chooser.h> #include <Fl/Fl.h> #include <Fl/fl_message.h> #include <Fl/Fl_Double_Window.h> #include <Fl/fl_draw.h> class TiffWidget : public Fl_Widget { int w,h; unsigned char* image; void draw() { fl_draw_image_mono(image,5,5,w,h); }; public: TiffWidget(int x, int y, int imageW, int imageH, unsigned char* imageD) : Fl_Widget(x,y,imageW,imageH) { w = imageW; h = imageH; image = imageD; }; }; // simple main to try out these two pieces int main(int argc, char** argv) { unsigned char image[300*300]; int i, j, w = 300, h = 300; for (i = 0; i < h; i++) for (j = 0; j < w; j++){ if (i+j>300) image[i*w + j] = 5 ; else image[i*w + j] = 250 ; } Fl::visual(FL_RGB); // make a window Fl_Double_Window window(w+10,h+10,"hello"); TiffWidget tw(5,5,w,h,image); window.end(); window.show(); Fl::run(); return 1; } BTW, the code above was modified from an example available at http://www.cs.wisc.edu/graphics/Courses/cs-638-1999/Examples/TIFF/TiffTest/ Even this code can not display the image file at all. Anyway, the same code does work under linux. I don't know why. I use cmake to generate the visual studio project file. If you need, I can provide the CMakeLists.txt file. Thanks! | ||
mike 07:55 Sep 24, 2004 | Your test program runs just fine on Linux and OSX. If memory serves, you were having problems on Windows, right? Did you try the program on another system already? | ||
innersmile.gmail 07:10 Oct 05, 2004 | Yes, I also tried on linux. It's OK on linux. | ||
mike 13:07 Oct 18, 2004 | Not sure what is going wrong, as it works OK on my WinXP system. Can you try with another Windows PC (with a different vendor's graphics card) to see if this is a general problem? | ||
mike 06:35 Nov 21, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | make install doesn't create link between FL/gl.H and FL/gl.h |
Version: | 1.1.5rc1 |
Created By: | duncan.thermal.esa |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
duncan.thermal.esa 01:32 Sep 22, 2004 | make install doesn't create link between FL/gl.H and FL/gl.h on systems where file names are case sensitive, compilations requiring fltk/OpenGL fail to find the include file | ||
mike 08:49 Sep 24, 2004 | The FL/gl.h header is only lowercase, so we won't make this change. (just like the GL/gl.h header...) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Scroll incorrectly paints the background |
Version: | 1.1.4 |
Created By: | alexey.parshin |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
alexey.parshin 12:35 Sep 23, 2004 | Fl_Scroll uses draw_clip() to paint its background. Unfortuantely, the background is paint using fl_rectf() and not draw_box(). That allows it to ignore the redefinition of box type. void Fl_Scroll::draw_clip(void* v,int X, int Y, int W, int H) { fl_clip(X,Y,W,H); Fl_Scroll* s = (Fl_Scroll*)v; // erase background as needed... switch (s->box()) { case FL_NO_BOX : case FL_UP_FRAME : case FL_DOWN_FRAME : case FL_THIN_UP_FRAME : case FL_THIN_DOWN_FRAME : case FL_ENGRAVED_FRAME : case FL_EMBOSSED_FRAME : case FL_BORDER_FRAME : case _FL_SHADOW_FRAME : case _FL_ROUNDED_FRAME : case _FL_OVAL_FRAME : case _FL_PLASTIC_UP_FRAME : case _FL_PLASTIC_DOWN_FRAME : if (s->parent() == (Fl_Group *)s->window() && Fl::scheme_bg_) { Fl::scheme_bg_->draw(X-(X%((Fl_Tiled_Image *)Fl::scheme_bg_)->image()->w()), Y-(Y%((Fl_Tiled_Image *)Fl::scheme_bg_)->image()->h()), W+((Fl_Tiled_Image *)Fl::scheme_bg_)->image()->w(), H+((Fl_Tiled_Image *)Fl::scheme_bg_)->image()->h()); break; } default : // Original code: // fl_color(s->color()); // fl_rectf(X,Y,W,H); // Replacement (simplified): s->draw_box(FL_FLAT_BOX, X, Y, W, H, s->color()); break; } .... } | ||
mike 12:48 Sep 23, 2004 | Please test against 1.1.5rc2 and let us know if you still experience difficulties. | ||
alexey.parshin 13:25 Sep 23, 2004 | 1.5rc2 uses exactly the same code as I described for 1.4. | ||
mike 10:27 Sep 24, 2004 | Since we rely on a solid background for scrolling, etc., we won't change the current behavior in 1.1.x. Resubmit for 2.0 if you want different behavior in a future release, and please include a patch that efficiently handles scrolling of non-solid backgrounds... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Fl_Image data pointer |
Version: | 1.1.2 |
Created By: | mark.t |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mark.t 12:58 Sep 30, 2004 | The Fl_Image data() function returns a pointer based upon the 'char' datatype, but the image data is intrinsically 'unsigned char'. | ||
mike 13:58 Sep 30, 2004 | Actually, that's not the case for at least the Fl_Pixmap image type, and the data() method returns an array of pointers... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | Multithreading issue with FLTK |
Version: | 1.1.5rc3 |
Created By: | rajnish.ncbs.res |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
rajnish.ncbs.res 04:07 Oct 02, 2004 | Multithreading.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rajnish.ncbs.res 04:07 Oct 02, 2004 | Operating system : Windows XP Compiler : Vc++ 7 (.Net) I observed that creation of FLTK objects and Fl::Run() has to be on same thread. If you create FLTK object(Window) on one thread and call Fl::run or while{ Fl::wait(0.2)} on diferent thread then window gets hanged. Even if you declare global FLTK object so that it will be accessible to all threads and then call Fl::run() on different thread then also it does not work. | ||
mike 05:48 Oct 02, 2004 | This is a limitation of windows GDI - only the main thread can create windows and do graphics operations. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | openGL on Dev-C++ |
Version: | 1.1.4 |
Created By: | zeji.king.gmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zeji.king.gmail 07:26 Oct 04, 2004 | Hi, just a silly question. I have downloaded the FLTK2.DevPack from eFLTK web. I tried to use Dev-C++ to create an FLTK_OPENGL program, but I got many errors of compilation. Then I changed to FLTK1.1, the fltk1.1.2.DevPak downloaded from Dev-Cpp website. Now I can create and compile FLTK1.1 program successfully. However, once I added the openGL code, there are many linker errors. I am wondering how the parameters should be set in the project compiler setting? I am new of using GCC and dev-c++. Thanks | ||
mike 12:05 Oct 18, 2004 | Please post this question to the fltk.general newsgroup. We generally have no experience with DevCPP, so there isn't much help we can provide you with here in STR-land... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | installation problem - X11 libraries not found |
Version: | 1.1.5rc3 |
Created By: | cdwang.ozemail.com |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cdwang.ozemail.com 18:45 Oct 05, 2004 | I'm trying to install FLTK-1.1.5rc3 and in the middle of configure i got this error message. checking for X... no configure: error: Configure could not find required X11 libraries ./configure: line 8239: exit: aborting.: numeric argument required ./configure: line 8239: exit: aborting.: numeric argument required make: *** [makeinclude] Error 255 I am running a Fedora Core 2 with kernel 2.6.8-1.521. I installed FLTK twice in my other computers who run FC2 as well and I have not had this problem. So why do I get this one now? | ||
mike 06:55 Oct 06, 2004 | Do you have the xorg-x11-devel package installed? If not, you'll need that to compile FLTK... | ||
mike 13:30 Oct 18, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Example Programs |
Summary: | Missing JPEG/PNG workspace in the FLTK.dsw workspace |
Version: | 1.2-current |
Created By: | rtrocca.libero |
Assigned To: | rokan |
Fix Version: | 2.0-current |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rtrocca.libero 06:41 Oct 09, 2004 | When opening the FLTK.dsw project file in VC6 the jpeg and png projects are missing fromthe "global" project. Therefore, when compiling, everything goes ok, but the pixmap_borwser sample is unable to open/preview jpg images anf it somebody attempts to open one, it crashes. This applies with FLTK 1.2 CVS but NOT with FLTK 1.1.5rc3. Riccardo | ||
fabien 01:31 Mar 23, 2006 | please check on fltk 2.0 and reopen a STR if necessary. AFAIK, There will be no 1.2 but 2.0 | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #560 |
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | menu over tooltip crashes on win32 (2K) |
Version: | 1.1.5rc3 |
Created By: | jason.bryan |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
jason.bryan 11:07 Oct 11, 2004 | fltk_tooltip_menu_bug.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jason.bryan 11:07 Oct 11, 2004 | my app started crashing sometimes only on windows (2K) in the following situation: i have a menubar above a few buttons, and the buttons have tooltips. when i click on the menu and start navigating to choose an entry, *sometimes* the app will crash when the mouse, while still technically in the menu, is still "over" a button. i managed to create an example program which shows this behavior, the source of which i have attached. there is also a precompiled version here: http://www.osc.edu/~jbryan/fltk_tooltip_menu_bug.exe for me, it crashes about 80% of the time if i start it up, click on "File", and then move the mouse down through the menu options. specifically, it DOES NOT crash if you let a tooltip popup first, and then navigate the menu. too bad i didn't find it before 10/8, but hopefully you can still find a patch... ;-) | ||
mike 11:18 Oct 11, 2004 | Dupe of #560. Thanks for the example, will attempt to duplicate... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Compilation fails to find jpeglib.h on WinXP/Cygwin |
Version: | 1.1.5rc3 |
Created By: | duncan.thermal.esa |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
duncan.thermal.esa 01:07 Oct 12, 2004 | Probably related to STR #550 I downloaded fltk-1.1.5rc3 to my WinXP / Cygwin configuration. I also have mingw/msys installed, but I'm not actively using it. After running configure with default values, make failed when it tried to build Fl_JPEG_Image.cxx, saying that it couldn't find jpeglib.h I was surprised, because I remembered an announcement about jpeg, png and zlib support now being bundled with fltk, but supposed that it must be for fltk 2.0. Therefore I decided to update the Cygwin installation, and explicitly selected the 3 library packages. After restoring fltk-1.1.5rc3 to its virgin state from the tar file, I ran configure, but make still died in the same place. I went through the configure options more carefully and reran with configure --enable-localjpeg --enable-localpng --enable-localzlib and everything build and installed. | ||
duncan.thermal.esa 01:09 Oct 12, 2004 | I failed to mention that after the Cygwin update, jpeglib.h existed, and was readable, in several places, including /usr/include/jpeglib.h | ||
mike 07:00 Oct 12, 2004 | Did you use the configure script or ??? | ||
duncan.thermal.esa 07:34 Oct 12, 2004 | Yes, I used the configure script 3 times, but the first 2 without specifying any options. Only the 3rd time with options succeeded. After I'd updated the Cygwin environment, I removed the original fltk-1.1.5rc3 directory and untar'd it again, just to make sure that I rebuilt everything. Again a plain 'configure' and again a failure to find jpeglib.h. After reading the options I re-ran configure with the --enable-local{jpeg,phg,zlib} options and it built correctly. | ||
mike 13:16 Oct 18, 2004 | Well, that sounds like a problem with your Cygwin install; there isn't anything we can do to help you since everything just works here... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | Documentation examples errors |
Version: | 1.1.2 |
Created By: | zdenek_novotny.tpca-cz |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
zdenek_novotny.tpca-cz 00:13 Oct 14, 2004 | Hello. I tryed to learn FLTK from PDF Documentation 1.2 downloaded here. I tryed to compile basic editor example in chapter 4. I am using DevC++ version 4.9.9.0. I cannot compile the example. Compiller gives me list of 20 and more errors. There is a lot of undefined objects cause the example does not keep Case sensitive rules. There is also problem with declaration of objects cause they are not declared before use. Is there somewhere archive of proven examples and maybe some better documentation? Regards Zdenek Novotny | ||
mike 08:41 Oct 14, 2004 | We don't provide support for DevCPP-specific issues, but perhaps you could post to the fltk.general newsgroup to get an answer? | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | Gl_Window does not work with Mesa-6.2 libraries on Windows XP |
Version: | 1.1.5rc3 |
Created By: | larry.finger.lwfinger |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
larry.finger.lwfinger 11:05 Oct 17, 2004 | Under Windows XP and Visual Studio 6.0, the cube example will work OK if linked with the standard lib and DLL files for glu32, glut32 and opengl32. If Mesa-6.2 libraries and DLLs are used, the graphics window is never displayed. Whatever is in the background remains in the window. This failure occurs in all Gl_Window programs including cube and CubeView from the test suite, and my program as well. | ||
mike 07:38 Oct 18, 2004 | I've never used Mesa on Windows, and I don't believe that any of the developers has tested Mesa on Windows with FLTK, so I'm guessing you are running into a bug with Mesa on Windows, or there is some assumption about Mesa only being used with X11 in the FLTK code. I'll leave this STR open for now, however I don't think there is much of a chance that we will be able to fix this... | ||
mike 06:35 Nov 21, 2004 | Unable to test or fix. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | bad view when minimal widget size 10 vs 14(as default) |
Version: | 1.1.5rc3 |
Created By: | podarok_mail.rambler |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
podarok_mail.rambler 14:19 Oct 18, 2004 | fltk.JPG 108k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
podarok_mail.rambler 14:19 Oct 18, 2004 | when I change default Widget size in sources int FL_NORMAL_SIZE = 10; (defaukt is 14) in fluid very bed look in tree view. try it! why? - fltk looks better vit a small font for my work | ||
mike 16:36 Oct 18, 2004 | FLUID was not designed to use small fonts like that. We have no plans of supporting non-standard font sizes with FLUID. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Example Programs |
Summary: | test\keyboard.cxx doesn't compile with Watcom - prototype needed |
Version: | 1.1.5rc2 |
Created By: | mnieuw.zap.a2000 |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mnieuw.zap.a2000 11:55 Oct 21, 2004 | For OpenWatcom 1.3, a prototype is needed in test\keyboard.cxx . Below is the diff from 1.1.5. Mat Nieuwenhoven @@ -45,6 +45,10 @@ void shift_cb(Fl_Button*, void*) {} void wheel_cb(Fl_Dial*, void*) {} +#if defined (__WATCOMC__) +int handle(int); +#endif + // this is used to stop Esc from exiting the program: int handle(int e) { return (e == FL_SHORTCUT); // eat all keystrokes | ||
mike 07:04 Oct 22, 2004 | Um, you need to have a prototype for the handle() function right above the function? That sounds like a compiler bug to me! | ||
mnieuw.zap.a2000 12:38 Oct 22, 2004 | It wasn't complaining about the function itself, but about the use of it in the first line of main: Fl::add_handler(handle); However, that was with 1.1.5rc2 . I have just retested with a clean 1.1.5, and the error has gone. So this fix isn't needed and this can be closed. Thanks, Mat | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #560 |
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | First tooltip + menu item = crash |
Version: | 1.1.5 |
Created By: | tgeijten.hotmail |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
tgeijten.hotmail 07:13 Oct 26, 2004 | If a menu window is activated while tooltip window is about to display for the first time, FLTK crashes. It can be reproduced easily as follows: 1. Run FLUID. 2. Shortly hover over the main area. 3. Select a menu item BEFORE the tooltip appears. 4. Application crashes on tooltip timeout. I have reproduced this only on WIN32 so far, but it doesn't look like an OS-specific bug to me. | ||
mike 13:33 Oct 27, 2004 | It *is* OS-specific, dupe of STR #560, so far I've been unable to reproduce the problem. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | i can't compile on SPARC Solaris 8 |
Version: | 1.1.5 |
Created By: | aka_.bk |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aka_.bk 05:27 Oct 27, 2004 | i tried to compile on SPARC Solaris 8 and after: >./configure CC=c++ >make i got following message: Linking fluid... Undefined first referenced symbol in file fl_scandir ../lib/libfltk.a(filename_list.o) ld: fatal: Symbol referencing errors. No output written to fluid collect2: ld returned 1 exit status make[1]: *** [fluid] Error 1 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | i can't compile on SPARC Solaris 8 |
Version: | 1.1.5 |
Created By: | aka_.bk |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aka_.bk 05:43 Oct 27, 2004 | i tried to compile on SPARC Solaris 8 and after: >./configure CC=c++ >make i got following message: Linking fluid... Undefined first referenced symbol in file fl_scandir ../lib/libfltk.a(filename_list.o) ld: fatal: Symbol referencing errors. No output written to fluid collect2: ld returned 1 exit status make[1]: *** [fluid] Error 1 | ||
mike 13:31 Oct 27, 2004 | Please post this question on the fltk.general newsgroup... | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | during FL_KEYDOWN, Fl::event_key(Fl::event_key()) returns zero |
Version: | 1.1.5 |
Created By: | wilson.afn |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
wilson.afn 09:28 Oct 28, 2004 | baz.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wilson.afn 09:28 Oct 28, 2004 | It seems in 1.1.4, after pressing, e.g., "PgUp" on the numeric keypad, Fl::event_key() returned FL_KP+'9'. In 1.1.5 it (incorrectly, IMO) returns FL_PageUp. I assert that event_key() should tell me what key was pressed, and not try to interpret the event for me. Furthermore, doesn't this break binary compatibility with earlier 1.1.x versions? The example illustrates the problem. Its behaviour is different under the two versions. (Actually, I see now that test/keyboard also shows this problem; I didn't have to go to all this effort. ;) | ||
wilson.afn 09:38 Oct 28, 2004 | Uh. I forgot... FC2, GCC 3.3.3. | ||
mike 11:01 Oct 28, 2004 | In response to the poll we did, we changed how the keypad works in 1.1.5. Basically, before 1.1.5 we ignored the numlock state and always treated the keypad keys as if numlock was enabled. In 1.1.5 we look at the numlock state and return either the special key or FL_KP + 'x'. This *is* different, however it is not a binary compatibility issue since we provide no guarantees WRT the numeric keypad since not all keyboards have one. The new behavior ensures that applications that are written to look for FL_PageUp, etc. will get the correct key code whether the user pressed Page Up, "9" on the keypad, or any other key that is mapped to that function. If you are developing an X11-only application, you can look at the fl_xevent variable to get the raw keycodes for any 1.1.x release. | ||
wilson.afn 11:50 Oct 28, 2004 | So that's what that "honoring numlock" poll was about! I wish I had understood. I use the 1,3,7, and 9 keys as "diagonal cursor keys". When I first encountered the new behaviour, I wasn't even sure FLTK was "broken". I just didn't remember having to engage NumLock to get these keys to work, and I had to redownload/build 1.1.4 to confirm that I wasn't going nuts. Fl::event_key(int)'s now-inconsistent behavior provides me a workaround. (A similar workaround would served the "pro-honor-NumLock" crowd.) | ||
spitzak 17:11 Oct 30, 2004 | I have to say I preferred the older "always numeric" behavior. Just like the original poster, this broke my use of the keypad as a seperate set of arrows, the user now has to make sure numlock is turned on for our application to work. However the poll went pretty strongly the other way... | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | WIN32 |
Summary: | Fl::set_font crashes when adding 129th font |
Version: | 1.1.5 |
Created By: | mnieuw.zap.a2000 |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mnieuw.zap.a2000 02:46 Nov 02, 2004 | Build Fltk-1.1.5 with OpenWatcom 1.3. Program test/fonts crashed. Debugging revealed it crashed when adding the 129th font (the system has a lot of fonts) in fl_set_font.cxx, method Fl:set_font. I think this is independent of OS or CPU. Mat Nieuwenhoven | ||
mike 17:24 Nov 19, 2004 | Well, as we don't have Watcom on any system, there is no way to test where it is crashing on our end. Could you possibly run a debugger and show us where the crash is occurring on your system? | ||
mike 19:29 Dec 05, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Can not Compile Program |
Version: | 1.1.5 |
Created By: | maka.powerdot |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
maka.powerdot 01:48 Nov 05, 2004 | main.cpp 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
maka.powerdot 01:48 Nov 05, 2004 | I compiled FLTK with Visual C++ 6.0, all the test demo files are working; but I can not Compile my own Hello World Application. I linked fltk.lib and wsock32.lib as it is requited, but I receive following linker error Compiling... main.cpp Linking... MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _tolower already defined in LIBCD.lib(tolower.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _free already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _realloc already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _malloc already defined in LIBCD.lib(dbgheap.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _sprintf already defined in LIBCD.lib(sprintf.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _exit already defined in LIBCD.lib(crt0dat.obj) MSVCRTD.lib(MSVCRTD.dll) : error LNK2005: _memmove already defined in LIBCD.lib(memmove.obj) LINK : warning LNK4098: defaultlib "MSVCRTD" conflicts with use of other libs; use /NODEFAULTLIB:library fltkd.lib(Fl_x.obj) : error LNK2001: unresolved external symbol __imp___TrackMouseEvent@4 Debug/TEST.exe : fatal error LNK1120: 1 unresolved externals Error executing link.exe. TEST.exe - 9 error(s), 1 warning(s) Could you please help me with it ? cheers, | ||
mike 14:44 Nov 09, 2004 | Please ask this question on the fltk.general newsgroup. We can't help you here... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | ZynAddSubFX hangs on MacOSX |
Version: | 1.1.5 |
Created By: | bennyp.rogers |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
bennyp.rogers 12:42 Nov 05, 2004 | zyntestfriday 6k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
bennyp.rogers 12:42 Nov 05, 2004 | The Software Synthesiszer ZynAddSubFX compiles on MacOSX, but the window just hangs attatched is the shell output from a VNC GDB session with the developer of ZASFX, Paul O Nasca also available at http://bennyp.thefreebizhost.com/zyntestfriday | ||
bennyp.rogers 12:42 Nov 05, 2004 | oh i forgot, Check out Thread 4 | ||
mike 18:45 Nov 05, 2004 | Please contact the author of ZynAddSubFX... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | an ambiguous call of sqrt in fltk-1.1.4/test/Threads.cxx to VS03_include/Math.h |
Version: | 1.1.4 |
Created By: | quang_buiduy.yahoo |
Assigned To: | mike |
Fix Version: | 1.1.5 |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
quang_buiduy.yahoo 15:26 Nov 05, 2004 | errorForBinBuild.txt 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
quang_buiduy.yahoo 15:26 Nov 05, 2004 | Compiling... cl : Command line warning D4029 : optimization is not available in the standard edition compiler threads.cxx \Graphics\fltk-1.1.4\test\threads.cxx(64) : error C2668: 'sqrt' : ambiguous call to overloaded function C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(626): could be 'long double sqrt(long double)' C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(578): or 'float sqrt(float)' C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\math.h(200): or 'double sqrt(double)' while trying to match the argument list '(int)' Build log was saved at "file://c:\Graphics\fltk-1.1.4\visualc\Release\BuildLog.htm" threads - 1 error(s), 1 warning(s) | ||
mike 18:46 Nov 05, 2004 | Please test against the current release; I believe this was already fixed, if not submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 5 - Critical, e.g. nothing working at all |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | segmentation fault when reading value of inputbox |
Version: | 1.1.5 |
Created By: | m_dheeraj.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
m_dheeraj.hotmail 12:49 Nov 06, 2004 | Hi, I am using fltk 1.1.5 on cygwin. When I try to read an input value through button click the application hangs for a while and then crashes with segmentation fault. const char* meterID = inputID->value(); What could be the reason for the same? thank you, dheeraj | ||
mike 19:37 Nov 06, 2004 | Please post a short example program that demonstrates the problem. | ||
m_dheeraj.hotmail 19:44 Nov 06, 2004 | Hi, This is what I have been trying to do. Please download/view the files/project at: http://www.recallinc.com/Meter.zip Thank you for your kind reply, Dheeraj | ||
mike 20:26 Nov 06, 2004 | We need a *small* example that demonstrates the problem. We cannot debug your application. | ||
m_dheeraj.hotmail 20:36 Nov 06, 2004 | Hi, Apologies. Here is shortened version. Error is on clicking Load button. /* START OF CODE */ #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Button.H> #include <FL/Fl_Input.H> #include <FL/Fl_Output.H> #include <stdlib.h> #include <string.h> #include <stdio.h> int formWidth = 600; int formHeight = 480; int buttonWidth = 120; int buttonHeight = 30; /* Input Boxes attributes */ int inputWidth = 200; int inputHeight = 30; /* Output Boxes attributes */ int outputWidth = 200; int outputHeight = 30; Fl_Input *inputID; Fl_Output *outputName; void cbLoad( Fl_Widget* o, void*); int main(int argc, char **argv) { Fl_Window *mainWindow = new Fl_Window(formWidth,formHeight); Fl_Button *loadButton = new Fl_Button(((formWidth - buttonWidth)/2), 440, buttonWidth, buttonHeight, "Load"); loadButton->callback((Fl_Callback*) cbLoad); loadButton->type(FL_NORMAL_BUTTON); Fl_Input *inputID = new Fl_Input(120, 40, inputWidth, inputHeight, "Enter Meter ID : "); inputID->value("Enter Meter ID"); Fl_Output *outputName = new Fl_Output(120, 90, outputWidth, outputHeight, "Name : " ); mainWindow->end(); mainWindow->show(argc, argv); mainWindow->redraw(); return Fl::run(); } void cbLoad(Fl_Widget* o, void*) { /* Crashes with either statement */ const char* meterID = inputID->value(); // inputID->value("Clicked Load"); } /* END OF CODE */ Thanks, Dheeraj | ||
m_dheeraj.hotmail 21:09 Nov 06, 2004 | Hi, The program crashes with the following also: Replacing the cbLoad() above. void cbLoad(Fl_Widget* o, void*) { outputName->value(inputID->value()); } | ||
m_dheeraj.hotmail 00:20 Nov 07, 2004 | Hi, Same behaviour / segmentation fault on Linux too. Thanks, Dheeraj | ||
mike 05:08 Nov 07, 2004 | OK, the problem isn't with FLTK but with your code - you are declaring inputID global and then redeclaring it local in your main() function - the global variable is never initialized. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | gl redraws interact badly with the mouse cursor in Linux |
Version: | 1.1.4 |
Created By: | shaiay.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
shaiay.yahoo 12:21 Nov 09, 2004 | The problem: It seems redrawing in an Fl_Gl window is interacting badly with the mouse cursor. The drawing does not occur under the cursor, so redrawing the window with the cursor over it does not redraw the square arounf the cursor. This problem can be obsered in all the gl demos of fltk if the cursor is placed over the window while something is drawn there. (i.e. spin the puzzle in the puzzle demo with the cursor MOVING over the puzzle to observe the flickering). This problem occurs in Linux -- both fedora core 1 with nvidia accelerated X and on another machine with fedora core 2 and intel accelarated X. On WinXP with an nvidia card and driver the problem does not show up. I have no access to a mac | ||
mike 14:41 Nov 09, 2004 | This is almost certainly a bug or misconfiguration with the X drivers on your systems, as I do not see similar results on FC2 or RH9 with ATI and NVidia cards. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Contents of Double_Window will not repaint when displayed over Gl_Window in WinXP only |
Version: | 1.1.5rc2 |
Created By: | rjs.uab |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rjs.uab 13:44 Nov 09, 2004 | I have an Fl_Gl_Window fullscreen'd with a global idle() method added via Fl:add_idle(). When the user commands the program to enter the configuration mode, I use Fl:remove_idle() to stop the idle() method from being called and repainting the Gl_Window, then I display an Fl_Double_Window modally over top of the Gl_Window. In Windows 2000, all is well, but in Windows XP (SP2 Home) the Double_Window and any child windows (also made modal) will paint correctly when opened but then controls within them will not update in response to events. However, hide()ing and show()ing the window again causes it to be repainted, and shows that the controls inside the window were in fact responding to events, as the changes the user made are saved, but the controls were not repainting themselves. I do have a workaround. Hiding the Gl_Window FIRST and then displaying the Double_Window works fine (it must be done in this order, otherwise the bug is still present), although this causes the GL graphics context to be lost, so I move the Gl_Window almost off the screen (two pixels worth are still visible) with the position() method, THEN display the Double_Window. This also results in repainting working as expected. However, it would be nice to not have to hide the Gl_Window. | ||
mike 19:11 Nov 19, 2004 | I suspect this is a driver-specific bug and not in a bug in FLTK. Please try a newer driver for your graphics card, or another XP system with a different vendor's card. | ||
mike 19:29 Dec 05, 2004 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #577 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | visualc .dsw file does not build software correctly |
Version: | 1.1.5 |
Created By: | e0richt.yahoo |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
e0richt.yahoo 19:31 Nov 09, 2004 | the DSW file in visual c++ creates libraries fltkxxx.lib but as the final link it wants to open xxx.lib files not fltkxxx.lib files. | ||
mike 19:48 Nov 09, 2004 | Dupe of 577 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | fltk-UTF8-1.1.4-3: pasting does not work on linux 6.1, possibly 7.3 |
Version: | 1.1.4 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 19:20 Nov 10, 2004 | fltk-utf8-1.1.4-3 apps under Redhat 6.1 and 7.3 can't be pasted into, but regular fltk-1.1.4 works fine. So this is specific to the UTF8 modifications. I found the culprit; in Fl_x.cxx, the code for Fl::paste() has: vvvvvvvvvvvvvvv XConvertSelection(fl_display, property, fl_XaUtf8String, property, fl_xid(Fl::first_window()), fl_event_time); If I change that 'fl_XaUtf8String' to the regular FLTK-1.1.4's value of 'XA_STRING', then pasting works again. I'm not sure of a way to have FLTK auto-detect if the local OS does not support UTF8 correctly.. I did poke at this quite a bit, and left a paper trail on fltk.bugs; From: Greg Ercolano, msgs dated 11/05/2004 and 11/06/2004. | ||
mike 20:18 Nov 10, 2004 | Since this isn't against an official release of FLTK, I'd prefer that you send this report to the author instead. Thanks! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | Simple event handling window that extends Fl_Window |
Version: | 1.1.5 |
Created By: | constantjunk.hotmail |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
constantjunk.hotmail 09:20 Nov 11, 2004 | EventWindow.zip 4k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
constantjunk.hotmail 09:20 Nov 11, 2004 | Program will not compile and gives the following error at compile time ld: Undefined symbols: EventWindow::EventWindow[in-charge](int, int) vtable for EventWindow I have declared an EventWindow(int, int) function in the EventWindow class | ||
mike 15:40 Nov 11, 2004 | Please post this question to fltk.general. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Counter doesn't work well with FL_PLASTIC_UP_BOX |
Version: | 1.1.5 |
Created By: | sennahb.gmx |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sennahb.gmx 06:54 Nov 14, 2004 | This STR is like STR #467 !!! Setting the boxtype of Fl_Counter to FL_PLASTIC_UP_BOX does not set the counter's text field to FL_PLASTIC_DOWN_BOX ! in the method "Fl_Counter::draw()", after the lines: "if (boxtype[0] == FL_UP_BOX) boxtype[0] = FL_DOWN_BOX; if (boxtype[0] == FL_THIN_UP_BOX) boxtype[0] = FL_THIN_DOWN_BOX;" add: "if (boxtype[0] == FL_PLASTIC_UP_BOX) boxtype[0] = FL_PLASTIC_DOWN_BOX;" | ||
mike 17:15 Nov 19, 2004 | Won't fix. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | No more FLTK flicker .. would you like to try it? |
Version: | 1.1.5 |
Created By: | mariwan |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
mariwan 14:19 Nov 14, 2004 | Fl.cxx 31k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mariwan 14:19 Nov 14, 2004 | In fact it is not a problem. It is improvement of FLTK to be without flickering. Double buffered widget. As I announced in the fltk.development. I did not know the possibility of sending files to FLTK in this way. Please test this and give your notes. Mariwan Note: 1-The modification is just in Fl.cxx version 1.1.5 2-For MacOS not implimented. 3-You will not be in need for Fl_Double_Window 4-I think it will be better to make this modification as an option not as "Default". I mean the "Double Buffered Widget"... | ||
anonymous 08:47 Nov 15, 2004 | Change is effectively the same as making all windows Double. This is not necessarily a bad idea. It looks like the contributed code is creating a new one on every draw and it should be destroying it after each drawing, but I don't see the destructor, is that a mistake? May want to consider using a single off-screen image rather than a different one per window. The memory savings could be considerable (especially on X where this memory is in the server) despite the fact that windows must be completely redrawn since the old image is not preserved (actually it could preserve it in the common case where only one window is being redrawn). It's true that this will make update completely flicker-free, and on modern machines the speed loss is unnoticable. In fltk2 it may mean that a lot of other tricks with clipping to reduce the flicker could be ignored. | ||
mariwan 10:24 Nov 15, 2004 | Dear Bill Spitzak, Thanks for your notes.. It was very important for me. I would like to say :- 1-The speed of program been produced by the modification is not affected at all by the modification of Fl.cxx .. As I used the modified Fl.cxx on many diff. computers from P1 166MHz to P4 2.8MHz both on windows & Linux. 2-To avoid the complexity of (drawing/redrawing) and thinking about which one should be drawn which one not, I use to redraw everything in an offscreen and then copy it to the main screen. For ex. if you test a program like Xforms Emulation. You have the clock widget and the timer that should be redrawn all the time … this is very easy with my modification as the new widget will not be redrawn directly but via offscreen. 3-As you see in the code Fl.cxx you have both commands DeleteDC(_sgc); DeleteObject(hpix); .. It is little bit surprising me if you say that the offscreen is not destroyed by the end of the process. … could you please explain that for me.. 3-If FLTK could provide this technique, as an option will be the best choice. Mine is activated by default.. Which one you prefer? Or what is your opinion ? Mariwan | ||
spitzak 00:56 Nov 17, 2004 | You are right the destructors are there. I'm not sure why I could not see them when I was looking at the diff in the code. I am thinking this is the right idea. The current 2.0 code has the option of setting windows to double buffered mode, but it creates a different pixmap for each window. Reusing the same one is probably much more efficient. I would like to keep that one around for the common case of redrawing the same window several times. And it can default to double-buffered mode always, or we could even drop the single-buffered support completely. The X version uses xdbe, which seems to work, but it is unclear if it does anything better than a pixmap. Also it has an annoying bug that means I must nest the outermost window inside another window to avoid it flashing black occasionally. Both newer X and Windows (and OS/X) support true double-buffered drawing natively, so it is probably best to simulate this. | ||
mike 08:27 Nov 20, 2004 | I've looked at the code. I'm not sure I like the change since a) we already have a double-buffered window class for apps that want this functionality, and b) it looks like this change will cause all redraws to be full redraws, which will negatively affect performance. Anyways, I will do some testing and see what effect this has with various applications. | ||
mike 09:20 Nov 20, 2004 | I've applied the change to a copy of my FLTK repo; the menubar demo (which uses a single-buffered window) still shows tearing and other single-buffered artifacts. CPU usage is increased, and I've added printf's to see that it is, in fact, drawing into the off-screen buffer. So, I guess I'm not inclined to apply this patch to CVS at this time since it does not eliminate flicker (while the current Fl_Double_Window stuff does) | ||
mike 19:31 Dec 05, 2004 | Please open another STR if you come up with a new patch... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #577 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | jpeg.lib |
Version: | 1.1.5 |
Created By: | peter.apple.gmail |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
peter.apple.gmail 08:36 Nov 17, 2004 | I tried to build FLTK 1.1.5 in VS. NET 7.0, but got a a linking error of "jpeg.lib not found". Are there any other dependencies I should build first before building FLTK? Thanks for your help. | ||
mike 17:14 Nov 19, 2004 | Dupe | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Windows version doesn't display two Polish chars correctly |
Version: | 1.1.5 |
Created By: | krzysiek.pawlik.people |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
krzysiek.pawlik.people 05:38 Nov 21, 2004 | fl_input_.cxx 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
krzysiek.pawlik.people 05:38 Nov 21, 2004 | Windows build of FLTK doesn't display two Polish characters cirrectly (those are: ś and ź - select ISO-8859-2 to see them correctly). Linux build doesn't have this problem. Under Windows those two are displayed with some blank space after them and next chars are not displayed at all. Proposed solution: I've traced the problem to be in Fl_Input_.cxx file in function Fl_Input_::expandpos(): this function uses 0xA0 constant to decide when the characters width is 4 or 1. 0xA0 is 160 in decimal, ś is 156 and ź is 159. Solution: changed 0xA0 to 0x9C. Results: everything works as expected. Could this be fixed in future versions? | ||
krzysiek.pawlik.people 05:39 Nov 21, 2004 | Attached a file that contains fixed version of function. | ||
mike 05:42 Nov 21, 2004 | Those characters are outside the valid ISO-8859-1 characters, which is the only character set that FLTK 1.1.x officially supports. You want to use FLTK 2.0 which has support for this sort of thing. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Unassigned |
Summary: | scale selected images? |
Version: | 1.2-current |
Created By: | dougtoppin.hotmail |
Assigned To: | rokan |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dougtoppin.hotmail 18:19 Nov 26, 2004 | Running flphoto 1.2 on my Red Hat Linux 9 system. When I try to select multiple images and scale them using the 'selected images' pulldown, flphoto will them prompt me for whether or not with "Are you sure you want to scale and save all of the selected images?". If I select 'yes', it brings up a "JPEG Parameters" window asking for the options I want (I'm selecting Progressive Display and Optimize for Size). It also brings up the "Scale" window but I can't get into it because the focus is being held by the JPEG Parameters window so I can't enter the scale that I want. When I click ok in the JPEG Paramters window, flphoto immediately begins saving the selected images and I do not get a chance to set the scale that I want. I'm probably missing something simple here. Can someone tell me what I'm doing wrong? tks Doug | ||
mike 06:49 Nov 29, 2004 | We do not provide support for flPhoto here; please post your message to the fltk.general group, or email to "flphoto@easysw.com". (FWIW, I believe this bug is fixed in CVS and will be included in the 1.3 release as soon as I get around to it...) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | problems under cygwin |
Version: | 1.1.6 |
Created By: | tgrab.acn.waw |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
tgrab.acn.waw 23:41 Dec 02, 2004 | You cannot compile fltk under cygwin after simply running configure without option --enable-cygwin. If you switch on --enable-localjpeg and --enable-localpng options, it compiles but than some executable programs require mgwz.dll, which is not teh standard MS Windows library. If you switch on only --enable-localjpeg and libpng option is manualy removed from config.h it works (of course png is disabled) | ||
mike 06:11 Dec 03, 2004 | The "--enable-cygwin" option enables the use of the Cygwin libraries. DO NOT USE --enable-cygwin if you do not want to have binaries dependent on the Cygwin libraries. The default is to not use the Cygwin libraries. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Unassigned |
Summary: | not create subclass - error multiple of typeinfo, DEV C++ MinGW (VisualC++ 6 OK but FLTK1.1.6), OS: Windows(attempt 98,2000,ME) |
Version: | 1.1.4 |
Created By: | pavel.tomanek.dlg |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
pavel.tomanek.dlg 04:32 Dec 05, 2004 | #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Box.H> class Box_class : public Fl_Window { public: Box_class(int x, int y, int w, int h, const char *t ) ; } ; Box_class::Box_class( int x, int y, int w, int h, const char *t ) : Fl_Window(x ,y, h, w, t ) { } int main (int argc, char *argv[]) { Fl_Window *window ; Box_class *box; window = new Fl_Window (300, 400) ; box = new Box_class (20, 40, 260, 100, "Hello World!") ; window->end (); window->show (argc, argv); return Fl::run(); } make.exe -f "D:\Z-Pavel\jazyk-C\Dev-Cpp\FLT2\Makefile.win" all g++.exe mainx.o P2_private.res -o "P2.exe" -L"C:/DEV-CPP/lib" -mwindows -lfltk -lole32 -luuid -lcomctl32 -lwsock32 -lm C:/DEV-CPP/lib/libfltk.a(Fl.o)(.data$_ZTI9Fl_Window+0x0):Fl.cxx: multiple definition of `typeinfo for Fl_Window' mainx.o(.rdata$_ZTI9Fl_Window+0x0):mainx.cpp: first defined here C:/DEV-CPP/lib/libfltk.a(Fl.o)(.text$_ZTS9Fl_Window+0x0):Fl.cxx: multiple definition of `typeinfo name for Fl_Window' .... | ||
mike 19:26 Dec 05, 2004 | Please post this to the fltk.general group; this sounds like a mismatch between the compiler used to create the library and the compiler used to compile your code... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | OpenGL |
Summary: | OpenGL and handling of the back buffer |
Version: | 1.1.6 |
Created By: | dicuccio.comcast |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dicuccio.comcast 11:53 Dec 06, 2004 | In some circumstances, Fl_Gl_Window::flush() will leave the OpenGL subsystem in an undefined state. The function will be called, make_current() is called, rendering is set to the back buffer, and no rendering will be done before swap_buffers() is called. This leaves a potentially damaged back buffer being swapped to the front. This is particularly noticeable with the Intel Extreme Graphics chipset, with GL_SWAP_TYPE = NODAMAGE, but is present in all pathways. The patch below corrects the problem. The (hackish) solution is to swap *twice* to insure that the back buffer remains as the back buffer; a more balanced solution is not to invalidate the window at all. diff -urN fltk-1.1.6.orig/src/Fl_Gl_Window.cxx fltk-1.1.6/src/Fl_Gl_Window.cxx --- fltk-1.1.6.orig/src/Fl_Gl_Window.cxx 2004-09-09 17:34:46.000000000 -0400 +++ fltk-1.1.6/src/Fl_Gl_Window.cxx 2004-11-29 10:48:18.986260200 -0500 @@ -285,14 +285,21 @@ if (SWAP_TYPE == NODAMAGE) { // don't draw if only overlay damage or expose events: - if ((damage()&~(FL_DAMAGE_OVERLAY|FL_DAMAGE_EXPOSE)) || !save_valid) - draw(); + if ((damage()&~(FL_DAMAGE_OVERLAY|FL_DAMAGE_EXPOSE)) || !save_valid) { + draw(); + } else { + swap_buffers(); + } swap_buffers(); } else if (SWAP_TYPE == COPY) { // don't draw if only the overlay is damaged: - if (damage() != FL_DAMAGE_OVERLAY || !save_valid) draw(); + if (damage() != FL_DAMAGE_OVERLAY || !save_valid) { + draw(); + } else { + swap_buffers(); + } swap_buffers(); } else { // SWAP_TYPE == UNDEFINED | ||
mike 13:41 Dec 16, 2004 | OK, does changing the code so that the swap_buffers() is only done when draw() is called fix things? Calling swap_buffers() twice could cause a block until the next vertical retrace on some hardware... | ||
mike 14:22 Feb 06, 2005 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #631 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | The back key does not function correclty in Fluid - Help |
Version: | 1.1.6 |
Created By: | onoonan.sympatico |
Assigned To: | mike |
Fix Version: | 1.1.7 |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
onoonan.sympatico 20:01 Dec 07, 2004 | After compiling FLUID I found that I can only go into a topic and not return out of it by pressing the back key. When you press the back key the browser jumps to the top of the page that you are viewing. I have substituted the Fl_Help_View.cxx and Fl_Help_View.h file from FLTK version 1.1.4 and this has fixed the problem. I did have to rebuild the fltk.lib for the change to take effect. Note: I am using Dev C++ to compile the library with GCC 3.4.2 | ||
mike 07:50 Dec 08, 2004 | Dupe of STR #631 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | Fl_Select_Browser::value() error |
Version: | 1.1.6 |
Created By: | twd20.eng.cam.ac |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
twd20.eng.cam.ac 04:57 Dec 12, 2004 | occasionally (approx one time in 10) Fl_Select_Browser::value() returns 0 when called from within a callback attached to the browser widget. The callback is being called in response to a mouse click on an item in the list so should return an int > 0 (which it does do 9 times out of 10). Some kind of logic race? (e.g. mouse button is up before event is handled?) The code is doing quite a bit of work in Fl::add_timeout() functions so mouse up before event handle is quite likely. OS is Linux (ie events coming from XFree86) | ||
mike 05:37 Dec 12, 2004 | Please attach a short example program which demonstrates this problem. | ||
mike 14:20 Feb 06, 2005 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Shortcut handling on MacOS X (with FL_SHIFT) |
Version: | 1.1.6 |
Created By: | geuzaine.acm.caltech |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
geuzaine.acm.caltech 23:58 Dec 15, 2004 | On MacOS X, a shortcut that uses the Shift key (e.g., FL_SHIFT+'c') is executed even when the focus is on an input text field. On Linux and Windows, FL_SHIFT+'c' always leads, as expected, to a capital 'C' being inserted in the input field (the shortcut being ignored). | ||
mike 04:31 Dec 16, 2004 | Please attach a short example program that demonstrates the problem. | ||
geuzaine.acm.caltech 10:08 Dec 16, 2004 | After further testing, it appears that the problem only exists when using Fl_Sys_Menu_Bar. Here is a small example that demonstrates the problem: #include <stdio.h> #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Input.H> #include <FL/Fl_Sys_Menu_Bar.H> void cb(Fl_Widget* w, void*) { printf("Received Shift shortcut\n"); } Fl_Menu_Item table[] = { {"Shortcuts", 0, 0, 0, FL_SUBMENU}, {"Shift+c", FL_SHIFT+'c', (Fl_Callback *)cb, 0}, {"Shift+n", FL_SHIFT+'n', (Fl_Callback *)cb, 0}, {"Shift+o", FL_SHIFT+'o', (Fl_Callback *)cb, 0}, {0}, {0} }; int main(int argc, char **argv) { Fl_Window *window = new Fl_Window(400,400); Fl_Sys_Menu_Bar *menu = new Fl_Sys_Menu_Bar(1, 1, 1, 1); menu->menu(table); menu->global(); Fl_Input *input = new Fl_Input(70,60,300,30,"Input:"); window->end(); window->show(argc,argv); return Fl::run(); } | ||
mike 13:41 Feb 24, 2005 | Reassigning to Matthias, our resident Mac expert. Matthias - is this a limitation of the sys menu support on OSX? | ||
matt 14:10 Mar 09, 2005 | Yes, this is a limitation of Fl_Sys_Menubar. I'll leave this item open, hoping that I will come up with some improvements of this file that I wrote on OS 9 for fltk 1.0.9 (grin). | ||
matt 00:22 Jun 01, 2005 | Carbon sends the keyboard events to the System Menu Bar before FLTK gets a look at them. So, no, I don't beleive we can change that behavior. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Fl_Multi_Label missing in documentation |
Version: | 1.1.6 |
Created By: | sanel.z |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
sanel.z 04:40 Dec 17, 2004 | I didn't find in alphabetical list of classes Fl_Multi_Label class. I's it missing or it will be removed from distribution ? Sanel | ||
mike 12:36 Feb 05, 2005 | It is deprecated and is gone in 2.0. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | Configure stops at "Checking for X overlay visuals..." |
Version: | 1.1.6 |
Created By: | anonymous |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
anonymous 07:38 Dec 17, 2004 | When I do a configure (with default parameters), the configure does a bunch of "Checking ...", but gets stuck on: "Checking for X overlay visuals..." Any suggestions? | ||
mike 11:23 Feb 05, 2005 | Please repost as a non-anonymous user. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Widgets with alpha-channel seem to use just 1-bit |
Version: | 1.1.6 |
Created By: | hugo.framework-studios |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
hugo.framework-studios 10:02 Dec 19, 2004 | alpha.jpg 13k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
hugo.framework-studios 10:02 Dec 19, 2004 | When using a PNG to create a transparent widget on Windows XP, it seems the 8-bit alpha-channel of the PNG is converted to 1-bit (?). In other words, I cannot make half-transparent pixels but I can make a pixel eighter 100% transparent or 100% visable. | ||
mike 11:24 Feb 05, 2005 | This is a side-effect of how transparency is handled in the current release and the limitations of the X11, QuickDraw, and GDI APIs. FLTK 2.0 adds support for full alpha blending on platforms that support it. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Fl_Menu_Item::callback(<stuff>) documentation |
Version: | 1.1.6 |
Created By: | wilson.afn |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wilson.afn 07:41 Dec 20, 2004 | The callback method documentation for Fl_Menu_Item seems to have diverged a bit from reality. The documentation mentions two ways to set the callback. But in FL/Fl_Menu_Item.H we find four inline methods. (Comments and elision are mine): struct ... Fl_Menu_Item { : void callback(Fl_Callback* c, void* p) {...} // Note missing default value void callback(Fl_Callback* c) {...} // Very useful (but undocumented) void callback(Fl_Callback0*c) {...} // Obfuscated source, but OK void callback(Fl_Callback1*c, long p=0) {...}// Undocumented (but has default!) : } Elsewhere, Fl_CallbackN are defined (using the profoundly confusing C++ syntax) as functions accepting an Fl_Widget* argument (along with sometimes an optional long integer) and returning nothing. I have no clue how the compiler would resolve between the last two methods when a solo Fl_Widget* were passed. | ||
mike 15:08 Feb 05, 2005 | There are various ways to set the callback function and data that is passed to the function, however the documented methods are all that we support (the others are for back compatibility only) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | MacOSX 10.3.7 trouble making fluid |
Version: | 1.1.6 |
Created By: | bennyp.rogers |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
bennyp.rogers 21:02 Dec 22, 2004 | I am having trouble building fltk 1.1.6 on my system. === making src === === making fluid === Linking fluid... ld: Undefined symbols: Fl_Widget::label(char const*) make[1]: *** [fluid] Error 1 | ||
mike 12:34 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Text_Display doesn't honnor FL_NO_BOX box |
Version: | 1.1.6 |
Created By: | ariock.skynet |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ariock.skynet 14:43 Dec 23, 2004 | Fl_Text_Display doesn't honnor FL_NO_BOX box: it always redrew over the background. It can be partially corrected by adding a "if(box()==FL_NO_BOX) return;" in the clear_rect method but it's not a good fix as this will not honnor any FRAME box type and keeps drawinf white border around + erase background under letters... | ||
mike 10:59 Feb 05, 2005 | Since we need to clear the background when drawing text (otherwise you'll get smearing), we will not change this behavior. This widget isn't designed to operate in a "transparent" mode. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | index.html Manual needs correction |
Version: | 1.1.6 |
Created By: | robert.arkiletian |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
robert.arkiletian 21:35 Dec 26, 2004 | The Manual which comes with 1.1.6 under /usr/local/share/doc/fltk/index.html is incorrect. It lists the sections of part 3 "Common Widgets and Attributes" as * Colors * Box Types * Labels and Label Types * Images * class Fl_Pixmap Also, online section 6 "Handling Events" and 9 "Programming with Fluid" are incorrect. It seems like it's just the one index.html file that needs fixing. Robert Arkiletian | ||
mike 16:04 Feb 05, 2005 | The index.html file in the source distro is a place holder for the raw HTML files. The real documentation is generated using HTMLDOC. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | autoconf and makeinclude.in errors |
Version: | 1.1.6 |
Created By: | jedimasterthrash |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jedimasterthrash 09:49 Dec 30, 2004 | Running on Sun OS 5.8 sparc sun-fire, gcc 2.95 When I run "make", I get the following error: make: Warning: Can't find `makeinclude': No such file or directory make: Fatal error in reader: Makefile, line 26: Read of include file `makeinclude' failed When I run "configure" manually (which seems to work), and then run "make", I get the following error: sh: autoconf: not found *** Error code 1 The following command caused the error: autoconf make: Fatal error: Command failed for target `configure' Thanks, JMT. | ||
mike 21:59 Dec 30, 2004 | More than likely you are running into a bug with the Sun make program; try using the (far superior) GNU make program (usually installed as "gmake" with the Sun packages) to build the software. We have no plans of changing the build system in 1.1.x. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #681 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Check_Browser: Items and Checkboxes not deactivated |
Version: | 1.1.6 |
Created By: | daniel.polski.vst |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
daniel.polski.vst 11:17 Jan 03, 2005 | When setting an Fl_Check_Browser->deactivate() only the frame and the label get greyed out. Items and checkboxes keep the "activated" color. Tested under winXP. | ||
mike 12:17 Jan 03, 2005 | Dupe of #681 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Documentation |
Summary: | #include <FL/fl_ask.H> goes against documentation |
Version: | 1.1.6 |
Created By: | renez.xs4all |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
renez.xs4all 03:10 Jan 04, 2005 | http://www.fltk.org/doc-1.1/basics.html#3_5 The proper way to include FLTK header files is: #include <FL/Fl_xyz.H> However under Linux version 2.4.28 (root@win2k) (gcc version 3.3.4 20040623 (Gentoo Linux 3.3.4-r1, ssp-3.3.2-2, pie-8.7.6)) #12 Fri Dec 24 15:00:27 CET 2004 b with fltk 1.1.6 #include <FL/Fl_ask.h> doesn't work instead #include <FL/fl_ask.h> is needed. | ||
mike 16:14 Feb 05, 2005 | The basics stuff is for widget includes; fl_ask.H is a convenience function. Since the whole include file scenario is different in 2.0, we're going to close this one out... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | fl_ask doesn't always show Yes and No button e.g. they overlap |
Version: | 1.1.6 |
Created By: | renez.xs4all |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
renez.xs4all 03:46 Jan 04, 2005 | if you run fl_alert/fl_message first and then fl_ask, inconsistenly, shows not the fl_ask Yes box (see programs below. fltk 1.1.6 Linux version 2.4.28 (root@win2k) (gcc version 3.3.4 20040623 (Gentoo Linux 3.3.4-r1, ssp-3.3.2-2, pie-8.7.6)) #12 Fri Dec 24 15:00:27 CET 2004 XFree version 4.3.0.1 fluxbox 0.9.11 compile cmd: g++ -I/home/renez/include -I/usr/X11R6/include -g \ -L/home/renez/lib -Wl,-rpath,/home/renez/lib \ -L/usr/X11R6/lib -lfltk -lm -lXext -lX11 -lsupc++ \ good.cxx -o good tested same programs 1) on the above platform with 1.1.4 same result 2) Under windows 2k no problem valgrind reports possible problem 12042 bytes in 2 blocks are possibly lost in loss record 75 of 75 at 0x1B904C2D: malloc (vg_replace_malloc.c:131) by 0x1B9C680B: XListFonts (in /usr/X11R6/lib/libX11.so.6.2) by 0x1B97A03D: find(int, int) (fl_font_x.cxx:122) by 0x1B97A0DC: fl_font(int, int) (fl_font_x.cxx:202) might have nothing to do with the problem described because the same happens in the wrong and the good version wrong.cxx shows after maximum of 3 runs only the NO button. pressing the button down and holding shows the YES button. Strangely enough not running the wrong program for some time (3 To 5 minutes), Starting it again shows correct behaviour but after maximum of 3 runs shows wrong behaviour again. However running good after wrong always shows correct behaviour. ---wrong.cxx---- #include <FL/Fl.H> #include <FL/fl_ask.H> int main(int argc, char **argv) { fl_alert("Hello message"); fl_ask("Hello Ask"); return 0; } ----good.cxx-----this always works #include <FL/Fl.H> #include <FL/fl_ask.H> int main(int argc, char **argv) { fl_ask("Hello Ask"); fl_alert("Hello message"); return 0; } ------------------------- | ||
mike 10:53 Feb 05, 2005 | I am not able to reproduce the problem on Fedora Core 3. Perhaps you are running into a Window Manager problem or something... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | How to derive Key information?? |
Version: | 1.1.0 |
Created By: | aredfish.ksmail.seed.net |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aredfish.ksmail.seed.net 19:14 Jan 05, 2005 | Hello, I have two questions that I can not breakthrough 1. When I press my hardware keypad,my FL_Input component will show ASCII related to my keypad. But I do not want my hardware keypad information show in my FL_Input component !! 2. When I press my hardware keypad I want to know which key has been pressed. I do not know how to do this !! Hope someone could help me!! Thanks ! Julian | ||
spitzak 00:59 Jan 06, 2005 | This is in Fl::event_key(). It will be FL_KEYPAD+'1', FL_KEYPAD+'2', etc. You have to subclass the Fl_Input and make a handle() that checks for this. The default action of the Fl_Input is to insert the characters. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Build Files |
Summary: | Why does 1.1.6 require jpeg, png and z libraries? |
Version: | 1.1.6 |
Created By: | james.a.frailey.lmco |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
james.a.frailey.lmco 13:24 Jan 06, 2005 | I've migrated from FLTK-1.1.4 to 1.1.6 and have noticed that now libfltk_images.a library has dependencies to libfltk_jpeg.a, libfltk_png.a and libfltk_z.a. I am building on AIX, HPUX and WIN32. The problem is all my applications have suddenly become much larger in size linking with these additional libraries. Am I doing something wrong, or has something changed in the libfltk_images library from 1.1.4 that requires these additional libraries? | ||
mike 11:01 Feb 05, 2005 | We made the decision to include the PNG, JPEG, and Z libraries with FLTK so that those image file formats are consistently supported (i.e. a LOT of users simply didn't know how to do it themselves...) 1.1.7 will include options to specifically exclude the corresponding image libraries from the build - the default will be to include support for JPEG and PNG files. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Need a way to clear the FLTK event queue without processing events |
Version: | 1.1.6 |
Created By: | levoy.cs.stanford |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
levoy.cs.stanford 14:51 Jan 07, 2005 | Dear FLTK, I am having trouble implementing an abort button in a certain situation. I have posted my problem to the usage bboard, but I suspect that there is no way to implement what I am seeking, so it therefore becomes a feature request. I have a slider, and as the user moves the slider, it triggers callbacks, which perform calculations and modify the display. Occasionally, a calculation will "run away", and I would like my user to be able to release the slider and press an abort button. Unfortunately, if in order to check the abort button I call check(), then FLTK begins processing any and all callbacks. If the user is pressing the abort button, it aborts as expected. However, if the user still has the slider pressed, my call to check() will invoke the slider's callback, which will in turn invoke a new calculation (recursively). This is not safe in my program, and I cannot bring my calculation to a state where recursive calls are safe. So how should I implement the abort function? I have tried modifying my slider callback to not invoke a new calculation, but FLTK continues to update the slider position, so it becomes out of synch with what has been calculated, thereby presenting an inconsistent display to the user. I have tried invoking slider->deactivate() to disable the slider before beginning each calculation, but this makes the widget flash off and on. I have tried slider->set_output() to disable the slider before beginning each calculation, but either FLTK ignores this if the slider is currently pressed, or FLTK has already entered additional slider events in the queue, representing the user's continued attempt to move the slider; these get erroneously processed when I check() for the abort button, leading again to recursive calculations. What I'd like is a way to clear the event queue in FLTK, without having FLTK process the events being cleared. There seems to be no way to do this in FLTK. Alternatively, I'd like a way to search the queue for a particular event (the abort button) without processing any other events, but there seems to be no way to do this either. Any ideas out there? Am I missing an obvious solution? -Marc Levoy Associate Professor Computer Science Department Stanford University | ||
spitzak 17:40 Jan 16, 2005 | Before calling check(), set some sort of variable so that you will ignore the slider callback if it happens. Then when your calculation is done, check if the slider is at a new value and start it again if so. If the abort callback happens, set the slider back to the value that was actually calculated. So the user will let go of the slider, hit the abort, and see the slider jump back to the previous value. | ||
mike 10:39 Feb 05, 2005 | Aside from the method that Bill proposed, you could also subclass the slider widget and add an "ignore" flag that you handle() event uses, or simply deactivate() the widget, call Fl::check(), and then activate() the widget. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | How to change active window by software?? |
Version: | 1.1.0 |
Created By: | aredfish.ksmail.seed.net |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aredfish.ksmail.seed.net 18:38 Jan 09, 2005 | Hello, There are two Fl_Input screen on my LCD panel, I want to change active window by software but not touchpanel. How can I do?? Which function should I use?? Thank you very much !! | ||
mike 10:54 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Browser_ has a bug which is hard to discover |
Version: | 1.1.6 |
Created By: | mariwan |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mariwan 01:51 Jan 11, 2005 | I think I found a problem, which is hidden in Fl_Browser_. It is hard to recognize it. In that widget there are 2 possibilities of selections. In file Fl_Browser_ :: fltk1.1.6 1-Done with the function fl_rectf(X, yy+Y, W, hh); line 357 This is to draw the filled rectangle. 2- Done with the function draw_focus(FL_NO_BOX, X, yy+Y+1, W, hh); 365 This is to draw the focus rectangle. Now: If you disable the first function by just “//” remarking the line and trying to run the examples. See the file_chooser project example, without having the first function and with having the second function. Try to scroll up and down with selecting an item. … ha’haaa .. problem .. the redrawing is not correct. The problem is clear: I do not think Fl_Browser_ needs to have draw_focus function. Each time you select an item, 2 times redrawing take affects. It is strange. I have discovered this problem when I was working on the Fl.cxx for making the FLTK based on the Double_Buffered technique . I saw that whenever you click one item, 2 times Fl_Browser_ redraw. But I could not know why. Now I report the bug. My suggestions: 1- I think the first function should not be in the base class, i.e the Fl_Browser_. Because if you want to make the selection in another way, you should make your own Fl_Browser.. Since Fl_Browser_ is the base class, I don’t think it is a good idea to specify the base class like this. 2- I don’ think that Fl_Broser_ , which is the base class , needs to have draw_focus function. Or it could be an option. I hope that the developer test what I discovered, as I see it a bug, and comment on this message. Mariwan | ||
mike 10:40 Feb 05, 2005 | The current focused item is not the same as the current selection. Pressing ENTER or SPACE will select the focused item. The widget is behaving as designed. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | If I create an actural window, how can I remove it's title bar . |
Version: | 1.1.6 |
Created By: | aaron.yahorng.com |
Assigned To: | mike |
Fix Version: | Not Applicable |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
aaron.yahorng.com 23:35 Jan 11, 2005 | hjh_tst_event.cxx 6k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aaron.yahorng.com 23:35 Jan 11, 2005 | Hello, I have a difficult problem I can not solve. Now I create an actural window and I want to remove it's title bar. How can I do?? Which function should I use?? Thank you very much !! | ||
mike 07:03 Jan 13, 2005 | You want to use the "noborder" method of Fl_Window. Please use the FLTK newsgroups to ask questions. These pages are for tracking bugs and feature requests... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | How can I remove the actual window '' title bar'' ? |
Version: | 1.1.6 |
Created By: | aaron.yahorng.com |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
aaron.yahorng.com 17:50 Jan 12, 2005 | Dear FLTK, Now I have a question and I don't know how to solve it? If I create a actual window and I want to remove it's title bar. How can I do? Which method should I use? | ||
mike 17:35 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Scrollbar increments incorrectly in specific situation |
Version: | 1.1.5rc1 |
Created By: | andhow.gmail |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
andhow.gmail 20:16 Jan 14, 2005 | The undesirable behavior is that the scrollbar increments .value() by increasing values (as opposed to a smooth linear increment) which makes the scrollbar appear to scroll too fast. It only does it in a specific situation though: while in the middle of dragging the scroll bar (before releasing the mouse), calling the Fl_Scrollbar::value(int,int,int,int) function using a changing "min" (parameter 3) number. While not letting go of the scrollbar, the increment added to the .value() will increase until the scroll bar is released. Now if dragged again, the increment will be at its normal value, and increase as it is dragged more. For this reason, I believe that some internal value used by the Fl_Scrollbar is not being set when value() is called during a drag. Due to the way scrollbars are usually used, this doesn't come up often, but I have a program where it can be observed very clearly. I can provide the .exe and source if needed. | ||
mike 11:16 Feb 05, 2005 | I don't see any way for us to detect or "fix" this behavior, so I am closing this report out. The widget is just not designed to handle this behavior or scenario. You can propose a change to this for FLTK 2.0, but it will not be fixed or changed in 1.1.x. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Multiple '@' symbols not displaying correctly |
Version: | 1.1.6 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 02:03 Jan 15, 2005 | Seems to be a problem with the @-> stuff.. 1) If you make an Fl_Box in fluid, then make the label: @-> test @-> two @-> three ..(note spaces), you will only will see on the screen: -> test ..the rest of the text becomes 'invisible'. 2) Similarly, if you make the label: @->test ..nothing appears, where I'd expect to see "->test". | ||
mike 11:19 Feb 05, 2005 | The current label drawing code handles a maximum of 2 symbols (1 at the beginning, 1 at the end); I don't see us changing that in 1.1.x since it will require a lot of rewriting for an edge case. FLTK 2.0's symbol support is much more advanced, and IIRC already handles this properly. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Menu_::clear(); |
Version: | 1.1.6 |
Created By: | bryano.europlex |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
bryano.europlex 10:38 Jan 19, 2005 | Essentially when I call Fl_Choice::clear(); It clears the menu alright, but, for some reason doesn't seem to clear things like that which gets returned by Fl_Choie::value(); Also, it would seem that subsequent add(); operations on this widget immediately after a clear(); results in the value returned by value(); incrementing. The same is true of calling menu(NULL) on this widget. The result is that it is practically impossible to repopulate a menu/choice widget, since the indexes _don't_ get reset by ::menu(NULL); or by ::clear(); Perhaps I'm just not understanding how to use the API properly, but, clear(); to my mind means "Remove and reset _everything_". Bryan. | ||
mike 17:22 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | problem whit fltk menuitem and gettext |
Version: | 1.1.6 |
Created By: | marco_monacelli.hotmail |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
marco_monacelli.hotmail 00:12 Jan 24, 2005 | now i develop a small project whit fltk, fluid and gnu gettext i18. for the orther control work fine the label of the control change at runtime, but for the menuitem it not change. sorry for mai bed english and thanks for all. | ||
mike 10:50 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support.See the flPhoto software on the links page for an example of how to use I18N with statically initialized strings. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | create buttons dynamically |
Version: | 1.1.6 |
Created By: | netto.ime.usp |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
netto.ime.usp 13:21 Jan 24, 2005 | Hello, I'm new using fltk and I would like to do a very simple task, but I'm having problems to perform it. I would like to create buttons dynamically, that is, while my program execute some tasks. For example: . . Fl_Window *window = new Fl_Window(800,400); window->end(); window->show(); ///execute tasks... mynewbox = new Fl_Box(FL_UP_BOX,40,40,50,50,"hello"); . . How can I attach mynewbox on the window? Thanks, Marco | ||
mike 17:31 Feb 05, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Test STR page |
Version: | 1.1.6 |
Created By: | miketest |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
miketest 10:32 Feb 01, 2005 | This is a test. Now is the time for all good men to come to the aid of their country. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #712 |
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | Fluid 1.1.6 External Editor Patch |
Version: | 1.2-feature |
Created By: | alvin |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
alvin 10:55 Feb 01, 2005 | fluid-1.1.6.patch-alvin.tar.gz 9k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
alvin 10:55 Feb 01, 2005 | Please find attached a Suversion patch for Fluid 1.1.6. This patch adds external editor support when editing code. When a code node is selected, the user is presented with the standard code dialog box. However, there is an additional "Editor..." button that when clicked will launch the editor command defined by the FLUID_EDITOR environment variable. If this variable is not set, then the user is presented with another dialog that instructs the user how to set this variable both in bash, csh and in Windows (NT and XP). This patch is a port of the original patch (see fltk.general for message Fluid External Editor Patch for URL). I simply just made this patch work. I retained (as best I could) the coding style of the original authours. Personally, I do not like this style and would gladly clean up the code to match FLTK's coding standard. One other thing, I am unable to test the WIN32 code. I do not have Windows at work nor at home. If possible, I'd like this patch to become part of Fluid. Fluid is a great GUI builder but I miss my syntax highlighting, code expansion, code folding abilities of my current editors (gvim, nedit, kate, etc.). Thank you for your time, Alvin Beach Email: abeach@deepvision.ca | ||
mike 17:33 Feb 05, 2005 | Dupe... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | FL_RELEASE message not sent after Fl_Menu_Button::popup() |
Version: | 1.1.5rc1 |
Created By: | andhow.gmail |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
andhow.gmail 07:27 Feb 10, 2005 | test.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
andhow.gmail 07:27 Feb 10, 2005 | FLTK is usually very good about always sending corresponding FL_ENTER/FL_LEAVE and FL_PUSH/FL_RELEASE, which is why I submit this tiny bug: If you tell a Fl_Menu_Button to popup (via Fl_Menu_Button::popup()) in response to a click (after an FL_PUSH) has been sent, after the modal loop, a corresponding FL_RELEASE is not sent. I attached a simple program that demonstrates this. Just click on the box in the middle and look at the console. | ||
mike 05:19 Feb 14, 2005 | Not sure what we can do here - modal windows (like menus) take all events, and if we sent release/lease events when Fl::modal() is called, then we'd probably have problems with existing apps and widgets that depend on the current behavior... | ||
andhow.gmail 07:31 Feb 14, 2005 | Perhaps sending the FL_RELEASE before entering the modal loop or after the modal loop has finished? | ||
mike 13:33 Feb 24, 2005 | The popup() method doesn't know if it has been called in response to a FL_PUSH event, so there is no way for us to do this safely. And again, sending extra events may cause side-effects. So, I think there is nothing we can do to "fix" it in FLTK; you should expect that you *won't* get a FL_RELEASE event after you do a menu. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #732 |
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Improved Icon support for Win32 and X11 (but no Mac yet...) |
Version: | 1.2-feature |
Created By: | joekrahn |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joekrahn 11:24 Feb 26, 2005 | Here is some source code for improved icon support under Win32 and X11. It is currently written as a Window subclass, and works well (for me) under FTK 1.1.6. I've included the header at the top of the attached source file. Features: Creates plain X11 Color+Mask Pixmaps, or ARGB NET_WM_ICONs when supported. Creates WinXP DIB icons with alpha, or standard DDB bitmask icons. In both cases, the icons are updated for active windows. Win32 note: Documentation states that CreateIcon requires DDBs. I am assuming that pre-WinXP GDI will return unsuccessfully if given a DIB, in which case the DIB is reduced to a DDB. But, this has not been tested. TODO: Some rules for supporting Large and Small icons, and possibly additional sizes. Move some init code, like checking for WM_NET_ICON support, into FLTK's init routines. Mac OSX support; a bit more involved due to icon resource organization. I couldn't find any example code for direct run-time icon building. Look here for documentation: http://developer.apple.com/documentation/Carbon/Reference/IconServices/icon_services/function_group_19.html Have fun, Joe Krahn | ||
mike 21:32 Mar 18, 2005 | Dupe | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #732 |
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Improved Icon support for Win32 and X11 (but no Mac yet...) |
Version: | 1.2-feature |
Created By: | joekrahn |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
joekrahn 15:22 Feb 26, 2005 | Fl_Icon_Window.cpp 12k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joekrahn 11:26 Feb 26, 2005 | Here is some source code for improved icon support under Win32 and X11. It is currently written as a Window subclass, and works well (for me) under FTK 1.1.6. I've included the header at the top of the attached source file. Features: Creates plain X11 Color+Mask Pixmaps, or ARGB NET_WM_ICONs when supported. Creates WinXP DIB icons with alpha, or standard DDB bitmask icons. In both cases, the icons are updated for active windows. Win32 note: Documentation states that CreateIcon requires DDBs. I am assuming that pre-WinXP GDI will return unsuccessfully if given a DIB, in which case the DIB is reduced to a DDB. But, this has not been tested. TODO: Some rules for supporting Large and Small icons, and possibly additional sizes. Move some init code, like checking for WM_NET_ICON support, into FLTK's init routines. Mac OSX support; a bit more involved due to icon resource organization. I couldn't find any example code for direct run-time icon building. Look here for documentation: http://developer.apple.com/documentation/Carbon/Reference/IconServices/icon_services/function_group_19.html Have fun, Joe Krahn | ||
mike 21:32 Mar 18, 2005 | Dupe | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | Test test test |
Version: | 1.1.6 |
Created By: | mike |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
mike 12:14 Feb 26, 2005 | DiffChooser.cxx 8k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mike 12:14 Feb 26, 2005 | This is a test. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | Not compilable with Metrowerks Codewarrior 8.3 |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Portale 11:06 Mar 16, 2005 | OS_X.sit 36k | ||
Portale 11:07 Mar 16, 2005 | config_h_Metrowerks.patch 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 11:06 Mar 16, 2005 | I still have to deal with Carbon, so Metrowerks Codewarrior 8.3 is still a tool I use (commiserate with me!). In order to make things compilable again, the config.h inside <fltk>/CodeWarrior/OS_X.sit needs two little modifications: 1) Undefining USE_QUARTZ and defining __APPLE_QD__ 2) An extra FD_ZERO definition that uses memset instead of bzero (this is a hack, but works for me) The attached OS_X.sit includes the changed config.h and compiles with the 1.1-current version. The patch of the config.h is also attached. | ||
mike 15:17 Apr 10, 2005 | The consensus among the core developers is to drop the Codewarrier files from the official releases, since none of us have or use Codewarrier... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Tabs do not allow more than one line of Tabs, nor scrolling. |
Version: | 1.2-feature |
Created By: | matt |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
matt 09:27 Mar 31, 2005 | Fl_Tabs are limited to a single row of tabs, with no option to 'scroll' the tabs left and right (see VisulC6) or have a pulldown menu extension. The developer team feels that overloaded Tabs are bad practice and a different user interface element should be used in those cases. This feature will not be added to FLTK1. As a side note: FLTK2 starts 'squshing' tabs when there are too many. Feasible, but still not a prefered option. | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | FLTK conflicts with math standard library |
Version: | 1.1.6 |
Created By: | alexhunt |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
alexhunt 09:23 Apr 01, 2005 | main.cc 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
alexhunt 09:23 Apr 01, 2005 | I have been having some problems with a library that uses FLTK for graphical purposes. This library creates a wraper around the standard math library so that the trigonometric functions accept parameters in degrees instead of radians. If have reproduced the fragment of code that gives the problem. The code compiles and executes correctly when the FLTK library is not specified in the compiler arguments. It follows the compile line that rises all errors: g++ -I/usr/local/include/FL -Wall -g -o cosine main.cc When removing "-I/usr/local/include/FL" the example compiles and works. My system: Debian Sarge Kernel 2.6.8 FLTK 1.1.6 Thank you in advance for any help. Source code: #include <iostream> #include <cmath> class Maths{ public: static double cos(double a){ std::cout << "This is a test of the cosine function: " << std::cos(a) << std::endl; return std::cos(a); }; }; int main() { std::cout << "Hello world" << std::endl; Maths::cos(3.0); return 0; } | ||
mike 18:34 Apr 01, 2005 | This is not a bug - DO NOT ADD THE FL DIRECTORY TO YOUR INCLUDE PATH! The correct procedure is to add the main include directory, i.e. "-I/usr/local/include", and then use "#include <FL/Fl_Foo.H>"... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Example Programs |
Summary: | Subwindow demo: Character drawing |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Portale 09:55 Apr 08, 2005 | Subwindow_text_drawing.png 6k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 09:55 Apr 08, 2005 | If I click into the window background area and type text, the letters are drawn next to the mouse pointer. I vote for leaving it as a funny easter egg :) Seen on Windows XP and OSX | ||
mike 10:28 Apr 08, 2005 | This is not a bug, but is actually the subwindow program is implemented. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | Fluid: end() in custom non-Fl_Group widgets |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Portale 07:45 Apr 11, 2005 | Box_with_end.fl 0k | ||
Portale 07:46 Apr 11, 2005 | Box_with_end.cxx 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 07:45 Apr 11, 2005 | If I create a CustomWidget which is a direct Fl_Widget subclass (e.g. Fl_Box), an "end();" statement is added at the end of the custom widgets' constructor as if it were a Fl_Group descendant. See attached example. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | Window repeatedly gets redrawn unnecessarily. |
Version: | 1.1.6 |
Created By: | seeger.cs.unc |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
seeger.cs.unc 10:59 Apr 12, 2005 | buggyWindow.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
seeger.cs.unc 10:59 Apr 12, 2005 | System: Windows XP professional When an Fl_Gl_Window is larger than its parent, it sometimes gets stuck in a state where it continually gets redrawn and takes up all the CPU time. I am able to trigger this state by typing CTRL-ALT-DELETE and then hitting the cancel button in the dialog but it also happens after running the program for a while for other unknown reasons. I can see that FLTK keeps grabbing WM_PAINT messages off the event queue but they do not seem to be generated by FLTK. Perhaps there is some problem in indicating to the OS what needs to be redrawn and the OS thinks it needs to draw the part of the Fl_Gl_Window that is hidden but this never gets drawn. | ||
mike 16:12 Apr 12, 2005 | This usage (widgets larger than the parent window) is now supported. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_File_Chooser value(const char*) not wholly functional |
Version: | 1.1-current |
Created By: | Natevw |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Natevw 08:34 Apr 16, 2005 | testfc.cxx 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Natevw 08:34 Apr 16, 2005 | Fl_File_Chooser does not select files that are passed using value() when the chooser is not loaded. When this occurs, the filename is not even shown in the input box, though the correct path is selected. For the purpose of saving files, if the string passed to value() doesn't start with a drive letter or slash, it should be appended to the chooser's current directory. The attached program should demonstrate what I mean. -nvw | ||
matt 01:06 Jul 26, 2005 | I looked through this for a while and I do see your issues. I decided to not fix these because they will have an impact on how the whole File_Chooser works and may disturb too many existing applications. For FLTK2, we must definetly rethink some of the interface. In detail: The first issue, '\' vs. '/', was already fixed by an earlier STR. The second issue, setting Fl_File_Chooser::value("..") is reduced to setting only a directory during '::show()' is true and unfortunate, but changing this would change the way that '::directory()' works. There is an easy workaround though. This sequence works well: chooser.directory("C:\\Windows\\System\\"); chooser.show(); chooser.value("C:\\Windows\\System\\COMMDLG.DLL"); ... Sorry for the inconvinence. Issue 3, to be able to first set a directory, then add a relative filepath, is again logical and nice to have, but would change the existing API too much IMHO. I'll leave this item open for a week or so, and if another developer comes up with a good solution, I'll be happy to implement that. Matthias | ||
Natevw 13:20 Jul 26, 2005 | I definitely agree...issues two and three can be taken care of easily enough on the application side. Perhaps we should make a note of issue #2 in the docs, but after that closing this STR is fine by me. thanks, -nate | ||
matt 07:17 Aug 15, 2005 | Close because the issues are easily avoidable, plus fixing them would be opening up trouble in existing software. Sorry! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | glitch in text display when using Xft |
Version: | 1.1-current |
Created By: | krishnakumar |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
krishnakumar 01:22 Apr 19, 2005 | xft_txt_disp_bug.png 10k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
krishnakumar 01:22 Apr 19, 2005 | When built with --enable-xft, the first line of all text display widgets is not displayed properly - the top portion is cut-off by a pixel. In the attached screenshot, the '.' in 'i' is not displayed at all. | ||
mike 05:46 Apr 19, 2005 | What linux distro, version of XFree/Xorg, etc? | ||
mike 06:35 May 12, 2005 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | FLUID |
Summary: | Fluid: New Input_Choice via menu, crash |
Version: | 1.1.7 |
Created By: | Portale |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 15:48 Apr 19, 2005 | Adding a new Input_Choice by clicking on the Widget Bin in fluid works fine. But Adding it via "New->Menus->Input_Choice..." makes fluid crash. Windows XP Visual Studio .NET | ||
Portale 15:50 Apr 19, 2005 | ahmm.. I meant version 1.1-Current. Sorry. | ||
Portale 00:00 May 23, 2005 | Please delete me. I am a duplicate of <http://www.fltk.org/str.php?L879> | ||
matt 07:44 Jul 14, 2005 | duplicate | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | One time mouse cursor's shape changes to pointer (hand) when moving over a link in Fl_Help_View, but another it doesn't. |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 23:52 Apr 25, 2005 | This happens mostly (if not only) to Fl_Help_View controls created AFTER the first one. The first one behaves almost perfect. I'm not sure if I have ever observed this behaviour with my first-created Fl_Help_View. | ||
mike 06:59 Apr 26, 2005 | What OS? What compiler? Can you attach a small example that shows this issue? | ||
Sadysta 11:29 Apr 26, 2005 | Currently I'm developing my software using MSVC 6 @ Win XP SP1. I'll be porting it to Linux and MacOS later. Actually it happens no matter what content is in Fl_Help_View. When I move mouse pointer over my app's window and then over Fl_Help_View it happens from time to time that the cursor doesn't change. On the other hand when it changes and I move mouse pointer out of Fl_Help_View and hold it over something else (e.g. Fl_Tabs) cursor's shape remains the same (hand). This is not a regular behaviour, however I experience it quite often. | ||
mike 08:38 May 12, 2005 | I can't reproduce this with the help demo or with FLUID. Can you attach a small example program that shows this behavior? Thanks! | ||
Sadysta 03:34 May 13, 2005 | Mangling with the links (especially double-clicking) causes w2 to show, but the cursor remains in pointer state. I don't think any other other widgets affect this. #include <FL/Fl.H> #include <FL/Fl_Window.H> #include <FL/Fl_Help_View.H> #include <FL/Fl_Hold_Browser.H> #include <FL/Fl_Menu_Button.H> const char text[] = "<a href=\"http://www.google.com\">Test</a>"; Fl_Window *w2; void cb (Fl_Widget *w, void *param) { w2 -> show(); } const char* hf (Fl_Widget *w, const char *uri) { Fl_Menu_Item menu[] = { {"Test1", 0, cb}, {"Test2"}, {0} }; menu -> popup (Fl::event_x(), Fl::event_y()); w2 -> show(); return NULL; } int main (int argc, char *argv[]) { Fl_Window w (0, 0, 640, 495, "flbug"); Fl_Help_View hv1 (0, 0, 640, 150); Fl_Hold_Browser hb (0, 155, 640, 150); Fl_Help_View hv2 (0, 310, 640, 150); Fl_Menu_Button mb (0, 465, 100, 25, "Test"); w.end(); Fl_Window w2 (30, 30, 640, 480, "flbug2"); ::w2 = &w2; w2.end(); hv1.value (text); hv2.value (text); hv1.link (hf); hv2.link (hf); hb.add ("Test"); w.show (argc, argv); return Fl::run(); } | ||
mike 06:28 May 19, 2005 | I think we're looking at a well-known issue on Windows where we do not get an event when the mouse moves to another window due to a focus change. Not sure what to do here - your program works perfectly on Linux and OSX. | ||
Sadysta 02:19 May 27, 2005 | I'm having some tough times on my university, but I'll look into FLTK code ASAP and investigate those issues. It has to be possible to make this work as this stuff works perfectly in all Win32 apps except those written in FLTK :) | ||
Sadysta 02:13 Jul 09, 2005 | Today I've observed the same behaviour on Linux OS. fltk-1.1.x-r4406, gcc 3.3.4, x.org 6.8.1, linux 2.6.12.2 (slack 10.1). | ||
matt 05:57 Jul 19, 2005 | Can't duplicate this on Mac OS X. Maybe it is related to the fact that you open popups? I'll try on another machine next week, but if I can't reproduce this at all, we may have to close this report. Maybe you have a more repeatable way that shows the faulty behavior every time? | ||
Sadysta 01:57 Jul 21, 2005 | >> Maybe it is related to the fact that you open popups? That's highly probable. But I really need to open those popups in my real app so they should cooperate correctly with Fl_Help_View. >> I'll try on another machine next week, but if I can't reproduce >> this at all, we may have to close this report. And the issue will remain... | ||
matt 15:30 Jul 23, 2005 | I am doing my best, clicking, moving and dragging the mouse for 20 minutes, but I can not repeat this problem (Windows XP, SP2, VC 6). Please send me more details on what to do and what exactly I should see. | ||
mike 18:03 Aug 07, 2005 | I'm afraid there is nothing we can do to fix this, since none of us can reproduce it. Sorry. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | I need a way to force window handle creation (e.g. to use with fl_xid()) without showing it to the user. |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 12:05 Apr 26, 2005 | More precisely I want to show a window with Win32 SW_SHOWNA or equivalent command. I don't want to get the focus. | ||
mike 13:18 Apr 26, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Obtain the hardware scancode of the pressed/released key |
Version: | 1.2-feature |
Created By: | lucas |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
lucas 06:31 Apr 28, 2005 | I'm working on a low level project that needs access to the keyboard scancode generated by a key press/release. Currently that is done using a global keyboard hook, but it world be nice if I could have a event_scancode() or whatever. I'm using the library on windows and linux, so instead of having different code for each case, if I just could get this from the fltk event handler would spare me some work. In windows the scancode is on every keyboard message, but on linux I confess I'm not sure (I didn't made that part of the code), but I know gdk returns that info, so I suppose fltk can do it too. Keep up the good work... | ||
matt 02:20 Jul 16, 2005 | fl_xevent will give you the original event structure which you can use to get the scan code. http://www.fltk.org/documentation.php/doc-1.1/osissues.html#osissues There is similar stuff for all other supportes OS's | ||
matt 02:21 Jul 16, 2005 | Feature is already available. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | glitch when drawing text with fonts containing complex glyphs. |
Version: | 1.1-current |
Created By: | krishnakumar |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
krishnakumar 03:50 May 02, 2005 | cursor_glitch.png 27k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
krishnakumar 03:50 May 02, 2005 | Platform: Mandrake 10.0 XFree86-4.3.1, Mandrake-10.1 Xorg The I-beam cursor is causing some weird drawing problems when used with fonts containing complex glyphs. The problem occurs when the cursor is moved back from the end of line to the start of line. It is more prominent when the line is longer. Unfortunately, it is a propreitery font so I cannot redistribute it. A screenshot showing the problem is attached. | ||
mike 08:50 May 02, 2005 | So long as we can't have access to the font, it will be impossible to test or fix this problem. If you can reproduce this problem with a freely available font, please re-submit your STR along with a test program, the font in question, and directions to reproduce it. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | wrong position when moving invisible window |
Version: | 1.1.6 |
Created By: | xovo |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
xovo 11:16 May 05, 2005 | test-050505-vismovebug.cpp 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
xovo 11:15 May 05, 2005 | Hi, when hiding a window and moving it into negative coordinates, a following show() will position it to 0,0 please find attached code thanks, Thomas OS: WinXP SP2 | ||
mike 18:36 May 05, 2005 | The window manager does not have to honor the application's request, which is what is happening here. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Compile time error on OS X Tiger |
Version: | 1.1.6 |
Created By: | cwingrav |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
cwingrav 08:25 May 12, 2005 | makeinclude 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cwingrav 03:49 May 07, 2005 | OS X requires linking with more libraries than Panther or previous. By adding '-lobjc' to the tail of the LDLIBS in makeinclude, it links. | ||
cwingrav 03:54 May 07, 2005 | Also, need to add -lobjc to the end of GLDLIBS. | ||
mike 06:37 May 07, 2005 | Please test using the current SVN snapshot and let us know if the problem persists... | ||
cwingrav 14:55 May 07, 2005 | Tried compiling fltk-2.0.x-r4315 and ran into the same problem. I had to change a few of my personal Makefiles also to deal with Tiger in the same manner. | ||
mike 06:18 May 09, 2005 | How about with the 1.1.x-r4315 snapshot? | ||
cwingrav 07:59 May 09, 2005 | Nope. It errors out with the following: ld: /usr/lib/crt1.o illegal reference to symbol: __objcInit defined in indirectly referenced dynamic library /usr/lib/libobjc.A.dylib make[1]: *** [fluid] Error 1 | ||
mike 13:23 May 09, 2005 | Hmm, strange, I can't reproduce this on a fresh install of 10.4 final. Do you also have Xcode 2 installed, or are you using the 10.3 compiler tools? | ||
cwingrav 13:42 May 09, 2005 | I wonder if it has something to do with a fink library it is linking to and should not. I did an upgrade install too from 10.3. That too might have something to do with it. | ||
cwingrav 13:48 May 09, 2005 | Hmmm, looks like there might be something to that. I took a look at the config.log file and it has for several "PATH:" entries, '/sw/bin' or '/sw/sbin' which are used for Fink. Don't know if that has to do with linking however. | ||
mike 07:33 May 12, 2005 | Yes, the Fink stuff will likely get in the way. What are "CC" and "CXX" set to in the makeinclude file? | ||
cwingrav 08:24 May 12, 2005 | gcc - apple uses gnu tools I'll attach the makeinclude file. | ||
mike 09:59 May 12, 2005 | OK, and if you type: which gcc does it show "/usr/bin/gcc", or something else? | ||
cwingrav 16:51 May 12, 2005 | Yup, goes to /sw/bin/gcc. This is set in the /sw/bin/init.csh which is set to run in the login files for Fink users. I set it so it appended the Fink directories as opposed to prepended. I then used a new shell and recompiled 2.0.x-r4315 and 1.1.x-r4315. Both had the same problem still. I'm guessing anyone with Fink is going to have problems unless they link to the libobjc.a library. | ||
mike 20:21 May 12, 2005 | We don't support the Fink GCC compiler, just the Apple GCC compiler. The Fink folks need to fix their port of GCC... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Image Support |
Summary: | Fl_Pixmap should have a protected copy constructor |
Version: | 1.1.6 |
Created By: | ariock |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ariock 23:06 May 10, 2005 | Fl_Pixmap should have a protected copy constructor has Fl_Image has one. The current lack of such a constructor makes it impossible to write: Fl_Pixmap img = Fl_Pixmap(xxx); and worst: Fl_Pixmap img[5] = {Fl_Pixmap(xxx), Fl_Pixmap(xxx), Fl_Pixmap(xxx), Fl_Pixmap(xxx), Fl_Pixmap(xxx)}; | ||
mike 07:07 May 11, 2005 | What prevents you from doing: Fl_Pixmap img(my_pixmap_data); ??? | ||
ariock 23:03 May 11, 2005 | The problem is not when you trie to create one instance, but when you want to create an array of them. The first example iwas just to demonstrate the problem... | ||
mike 06:07 May 12, 2005 | Why not use pointers, e.g.: Fl_Pixmap *array[] = { Fl_Pixmap(foo), Fl_Pixmap(bar) }; I *think* that should work in a static initializer... | ||
mike 06:08 May 12, 2005 | FWIW, we can't change the ABI at this point to support a copy constructor... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | [redhat9] Window size shrinks of resizable() set to a widget larger than window |
Version: | 1.1.6 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
greg.ercolano 19:56 May 14, 2005 | foo.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 19:56 May 14, 2005 | Actual situation: in the group, someone made a window with a scroll, and in the scroll, a box much larger than the screen's size, then made the box the resizable(). I noticed (under redhat9) setting the resizable made the 800x600 window shrink to a tiny postage stamp sized window. Code attached is a melted down version of the problem, w/out the scroll. | ||
mike 06:37 May 19, 2005 | I'm not even sure this is something we want to support. However, if we did handle this situation, what would be the proposed behavior? Limit the resizeable size to the containing parent? | ||
matt 16:02 Jul 23, 2005 | I don't think that there is anything wrong with FLTK here. Widgets inside an Fl_Scroll are not resizable by the nature of the Fl_Scroll, so setting a 'resizable()' is undefined. The sample program does not have a scroll, but sets a widget larger than its parent, which again is not supported in FLTK. What do the other developers think. | ||
mike 17:51 Aug 07, 2005 | Resizable stuff will only work for non-scrolled widgets. By definition, widgets inside a Fl_Scroll widget will be scrolled within the size of the widget, not resized. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | FL_DND_LEAVE event wrongly delivered |
Version: | 1.1.6 |
Created By: | oracle2025 |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
oracle2025 14:27 May 15, 2005 | fl_drag_test.cxx 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
oracle2025 14:27 May 15, 2005 | I tried to implement Drag and Drop inside my Application with the Help of FLU (http://www.osc.edu/~jbryan/FLU/), and it seems is if I found a rather annoying bug. Whenever two Widgets are close, it happens occasionally that a FL_DND_LEAVE event is delivered although the mouse-cursor is inside the widget. I checked FLU, but I'm fairly sure the issue is to be resolved in fltk. I'm attaching a little sample to expose the bug. It can be reproduced as follows: Drag from the upper File_Browser to the lower box, depending on the mouse speed, sometimes the drop Event is delivered, and sometimes not. Actually the FL_PASTE event is delivered correctly, but because there was a FL_DND_LEAVE event before, it assumes the paste occured outside the widget, and therefore it ignores the event. I'm developing in Linux-X11 I'will try to dig a little deeper, and might be able to come up with a solution if the fix is trivial. | ||
mike 18:10 May 15, 2005 | Please post an example program that does not use FLU, otherwise we won't be able to help you! | ||
oracle2025 04:46 May 16, 2005 | Hi, It seems as if I've found the issue, the Problem is that sometimes (if the widgets are very close) the events are delivered as follows: FL_DND_ENTER to Widget_2 FL_DND_LEAVE to Widget_1 I don't know whether fltk is supposed to order the ENTER and LEAVE events such that the FL_DND_LEAVE event comes first? However Flu_DND assumes that this is the case and fails in this situation. If fltk makes no guarantees about the order in which events are delivered, and I suppose this is the case, then this issue has to be resolved in FLU. | ||
mike 06:16 May 16, 2005 | I don't believe we have any documentation on the behavior, so FLU should handle it in any order, however it would make sense to send the FL_DND_LEAVE event *first* and then FL_DND_ENTER... | ||
mike 19:32 May 19, 2005 | OK, it appears that it will not be possible to change the current behavior, since we will not know whether to send FL_DND_LEAVE until we find a new widget that accepts FL_DND_ENTER. So, FLU needs to be fixed to handle the current behavior. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #722 |
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Sometimes FLTK misses the change of clipboard data. |
Version: | 1.1.6 |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 10:19 May 17, 2005 | It happens from time to time that when I cut/copy something (I mean text) in non-FLTK application and then try to paste it into my FLTK app's Text_Input or another widget capable of pasting, the text which used to be in the clipboard before my last cut/copy gets pasted or nothing is pasted at all. I haven't done an in-depth research on how FLTK clipboard mechanism works but it seems to be a serious bug to me. OS: Win XP | Compiler: MSVC6 PS. Of course the data in the clipboard is pasted normally in all the other, i.e. non-FLTK applications | ||
mike 13:20 May 17, 2005 | Dupe of STR #722 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | select incorect unreadable font with Xft in ru_RU.UTF-8 locale |
Version: | Will Not Fix |
Created By: | yuri |
Assigned To: | yuri |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
yuri 01:03 Dec 08, 2005 | setfont.png 6k | ||
yuri 01:04 Dec 08, 2005 | nosetfont.png 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yuri 22:59 May 19, 2005 | I use russian utf-8 locale. when application started it use incorect unreadable fonts when using Xft unless manualy call Fl::set_font for replacement default fltk fonts family. The same problem present in 2.0 snapshot. | ||
mike 05:12 May 20, 2005 | Reassigned to 2.0. 1.1.x does NOT support UTF-8. | ||
yuri 01:03 Dec 08, 2005 | no problem on fltk-2.0-svn. problem is in 1.1.x.!!! --------------------- i use gentoo gnome-2.10 in bash_profile: LANG=ru_RU.UTF-8 (or LANG=en_US.UTF-8 same problem) the problem is: when fltk configured --with-xft. No Symbols drawn! but fl_measure seems to work correctly. when changed to bash_profile: LANG=ru_RU (or LANG=en_US) all work well all work well in case manual font setting, fo example FL::set_font(FL_HELVETICA," Tahoma"); --------------------- | ||
greg.ercolano 16:44 Dec 10, 2005 | Yuri: I think Mike's point is that the version of FLTK you're using (1.1.x) does NOT support UTF8. So what you're doing will not work in 1.1.x. FLTK 2.0 is the only official version of FLTK that will support UTF8, and Mike is assigning the problem to that version, presumably, to ensure the problem doesn't exist there, if at all. To do what you want with 1.1.x, you would need to use the 'unofficial' version of FLTK 1.1.x that supports UTF8: http://developer.berlios.de/svn/?group_id=4036 | ||
yuri 02:01 Jan 16, 2007 | I've close this bug w/o resolution, because: !!!there is no bug in 2.0, all works good!!!!! This is not fltk, but my fontconfig problem with 1.1 fltk version. Solution for fltk-1.1 is patch fltk font families names (or maybe ajust fontconfig) , or don't use xft with fltk-1.1 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Fl_Window::copy_label() doesn't work on Mac OS X |
Version: | 1.1-current |
Created By: | Natevw |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Natevw 07:24 May 20, 2005 | testlabel.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Natevw 07:24 May 20, 2005 | If ::copy_label() is used on an Fl_Window under OS X, the label remains unchanged. The following code works fine under Windows, not OS X: thanks, and sorry I haven't had time to track this down myself, -nate | ||
Natevw 12:08 May 21, 2005 | Even though Fl_Window has it's own copy_label() function, the Fl_Widget::copy_label() function is used instead. This seems to causing the trouble, but why is this so? [GCC 4.0.0 on Tiger] -natevw | ||
Natevw 12:01 Jun 18, 2005 | This seems to be working as of r4393. -nate | ||
matt 01:26 Jul 15, 2005 | Tested in working in the current svn version. No need to fix. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Window short-time inactive when opening (OSX) |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Portale 14:16 May 26, 2005 | inactive_active_window.png 7k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 14:16 May 26, 2005 | Like in Bug http://www.fltk.org/str.php?L886 on slow macs, a short double draw can be observed. This time independantly from the scheme. When an FLTK window opens (no matter if it is a main window or subdialog) it first has an inactive window decoration (flat shadow) and immediately after that an active decoration (deep shadow). This is not typical for OSX apps. They always start directly with the active look. On normal Macs, this can only be observed, if you pay attention pedantically. I think that unlike STR #886, this is not so easy to change. Maybe we just live with this? :) | ||
matt 06:27 Jul 19, 2005 | I would not know what causes this. Creating a window is pretty straight forward and follows Apple example code. The only delay I see is calling some FLTK events, but that should not take much time at all... . | ||
mike 19:52 Aug 07, 2005 | I think we can just live with this... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | 1.1.x under mingw/msys does not build threads demo by default |
Version: | 1.1-current |
Created By: | ianmacarthur |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ianmacarthur 01:56 May 27, 2005 | If you configure the 1.1.x build under msys/mingw on win32, then the makeinclude file that is generated does not include the threads demo, even if you explicitly set --enable-threads on the configure. Indeed, since threads are always (are they?) enabled on win32, I guess the threads demo ought to be built by default. It's been this way since forever, of course, but I thought it might be nice to fix this before 1.1.x is frozen, just in case the lack of the threads demo confuses new users into thinking threads don't work on win32 or something. Cheers, -- Ian | ||
mike 19:57 Aug 07, 2005 | What is the THREADS variable set to in the makeinclude file? | ||
ianmacarthur 10:47 Aug 08, 2005 | The THREADS variable is empty in the makeinclude file, viz: THREADS = | ||
cable_guy_67 10:44 Aug 26, 2005 | I haven't seen if this has been dealt with or not. It does still seem to exist in svn 1.1.r-4537. As Ian mentioned threads should always be available if WIN32 is defined. This makes sense based on the check in test/threads.cxx where the check #if HAVE_PTHREAD || defined(WIN32) is done at line 30. In configure, after checking for pthread.h (and not finding it) is makes a check with -lpthead. This is the configure output when compiling with --disable-cygwin --enable-threads checking for pthread.h... (cached) no checking for pthread_create in -lpthread... (cached) yes checking for pthread_create using -pthread... no It would seem that is would be safe to give THREADS= a value (or not) based on this check. Either that, or somewhere below the THREADS= line that I commented out check if WIN32 is defined and put the proper value in THREADS at that point. This does make it work correctly for the following ./configure options in makeinclude --enable-cygwin --enable-threads THREADS = threads.exe --enable-cygwin --disable-threads THREADS = --disable-cygwin --enable-threads THREADS = threads.exe --disable-cygwin --disable-threads THREADS = Doing a diff on configure ( 1.1 is svn and ./ is changed ) $ diff ../../fltk-1.1/configure ./configure 2836a2837 > THREADS="threads.exe" 2839a2841 > THREADS= 2875c2877 < THREADS= --- > # THREADS= All this does it give a value when -lpthread is added to LIBS or empty in the else. HTH | ||
mike 13:51 Oct 12, 2005 | You don't have the pthread.h header file installed, which is why the threads demo is not built... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | Active popup menu causes Fl_Window::iconize() to malfunction |
Version: | 1.1.6 |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 02:42 May 27, 2005 | When I call w -> iconize() in one of Fl::add_fd() handlers while any popup menu is active, window w doesn't show despite it should or it shows in a weird minimized form characteristic for MDI applications and when I click restore button on its title bar it disappears. In both cases it doesn't show even if w -> show() or w -> iconized() is called again without any active popup menu. OS: WinXP | Compiler: MSVC6 | ||
matt 15:13 Jul 23, 2005 | Would it be possible to post a short, self contained source file that shows this behavior, so I can start the bug hunt? Thanks. | ||
mike 19:55 Aug 07, 2005 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | <font color="...">test <a href="test">test</a> test</font> |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 13:20 May 27, 2005 | In Fl_Help_View: <font color="...">test <a href="test">test</a> test</font> The last "test" is drawn with default color not the one supplied with <font> tag. It's not HTML 2.0-compliant behaviour AFAIK. | ||
Sadysta 13:31 May 27, 2005 | Quick and dirty solution: File: Fl_Help_Viex.cxx In: Fl_Help_View::draw() after: int underline; add: Fl_Color textcolor__; replace: else if (strcasecmp(buf, "FONT") == 0) { if (get_attr(attrs, "COLOR", attr, sizeof(attr)) != NULL) { fl_color(get_color(attr, textcolor_)); } with: else if (strcasecmp(buf, "FONT") == 0) { if (get_attr(attrs, "COLOR", attr, sizeof(attr)) != NULL) { textcolor__ = textcolor_; fl_color(textcolor_ = get_color(attr, textcolor_)); } after: else if (strcasecmp(buf, "/FONT") == 0) { add: textcolor_ = textcolor__; | ||
mike 18:55 May 27, 2005 | Fl_Help_View does not claim or desire to support the full functionality of HTML. In addition, we can't fix this problem in 1.1.x, since it would break binary compatibility. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | None |
Summary: | fltk-utf8-1.1.6 problems |
Version: | 1.1.6 |
Created By: | greg.ercolano |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 01:06 May 31, 2005 | Testing on redhat9 and Windows 2k w/VC 7.x .NET compiler. redhat9 ------- Compiles fine. Windows + VC 7.X ---------------- Fails to compile fltkutf8.dll, has these errors: http://seriss.com/people/erco/fltk/bugs/fltkutf8dll-1.1.6-windows-errors.html Looks like this problem was brought over from the original fltk 1.1.6 release, which was fixed by STR#726: http://www.fltk.org/str.php?L726 | ||
greg.ercolano 01:36 May 31, 2005 | Also: Looks like fltkutf8png.lib compiles OK, but ends up in the main fltk directory, instead of in the lib directory. Simple tweak to the fltk.dsw properties: Solution Explorer | libpng | Properties | Librarian | General ..just change: Output File: ..\fltkutf8png.lib ...to instead: Output File: ..\lib\fltkutf8png.lib | ||
mike 05:21 May 31, 2005 | Please contact the author of this patch, since we don't host it at fltk.org... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | Weird behaviour with two modal windows |
Version: | 1.1.6 |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
matt 03:25 Jul 20, 2005 | hello.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 11:47 Jun 02, 2005 | When I have two modal windows on the screen and third (non-modal) window is shown (in Fl::add_handler() handler) and then hidden (this time with use of Fl::add_timeout()) the most recent modal window disappears O_O I mean it's still active but it's not visible and only by pressing ESC can I bring focus back to the second modal window. | ||
mike 12:49 Jun 02, 2005 | What operating system? What compiler? | ||
Sadysta 13:13 Jun 02, 2005 | Jeez. I can't believe I forgot it once again. Sorry. OS: Windows XP, Compiler: MSVC6 This effect is 100% repeatable so I guess it's not likely to be caused by memory management issues. Currently I set the non-modal window's dimensions to (0,0) instead of hiding it as a workaround. But it should be fixed anyway. | ||
mike 15:37 Jun 02, 2005 | Have you tested this against the current 1.1.x snapshot? | ||
matt 03:25 Jul 20, 2005 | At least on OS X, I can't duplicate this. Would the attached code show this behavior? | ||
matt 15:00 Jul 23, 2005 | I can not duplicate this on XP SP2 and VC6 on my machine. Could you please verify the probem with the code I posted, or suppply your own short code that shows this behavior? Thanks. | ||
mike 19:48 Aug 07, 2005 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Can't build for FLTK/X11 on Mac OSX |
Version: | 1.1-current |
Created By: | chaircrusher |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
chaircrusher 12:38 Jun 09, 2005 | While the CMakeLists.txt included with fltk allows you to choose X11 on OSX, the code itself assumes that __APPLE__ targets are always going to use Carbon. I am in the process of fixing this locally and can submit a patch. In this particular case the program I'm working on runs on X11 (on Linux,SGI, etc) just fine and crashes on OSX/Carbon mysteriously, so I thought I could just build X11 on OSX and get the program running ... | ||
chaircrusher 15:13 Jun 09, 2005 | OK. I have a patch that works -- except for OpenGL support. I was working with the CMake files, not hacking the Makefiles, but anyway: 1. Configured with CMake to use X11. 2. Defined a new symbol, if you're on Mac OSX AND want to use X11. To get things to work -- and I'm on a deadline so it was quick and dirty -- most every place that there's an #ifdef __APPLE__ I had to change it to #if (defined(__APPLE__) && !defined(_APPLE_X11_) OpenGL I haven't had time to crack, because I need to delve into how the CMakeLists.txt files acquire and use the OpenGL libraries. What happens now is that it always picks up the OSX OpenGL library syntax, instead of the OpenGL libraries in /usr/X11R6/lib. I'm a little confused about how this should really work -- it seems like right now the choices (of 1.1-current) are Apple, Win32, and X11, and the Apple always uses the Carbon stuff. What needs to happen is to configure such that if you select X11, even if you're on a Mac OSX platform, it uses the vanilla X11. Since __APPLE__ gets defined whether you select X11 or Carbon, I had to do the fix ugly as described above. | ||
matt 01:50 Jul 15, 2005 | The Apple side of FLTK is using Quickdraw by default. The symbols is __APPLE_QD__. FLTK 1 can be compiled using Quartz instead via autoconf --enable-quartz. The associated symbol is __APPLE_QUARTZ__ . FLTK 2 also supports Apple X11. Since FLTK1 is nearing its completion with 1.1.7, I beleive that a backport of X11 support for FLTK 1 on Mac OS would introduce more trouble than worth. Your original statemnt was, that your Linux X11 versions of an app are working, but the Mac OS is not. It would be best to figure out if the trouble is FLTK related and fix the Quartz/Quickdraw code instead of introducing X11 code. Thanks you for your help. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | fltk_jpeg is out of data in power mac os 10.3.9 |
Version: | 1.1.6 |
Created By: | clearsky |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
clearsky 12:53 Jun 15, 2005 | I'm using fltk 1.1.6. When I tried to link to fltk_jpeg, the linker always complains: g++ -ggdb3 -L/sw/lib -lfltk -lfltk_gl -lfltk_images -lfltk_png -lz -lfltk_jpeg -o test test.o ld: table of contents for archive: /usr/local/lib/libfltk_jpeg.a is out of date; rerun ranlib(1) (can't load from it) After I call "ranlib libfltk_jpeg.a" and link again, it complains that symbols not found g++ -ggdb3 -c -I/sw/include -o test.o test.cpp g++ -ggdb3 -L/sw/lib -lfltk -lfltk_gl -lfltk_images -lfltk_png -lz -lfltk_jpeg -o test test.o ld: Undefined symbols: Fl_JPEG_Image::Fl_JPEG_Image[in-charge](char const*) Can someone tell me what is wrong with that? Thanks | ||
greg.ercolano 18:34 Jun 15, 2005 | [posted to STR from group] ld is probably detecting the datestamp on the lib is different than its datestamp at the time it was compiled. Question: when you compile fltk, do you copy the fltk libs to /sw/lib in such a way that maybe the modification time of the lib is changed? If so, you would probably need to rerun ranlib at that point in time to fix up the dates to match timestamps within the lib. When copying the libs around, make sure the -p flag is specified to cp(1) to ensure the modification time is preserved. I'm not sure if that's the problem, but it's likely timestamp oriented. It's also possible fltk's own install target in the Makefile is at fault here; the commands that install the lib don't seem to use the -p flag: --- fltk-1.1.6/makeinclude: [..] CP = cp --- fltk-1.1.6/src/Makefile: install: [..] -$(CP) $(LIBNAME) $(FLLIBNAME) $(GLLIBNAME) $(IMGLIBNAME) $(DESTDIR)$(libdir) Core folks: should there be a '-p' flag in there? === Replication === # make foo <-- normal build gcc -g -I.. -O2 -Wall -Wunused -Wconversion -fno-exceptions foo.cxx -o foo -L../lib -lfltk_images -lfltk -lfltk_png -lfltk_z -lfltk_jpeg -framework Carbon -framework ApplicationServices -lsupc++ /Developer/Tools/Rez -t APPL -o foo ../FL/mac.r # touch ../lib/libfltk.a <-- change date stamp # rm foo ; make foo <-- rebuild gcc -g -I.. -O2 -Wall -Wunused -Wconversion -fno-exceptions foo.cxx -o foo -L../lib -lfltk_images -lfltk -lfltk_png -lfltk_z -lfltk_jpeg -framework Carbon -framework ApplicationServices -lsupc++ ld: table of contents for archive: ../lib/libfltk.a is out of date; rerun ranlib(1) (can't load from it) <-- ERROR make: *** [foo] Error 1 # ranlib ../lib/libfltk.a <-- fix lib ranlib: file: ../lib/libfltk.a(Fl_lock.o) has no symbols ranlib: file: ../lib/libfltk.a(fl_overlay_visual.o) has no symbols [..] # make foo <-- rebuild gcc -g -I.. -O2 -Wall -Wunused -Wconversion -fno-exceptions foo.cxx -o foo -L../lib -lfltk_images -lfltk -lfltk_png -lfltk_z -lfltk_jpeg -framework Carbon -framework ApplicationServices -lsupc++ /Developer/Tools/Rez -t APPL -o foo ../FL/mac.r # <-- succeeds | ||
greg.ercolano 18:42 Jun 15, 2005 | PS. from the ranlib(1) man page: BUGS With the way libraries used to be created, errors were possible if the library was modified with ar(1) and the table of contents was not updated by rerunning ranlib(1). Thus the link editor, ld, warns when the modification date of a library is more recent than the creation date of its table of contents. Unfortunately, this means that you get the warning even if you only copy the library. | ||
matt 05:48 Jul 19, 2005 | I assume that the fltk-jpeg library is copied using Unix 'cp' instead of 'CpMac'. If that alone does not fix the problem, running 'ranlib' after copying the library should be a sufficient fix. I am not very good at changing autoconf related files, so maybe someone else could update and try this. btw: I didn't have problems on my 10.4 machine. Using --enable-loacljpeg will use FLTK's jpeg instead of an installed one, if I understood this crrectly. | ||
matt 13:54 Jul 22, 2005 | Ok, I can't duplicate this at all. Two things that I'd like to mention though: in the command line, you are missing "-framework Carbon -framework ApplicationServices" to create a working execuatble, plus the required Rez command to set the resource. Also, youseem to have the libraries inside '/sw/lib' instead of the default '/usr/local/lib' which by itslf is of course just as well. However, if you happened to copy 'libfltk_jpeg' via Unix 'cp' into that localtion, you would get exactly the error message as dscribed. The second error message could come from a missmatch between fltk_jpeg and fltk; maybe they were compield with different versions of the source? | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | None |
Summary: | Calling value("x") from within Fl_*Input callback hangs application |
Version: | 1.1.6 |
Created By: | duncan |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
duncan 06:31 Jun 26, 2005 | FlInputCallback.tgz 6k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
duncan 06:31 Jun 26, 2005 | http://www.fltk.org/newsgroups.php?s5399+gfltk.general+v5406+T0 I've re-engineered a version of the GeometryDialog class called DontWorkDialog that calls inputWidget->SetValue() from inside the callback if the value just entered is invalid. For example, enter "e" for the sphere radius and then hit return. The fl_alert() pops up OK, but the interface becomes unresponsive after that. This is fltk-1.1.6 on WinXP + SP2, running under Cygwin [The only log files I can find for Cygwin are dated 2002, but the cygwin setup.exe is dated 07/04/2004 - I thought I was far more up to date than that :-( ] | ||
matt 13:52 Jul 11, 2005 | I will try to check the code eventually, however, this is much too large of a project to test on the fly. Please try to keep all code in a single cxx file that I can test without taking apart any makefiles. Thanks. | ||
matt 07:16 Jul 14, 2005 | Hi Duncan, after looking through your code, I found a lot of string manipulation going on that goes into buffers that are not protected from overrun. As much as I like to help you debug your code, I simply do not have the time for it - sorry. Please look around for some bounds checking tools and try to use snprintf, strncpy, etc. instead of those functions without maximum buffer size count. Also, make sure that all buffers are large enough to hold all evetual text. I am pretty sure that this is not an FLTK problem because my code (and other samples I saw) do use the value(char*) feature successfully. I will be happy to reopen this bug report if you can send us some 20 line sample code that shows the behavior you see. Thanks, Matthias | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | None |
Summary: | How do I get a dialog closed? |
Version: | 1.1.6 |
Created By: | Kim |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Kim 20:55 Jun 26, 2005 | Looking for help - I have coded a dialog with OK and Cancel buttons that is launched from the main window, and the dialog does not disappear when I click OK or Cancel. I have miscoded something, but I am not sure what the correct code should be. * Current behavior: - user clicks the Settings button - callback is to settings() function - settings() launches the dialog, a Fl_Double_Window - user changes the Fl_Value_Slider on the Settings window - user clicks OK => OK button stays depressed, and settings dialog still showing, OR user clicks Cancel => Cancel button stays depressed, and settings dialog still showing - user clicks 'X' close button at upper right of window - settings dialog disappears. - speed variable correctly has new value if OK was clicked, proving that the callback for OK was invoked * Desired behavior: - the usual stuff - when OK or Cancel is clicked, the settings dialog goes away. * I could find no examples in the doc (although I saw the Replace window in the editor.cxx project in the test dir, but trying those ideas did not help). Does anyone have an example? * I tried reading the FLTK code and found the hide() function and a wait loop, but evidently I am not using it correctly. * Implementation follows (via Fluid). thanks very much!!! I am loving FLTK!! Kim ================================================================== static void cb_Cancel(Fl_Button* o, void*) { // hide the dialog owning this button // (is this correct??) o->parent()->hide(); } static void cb_OK(Fl_Return_Button* o, bool* v) { // need to reference v like this to get // fluid to declare the v arg. if do only // this: *v = true; then no v name decl!? bool* okFlagPtr = v; *okFlagPtr = true; // hide the dialog owning this button // (is this correct??) o->parent()->hide(); } static void settings(double& speed) { Fl_Double_Window* w; bool okFlag = false; Fl_Value_Slider* s = 0; { Fl_Double_Window* o = new Fl_Double_Window(218, 150, "Settings"); w = o; o->align(FL_ALIGN_BOTTOM); { Fl_Button* o = new Fl_Button(115, 100, 80, 25, "Cancel"); o->callback((Fl_Callback*)cb_Cancel); } { Fl_Return_Button* o = new Fl_Return_Button(25, 100, 80, 25, "OK"); o->callback((Fl_Callback*)cb_OK, (void*)(&okFlag)); } { Fl_Value_Slider* o = new Fl_Value_Slider(25, 40, 170, 30, "Speed (Letters per Second)"); o->tooltip("Change the display speed"); o->type(5); o->maximum(10); o->value(1); o->align(FL_ALIGN_TOP); o->value(speed); // start with passed-in value s = o; } //o->set_modal(); o->end(); } w->show(); while (w->shown()) { Fl::wait(); } if (okFlag) { speed = s->value(); duration = computeDuration(speed); } } ================================================================== | ||
mike 05:23 Jun 27, 2005 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #861 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Fl::wait ignores fds while waiting for Windows messages (includes fix for cygwin) |
Version: | 1.1.6 |
Created By: | keithr |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
keithr 21:36 Jun 27, 2005 | cygwin_select.diff.txt 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
keithr 21:36 Jun 27, 2005 | In fl_win32.cxx, Fl::wait() first does a zero-timeout select() for file descriptors, then (with the value passed in by Fl::run()) does an infinite wait for Windows messages, ignoring file descriptors until the next Windows message arrives. There is code to merge file descriptors into the Windows message queue using WSAAsyncSelect, but it is #ifdef'd out as non-working. With cygwin, the Windows message queue is available as a file, /dev/windows, so Windows messages can be combined into the select(). The attached patch implements this. I am porting a terminal emulator to fltk, so equal treatment for file descriptors and GUI events is critical (but the bug is marked low priority since it's not likely to be a big deal for others). | ||
mike 19:51 Aug 07, 2005 | Dupe of STR #861 | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | C++ builder personal edition compiler erors |
Version: | 1.1.6 |
Created By: | cdt |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
cdt 06:33 Jul 08, 2005 | File fl_get_system_colors.cxx 45 #if defined(__APPLE__) && defined(__MWERKS__) 46 extern "C" int putenv(const char*); 47 #endif "Fl_get_system_colors.cxx": E2322 Incorrect number format at line 45 "Fl_get_system_colors.cxx": E2188 Expression syntax at line 45 This sorted out by midifying source code as follows #if defined(__APPLE__) #if defined(__MWERKS__) extern "C" int putenv(const char*); #endif // __MWERKS__ #endif // __APPLE__ This makes fltk compatible with C++BuilderX compilers | ||
mike 17:05 Jul 12, 2005 | We WILL NOT make changes like this to support a buggy compiler. Report this as a serious bug to the Borland folks since the preprocessor syntax is required by the C and C++ standards! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | fl_ask functions can hang |
Version: | 1.1-current |
Created By: | Natevw |
Assigned To: | matt |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Natevw 23:10 Jul 08, 2005 | fl_ask_hang.patch.txt 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Natevw 23:10 Jul 08, 2005 | If an fl_ask.h function is called while the static message_form window is already in use by this class of function (eg through an Fl_Release callback), the library can enter an infinite loop. The most recent call will hide the window, leaving the previous calls waiting for an event from buttons that are no longer active. One possible workaround is to check if the dialog is still shown at the top of the loop. Patch attached. -natevw | ||
matt 13:38 Jul 11, 2005 | +1: I am not entirely sure how to get into this situation without modifying the library itself, but the patch does not seem to harm either, since shown() should always be set by show() | ||
mike 16:59 Jul 12, 2005 | You aren't supposed to call those functions recursively since they show modal dialog windows and the results are undefined - on some platforms you'll probably cause a crash. -1 on the proposed patch since it does not address the problem, it only hides it. | ||
matt 06:46 Jul 14, 2005 | Closed. Recursive use not supported. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Multiple |
Summary: | 0-dimension borderless windows still visible on X11 |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 01:25 Jul 11, 2005 | On MS Windows 0-dimension borderless windows are completely invisible. It would be useful if this behaviour was the same on X11. | ||
mike 17:01 Jul 12, 2005 | Since 0x0 windows are not generally supported, we will not try to support them in FLTK. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 1 - Request for Enhancement, e.g. asking for a feature |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | Show window without stealing focus |
Version: | 1.2-feature |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 05:50 Jul 13, 2005 | Equivalent of Win32's SW_SHOWNA is absolutely necessary !!!! | ||
Sadysta 05:50 Jul 13, 2005 | Of course I'm talking about X11. | ||
Sadysta 11:37 Aug 08, 2005 | I've been using a workaround on both Windows and Linux for quite a while but now I'm trying to port my app to Mac OS X and I'm unable to resolve this issue. So again: Fl_Window needs a method letting programmer to show it without stealing focus from currently active window. On all supported OSes. | ||
mike 12:37 Aug 08, 2005 | It is impossible to do this reliably on all OS's and window managers, so this will not be supported. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | Something is wrong with Fl::add_fd() under Linux |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Sadysta 02:44 Jul 21, 2005 | _add_fd.cxx 2k | ||
Sadysta 02:44 Jul 21, 2005 | Makefile 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 00:50 Jul 21, 2005 | Basically I have a situation in which callback specified in Fl::add_fd() with when==FL_READ is called and surprisingly there's no data on the watched socket. It's a normal network socket and there is neither data nor error so the recv() I'm calling in that callback blocks infinitely. The exact code works perfectly under Win32. I'll try to post short example in a moment. Problem is still there when using stable FLTK 1.1.6. GCC 3.3.4 Kernel 2.6.12.2 | ||
Sadysta 02:49 Jul 21, 2005 | I've posted source and Makefile demonstrating the bug, however it happens really rarely. In my original app (Unicom) it happens almost instantly when I do the following: select menu button, hold space and move mouse in the area of unfolded menu. This way items get selected over and over again and callback is called constantly. I don't know why it happens less often in this example, maybe I'm missing something or maybe the complexity of Unicom somehow accelerates this faulty behavior O_O Anyway - it happens, rarely, but it does. As soon as I add another select() in callback passed to Fl::add_fd() everything comes back to normal. | ||
Sadysta 02:55 Jul 21, 2005 | Oh, and if you want to check it out in Unicom use the Status menu button as it is the one I was using. Debug symbols are provided in the binary so you can seen what it is all about using gdb. At least one account must be added and active (Unicom restart is required after adding new account). | ||
Sadysta 02:56 Jul 21, 2005 | And delete lang.txt from bin/ directory to obtain English interface. | ||
Sadysta 03:13 Jul 21, 2005 | OK, I've observed something interesting. All this seems to happen when btncb() is called twice before the callback from Fl::add_fd() is triggered and then btncb() is called for the third time. It looks like this: fd send fd send fd2 recv fd2 send fd2 recv fd2 send fd recv (16 times or so) fd recv (16 times or so) // now is the interesting part fd send fd2 recv fd2 send NO "fd recv" APPEARS, app gets stuck // Here is the faulty bahavior, i.e. mycb() is called but it's getting blocked while recv()'ing as there is actually no data on the socket And what's really funny is that if I put select() into mycb() it says that there's no data and mycb() returns immediately. But instantly there's another call to mycb() as if the data which should've been on the socket duing last mycb() call has just come through. And everything works fine. Incredible and VERY annoying... | ||
Sadysta 03:32 Jul 21, 2005 | Oh, and as to both example and Unicom it must have sth to do with FLTK because I've just simulated sending data from code not from menu button callback and it works fine. | ||
mike 17:11 Aug 07, 2005 | I'm guessing you are running into a problem with pthreads on Linux, not FLTK. Try refactoring your example into two separate processes (either using fork() or as separate programs) and see if you still see the same problems. | ||
Sadysta 07:40 Aug 14, 2005 | The example seems to work fine when not using pthreads. However it's rather impossible to convert my whole application to the forking model ;/ | ||
mike 12:23 Aug 15, 2005 | Well, at this point I think you've run into a Linux (or glibc) bug, and not a bug in FLTK. I don't think there is anything else we can do to help you at this point, sorry... | ||
matt 13:09 Aug 15, 2005 | Sadysta, could you please provide a short example program that shows the behavior? I know, that is difficult with threads and sockets involved, but without seeing the error, we won't be able to fix it or verify it. I appreciate that you posted your chat program Unicom, but it is much to complex for me to start debugging it. Please understand. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #836 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Assertion failure in Microsoft Debug Runtime |
Version: | 1.1.6 |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 08:38 Jul 26, 2005 | This happens when user accidentally types any non-ASCII (or maybe non-ISO-8859-1 ?) character in any of FLTK widgets: http://sadysta.ath.cx/phpBB2/download.php?id=679 The problem is visible with Debug versions only as assertions are disabled in Release versions. -- OS: Windows XP, Microsoft VC++ .NET | ||
matt 09:38 Jul 26, 2005 | Since this happens inside the debugger, could you please trace this back into FLTK source code, so we know which line (and which function call) in FLTK causes the problem? It would also be helpful to know which character you entered, since neither my German nor my SU keyboard ever caused this assertion to be hit. Not sure if I can set my keyboard to Polish, but if I can, which key must I hit? | ||
Sadysta 12:54 Jul 27, 2005 | This is all my debugger spits out: > msvcr71d.dll!_chvalidator_mt(threadlocaleinfostruct * p=0x01014370, int c=-71, int mask=8) Line 68 + 0x2a C msvcr71d.dll!isspace(int c=-71) Line 133 + 0x2f C fltkdlld.dll!Fl_Input_::expand(const char * p=0x01059fe8, char * buf=0x0012f290) Line 62 + 0x2e C++ fltkdlld.dll!Fl_Input_::drawtext(int X=4, int Y=369, int W=447, int H=62) Line 183 + 0x16 C++ fltkdlld.dll!Fl_Input::draw() Line 45 C++ fltkdlld.dll!Fl_Group::update_child(Fl_Widget & widget={...}) Line 546 C++ fltkdlld.dll!Fl_Group::draw_children() Line 529 + 0x2d C++ fltkdlld.dll!Fl_Group::draw() Line 539 C++ fltkdlld.dll!Fl_Group::update_child(Fl_Widget & widget={...}) Line 546 C++ fltkdlld.dll!Fl_Tabs::draw() Line 244 C++ fltkdlld.dll!Fl_Group::update_child(Fl_Widget & widget={...}) Line 546 C++ fltkdlld.dll!Fl_Group::draw_children() Line 529 + 0x2d C++ fltkdlld.dll!Fl_Group::draw() Line 539 C++ fltkdlld.dll!Fl_Window::draw() Line 101 C++ fltkdlld.dll!Fl_Double_Window::flush(int eraseoverlay=0) Line 297 C++ fltkdlld.dll!Fl_Double_Window::flush() Line 242 + 0x11 C++ fltkdlld.dll!Fl_X::flush() Line 84 + 0x1a C++ fltkdlld.dll!Fl::flush() Line 376 + 0x17 C++ fltkdlld.dll!Fl::wait(double time_to_wait=0.21899999999999931) Line 290 C++ fltkdlld.dll!Fl::wait() Line 302 + 0x11 C++ unicom.exe!WinMain(HINSTANCE__ * hThis=0x00400000, HINSTANCE__ * hPrev=0x00000000, char * lpCmdLine=0x00142397, int nShowCmd=1) Line 587 + 0x8 C++ unicom.exe!WinMainCRTStartup() Line 390 + 0x39 C kernel32.dll!77e814c7() shlwapi.dll!772c1a29() | ||
Sadysta 12:58 Jul 27, 2005 | So the character seems to be -71 ? Which is 10111001 in U2 binary and 0xB9 in hex. The character I'm typing can be seen right here: http://tarkvara.e-pood.ee/fpilt/cp1250windowscentraleuropean.gif at row B0 column 9, AFAIK it's called a-ogonek in Unicode. | ||
Sadysta 13:02 Jul 27, 2005 | The problem is heavily related to version of Microsoft's software. There wasn't such error with MSVC 6.0. When you switch your keyboard layout to Polish press Right Alt + a. | ||
mike 13:10 Jul 27, 2005 | Can you test the current 1.1.x snapshot and let us know if the problem persists? This sounds similar to STR #731 and/or STR #836 | ||
matt 13:12 Jul 27, 2005 | Oh, the problem may be that (*p & 255) returns a negative number, which then triggers the assert in the library. The reason, without investigating much further, since I don't have .net, could be that 'char' may be signed on .net . Two possible solutions come to mind: change 'char' from signed to unsigned using a global flag that is hidden somewhere, but that is incompatible with previous version. Or, we have to go through the source and find all calls to 'isspace' and similar character based functions, and cast to 'unsigned char'. For now, can you change line 64 (IIRC) of 'src/Fl_Input_.cxx' from if (wrap() && (p >= value_+size_ || isspace(*p & 255))) { to if (wrap() && (p >= value_+size_ || isspace((unsignde char)(*p & 255)))) { and see if that fixes at least part of your problem. BTW, this should crash using German characters (like U-Umlaut) as well. Interesting, that nobody reported this earlier! | ||
Sadysta 13:44 Jul 27, 2005 | Patching line 64 ad you told and line 594: 591 // result in sub-optimal update when such wrapping does not happen 592 // but it is too hard to figure out for now... 593 if (wrap()) 594 while (b > 0 && !isspace((unsigned char) index(b))) b--; solved the problem. I hope there are no more such surprises and that it's fixed in SVN ;) | ||
mike 15:35 Jul 27, 2005 | > Oh, the problem may be that (*p & 255) returns a negative number, which Actually, unless MS is not following the C/C++ standards, it is impossible for that expression to return a negative number since "255" is an integer, which promotes the entire expression to int. Since int must be at least 16 bits, and the "& 255" masks off everything above bit 7, you can't end up with a negative number... Anyways, this *is* a duplicate of STR #836. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Documentation |
Summary: | Fluid menu translation |
Version: | 1.1-current |
Created By: | effelsberg |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
effelsberg 03:48 Jul 28, 2005 | As I mentioned in the development forum before, menus created by Fluid don't get translated if localization is set up as usual in main(). I like the hint of wrapping gettext to initialize localization on its first call more than my idea of replacing menu() with a series of add() calls. Nevertheless, this peculiarity really deserves to be mentioned in the Fluid docs. | ||
mike 19:38 Aug 07, 2005 | It isn't a peculiarity, menus are statically defined, which required a global initializer... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | None |
Summary: | fl_frame & fl_frame2 out of index in array (g = fl_gray_ramp()) |
Version: | 1.1.6 |
Created By: | paoloItaly |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
paoloItaly 02:13 Aug 02, 2005 | fl_boxtype.cxx : fl_frame() and fl_frame2() take the border color from active_ramp[24] or inactive_ramp[24], calling the fl_gray_ramp() function - this function return an invalid address memory uchar *fl_gray_ramp() {return (draw_it_active? active_ramp:inactive_ramp)-'A';} because of the displacement of -'A' from the effective memory address. - fl_frame() and fl_frame2() uchar *g = fl_gray_ramp(); ... fl_color(g[*s++]); it will point over the size of array. i've correct the problem in this mode uchar *fl_gray_ramp() {return (draw_it_active? active_ramp:inactive_ramp);} and fl_color(g[*s - 'A']); s ++; Regards | ||
mike 05:31 Aug 02, 2005 | That isn't the correct fix, since other box types will index using the character offsets, and your change effectively does nothing but move the subtraction into the loop. What OS, compiler, etc. are you using? When are you seeing this? With standard FLTK box types, or you own, or ??? | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Under Mac OS X Fl_Gl_Windows aren't always displayed |
Version: | 1.1.6 |
Created By: | derek.signalmarketing |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
derek.signalmarketing 14:32 Aug 31, 2005 | missingwindows.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
derek.signalmarketing 14:32 Aug 31, 2005 | Seemingly at random, Fl_Gl_Windows don't appear, but dragging another window over top of them will reveal them (as the dragged window's drop shadow moves over the hidden window). Fun for the whole family. Not really sure if this is fltk, or the usual osx flakiness. Removing the glPointSize() or glEnable(GL_POINT_SMOOTH) from the draw method seems to make it go away, so it could be timing related. The attached app just sets up a bunch of Fl_Gl_Windows. It's very infrequent that at least a couple don't appear. Also, it takes a rather extreme amount of time to display them all, on linux as well. (we're not waiting for a retrace per window, are we?) | ||
derek.signalmarketing 14:33 Aug 31, 2005 | "It's very infrequent that at least a couple don't appear." Should have read otherwise. Almost every run results in a few missing windows. | ||
matt 13:19 Sep 25, 2005 | Hmm, first of all, I do not get missing GL windows, but that doesn't mean that there is no bug of course ;-) I do get superslow performance when creating all the GL window, just like you say (scary!), but runnign through gprof did not reveal the bottleneck. Investigating. | ||
matt 01:04 Sep 26, 2005 | OK, I was able to duplicate the behavior by adding calls after aglSetInteger. Searching the web has not given me any help in figuring out what's going on. I assume we get the redraw() message before the actual context exists... . Does anyone else know what may be happening here? | ||
matt 07:08 Oct 31, 2005 | I ahd a discussion with the Apple Gurus on the companies mailing list and it seems to be like this: OpenGL in combination with QuickDraw does have its problems. There is nothing we FLTK'ers could do. However, compiling with --enable-quartz does improve performance quite a bit, still not being "great" by a long shot. Anyway, there seems to be no solution at this point. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | RPM build error on recent Mandriva systems |
Version: | 1.1.6 |
Created By: | gostal.foi |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
gostal.foi 01:20 Sep 02, 2005 | fltk.spec 3k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
gostal.foi 01:20 Sep 02, 2005 | Building RPM packages on recent Mandriva systems e.g. 10.2 aka Limited Edition 2005 from the tarball doesn't work due to a modification Mandriva has introduced in the rpmbuild program. The reason for this is that Mandriva want's to run 32-bit code on 64-bit machines. The modification to rpmbuild results in the following error: ----------------------------------------------------- The following files ought to be marked as %multiarch: /var/tmp/fltk-1.1.6/usr/bin/fltk-config ----------------------------------------------------- Detailed info regarding this can be found at the URL: http://archives.mandrivalinux.com/cooker/2005-01/msg01999.php In essence a line has to be added in the spec file section %install for every binary that is arch dependent and maybe also in the relevant %files section depending on how old the system is. Arch dependent headers has to be taken care of in a similar way. Adding the line: %multiarch_binaries $RPM_BUILD_ROOT%{_bindir}/fltk-config to the %install section of the supplied spec file solves the problem and working RPMs are built. The modified spec file is attached. The spec file in Mandriva's src.rpm package shows more modifications. The exact effect of these I don't know. Gösta Ljungdahl gostal@foi.se | ||
mike 13:48 Oct 12, 2005 | This change is not portable to other Linux distributions - you'll just have to maintain your own spec files or disable this Mandriva change to rpm... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Build failure on WIN32 under Cygwin |
Version: | 1.1-current |
Created By: | larry.finger.lwfinger |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
larry.finger.lwfinger 21:25 Sep 12, 2005 | When building libfltk-png.a under cygwin using Windows native mode (i.e. -mwindows -mno-cygwin switches), gcc does not find the zlib headers. The compiler options in 'makeinclude' need to have a switch '-I../zlib' added. | ||
a.rburgers.quicknet 13:45 Sep 26, 2005 | Did you install the mingw-zlib package for cygwin? fltk with gcc -mno-cygwin builds fine under cygwin for me. | ||
larry.finger.lwfinger 16:59 Sep 26, 2005 | No I had not previously installed zlib under Cygwin; however, I have installed it now without any effect. I still need to add a "-I../zlib" switch to the C++ and C compiles lines. | ||
mike 14:14 Oct 12, 2005 | Are you using the configure script or the files in the makefiles subdirectory? | ||
larry.finger.lwfinger 20:08 Oct 12, 2005 | I use ./configure with no additional switches or options. Larry | ||
mike 06:56 Oct 30, 2005 | All I can guess is that you have the libraries installed but not the headers. I am unable to duplicate this... Closing this out, as I don't want to hold up 1.1.7 for this... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Use of strdup highly dangerous. |
Version: | 1.1.6 |
Created By: | SirShane |
Assigned To: | mike |
Fix Version: | Unassigned |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
SirShane 20:51 Oct 06, 2005 | The use of the non-standard C function strdup is very dangerous. On Win32, this function can allocate memory on a different heap than is expected. When this happens, calling free( ) on a pointer returned by strdup will cause a crash. This is exactly what is happening to me with VC++ .NET 7.1. I was able to correct this issue by creating a function called fl_copy_string( const char* str ) in flstring.c which just copies a string with a call to malloc and memcpy. I just replaced all calls to strdup with it. If you would like me to submit a patch, I can do that. In any case, I highly recommend that FLTK stop using strdup. Thanks for a great library. :) Shane Parker Enygma Arts | ||
mike 06:33 Oct 07, 2005 | strdup() is hardly non-standard. It would be useful to know whether you are using the static or DLL library on Windows - unfortunately, depending on how you mix your DLLs and static libraries on Windows, you *can* end up mixing different heaps... This just doesn't happen on other OS's... :( | ||
SirShane 08:18 Oct 07, 2005 | As I understand it, the strdup function is not a real part of the official C standard library. It is very commonly available however on most implementations of it. However, I still recommend not using it. It is very simple to fix FLTK so that it doesn't call strdup, and then this type of problem won't happen anymore. | ||
SirShane 08:21 Oct 07, 2005 | To answer your question, I am using the static library for Windows. Both the static library and the application are compiled to use the Multi-threaded Debug DLL. As I stated earlier, replacing strdup with a function that uses malloc resolves the issue. | ||
mike 15:37 Oct 07, 2005 | strdup() is part of ISO C, ISO C++, and POSIX. It was not part of ANSI C. AFAIK, ALL strdup() implementations use malloc() since the return value is documented as needing to be freed using free(). | ||
SirShane 17:11 Oct 07, 2005 | If that is the case, why does strdup have such unpredictable behaviour under Win32? I should mention that this same type of problem as occurred on more than one occassion for me, not just with FLTK. If FLTK will not be patched to no longer use strdup, I will have to manually patch it myself every time I use it. This is fine, but I figure not using strdup is a better way to go in general since it is so unpredictable, at least on Win32. | ||
mike 17:26 Oct 07, 2005 | I, personally, have *never* had problems with strdup() on Windows, and I do a LOT of software that runs on Windows. Best bet is that your code is overwriting the heap someplace... | ||
SirShane 19:05 Oct 07, 2005 | I've thought of that possibility. However, in this particular case, the crash is occuring after main( ) has finished executing regardless of whether or not main actually has any useful code in it. In fact, it happens in the deconstructor of the Fl_Preferences::RootNode class. Even if my main( ) function is totally blank with no code in it, I still get the crash as long as FLTK is linked into my application. I can only assume that Fl_Preferences is being constructed when the application starts because there is a static copy of it someplace (I could be wrong about this - I didn't actually check the FLTK code to see if there is a static instance of the Fl_Preferences class), and when it gets deconstructed on exit the crash occurs. All I know for sure is that my stack trace shows there is call to the deconstructor of the Fl_Preferences::RootNode class on exit, and that the actual code calls free on a pointer returned by strdup, which causes the error. I realize this is not a common bug. I, too, have used strdup in many situations where I got away with it. However, I've also run into several occassions where it causes crashes. This is not a heap error caused by my application. This issue can be resolved with a 5 minute patch. I think i've argued my point enough. So, i'll just ask again. Can we patch FLTK so that this type of issue is gauranteed to not happen again, or am I just wasting my time? Thanks. :) -Shane | ||
mike 23:13 Oct 07, 2005 | OK, let's start by having you test the current 1.1.x snapshot - Fl_Preferences has had some bug fixes which might resolve this issue. Given the scenario you describe, the problem *cannot* be caused by strdup(). Please try the current 1.1.x snapshot and let us know if the problem persists. | ||
joekrahn 11:22 Oct 09, 2005 | In Win32, strdup() is one of those underscore-prefixed functions, which means it should be a normal supported function, but it is only semi-supported, because Microsoft loves to deviate from any concept of public standards. So, it would not be too surprising if certain versions of VC libs had a problem with strdup(). For cross-platform code, avoidance of Win32's "Compatibility" functions seems like a bad approach. It's better to figure out the root cause, and see how to avoid it. If possible, I suggest posting a tiny example code that demonstrates the problem. If not, can you determine if the problem relates to using single/multi thread libs, and/or release/debug libs? | ||
SirShane 19:51 Oct 09, 2005 | It's not really easy for me to just post a block of code that will cause this problem to happen consistently. On one machine, the code works without any problems, and on another machine it crashes when trying to call free on a pointer returned from strdup. This would suggest that the computer where it crashes is using a different version of the VC runtime libraries or something with a weird version of strdup. It happens with FLTK 1.1.X, as long as I have it linked into my application - regardless of whether or not any real code gets executed in main( ) (It'll crash inside the static class deconstructors). I can compile my application in release mode sometime and see if that fixes it (I'm using debug mode, with debug FLTK static libs), but not right now... and even if that does fix it, the issue still remains. I never liked strdup in the first place because it's an external C function that allocates memory for you, *somewhere*, possibly on some other heap or whatever. It's just not a safe function to use in my opinion - regardless of whether or not it's considered standard. Maybe this wouldn't even be a problem at all if Microsoft wasn't so retarded and only used one set of runtime libraries. | ||
mike 07:20 Oct 10, 2005 | Again, PLEASE TEST USING THE CURRENT 1.1.X SNAPSHOT. There have been several Fl_Preferences-related bug fixes related to accessing memory after it was freed, and *that* might be the cause of your problems. In short, test the current 1.1.x snapshot. Tell us if the problem persists. | ||
SirShane 08:31 Oct 10, 2005 | I stated above that it happens with the 1.1.X snapshot. | ||
mike 10:11 Oct 10, 2005 | OK, sorry I missed that in your responses. The next step is to provide a test program which demonstrates the problem, or run your program through a memory/heap debugger to detect when the corruption occurs. | ||
d.zimmer 08:46 Oct 15, 2005 | FYI, this is from the MSDN library: "The _strdup function calls malloc to allocate storage space for a copy of strSource and then copies strSource to the allocated space.... Because _strdup calls malloc to allocate storage space for the copy of strSource, it is good practice always to release this memory by calling the free routine on the pointer returned by the call to _strdup." IMO, the assertion that strdup is non-standard and dangerous is off-base. The only hint I could suggest is to search for mixing of C-runtime libraries with everything linked against your program... even a debug/release mismatch. I have run into malloc/free crashes in this situation before, and ensuring that all linked libraries use the same C-runtime has always solved the problem (even if it requires the use of the release C-runtime when doing a debug build). | ||
mike 06:52 Oct 30, 2005 | Closing this since we are not going to stop using strdup() and more than likely the problems you are seeing are a MS-specific issue with mixing of libraries or incorrect DLL usage. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | FLUID 1.1.x ignores image and deimage settings for Fl_Window |
Version: | 1.1-current |
Created By: | joekrahn |
Assigned To: | matt |
Fix Version: | 1.1.7 (SVN: v4716) |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joekrahn 11:41 Oct 09, 2005 | FLUID ignores image and deimage for Fl_Windows. The image should be included, because it is drawn as the Window's background. (I think that's a change from earlier 1.x). The deimage value should also be included, because it can still be useful for supplying a Window icon. Even for a widget could never use a given image, a user can access it for some other purpose. The relevant code is the is_window() test here: void Fl_Widget_Type::setimage(Fluid_Image *i) { if (i == image || is_window()) return; ... void Fl_Widget_Type::setinactive(Fluid_Image *i) { if (i == inactive || is_window()) return; ... | ||
mike 03:37 Dec 14, 2005 | Put this STR in the correct category... | ||
matt 07:15 Dec 14, 2005 | Actually, the image that you would assign to an Fl_Window via FLUID would be overridden by Fl_Window::show() which sets the image the the current scheme's background image, effectively erasing your image event if no scheme is selcted. You have to manually load you image into the Fl_Window after Fl_Window::show(). FLUID can not do that for you. Havinf a user provided image override the scheme image is not easily done because we are simply missing space to save an additional flag for this, and we cannot add a byte to the class without ruining binary compatibility. | ||
matt 07:17 Dec 14, 2005 | We should probably hide or deactivate the image fields though when a window is selected, unless we find a way to allow images in windows. | ||
matt 07:37 Dec 16, 2005 | I have grayed out the image fields in FLUID. FLTK does not support an image insed an Fl_Window without a little trickery (setting the image after calling Fl_WIndow::show(), or, even easier, creating a Group that has the same size as the Window and setting the image there. See: test/tiled_image.cxx). It would certainly be nicer to allow backdrop images and schemes at the same time, and I am sure this feature will be in 2.0. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | STR #1019 still kind of there? |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 13:21 Nov 09, 2005 | The bug http://www.fltk.org/str.php?L1019 was fixed and does not seem to be there anymore in the way it was described in that STR. However, if I: - Open FLUID - Open a widget Properties dialog - Mark the text in the "Label" field - Press Ctrl+c - Close FLUID completely ..the clipboard is empty when I try to paste something into another application. Windows XP Home, Visual Studio .NET | ||
mike 13:23 Nov 09, 2005 | I think this is the intended/expected behavior, since the application supplies the clipboard data... | ||
daniel.polski.vst 03:28 Nov 10, 2005 | Just a note, with other apps (under WinXP) it seems to work to copy data as the first poster describes. (For example copy something from a web browser, close it, then paste the copied data into a text document). | ||
matt 01:42 Dec 01, 2005 | This can be solved, but it is not easz. FLTK handles CLipboard event as thez come in. Once FLTK is gome, the clipboard handler is obviouslz gone as well, so we must make sure that the clipboard is published to world, before the application is closed. | ||
Portale 02:34 Dec 01, 2005 | Ok, in that case I am willing to live without this "fix" in Fltk 1.1.x ;) I was not aware that in addition to the system clipboard, there is an application specific clipboard on Windows or OSX. | ||
mike 17:55 Dec 13, 2005 | We'll revisit this for 2.0... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | None |
Summary: | I'd like Fl_Text_Display to be able to display images. |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 13:05 Nov 10, 2005 | I'd like Fl_Text_Display to be able to display images. No more, no less. | ||
mike 13:10 Nov 10, 2005 | This definitely won't happen in 1.1.x, and I doubt that 2.0. Fl_Help_View can already display images and is available in 1.1.x... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Building FLTK under MacOSX for the X11 environment is broken. |
Version: | 1.1-current |
Created By: | hjohnson |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
hjohnson 08:33 Nov 11, 2005 | fltk-svn.patch 22k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
hjohnson 08:33 Nov 11, 2005 | Two of the three __APPLE__ build types are supported, but X11 builds on __APPLE__ fail because improper default includes are occuring. There is no "code" change involved, it is compilation directive that need enhancement. The most common change is as follows: -# elif defined(__APPLE__) +# elif defined(__APPLE__) && !defined(_APPLE_X11_) A patch between my tree and the main repository is attached. THIS CONTAINS OTHER ENHANCEMENTS TO BUILD SYSTEM. This is the quickest and least amount of code change. A cleaner and more extensible solutions would be to #define new conditionals in Fl.H for each of the desired windowing ABI's. #if defined(WIN_32) #define FLTK_NATIVE_WINDOWS 1 #if !defined(_APPLE_X11_) && defined(__APPLE__) # if defined(__APPLE_QUARTZ__) # define FLTK_APPLE_QUARTZ 1 # elif defined(__APPLE_QD__) # define FLTK_APPLE_QUICKDRAW 1 # else # error: No valid windowing ABI found! # endif #else #define FLTK_X11 1 #endif Regards, Hans J. Johnson hans-johnson@uiowa.edu | ||
mike 06:46 Dec 14, 2005 | Reassigned to 1.1.x, which is where we will make these changes. | ||
matt 07:17 Dec 30, 2005 | Hi Hans, thanks for your patch and the work that you have put into this. I really would like to get the X11 port going. I spent more than a few hours figuring it out (and I believe I have a pretty good overview since I did the Mac port already). Anyway, here's why I can not do this port in this release (lengthy techie stuff follows): FLTK is designed to stay compatible, no matter which version of the header files you used to compile your app (binary compatibility). This means among other things, that if yoou dynamically link with FLTK on Mac, it doesn't matter if FLTK was compiled using Quickdraw or Quartz. And it should not matter if it was compiled for X11 either, it should just link. But here lies the problem. There are a few private structures in the headers that need to know the API. I managed to get away with casting to fix the Carbon/Quickdraw differences, but X11 is far too different. I tried unions, but had to give up. So in short, it is not healthy for our code to introduce the changes required to allow universal X11 compatibility on Mac. If you throw away binary compatibility, you are fine, but that would be limited to your private copy of FLTK. Sorry, Matthias | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | unable to do standard output to console under windows |
Version: | 1.1-current |
Created By: | yalding |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
yalding 22:47 Nov 14, 2005 | testfltk.cc 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
yalding 22:47 Nov 14, 2005 | Hi there: I am using fltk 1.1.x-r4633 under windows mingw32. I could not make standard output work, even I put std::out in the first line of main() function: int main (){ std::cout<<"HelloWorld"<<std::endl; SimpleWindow win(300,200,"SimpleWindow"); return Fl::run(); } The entire file is attached. Note that I am just editing an example from http://www3.telus.net/public/robark/#basics In my linux box with fltk1.1.6, it works fine. Though I install 1.1.6 in windows under mingw, it has the same problem. PS: in 1.1.6, after ./configure --prefix=D:/fltk-1.1.6 and make, it will not generate the include directory althouth its fltk-configure --cflags generate $(fltk-dir)/include. The 1.1.x-rc4633 works fine on this point, I assume it is a fixed bug. regards, yalding | ||
mike 07:47 Nov 15, 2005 | Welcome to the world of Windows. console applications can communicate with the console. WIN32 applications cannot communicate with the console. You need to build your program as a console application. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | wrong include directory |
Version: | 1.1-current |
Created By: | elix |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
elix 06:23 Nov 21, 2005 | In Fl_File_Chooser.H occurs an error because the incorrect include directory (compiled with MSCV 6.0, without additional compiler settings for headers): #include <FL/Fl.H> #include <FL/Fl_Double_Window.H> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <FL/Fl_Group.H> #include <FL/Fl_Choice.H> #include <FL/Fl_Menu_Button.H> #include <FL/Fl_Button.H> #include <FL/Fl_Preferences.H> #include <FL/Fl_Tile.H> #include <FL/Fl_File_Browser.H> #include <FL/Fl_Box.H> #include <FL/Fl_Check_Button.H> #include <FL/Fl_File_Input.H> #include <FL/Fl_Return_Button.H> #include <FL/fl_ask.H> has to changed to #include "Fl.H" #include "Fl_Double_Window.H" #include <stdio.h> #include <stdlib.h> #include <string.h> #include "Fl_Group.H" #include "Fl_Choice.H" #include "Fl_Menu_Button.H" #include "Fl_Button.H" #include "Fl_Preferences.H" #include "Fl_Tile.H" #include "Fl_File_Browser.H" #include "Fl_Box.H" #include "Fl_Check_Button.H" #include "Fl_File_Input.H" #include "Fl_Return_Button.H" #include "fl_ask.H" The same in Fl_File_Input.h: #include <FL/Fl_Input.H> becomes #include "Fl_Input.H" . | ||
Portale 00:33 Nov 22, 2005 | Did you use fltk 2.0 or 1.x? It looks more like 1.x About the include pathes: If you add the main directory of your fltk source (_not_ the "FL" subdirectory) to the settings (Project Settings->C/C++->Category:Preprocessor->Additional include directories), the headers should be found. | ||
mike 05:44 Nov 22, 2005 | As Alessandro pointed out, your include path is wrong. This will not be changed. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Image Support |
Summary: | Fl_RGB_Image::copy return type |
Version: | 1.2-current |
Created By: | schwarzi |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
schwarzi 01:08 Nov 22, 2005 | Fl_RGB_Image_copytype.diff 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
schwarzi 01:08 Nov 22, 2005 | For reasons of type safety I think it would be better if the Fl_RGB_Image::copy() members would return an Fl_RGB_Image pointer instead of Fl_Image*. Internally, the copied image already IS an Fl_RGB_Image. Practical background: For additional code relying on the special proporties of the Fl_RGB_Image->data() it looks advisable if the copy of an existing Fl_RGB_Image had the same type instead of the base class (Fl_Image), wich could also be an Fl_Bitmap, Fl_Pixmap, Fl_Shared_Image, Fl_Tiled_Image with totally different codings contained in its data(). The patch (against fltk-1.1.x-r4647m) attached fixes this. Changes are only in the return type definitions in Fl_RGB_Image::copy() in Fl_Image.{cxx,H}, the whole fltk package compiles fine without further changes. | ||
schwarzi 01:35 Nov 22, 2005 | Sorry, just noted an incompatibility issue: The patch proposed doesn't work on WIN32 using MSVC6, it results in error C2555. Really bad. It does work with gcc (for example, 3.2.3, solaris). Possibly one could workaround this issue by removing the copy method from the Fl_Image class and give all Fl_XXX_Image classes their own copy methods. But this certainly breaks too much existing code. | ||
mike 05:43 Nov 22, 2005 | Rejecting this patch... First, we can't change the API in 1.1.x - that would break binary compatibility. Second, Fl_Image::copy() is virtual and is designed to always return the base class because the classes are interchangeable - the accessor methods allow the caller to determine the type of image data that is stored in the image object. You could argue for a fully opaque implementation which has a virtual data() method that returns data in a specific format, however that too will have to wait for 2.0 and will involve different changes. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Output widgets using frames not being redrawn correctly |
Version: | 1.1-current |
Created By: | srajko |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
srajko 20:35 Nov 25, 2005 | When certain output widgets (I've observed this with Fl_Output and Fl_Value_Output) using frames (for example BORDER_FRAME or PLASTIC_UP_FRAME) get a new value, the new value is drawn over the old value (i.e., the display of the old value is not erased). So, if the old value was "0" and it is changed to "1", the display has a "1" drawn over a "0". The code for Value_Output, for example, does the following: if (damage()&~FL_DAMAGE_CHILD) draw_box(b, color()); else { fl_color(color()); fl_rectf(X, Y, W, H); } ... draw new value So, it seems like it expects draw_box to "erase" the interior of the box, which doesn't happen with at least a few types of frames (they only redraw the border). I didn't know whether the proper change would be to make draw_box redraw/erase the interior even for frames, or to make the output widgets not depend on the interior being erased by draw_box, so no patch - sorry. | ||
mike 18:38 Nov 26, 2005 | You must use a box boxtype, not a frame type, to get the background redrawn properly. This is not a bug and will not be fixed. | ||
END OF ASSIGNMENT |
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | FLTK is incompatible with Alltray |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 11:23 Nov 29, 2005 | When I'm running a FLTK application using Alltray (http://alltray.sourceforge.net) strange things are happening. Menus just don't appear or they're completely misplaced and when I close all windows the application won't quit. Instead it keeps hanging somewhere in the background with only one way to get rid of it (the 'kill' command). Applications using bare X11 or one of the popular toolkits (e.g. Qt or GTK) are acting perfectly fine w/Alltray. Would you be so kind and take a look at this annoying issue? OS: Linux 2.6.12.2, XOrg 6.8.1, KDE 3.4.1/GNOME 2.6 Compiler: GCC 3.3.4 | ||
mike 18:47 Dec 13, 2005 | I just tested sudoku from current 1.1.x with alltray 0.65 and everything seems to work as expected... | ||
Sadysta 01:42 Dec 14, 2005 | Does sudoku have menus? | ||
matt 01:56 Dec 14, 2005 | Yes, it does. How could that be related? Having a menu does not create any additional X11 structures until the menu is popped up. However after handling the menu, they are all destroyed again. I beleive that Alltray is messing big time with Window positions and sze. Some virtual desktops do that and mess up all kinds of applications. Unfortunatly, I am not aware how the different toolkit solve this. I know that there are many ways that virtual desktops work and that they are not always compatible. | ||
mike 04:12 Dec 14, 2005 | AllTray does some monkey business with overloading Xlib functions, so it is certainly possible that it isn't implementing something that FLTK needs, however everything seems to work properly on FC4... | ||
Sadysta 13:33 Dec 16, 2005 | Consider this issue as obsolete. I was trying to use Alltray to put my app into tray on Linux. Luckily I've managed to do it without Alltray. | ||
matt 13:40 Dec 16, 2005 | It is still unfortunate that it doesn't work. Maybe one day, we will find the time... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | OSX 10.4 FLTK multithreading does not work |
Version: | 1.1.6 |
Created By: | veplaini |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
veplaini 02:42 Nov 30, 2005 | I am a Csound developer; Csound uses fltk 1.1.6 for its widget opcodes. I have been struggling with multithreading using FLTK and csound. The problem seems to boil down to Fl::wait() [both versions of the method] never returning when it is called from a different thread to that in which the Fl objects (window etc) have been created. This is very critical in Csound, as in many applications multithreading is required. | ||
mike 06:30 Nov 30, 2005 | Please continue your discussions on the fltk.bugs group - the present discussions seem to indicate that the problem is in your code - only the main thread can call Fl::wait(), as it is the only thread that has access to the event data on all operating systems. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | strange DnD messages (Windows) |
Version: | 1.1-current |
Created By: | xovo |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
xovo 13:11 Dec 02, 2005 | test-051026-dnd.cpp 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
xovo 13:11 Dec 02, 2005 | Hi all, enter/leave messages are strange in combination with DnD mode. Please see the attached program with a window and two widgets A and B. 1) within widget A, press the button - we are entering DnD mode. The widgets gets a LEAVE (even double). Following is a DND_ENTER (one probably has to move the mouse a bit). Now still inside widget A, release the button. We get a DND_LEAVE, but no following ENTER, as one would expect. 2) the same applies to pressing the button inside A, dragging into B and releasing there. We never get a ENTER for B. 3) Pressing the button inside A and dragging to the window outside the widgets. When releasing there, we get spurious DND_LEAVES for both widgets. greetings, Thomas | ||
mike 18:11 Dec 13, 2005 | No FL_ENTER event because Windows doesn't pass us an "enter" message, and DND dropping does not qualify for ENTER events. I doubt there is anything we can do about this for 1.1.x... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | insert() method for Fl_Menu_ would be of much use to me |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | 1.0.x |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 10:02 Dec 10, 2005 | Sometimes I want to insert new item before those which are already in the menu. AFAIK it's currently impossible. | ||
mike 13:53 Dec 10, 2005 | The 1.1.x API is frozen, so we will not be adding any new methods to Fl_Menu_ or Fl_Menu_Item. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Windows shown while popup menu was active disappear when the menu is dismissed. |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Sadysta 13:05 Dec 11, 2005 | test.cxx 1k | ||
Sadysta 13:12 Dec 11, 2005 | bug.GIF 39k | ||
Sadysta 13:17 Dec 11, 2005 | test2.cxx 1k | ||
Sadysta 13:22 Dec 11, 2005 | bug2.PNG 31k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 13:05 Dec 11, 2005 | This is resumption of STR #889. This time I've prepared a short demonstration code and some screenshots to make things 100% clear. | ||
Sadysta 13:17 Dec 11, 2005 | Furthermore, once the second window disappers in the abnormal manner I can't show it anymore, even with a call to show() in a callback. See test2.cxx. | ||
Sadysta 13:22 Dec 11, 2005 | And yet another weirdness. When I use iconize() instead of show() the second window doesn't even make it to the taskbar. Instead it acts like a MDI client window O_O See bug2.png. | ||
Sadysta 13:28 Dec 11, 2005 | If you happen to experiment on the code I've provided please notice that you have to kill the application manually if the second window disappears as in this circumstances Fl::run() won't return even if you close the first window. | ||
Sadysta 13:31 Dec 11, 2005 | I've seen this behaviour multiple times on Windows. I'm not sure right now if it's present on Linux because I use it less often than Windows. I'll check it out ASAP. | ||
alvin 14:51 Dec 11, 2005 | I have tried test.cxx on SUSE 9.3. I notice the following: test.cxx: A window appears with a menu button. When clicked a submenu is shown with entries: One, Two, Three. When I click on One, another window is shown. I close the new window by clicking on the top-right X. When I click on the One menu item a second time, the new window does not reappear. test2.cxx: A window appears same as with test.cxx. Just as in test.cxx, when I click the One menu item a new window appears. However, when I close the window by pressing the top-right X, the window disappears and then immediately reappears. This happens only for the first time I click on the One menu item. All other subsequence attempts behave as expected. The window appears when One is clicked. Then disappears when the top-right X is clicked. The window will not reappear until the One menu item is clicked. Hope this helps, Alvin SUSE 9.3 Pro Kernel 2.6.11.4-21.9 Xorg 6.8.2 | ||
Sadysta 21:57 Dec 11, 2005 | As the matter of fact you were not supposed to click on "One" (especially in test.cxx). You should've opened the popup menu and waited 3 seconds until the second window was shown via timeout trigger and then dismissed the popup menu only to find out that the second window would also disappear. Thanks for testing, anyway :) | ||
alvin 07:54 Dec 12, 2005 | My bad. This time I ran test.cxx (compiled using 1.1.x r4690). I started 'test' and within 3 seconds a 2nd window with the title "Test 2" appeared. I closed the first window (the one with the Menu button). However, "Test 2" remained. The program terminated when I closed "Test 2". With test2.cxx, the "Test 2" window appears within 3 seconds. The only difference with test.cxx is that if I select Menu->One in the first window, the "Test 2" entry on my taskbar flashes. This does not happen with test.cxx. | ||
Sadysta 10:04 Dec 12, 2005 | Ehh... I'm afraid you don't understand. You should not only start the application, but also open the popup menu (i.e. click on menu button exactly once) and then wait for the second window to appear. After it's shown you can dismiss the popup menu or even close the first window. Anyway, the second window will disappear along, despite it shouldn't. | ||
Sadysta 10:14 Dec 12, 2005 | I've just checked the issue on Linux myself and it's not present. So it's possible that you might've done everything perfectly fine and you didn't notice the fault because it simply wasn't there. I'm sorry if this is the case. I just had an impression that you weren't quite following my instructions. | ||
Sadysta 10:15 Dec 12, 2005 | Still, the problem IS PRESENT ON WINDOWS! It's 100% repeatable and really serious! | ||
matt 16:15 Dec 12, 2005 | OK, I found the problem. Posting new windows while a popup menu is up makes FLTK beleive, that the new window is a submenu of the posted menu. This is so because the original author assumed that no callbacks are called during a menu popup phase. Until recently, all other events were actually blocked while a popup was handled, but users did not like that. Well, to make it short, I have no spontaneous fix. Maybe someone else has time to look at this. The offending code is line 1299 in Fl_win32.cxx: } else if (Fl::grab()) parent = fl_xid(Fl::grab()); ^^^^^^^^^^^^^^^ do we really need this? But to not let you hang in the rain. Putting you "show()" code between these lines will at least help as a quick fix: Fl_Window *wg = Fl::grab(); Fl::release(); w -> show(); Fl::grab(wg); | ||
mike 17:05 Dec 13, 2005 | I think we really do need the current code, and the workaround you posted is the only possible fix for FLTK 1.1.x. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | Ctrl+C doesn't work instantly in Fl_Text_Display |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Sadysta 01:03 Dec 13, 2005 | test.cxx 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 01:03 Dec 13, 2005 | OS: Windows XP Compiler: MSVS.NET and Mingw GCC (both give the same results) I remember that once there was a problem with pasting things from external applications to fltk ones, but now the situation is inversed. Sometimes I have to hold Ctrl and press C several times to make the text selected in Fl_Text_Display to get copied to the clipboard. I've attached a really simple demo program. All you need to do is to try to copy some text from Fl_Text_Display to the clipboard. | ||
mike 17:06 Dec 13, 2005 | Cygwin has historically had lots of problems like this; I really don't think this is a FLTK issue but rather a Cygwin issue (with the included WIN32 libraries) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | Overflow for X11 internal 16 bit screen coordinates |
Version: | 1.1-current |
Created By: | thorha |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
thorha 13:37 Dec 20, 2005 | fix_overflow.diff 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
thorha 13:37 Dec 20, 2005 | I have a map application ported from Windows to Linux. When I zoom in, suddenly the screen is covered by a single polygon. I have tracked this down to overflow in conversion from double to COORD_T in the fl_transformed_vertex function in fl_vertex.cxx. The WIN32 coordinates is 32bit and the X11 versjon uses 16bit, this explains the difference. To fix this I have made clamping code in fl_transformed_vertex, see attached diff for details. Regards Thor Henning Amdahl | ||
wilson.afn 10:01 Dec 22, 2005 | Were I flying my Cessna from (0.0, 0.0) to (HUGE_VAL, 32767.0) using one of Thor's maps, I'm not sure I'd want to be directed 45 degrees off course. Earlier, I lodged a similar complaint, but after analyzing the problem it became clear that if FLTK is to keep it's L, it doesn't belong in the clipping business. I think Thor has the same problem problem in WIN32, but it remains hidden until he zooms in a little further and overflows 32 bits (or he tries to display a map using a Windows "thunking" graphics driver that discards half his bits). Note to Thor: I recommend Liang-Barsky, "A new Concept and Method for Line Clipping", ACM Transactions on Graphics, V3(1) January, 1984 pp. 1-22. There is considerable magic involved, and this was the only method I could absorb with my feeble brane. Or OpenGL. | ||
matt 11:23 Dec 28, 2005 | Please use OpenGL for rendering extensive graphics in floating point space. FLTK's drawing functions are meant to create widgets and symbols, but they do not replace a dedicated graphics library. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Unable to conveniently escape @ characters in Fl_Tooltip |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 06:36 Dec 26, 2005 | In Fl_Tooltip.cxx, remove line 223: if (wid==widget_ /*&& x==X && y==Y && w==W && h==H*/ && t==tip) return; It only makes things more complicated. Because of it there is no convenient way to set tooltips using code like below: { static string tt; tt = Replace (contacts[n] -> desc, "@", "@@"); Fl::enter_area (..., tt.c_str()); } because they won't change as .c_str() has constant address in memory most of the time. | ||
Sadysta 06:58 Dec 26, 2005 | Hmm... actually, as I've just checked, simple removal doesn't solve the problem. There are several methods to resolve this issue. The simplest (and probably the best) one involves making strdup() of tooltip and using strcmp (t, tip) instead of t == tip comparsion! | ||
Sadysta 07:05 Dec 26, 2005 | Yet another simple and elegant solution is to append , 0, 0 to line 99 of Fl_Tooltip.cxx (fl_draw() call in Fl_Tooltip::draw() method). I would opt for this one! | ||
matt 05:06 Dec 29, 2005 | Set the tooltip to NULL, then once more to b the string you want to set it to. This does not take much time and fixes your problem. Apart from that, string.c_str() is only available during the call itself, so you do good in using strdup() or something to make your string permanent - in which case you would most likely get a new address anyways. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | fltk 1.1.7 + fc4 x86_64 = lost precision errors |
Version: | 1.1.7 |
Created By: | greg.ercolano |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 11:49 Jan 27, 2006 | 1.1.7 won't build completely on a stock FC4 x86_64; The libs built fine, but the test programs failed with precision loss errors when converting to 'int's. During the default 'make', it stopped here: Compiling cursor.cxx... cursor.cxx: In function ‘void choice_cb(Fl_Widget*, void*)’: cursor.cxx:42: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [cursor.o] Error 1 I fixed the error, but it failed in a different file, same error. I fixed each one by just changing 'int' to 'long' in all cases. Not sure if that's the correct solution, but the errors went away. and I was able to build the whole thing. No other errors or warnings. Here's the complete list of all 5 files that croaked: Compiling cursor.cxx... cursor.cxx: In function ‘void choice_cb(Fl_Widget*, void*)’: cursor.cxx:42: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [cursor.o] Error 1 +++ Compiling input.cxx... input.cxx: In function ‘void color_cb(Fl_Widget*, void*)’: input.cxx:64: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [input.o] Error 1 +++ Compiling keyboard.cxx... keyboard.cxx: In function ‘int main(int, char**)’: keyboard.cxx:108: error: cast from ‘void*’ to ‘int’ loses precision keyboard.cxx:112: error: cast from ‘void*’ to ‘int’ loses precision +++ Compiling line_style.cxx... line_style.cxx: In member function ‘virtual void test_box::draw()’: line_style.cxx:57: error: cast from ‘void*’ to ‘int’ loses precision line_style.cxx:58: error: cast from ‘void*’ to ‘int’ loses precision line_style.cxx:59: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [line_style.o] Error 1 +++ Compiling scroll.cxx... scroll.cxx: In function ‘void type_cb(Fl_Widget*, void*)’: scroll.cxx:74: error: cast from ‘void*’ to ‘int’ loses precision scroll.cxx: In function ‘void align_cb(Fl_Widget*, void*)’: scroll.cxx:90: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [scroll.o] Error 1 | ||
greg.ercolano 18:28 Jan 27, 2006 | Meka[ni] wrote: > You're gcc --version is 4.0.x, right? Correct -- it's the default 4.0.0-8 compiler that comes with FC4/64: [root@superior ~]# g++ --version g++ (GCC) 4.0.0 20050519 (Red Hat 4.0.0-8) [..] > Try adding -fpermissive to CFLAGS/CXXFLAGS. > It solved my problems (while I was using gcc-4.0.1). Yes, that worked for me too; starting with the unmodified release, the following compiles OK from my CSH oriented login: ( setenv CFLAGS '-fpermissive'; setenv CXXFLAGS "-fpermissive"; make ) ..or with a BASH login shell, I imagine this is the equivalent: ( export CFLAGS='-fpermissive'; export CXXFLAGS="-fpermissive"; make ) | ||
ianmacarthur 06:48 Jan 29, 2006 | Greg said: Compiling cursor.cxx... cursor.cxx: In function ‘void choice_cb(Fl_Widget*, void*)’: cursor.cxx:42: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [cursor.o] Error 1 I fixed the error, but it failed in a different file, same error. I fixed each one by just changing 'int' to 'long' in all cases. Not sure if that's the correct solution, but the errors went away. and I was able to build the whole thing. ---- Hmm, I wonder if that is the right thing to do? On a 64 bit platform, a void* will be 64 bits - and on *most* of those a long will also be 64 bits, and all will be well... But on the MS platforms, they have decided that (for backwards compatability reasons) a long will still be 32 bits, and you need to use a long long to get 64 bits. So this fix probably is not portable enough. Or maybe it is? I dunno... -- Ian | ||
matt 19:40 Apr 26, 2006 | The problem by now is well known and well discussed. Unfortunatly we will not be able to fix this issue in FLTK 1.1.x . The "error" can be converted into a "warning" on gcc by setting the approprite compile time flag. However, the risk of pointers larger than 4GB getting damaged remains. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Config Files |
Summary: | fltk-config is not setup correctly when cross compiling |
Version: | 1.1.7 |
Created By: | wrt |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wrt 07:06 Jan 30, 2006 | hi there i'm cross compiling fltk from linux for windows with a mingw cross compiler. most thinks work fine, except setting up fltk-config. As a consequence(?) linking of fluid and the samples fails. This is both not really a problem cause i just copy the fltk-config file from a native windows installation, but it definitly is a bug so i report it here. here's the output from the "cross compiled" version of fltk-config --ldflags: "-L/path/to/fltk/lib -L/usr/X11R6/lib -lfltk -lXext -lX11" however, to create a working windows exe it should be like this: "-L/path/to/fltk/lib -mwindows -mno-cygwin -lfltk -lole32 -luuid -lcomctl32 -lwsock32 -lsupc++" the simplest way would propably be to have a configure option --enable crossmingw or something like that. Of course a cleaner way would be to check the build option given to configure. cu wrt | ||
mike 08:08 Jan 31, 2006 | Since FLTK currently cannot be cross-compiled, we will not support this configuration. You'll need to hand-edit your fltk-config to do what you want... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Closing dialog with esc & pressed buttons |
Version: | 1.1-current |
Created By: | Portale |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Portale 05:44 Feb 05, 2006 | Many_pressed_Buttons.png 15k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Portale 05:44 Feb 05, 2006 | When I open a dialog like the file chooser, press a button and before releasing that button press escape the dialog closes wit hthe button still in the "pressed" state. When I reopen the dialog, the button is still pressed. See the attached screenshot. There are many pressed buttons. | ||
matt 15:19 Mar 30, 2006 | Hihi, looks funny. I am not sure how to fix this without making 'hide()' into an expesive operation. Any suggestions? | ||
matt 17:25 Apr 27, 2006 | RFC: I do not know how to fix this. Earlier versions of fl_throw_focus have sent an FL_LEAVE and FL_UNFOCUS message, but were abandoned for some reason. Sending an FL_RELEASE event seems wrong as well. Disabeling ESC->hide() while the mouse is down is out of the question too. Any suggestions appreciated. | ||
matt 07:59 May 08, 2006 | I have not found a way of fixing this without changing some base behavior of the library. Also, it seems that originally, hiding a window would send FL_LEAVE events which might be used for the purpose of releasing the buttons, but accroding to Bill, this cause other undesired behavior. Since this seems to be a rather rare problem, I'd prefer to leave it the way it is, avoiding new trouble from fixing/breaking what is. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Example Programs |
Summary: | fl_button and browser clipping does not work after resize; Win2000 |
Version: | 1.1.7 |
Created By: | Dima |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Dima 01:09 Feb 13, 2006 | Occurred in test programs. file_chooser.exe (and many others), when resized to a minimal size, the buttons do not clip their captions in the button regions (possibly it is a feature?) fast_slow.exe: resize window to a minimal size - the browser starts to draw over the window region - seems it has no clipping. By the way, no scrollbars are becoming enabled either. | ||
mike 05:11 Feb 13, 2006 | None of these demos use scrollable areas or clipping on the labels (which is enabled only on request)... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Example Programs |
Summary: | color_chooser.exe - does not work correct when run without parameters |
Version: | 1.1.7 |
Created By: | Dima |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Dima 01:11 Feb 13, 2006 | Color_chooser.exe - does not print any help text when run without parameters, it just quits. | ||
mike 05:08 Feb 13, 2006 | Windows applications do not have a console so there is no chance to display any help text. Build with debugging enabled to see the help text. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | WIN32 |
Summary: | fl_scandir don't wildcards on Windows |
Version: | 1.1.7 |
Created By: | srecio |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
srecio 00:08 Feb 15, 2006 | If the name given to fl_scandir contains wildcards, GetFileAttributes fails and the name is considered as a directory ("\*" will be added). So the internal name is incorrect (ex: for "*.txt", "*.txt\*") and the function never finds files entries. Suggested correction: if (!is_dir) { /* this file may still be a directory that we need to list */ DWORD attr = GetFileAttributes(findIn); if ((attr!=0xFFFFFFFF) && (attr&FILE_ATTRIBUTE_DIRECTORY)) strcpy(d, "\\*"); } Best regards, | ||
mike 07:12 Feb 15, 2006 | Wildcards are not supported by fl_scandir(), never have been, never will be... | ||
matt 07:22 Feb 15, 2006 | You can however use fl_filename_match on the list that you get from this call and sort out the files that you are interested in. http://www.fltk.org/documentation.php/doc-1.1/functions.html#fl_filename_match | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Build Files |
Summary: | Recurring error in linkage |
Version: | -current |
Created By: | StoneCypher |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
StoneCypher 01:43 Feb 20, 2006 | After manually adding several libraries to various projects in the IDE tree, the following linker error is the only error I have (except in the OpenGL examples, which is fine, because there's no OpenGL on this machine): HelpView.obj : error LNK2019: unresolved external symbol _newstring referenced in function "public: int __thiscall fltk::HelpView::load(char const *)" (?load@HelpView@fltk@@QAEHPBD@Z) It's not entirely clear where that's coming from, as _newstring doesn't occur anywhere in the solution, but there are several calls to newstring in that function, so it may just be a name mangling issue. The following two bugs may or may not be related, I can't tell: http://www.fltk.org/articles.php?L370+I0+TFAQ+P1+Q http://www.fltk.org/articles.php?L486 Also, I'm not sure whether I'm running 2.0-feature or 2.0-current (there's no indication I can find either way.) This is built against fltk-2.0.x-r4784 on Feb 20, '06. | ||
mikko 11:44 Feb 20, 2006 | What IDE? VC6 or vc.net? Did you remember take clean rebuild? Does "image" test work? It links against fltk_images, where HelpView/HelpDialog is located at the moment | ||
mike 08:29 Feb 02, 2007 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Drawing glitch in Radio Button |
Version: | 1.1-current |
Created By: | Sadysta |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Sadysta 02:28 Mar 03, 2006 | radio.gif 10k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Sadysta 02:28 Mar 03, 2006 | Very similar problem to that described in STR #1133. When I check one of the radio buttons, the other is getting unchecked but its text seems to be drawn twice, resulting in this thickening effect. | ||
fabien 01:17 Mar 23, 2006 | please precise on wich platform, which version you notice that, please send an example with your fluid design if possible. | ||
Sadysta 11:58 Mar 25, 2006 | As the matter of fact, this problem is not FLUID-specific. I observe it in my own application no matter if I create radio buttons with FLUID or manually. This is my design: ------- window -------- | ------ group ------ | | | radio1 radio2 | | | ------------------- | ----------------------- Platforms: So far just Windows XP, using MSVS.NET. Will test it on Linux soon enough though. | ||
Sadysta 09:10 Mar 28, 2006 | It's all the same on Linux. | ||
matt 05:38 Apr 01, 2006 | The code that draws this button is in src/Fl_Light_Button. The code actually behaves correctly. The problem though is, that check boxes generally do not have a box type, so the label gets overdrawn multiple time without redrawing the backgroud. When using an atialieased font, the text appears bolder with every redraw. The only fix that comes to mind is finding the next parent that has a boxtype set and calling "draw" on it before drawing the text. Maybe anyone has a better suggestion? | ||
fabien 08:38 Apr 04, 2006 | good idea, let you the hand on that bug, matt ;) | ||
matt 20:56 Apr 27, 2006 | OK, I have tried out various fixes, bu none of them are satisfying. They are either much too slow or just don't give the expected result. The two workarounds are: 1: don't use antialiased fonts 2: or, set the box type to FLAT_BOX, which of course only works if the background has no pattern (is not "plastic" scheme) | ||
matt 03:13 Aug 17, 2006 | This problem can not be fixed without changing the ABI. Better yet we allow alpha channels etc. . But that is far more than just a "fix". | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | fl_draw.H "contains" macros (variables in CAPITALS) |
Version: | 1.1.6 |
Created By: | loco |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
loco 19:31 Mar 09, 2006 | on line 183 of fl_draw.H, there is: FL_EXPORT void fl_scroll(int X, int Y, int W, int H, int dx, int dy, void (*draw_area)(void*, int,int,int,int), void* data); The problem is that the first four arguments are in CAPITALS which pose a problem for those of us who use macros extensively. I had to hack my defines so that my values for X Y W get undefined before I include "fl_draw.H" otherwise I get all sorts of compiler errors. My particular situation is that I am developing a small app in OpenGL and FLTK for the front end, and use macros for X Y Z W across the program. Not sure how widespread this problem is - I ran into it in this file only however. FLTK Version 1.1.6 (running on Mac OS X 10.4.5) --------- Proposed fix: FL_EXPORT void fl_scroll(int x, int y, int w, int h, int dx, int dy, void (*draw_area)(void*, int,int,int,int), void* data); | ||
mike 05:57 Mar 10, 2006 | Picking single letter (or even two-letter) macro names is going to cause a LOT of problems with libraries other than FLTK. This isn't something we will be fixing, as we need to use X, Y, W, and H in several places to avoid hiding the x, y, w, and h methods for widgets, for example. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 1 - Specific to a machine/printer |
Subsystem: | Core Library |
Summary: | compile error on x86_64 |
Version: | 1.1.7 |
Created By: | wernerflamme |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wernerflamme 02:08 May 15, 2006 | Hi, yesterday I compiled 1.1.7 at home (i586) and everything worked fine. Today I'm in the office (x86_64) and get an error: ---snip--- Compiling cursor.cxx... cursor.cxx: In function ‘void choice_cb(Fl_Widget*, void*)’: cursor.cxx:42: error: cast from ‘void*’ to ‘int’ loses precision make[1]: *** [cursor.o] Error 1 ---pins--- I'm working with SuSE Linux 10.0 on both machines, at home with smp kernel, in the office with default kernel - version 2.6.13-15.8 in both cases. gcc version 4.0.3 20051024 (prerelease) (SUSE Linux). Regards, Werner | ||
matt 03:13 May 20, 2006 | As discussed before, this "bug" can not be fixed without breaking binary compatibility, hence it will not be fixed in FLTK 1.1 . Please download the latest snapshopt of FLTK where this error will only be reported as a warning. This warning will not have any impact on your code unless you specifically compile your application to run on 64 bit machines requesting more than 4GB of space. VERY few applications are currently compiled in that way. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Nested Fl_Scroll buggy |
Version: | 1.1.7 |
Created By: | danh |
Assigned To: | matt |
Fix Version: | Will Not Fix (SVN: v5324) |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
danh 11:15 May 31, 2006 | Nest an Fl_Scroll inside another. Make sure that both regions to scroll so the bars are usable. The following is an example: int main(int argc, char** argv) { Fl_Window window(5*75,400); window.box(FL_NO_BOX); Fl_Scroll scroll(0,0,5*75,400); Fl_Button bn(0,0,5,600); Fl_Scroll inscroll( 40, 40, 100, 100); inscroll.box(FL_FLAT_BOX); Fl_Button sbn(0, 0, 200, 200); sbn.box(FL_FLAT_BOX); inscroll.end(); window.show(argc,argv); return Fl::run(); } Now you scroll the outer one. Now try to scroll the inner one. It seems like the first attempt at manipulating the inner scroll bars does not work because their positions are not up to date. After you have clicked once in the inner scroll region, everything seems ok. | ||
matt 04:30 Jul 18, 2006 | Unfortunatly, the virtual function "position()" was overridden to mean positioning subwidgets inside the scroll, and not positioning the scroll inside its parent. This is kind of logical at first sight, but blocks exactly what is described here: putting a scroll inside a scroll. I am sure I will find a way to patch this without breaking binary compatibility, but I will need a few quiet hours for that. Please be patient. | ||
matt 03:10 Aug 17, 2006 | This bug can not be fixed without changing the ABI and genral interface. Please do not nest Fl_Scrolls. I updated the documentation accordingly. The problem lies within the reuse of the "position" method for other purposes. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | automated safe widget deletion inside the library |
Version: | 1.1-current |
Created By: | Albrecht.Schlosser |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Albrecht.Schlosser 07:23 Jun 02, 2006 | If Fl::check() is called from a callback _after_ one or more call(s) to Fl::delete_widget(), then the scheduled widgets will be deleted immediately. Fl::check() calls Fl::wait(0.0), which in turn calls do_widget_deletion(). IMHO, this is not intended, and this is a bug. Proposal: Fl::check() shouldn't call Fl::wait(0.0), or Fl::wait() should not call do_widget_deletion(), if time_to_wait equals 0 (zero). Albrecht | ||
matt 00:26 Jun 14, 2006 | I have updated the documentation to reflect the problem for now. I plan to use an automated deletion scheme though that will make delete_widget obsolete, which is why I changed the Summary of this bug. | ||
matt 07:04 Aug 17, 2006 | This was not supposed to be here ;-) | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | overlay_rect functionality not working with quartz |
Version: | 1.1.7 |
Created By: | fabien |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
fabien 14:45 Jun 06, 2006 | To notice this bug, execute mandelbrot and select an area. The area selection does not work : the overlay rect is never cleared. This is because on quartz, the pen mode change has no effect. (see fl_overlay.cxx) However, this functionality works fine under Quickdraw. It seems that we should use overlay window instead of pen mode api in quartz. | ||
matt 02:44 Jun 07, 2006 | This is a well known bug with Quartz on OS X which does not support XOR drawing. I have two ideas to fix this. 1: fl_overlay_rect is limited to exactly that, a single rect. So we can actually read the screen coontents before drawing the rect, store it, and write it back at overlay-clear. This will not work if the contents under the rect changes, but neither will the original implementation. 2: we could change the associated window into an overlay window and then... - ah, never mind, too complicated, too many side effects, and, who knows, a user might use an Overlay_Window *and* the overlay_rect. | ||
matt 00:07 Jun 14, 2006 | After sinking another four hours into reading amnuals, mailing lists and messages, I decided to abandon this. It is ossible to read screen pixels back in Quartz and use the data to generate XOR rendering. However, there is no guarantee that the rendered pixels have any significant relation.. (Just imagine a shadow from another window on top of ours - the pixels we read back are obviouls darker than the ones we rendered). So unless someone finds a way to read Pixel Data from a CGContext back buffer, this will not be solved. PS: Reading pixels via OpenGL is IMHO far to involved for this little used and undocumented feature. Reading through the Display Base Address usesolled CPU I/O through the PCI bus, which is at least as involved and incredibly slow. Users should use Fl_Overly_Window to get the sameand better results. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Tabs value() method is not const. |
Version: | 1.1.7 |
Created By: | dj |
Assigned To: | matt |
Fix Version: | None (SVN: v5319) |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dj 12:54 Jul 28, 2006 | The documentation for FLTK 1.1.7 describes the Fl_Tabs value methods like this: Fl_Widget* Fl_Tabs::value() const int Fl_Tabs::value(Fl_Widget*) Gets or sets the currently visible widget/tab. However,the "get" method is defined like: Fl_Widget *value(); Which is missing the const keyword. | ||
matt 01:44 Aug 17, 2006 | Fixed in Subversion repository. | ||
mike 05:54 Aug 17, 2006 | The committed fix is not binary compatible and must be reverted. The documentation should be changed instead to match the implementation. (adding/removing const changes the mangled name...) | ||
matt 23:54 Aug 17, 2006 | I reverted the change and updated the documentation. | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #1405 |
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | Make fails due to missing -L/usr/X11R6/lib |
Version: | 1.1-current |
Created By: | dvdkhlng |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
dvdkhlng 11:14 Sep 04, 2006 | configure.patch.gz 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
dvdkhlng 11:14 Sep 04, 2006 | Compilation fails in the current snapshot 1.1.x-r5385. It did work flawlessly with 1.1.7. Problem seems to be that during shared-object linking no library-paths are passed to the linker command line. I fixed that by adding $(LDFLAGS) to DSOCOMMAND. The attached patch to configure.in fixes the problem for me. I'm not a autoconf export, so maybe that's not the best solution. (but maybe it is) | ||
mike 18:23 Sep 04, 2006 | This patch is not correct - LDFLAGS *cannot* be used when making DSOs on all platforms, since LDFLAGS is for the linker program while DSOFLAGS is for the shared library creation program. They are not always the same. We need to add X_LIBS to DSOFLAGS. | ||
mike 18:26 Sep 04, 2006 | Dupe of STR #1405. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Problem in Fl_Gl_Window::make_current() on Apple Mac OS X (r5329) |
Version: | 1.1-current |
Created By: | strecken |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
strecken 09:19 Sep 13, 2006 | Wrong detection of GL buffer rect. The y0-coord becomes to 0. I changed line153: "if (parent()) {"; inserted in line154: "Rect wrect; GetWindowPortBounds( fl_xid(this), &wrect );" ; and changed line155: "xywh[1] = wrect.bottom - h()- y();". I only used the same mechanism as in Fl_GL_Choise.cxx:fl_set_gl_context,fl_create_gl_context. | ||
matt 04:02 Sep 15, 2006 | Thanks for submitting a patch with your STR! Do you happen to have a short test app where I can verify the fix? | ||
matt 07:00 Sep 17, 2006 | Hmm, I just played around with "test/cube" which has two GL subwindows and added "make_current" to the game. Everything looks normal to me. Please give me a short sample program which exposes the problem. Thanks. | ||
matt 01:26 Sep 23, 2006 | Please post some sample code. | ||
mike 18:04 Oct 10, 2006 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | fltk1.1.7: Fl_Browser: does not handle events if inside class |
Version: | 1.1.7 |
Created By: | jharms |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jharms 01:00 Sep 22, 2006 | i try to make a special drop-down-input-widget. i derived Fl_Group and created Fl_Input, Fl_Button, Fl_Select_Browser. The Browser does not handle any event. what is wrong? attachments: working example jens | ||
matt 01:27 Sep 23, 2006 | Is this issue sufficiently solved? | ||
jharms 06:48 Sep 24, 2006 | my fault. str sufficiently solved! | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | fltk1.1.7: Fl_Browser: does not handle events if inside class |
Version: | 1.1.7 |
Created By: | jharms |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jharms 01:03 Sep 22, 2006 | i try to make a special drop-down-input-widget. i derived Fl_Group and created Fl_Input, Fl_Button, Fl_Select_Browser. The Browser does not handle any event. what is wrong? because last time i got: "unable-to-save-file-to-str": (includes excluded) class BI : public Fl_Group { public: Fl_Input *inp; Fl_Select_Browser *br; Fl_Button *but; static void but_cb(Fl_Widget*w, void* data) { BI *b = (BI*) data; } // Custom resize behavior -- input stretches, menu button doesn't inline int inp_x() { return(x() + Fl::box_dx(box())); } inline int inp_y() { return(y() + Fl::box_dy(box())); } inline int inp_w() { return(w() - Fl::box_dw(box()) - 20); } inline int inp_h() { return(h() - Fl::box_dh(box())); } inline int menu_x() { return(x() + w() - 20 - Fl::box_dx(box())); } inline int menu_y() { return(y() + Fl::box_dy(box())); } inline int menu_w() { return(20); } inline int menu_h() { return(h() - Fl::box_dh(box())); } public: BI(int x,int y,int w,int h,const char*l=0) : Fl_Group(x,y,w,h,l) { Fl_Group::box(FL_DOWN_BOX); align(FL_ALIGN_LEFT); // default like Fl_Input inp = new Fl_Input(inp_x(), inp_y(), inp_w(), inp_h()); inp->box(FL_FLAT_BOX); // cosmetic but = new Fl_Button( menu_x(), menu_y(), menu_w(), menu_h(),"1" ); but->box(FL_FLAT_BOX); // cosmetic but->callback(but_cb,(void*)this); br = new Fl_Select_Browser( inp_x(), inp_y()+inp_h() + 5, inp_w()+menu_w(), inp_h() * 5 ); br->add("one"); br->add("two"); br->select(2); br->type( FL_MULTI_BROWSER ); end(); } }; int main(int argc, char **argv ) { Fl_Double_Window win(720,486); BI *ch = new BI( 140,50,100,30 ); win.show(); Fl::run(); return 0; } | ||
greg.ercolano 01:37 Sep 22, 2006 | Bugs in your code; your browser is not inside the group. Your group's xywh is 140,50,100,30 Your browser xywh is 142,83,96,130 ..the y position of the browser is 3 pixels below the bottom of the group. When you derive a class from a group, you need to be sure all the widgets you create are /within/ the xywh of the group. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Background color of Fl_Choice has changed from white to light grey |
Version: | 1.1-current |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 01:41 Sep 23, 2006 | I've upgraded to the latest snapshot release 5479 and find that the background color of choice menus has suddenly changed to a light grey. They now look like being in inactive state. FLTK: 1_1_x-r5479 Compiler: g++, gcc OS: Windows, OS X | ||
mike 05:22 Sep 23, 2006 | This is intentional. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | seg.fault in fl_measure, configure --enable-xft, ubuntu 6.06 |
Version: | 1.1.7 |
Created By: | jharms |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
jharms 07:34 Sep 26, 2006 | Program received signal SIGSEGV, Segmentation fault. 0x0805324e in fl_height () at fl_font_xft.cxx:166 166 if (current_font) return current_font->ascent + current_font->descent; hope, it's not my fault. if i do fl_font(0,16) first, everything is ok. (configure --enable-xft) int main(int argc, char **argv ) { Fl_Double_Window win(720,486); char *label = "Nachname"; int w,h; ::fl_measure(label,w,h, 0); // segmentation fault printf("%d %d\n", w , h ); win.show(argc,argv); Fl::run(); return 0; } | ||
mike 08:00 Sep 26, 2006 | Since you haven't shown a window yet, none of the font subsystem is initialized. So, yes, it *is* your fault. Just move the fl_measure() after the win.show(). Also, fl_measure expects the width and height variables to be initialized (0 for the width will give you the un-wrapped text width...) | ||
matt 08:22 Sep 26, 2006 | Also, you have not set a font or size yet. There is nothing to measure. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Scroll contents disappear |
Version: | 1.1-current |
Created By: | derek.signalmarketing |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
derek.signalmarketing 12:41 Sep 26, 2006 | scroll.cpp 0k | ||
derek.signalmarketing 14:02 Sep 26, 2006 | scroll2.cpp 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
derek.signalmarketing 12:41 Sep 26, 2006 | The labels on Fl_Groups don't get redrawn on scroll, and don't seem to be included in the Fl_Scroll's size. | ||
greg.ercolano 13:22 Sep 26, 2006 | You're not end()ing your groups: for (i = 0; i < 20; i++) - new Fl_Group(10, 50 * i, 100, 30, "Word"); + { + Fl_Group *g = new Fl_Group(10, 50 * i, 100, 30, Word"); + g->end(); + } The way you had it, each group was getting put 'inside' the other, but outside the xywh extents of the group, causing them all to be clipped and not drawn. | ||
derek.signalmarketing 13:34 Sep 26, 2006 | Sorry, I guess I tried too hard to make my test case "minimal". It still occurs with the g->end(); though... (Also, shouldn't the first group's label still be visible after the scroll if that's the problem?) | ||
derek.signalmarketing 13:40 Sep 26, 2006 | oops, it really is ok with g->end() in place and whether the first one should or should not display in the nested case really doesn't matter to me at all, so I'm outta here. :) | ||
derek.signalmarketing 14:07 Sep 26, 2006 | Sorry, I'm still confused. I'd thought the problem with my real program was entirely caused by the fact that I'd not realized that a widget with X = 0, Y = 0 inside the group will by default have a label that is outside the group, and therefor clipped away by the Fl_Scroll... I still keep losing the label of the topmost Fl_Group in the Fl_Scroll - scroll2.cpp hopefully shows this. Is this a bug, or am I further confused? | ||
greg.ercolano 18:28 Sep 26, 2006 | Devs: suggest this bug be closed and the OP move the new question to fltk.general so that it will be answered there, as these are usage errors and not bugs. Derek, if you can repost your 2nd question to the fltk.general newsgroup, I (or someone) will answer it there, so that others can benefit from the question/response, so it isn't in the bug system. | ||
mike 18:31 Sep 26, 2006 | What Greg said... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | OpenGL |
Summary: | gl_draw() produces incorrect text output |
Version: | 1.1.7 |
Created By: | parshin |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
parshin 04:12 Oct 13, 2006 | OS: Gentoo Linux WM: metacity+XGL, beryl+XGL Symptoms: The attempt to draw text with (for any font, size, or text): gl_font(FL_HELVETICA,14); gl_draw("Test",10,10); gives the output of a set of four rectangles (one per character). The first attempt prints "bad font --adobe-.....-75-75-.....". xfontsel shows no fonts available for 75 dpi. Simple switch to the window manager without XGL brings everything to normal. | ||
mike 05:54 Oct 13, 2006 | If the standard X server works, the XGL server is broken - please report the bug to the XGL folks... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Fl_Window::show() does not use Fl::screen_* API for extents |
Version: | 1.1-current |
Created By: | mike |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mike 06:30 Oct 17, 2006 | [Filed on behalf of Edzard Egberts] The FLTK mechanism to assure that windows are shown in the visible area doesn't consider the enlarged desktop, when using more than one monitors. This appears as a problem, when a window is directly placed on the second screen - the coordinates will be changed to first screen, when show() is called. I patched the problem by overwriting show(): void E_G_Window::show() { int tx= x(), ty= y(); Fl_Double_Window::show(); if ((tx!= x() || ty!= y()) && Fl::screen_count()> 1) for (int i= 0; i< Fl::screen_count(); ++i) { int px, py, pw, ph; Fl::screen_xywh(px, py, pw, ph, i); if ((tx >=px && tx < px+pw) && (ty >=py && ty < py+ph)) { // Something of the window is visible somewhere position(tx, ty); return; } } } | ||
mike 05:29 Oct 29, 2006 | I'm looking at the existing make_xid() code (which handles positioning of the window), and at least the X11 code already uses the screen_xywh() API. What platform are you seeing this on? | ||
mike 16:31 Nov 14, 2006 | Unable to reproduce, code looks OK. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | MacOS |
Summary: | Re-parenting of Fl_Windows under OSX on 1.1.x |
Version: | 1.1-current |
Created By: | ianmacarthur |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
ianmacarthur 12:03 Oct 25, 2006 | test3.cxx 4k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ianmacarthur 12:03 Oct 25, 2006 | I have an application that "re-parents" an Fl_Group filled with widgets *from* an enclosing group in one window *to* another enclosing group in another window. That seems to work well on my three main platforms (linux, OSX and win32). However, I wanted to make the "object" that was being moved around an Fl_Window rather than an Fl_Group, for various reasons, so I substituted Fl_Window for Fl_Group in my test code, and fixed up the origins of the enclosed widgets (windows and groups have different ideas bout origins...) and recompiled. Now, that pretty much seems to work - except on OSX, where the "object" disappears from the first window, but does not re-appear in the second window... Two questions: - Is moving the "object" between windows like that expected to work? - What might I try on OSX to get it working? This is with svn 1.1.x (nearly 1.1.8, I guess!) on various platforms, the problem is observed on OSX 10.4.8... Attached demo program is a cut down test harness that "works" on linux but does not work on OSX - by changing the #define at the top, you can switch between the "group" and "window" behaviours for comparison. -- Ian | ||
matt 12:46 Jan 18, 2007 | Yes, this is absolutely leagal and supposed to work as expected on all platforms. However, OS X has no concept of subwindows, so I had to come up with somwhat of a hack. This hack does not take reparenting into account. You may try to "hide()" the Window before removing it, then reparent it, then call "show" on it again, but I am not sure if that actually fixes it. If not, all you can do is delete the whole Window tree and recreate it in the new location. | ||
ianmacarthur 08:06 Jan 19, 2007 | I had tried all manner of variants on the theme of showing/hiding the window to be reparented, but to no avail. Rather than rebuilding the entire window and its widget hierarchy each time, I have a workaround that keeps all the children in a group, and I create/destroy new windows to parent that group, and move the group (and all its widgets intact) between them. This seems to be working OK, but is a bit more hassle than simply reparenting the window would be! I don't think there is a better fix, so this will do fine. | ||
mike 16:21 Jan 21, 2007 | Based on Ian's feedback, I'm closing this STR without resolution. It should be possible to just reparent a regular group under another window rather than reparenting a window widget... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | problem in compliering the library in VC2005 pro |
Version: | 1.1.4 |
Created By: | Ricky2006 |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Ricky2006 03:25 Nov 07, 2006 | 1>------ Rebuild All started: Project: fltkforms, Configuration: Debug Win32 ------ 2>------ Rebuild All started: Project: fltkimages, Configuration: Debug Win32 ------ 2>Deleting intermediate and output files for project 'fltkimages', configuration 'Debug|Win32' 1>Deleting intermediate and output files for project 'fltkforms', configuration 'Debug|Win32' 2>fltkimages : warning PRJ0009 : Build log could not be opened for writing. 2>Make sure that the file is not open by another process and is not write-protected. 1>Compiling... 2>Compiling... 2>Fl_PNM_Image.cxx 1>forms_timer.cxx 2>..\src\Fl_PNM_Image.cxx(59) : warning C4996: 'fopen' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(234) : see declaration of 'fopen' 2> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_PNM_Image.cxx(133) : warning C4996: 'fscanf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(249) : see declaration of 'fscanf' 2> Message: 'This function or variable may be unsafe. Consider using fscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_PNM_Image.cxx(138) : warning C4996: 'fscanf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(249) : see declaration of 'fscanf' 2> Message: 'This function or variable may be unsafe. Consider using fscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_PNM_Image.cxx(139) : warning C4996: 'fscanf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(249) : see declaration of 'fscanf' 2> Message: 'This function or variable may be unsafe. Consider using fscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_PNM_Image.cxx(140) : warning C4996: 'fscanf' was declared deprecated 1>..\src\forms_timer.cxx(84) : warning C4996: 'sprintf' was declared deprecated 1> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(345) : see declaration of 'sprintf' 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(249) : see declaration of 'fscanf' 2> Message: 'This function or variable may be unsafe. Consider using fscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 1> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 1>..\src\forms_timer.cxx(87) : warning C4996: 'sprintf' was declared deprecated 1> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(345) : see declaration of 'sprintf' 1> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>Fl_PNG_Image.cxx 1>forms_pixmap.cxx 2>Fl_JPEG_Image.cxx 2>fl_images_core.cxx 2>Fl_Help_Dialog.cxx 1>forms_fselect.cxx 2>Fl_GIF_Image.cxx 2>..\src\Fl_GIF_Image.cxx(83) : warning C4996: 'fopen' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(234) : see declaration of 'fopen' 2> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_GIF_Image.cxx(342) : warning C4996: 'sprintf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(345) : see declaration of 'sprintf' 2> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_GIF_Image.cxx(344) : warning C4996: 'strcpy' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(73) : see declaration of 'strcpy' 2> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>Fl_File_Icon2.cxx 2>..\src\Fl_File_Icon2.cxx(132) : warning C4996: 'fopen' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(234) : see declaration of 'fopen' 2> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(135) : warning C4996: 'strerror' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(125) : see declaration of 'strerror' 2> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(307) : warning C4996: 'sscanf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(311) : see declaration of 'sscanf' 2> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(430) : warning C4996: 'sscanf' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(311) : see declaration of 'sscanf' 2> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(515) : warning C4996: 'strnicmp' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(215) : see declaration of 'strnicmp' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strnicmp. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(516) : warning C4996: 'strnicmp' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(215) : see declaration of 'strnicmp' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strnicmp. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(517) : warning C4996: 'strnicmp' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(215) : see declaration of 'strnicmp' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strnicmp. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(663) : warning C4996: 'getenv' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdlib.h(483) : see declaration of 'getenv' 2> Message: 'This function or variable may be unsafe. Consider using _dupenv_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(665) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(667) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(676) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(684) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(695) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(701) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(710) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(731) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(746) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(760) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(767) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(844) : warning C4996: 'fopen' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(234) : see declaration of 'fopen' 2> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(862) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(890) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(899) : warning C4996: 'access' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\io.h(315) : see declaration of 'access' 2> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _access. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(931) : warning C4996: 'strcpy' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(73) : see declaration of 'strcpy' 2> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(932) : warning C4996: 'strcpy' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(73) : see declaration of 'strcpy' 2> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>..\src\Fl_File_Icon2.cxx(937) : warning C4996: 'strcat' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\string.h(78) : see declaration of 'strcat' 2> Message: 'This function or variable may be unsafe. Consider using strcat_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>Fl_BMP_Image.cxx 1>forms_free.cxx 2>..\src\Fl_BMP_Image.cxx(88) : warning C4996: 'fopen' was declared deprecated 2> C:\Program Files\Microsoft Visual Studio 8\VC\include\stdio.h(234) : see declaration of 'fopen' 2> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 2>Generating Code... 1>forms_compatability.cxx 1>forms_bitmap.cxx 1>Generating Code... 1>Creating library... 2>Creating library... 2>Creating browse information file... 1>Creating browse information file... 2>Microsoft Browse Information Maintenance Utility Version 8.00.50727 2>Copyright (C) Microsoft Corporation. All rights reserved. 2>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\Debug\BuildLog.htm" 2>fltkimages - 0 error(s), 36 warning(s) 3>------ Rebuild All started: Project: fltk, Configuration: Debug Win32 ------ 3>Deleting intermediate and output files for project 'fltk', configuration 'Debug|Win32' 3>fltk : error PRJ0008 : Could not delete file 'e:\mylib\fltk-1.1.4\visualc\debug\RSP00010D37443900.rsp'. 3>Make sure that the file is not open by another process and is not write-protected. 3>fltk - 1 error(s), 0 warning(s) 4>------ Rebuild All started: Project: fltkgl, Configuration: Debug Win32 ------ 4>Deleting intermediate and output files for project 'fltkgl', configuration 'Debug|Win32' 4>fltkgl : error PRJ0008 : Could not delete file 'e:\mylib\fltk-1.1.4\visualc\debug\RSP00010D37443900.rsp'. 4>Make sure that the file is not open by another process and is not write-protected. 4>fltkgl - 1 error(s), 0 warning(s) 1>Microsoft Browse Information Maintenance Utility Version 8.00.50727 1>Copyright (C) Microsoft Corporation. All rights reserved. 1>Could not delete temporary file. Make sure that you do not have it open and that the file is not protected. 1> 1>Could not delete temporary file. Make sure that you do not have it open and that the file is not protected. 1> 1>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\Debug\BuildLog.htm" 1>fltkforms - 0 error(s), 2 warning(s) 5>------ Rebuild All started: Project: fltkdll, Configuration: Debug Win32 ------ 5>Deleting intermediate and output files for project 'fltkdll', configuration 'Debug|Win32' 6>------ Rebuild All started: Project: fluid, Configuration: Debug Win32 ------ 6>Deleting intermediate and output files for project 'fluid', configuration 'Debug|Win32' 6>Compiling... 5>Compiling... 6>about_panel.cxx 5>vsnprintf.c 5>e:\mylib\fltk-1.1.4\src\vsnprintf.c(82) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\vsnprintf.c(90) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\vsnprintf.c(95) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>scandir.c 6>align_widget.cxx 5>e:\mylib\fltk-1.1.4\src\scandir_win32.c(50) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\scandir_win32.c(52) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\scandir_win32.c(53) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\scandir_win32.c(68) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\scandir_win32.c(71) : warning C4996: 'strcat' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(78) : see declaration of 'strcat' 5> Message: 'This function or variable may be unsafe. Consider using strcat_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>alignment_panel.cxx 5>numericsort.c 5>flstring.c 5>Generating Code... 5>Compiling... 5>glut_font.cxx 6>code.cxx 5>glut_compatability.cxx 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(58) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(90) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(121) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(264) : warning C4996: 'fopen' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 6> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(270) : warning C4996: 'fopen' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 6> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\code.cxx(335) : warning C4996: 'fopen' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 6> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>factory.cxx 5>gl_start.cxx 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(195) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(230) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(779) : warning C4996: 'strcpy' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 6> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(780) : warning C4996: 'strcat' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(78) : see declaration of 'strcat' 6> Message: 'This function or variable may be unsafe. Consider using strcat_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(819) : warning C4996: 'stricmp' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(213) : see declaration of 'stricmp' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _stricmp. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\factory.cxx(959) : warning C4996: 'stricmp' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(213) : see declaration of 'stricmp' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _stricmp. See online help for details.' 5>gl_draw.cxx 6>file.cxx 5>forms_timer.cxx 5>e:\mylib\fltk-1.1.4\src\forms_timer.cxx(84) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\forms_timer.cxx(87) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>forms_pixmap.cxx 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(44) : warning C4996: 'fopen' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 6> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(136) : warning C4996: 'fopen' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 6> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(410) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(414) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(418) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(422) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(436) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\file.cxx(442) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>Fl_Function_Type.cxx 6>c:\program files\microsoft visual studio 8\vc\include\stdio.h(333) : error C2375: 'fl_vsnprintf' : redefinition; different linkage 6> e:\mylib\fltk-1.1.4\src\flstring.h(78) : see declaration of 'fl_vsnprintf' 6>Fl_Group_Type.cxx 5>forms_fselect.cxx 6>Fl_Menu_Type.cxx 5>forms_free.cxx 5>forms_compatability.cxx 6>Fl_Type.cxx 6>c:\program files\microsoft visual studio 8\vc\include\stdio.h(333) : error C2375: 'fl_vsnprintf' : redefinition; different linkage 6> e:\mylib\fltk-1.1.4\src\flstring.h(78) : see declaration of 'fl_vsnprintf' 6>Fl_Widget_Type.cxx 5>forms_bitmap.cxx 5>Fl_XPM_Image.cxx 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(271) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(310) : warning C4996: 'strcpy' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5>e:\mylib\fltk-1.1.4\src\fl_xpm_image.cxx(57) : warning C4996: 'fopen' was declared deprecated 6> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 5> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(468) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1382) : warning C4996: 'strncpy' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(156) : see declaration of 'strncpy' 6> Message: 'This function or variable may be unsafe. Consider using strncpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>Fl_XBM_Image.cxx 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1801) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1819) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1836) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1844) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1857) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1859) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1861) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1863) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1865) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1877) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1879) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1881) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1934) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_widget_type.cxx(1954) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_xbm_image.cxx(48) : warning C4996: 'fopen' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 5> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_xbm_image.cxx(60) : warning C4996: 'sscanf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 5> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_xbm_image.cxx(90) : warning C4996: 'sscanf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 5> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>Fl_Window_Type.cxx 5>Fl_x.cxx 5>c:\program files\microsoft visual studio 8\vc\include\stdio.h(333) : warning C4273: 'fl_vsnprintf' : inconsistent dll linkage 5> e:\mylib\fltk-1.1.4\src\flstring.h(78) : see previous definition of 'fl_vsnprintf' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(175) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(176) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(177) : warning C4996: 'sprintf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 6> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(779) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(781) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fl_window_type.cxx(785) : warning C4996: 'sscanf' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdio.h(311) : see declaration of 'sscanf' 6> Message: 'This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>fluid.cxx 5>Fl_Wizard.cxx 5>Fl_Window_iconize.cxx 5>Fl_Window_hotspot.cxx 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(100) : warning C4996: 'getcwd' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\direct.h(121) : see declaration of 'getcwd' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _getcwd. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(101) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(103) : warning C4996: 'chdir' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\direct.h(127) : see declaration of 'chdir' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _chdir. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(104) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(110) : warning C4996: 'chdir' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\direct.h(127) : see declaration of 'chdir' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _chdir. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(111) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(150) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(214) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(239) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(298) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(301) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(321) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(324) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(368) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(380) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(395) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(430) : warning C4996: 'getenv' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\stdlib.h(483) : see declaration of 'getenv' 6> Message: 'This function or variable may be unsafe. Consider using _dupenv_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(621) : warning C4996: 'stricmp' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(213) : see declaration of 'stricmp' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _stricmp. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(783) : warning C4996: 'strdup' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 6> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(868) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>e:\mylib\fltk-1.1.4\fluid\fluid.cxx(871) : warning C4996: 'strerror' was declared deprecated 6> c:\program files\microsoft visual studio 8\vc\include\string.h(125) : see declaration of 'strerror' 6> Message: 'This function or variable may be unsafe. Consider using strerror_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 6>Fluid_Image.cxx 6>c:\program files\microsoft visual studio 8\vc\include\stdio.h(333) : error C2375: 'fl_vsnprintf' : redefinition; different linkage 6> e:\mylib\fltk-1.1.4\src\flstring.h(78) : see declaration of 'fl_vsnprintf' 5>Fl_Window_fullscreen.cxx 5>Fl_Window.cxx 6>function_panel.cxx 5>Fl_Widget.cxx 5>Fl_visual.cxx 6>widget_panel.cxx 6>Generating Code... 5>Generating Code... 6>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\fluid__0\BuildLog.htm" 6>fluid - 3 error(s), 65 warning(s) 7>------ Rebuild All started: Project: subwindow, Configuration: Debug Win32 ------ 7>Deleting intermediate and output files for project 'subwindow', configuration 'Debug|Win32' 7>Compiling... 7>subwindow.cxx 7>Compiling manifest to resources... 7>Linking... 5>Compiling... 5>fl_vertex.cxx 7>LINK : fatal error LNK1104: cannot open file 'fltkd.lib' 7>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\subwindow_\BuildLog.htm" 7>subwindow - 1 error(s), 0 warning(s) 8>------ Rebuild All started: Project: shape, Configuration: Debug Win32 ------ 8>Deleting intermediate and output files for project 'shape', configuration 'Debug|Win32' 8>Compiling... 5>Fl_Value_Slider.cxx 8>shape.cxx 5>Fl_Value_Output.cxx 5>Fl_Value_Input.cxx 5>Fl_Valuator.cxx 5>e:\mylib\fltk-1.1.4\src\fl_valuator.cxx(118) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_valuator.cxx(122) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>Fl_Tooltip.cxx 5>Fl_Tiled_Image.cxx 8>Compiling manifest to resources... 8>Linking... 5>Fl_Tile.cxx 8>LINK : fatal error LNK1104: cannot open file 'fltkd.lib' 8>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\shape_\BuildLog.htm" 8>shape - 1 error(s), 0 warning(s) 5>Fl_Text_Editor.cxx 5>Fl_Text_Display.cxx 5>e:\mylib\fltk-1.1.4\src\fl_text_display.cxx(711) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>Fl_Text_Buffer.cxx 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(355) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(474) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(938) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(941) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(944) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(1712) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(1772) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(1872) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(2367) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(2482) : warning C4996: 'fopen' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 5> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_text_buffer.cxx(2498) : warning C4996: 'fopen' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 5> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>Fl_Tabs.cxx 5>fl_symbols.cxx 5>Fl_Slider.cxx 5>Fl_Single_Window.cxx 9>------ Rebuild All started: Project: scroll, Configuration: Debug Win32 ------ 9>Deleting intermediate and output files for project 'scroll', configuration 'Debug|Win32' 5>fl_show_colormap.cxx 9>Compiling... 5>fl_shortcut.cxx 9>scroll.cxx 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(126) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(127) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(128) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(129) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(144) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shortcut.cxx(152) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>Fl_Shared_Image.cxx 9>e:\mylib\fltk-1.1.4\test\scroll.cxx(107) : warning C4996: 'sprintf' was declared deprecated 9> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 9> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 9>e:\mylib\fltk-1.1.4\test\scroll.cxx(108) : warning C4996: 'strdup' was declared deprecated 9> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 9> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shared_image.cxx(107) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shared_image.cxx(224) : warning C4996: 'fopen' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(234) : see declaration of 'fopen' 5> Message: 'This function or variable may be unsafe. Consider using fopen_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shared_image.cxx(286) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_shared_image.cxx(359) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>fl_shadow_box.cxx 9>Compiling manifest to resources... 9>Linking... 5>fl_set_fonts.cxx 9>LINK : fatal error LNK1104: cannot open file 'fltkd.lib' 9>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\scroll_\BuildLog.htm" 9>scroll - 1 error(s), 2 warning(s) 5>e:\mylib\fltk-1.1.4\src\fl_set_fonts_win32.cxx(71) : warning C4996: 'strcpy' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(73) : see declaration of 'strcpy' 5> Message: 'This function or variable may be unsafe. Consider using strcpy_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_set_fonts_win32.cxx(72) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_set_fonts_win32.cxx(74) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_set_fonts_win32.cxx(75) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_set_fonts_win32.cxx(77) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>Generating Code... 10>------ Rebuild All started: Project: resizebox, Configuration: Debug Win32 ------ 10>Deleting intermediate and output files for project 'resizebox', configuration 'Debug|Win32' 10>Compiling... 10>resizebox.cxx 10>Compiling manifest to resources... 10>Linking... 10>LINK : fatal error LNK1104: cannot open file 'fltkd.lib' 10>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\resizebox_\BuildLog.htm" 10>resizebox - 1 error(s), 0 warning(s) 11>------ Rebuild All started: Project: resize, Configuration: Debug Win32 ------ 11>Deleting intermediate and output files for project 'resize', configuration 'Debug|Win32' 11>Create .cxx and .h file with fluidd 11>'..\fluid\fluidd' is not recognized as an internal or external command, 11>operable program or batch file. 11>Compiling... 11>resize.cxx 11>c1xx : fatal error C1083: Cannot open source file: '..\test\resize.cxx': No such file or directory 11>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\resize_\BuildLog.htm" 11>resize - 1 error(s), 0 warning(s) 5>Compiling... 5>fl_set_font.cxx 5>Fl_Scrollbar.cxx 12>------ Rebuild All started: Project: radio, Configuration: Debug Win32 ------ 12>Deleting intermediate and output files for project 'radio', configuration 'Debug|Win32' 5>fl_scroll_area.cxx 12>Create .cxx and .h file with fluidd 12>'..\fluid\fluidd' is not recognized as an internal or external command, 12>operable program or batch file. 12>Compiling... 12>radio.cxx 12>c1xx : fatal error C1083: Cannot open source file: '..\test\radio.cxx': No such file or directory 5>Fl_Scroll.cxx 12>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\radio_\BuildLog.htm" 12>radio - 1 error(s), 0 warning(s) 5>fl_rounded_box.cxx 5>Fl_Round_Button.cxx 5>fl_round_box.cxx 5>Fl_Roller.cxx 5>Fl_Return_Button.cxx 5>Fl_Repeat_Button.cxx 5>fl_rect.cxx 13>------ Rebuild All started: Project: preferences, Configuration: Debug Win32 ------ 13>Deleting intermediate and output files for project 'preferences', configuration 'Debug|Win32' 13>Create .cxx and .h file with fluidd 13>'..\fluid\fluidd' is not recognized as an internal or external command, 13>operable program or batch file. 13>Compiling... 13>preferences.cxx 13>c1xx : fatal error C1083: Cannot open source file: '..\test\preferences.cxx': No such file or directory 5>Fl_Progress.cxx 13>Build log was saved at "file://e:\MyLib\fltk-1.1.4\visualc\preferences_\BuildLog.htm" 13>preferences - 1 error(s), 0 warning(s) 5>Fl_Preferences.cxx 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(220) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(242) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(264) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(336) : warning C4996: 'strdup' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\string.h(205) : see declaration of 'strdup' 5> Message: 'The POSIX name for this item is deprecated. Instead, use the ISO C++ conformant name: _strdup. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(523) : warning C4996: 'sprintf' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\stdio.h(345) : see declaration of 'sprintf' 5> Message: 'This function or variable may be unsafe. Consider using sprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_DEPRECATE. See online help for details.' 5>e:\mylib\fltk-1.1.4\src\fl_preferences.cxx(559) : warning C4996: 'access' was declared deprecated 5> c:\program files\microsoft visual studio 8\vc\include\io.h(315) : see declaration of 'access' 5> Message: 'The POSIX name for this item is deprecated. Inst | ||
mike 05:43 Nov 07, 2006 | We don't support VC2005 with FLTK 1.1.4. Please try the current 1.1.7 or 1.1.x snapshot releases for VC2005 support. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Small FL_GTK_THIN_UP_BOX draws outside its area |
Version: | 1.1-current |
Created By: | greg.ercolano |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
greg.ercolano 21:03 Dec 09, 2006 | gtkboxtest.png 1k | ||
greg.ercolano 21:03 Dec 09, 2006 | foo.cxx 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
greg.ercolano 21:03 Dec 09, 2006 | As w() approaches 0, the FL_GTK_THIN_UP_BOX begins drawing outside of its x/y/w/h bounds. See 2x magnified screenshot and test program attached. When w()=1 and w()=0, the box draws to the left of its x() boundary (the black rectangle). This is particularly noticeable in bar chart type applications based on Fl_Box. | ||
matt 13:13 Jan 18, 2007 | Yes, this is the case with all of the complex drawing box types and can not be avoided without adding a lot of bounds checking to the drawing functions, making them a lot more inefficient. If you need really small bxes, you must use another box type. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | modal windows on WIN32 |
Version: | 1.1-current |
Created By: | wavexx |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
wavexx 11:28 Jan 07, 2007 | test.cc 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
wavexx 11:28 Jan 07, 2007 | Modal windows on windows (verified only on '98) are a bit "funky". Love the fact that you can actually move background windows while a modal window is shown, but, is this feature intended? Anyway, this leads to a consistency bug: if a modal window is shown, you can bring into foreground "zombie" windows. If the UI creates non-modal windows _after_ a modal one, the non-modal window results being the top-most one, while this isn't true on OSX. Since I don't know if this is intended or not, show() should at least check that if other modal windows are shown and don't "overtake" visual space. Also, keyboard events are send "correctly" to the modal window. See attached test. | ||
mike 13:07 Jan 07, 2007 | Your example creates a modal window and then a non-modal window. Normally windows are modal for another "parent" window, typically the application or document window, and raising the modal window will also raise the parent window(s). By creating a modal window without a parent window, you effectively are creating what FLTK calls a non-modal window, but with modal event handling... | ||
Application: | FLTK Library |
---|---|
Duplicate Of: | STR #1475 |
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | Input fields: cursor hiding, may stay hidden |
Version: | 1.1-current |
Created By: | effelsberg |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
effelsberg 03:41 Jan 23, 2007 | When editing text inside a Fl_Input or Fl_Text_Editor the mouse cursor gets hidden. Unfortunately, it also gets hidden if it is placed on a different widget of the window or another FLTK window (with the editor widget still having the focus) and doesn't get restored when moving the mouse over widgets that don't affect the cursor's shape. Moving the cursor out of the window and entering it again doesn't help either. As a fix, the two affected widgets should process the enter and leave events to remember if the cursor is really inside the editable area before hiding it. | ||
ianmacarthur 05:11 Jan 23, 2007 | I think this is perhaps a duplicate of #1475? | ||
effelsberg 06:39 Jan 23, 2007 | Ah, indeed it is, sorry for the duplication. I searched for "cursor", because that's the name of the FLTK function, but the search function didn't came up with the STR (it doesn't search bodies, only subjects), "pointer" would have been a better keyword. Now I realize that "my solution" has already been discussed and rejected because of breaking the ABI. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | problem compiling fltk-1.1.7 with mingw (winxp) |
Version: | 1.1.7 |
Created By: | lupussy |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
lupussy 07:28 Feb 04, 2007 | Trying to compile fltk-1.1.7 with mingw (gcc 3.4.5) on winxp (SP2) I found some errors (e.g. "Compiling Fl_JPEG_Image.cxx... Fl_JPEG_Image.cxx:56:23: jpeglib.h: No such file or directory") in #include-assignments (missing subdirs). After correcting them compiling went fine (:-). Below is the outputfile mods.diff of 'diff' comparing original and corrected dirs: ###################################################### ## file: mods.diff ## ## --- ERRORS in fltk-1.1.7 --- ## ## copy file mods.diff into directory fltk-1.1.7 ## ## apply patch (in directory fltk-1.1.7) with command: ## patch -p1 < mods.diff ####################################################### --- fltk-1.1.7/fluid/fluid.cxx 2005-11-27 09:45:48.000000000 +0100 +++ fltk-1.1.7_WJB/fluid/fluid.cxx 2007-02-03 22:44:57.240747200 +0100 @@ -79,9 +79,9 @@ extern "C" { #if defined(HAVE_LIBPNG) && defined(HAVE_LIBZ) -# include <zlib.h> +# include <zlib/zlib.h> # ifdef HAVE_PNG_H -# include <png.h> +# include <png/png.h> # else # include <libpng/png.h> # endif // HAVE_PNG_H --- fltk-1.1.7/png/png.h 2005-02-24 16:55:12.000000000 +0100 +++ fltk-1.1.7_WJB/png/png.h 2007-02-03 22:42:11.272096000 +0100 @@ -343,7 +343,7 @@ #ifndef PNG_VERSION_INFO_ONLY /* include the compression library's header */ -#include "zlib.h" +#include "zlib/zlib.h" /* include all user configurable info, including optional assembler routines */ #include "pngconf.h" --- fltk-1.1.7/src/Fl_JPEG_Image.cxx 2005-08-29 16:39:02.000000000 +0200 +++ fltk-1.1.7_WJB/src/Fl_JPEG_Image.cxx 2007-02-03 18:54:13.958414400 +0100 @@ -53,7 +53,7 @@ extern "C" { #ifdef HAVE_LIBJPEG -# include <jpeglib.h> +# include <jpeg/jpeglib.h> #endif // HAVE_LIBJPEG } --- fltk-1.1.7/src/Fl_PNG_Image.cxx 2005-04-15 21:13:18.000000000 +0200 +++ fltk-1.1.7_WJB/src/Fl_PNG_Image.cxx 2007-02-03 22:40:36.676073600 +0100 @@ -43,9 +43,9 @@ extern "C" { #if defined(HAVE_LIBPNG) && defined(HAVE_LIBZ) -# include <zlib.h> +# include <zlib/zlib.h> # ifdef HAVE_PNG_H -# include <png.h> +# include <png/png.h> # else # include <libpng/png.h> # endif // HAVE_PNG_H | ||
mike 04:21 Feb 05, 2007 | This fix is incorrect and would break builds on all other environments. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | None |
Summary: | Fl::set_font crashes due to non-initialised table_size |
Version: | 1.1-current |
Created By: | rokan |
Assigned To: | mike |
Fix Version: | 1.1-current |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
rokan 03:59 Feb 09, 2007 | Fl::set_font(Fl_Font fnum, const char* name) crashes sometimes the app as it uses non-initialised static int table_size; Changing that to static int table_size = 0; solves the problem. | ||
rokan 04:03 Feb 09, 2007 | ...the file to modify is obviously fl_set_font.cxx | ||
mike 05:59 Feb 09, 2007 | What platform are you using? Can you provide a short example that demonstrates the problem? Static variables are initialized to 0 by default, so I would doubt this is the issue... | ||
rokan 08:23 Feb 09, 2007 | I might be wrong, but… The thing happened while I was experimenting to port cairo with VS2005 on windows. This problem occurred ONLY for release compilation (running "within" ide) running fonts demo, so i thought this will be initialization of some dynamic memory. To narrow the scope (as the debug info was not available), I was printing some info to a log file and realized that the function was called but never went to "table expansion" loop while (fnum >= table_size)... I was very surprised, but after putting this initialization everything seemed to work fine. It is possible that this is some bug in some initialization of dynamic vars somewhere else (in my modifications) and just mere recompilation changed memory footprint and hided the problem, but the fact that the program went to the Fl::set_font function and not to table expansion code (which I have not changed at all) seems to suggest otherwise - and is strange... You probably can close the STR, I will try to investigate it further but unfortunately it does not crashes any more so I will probably never know what this actualy is... | ||
mike 08:27 Feb 09, 2007 | If you discover a bug in the base FLTK code, please post a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Build Files |
Summary: | configure problem "U32" missing in config.h |
Version: | 1.1-current |
Created By: | Albrecht.Schlosser |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
Albrecht.Schlosser 09:33 Feb 12, 2007 | autoconf_U32.txt 2k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
Albrecht.Schlosser 09:33 Feb 12, 2007 | Version: fltk 1.1-r5691 svn / cygwin. Summary: configure does not define U32 (and U16) in config.h I've seen some more reports in fltk.general regarding this problem. It may be related to a recent upgrade of autoconf, but I'm not sure. Previous versions didn't show this problem. Configure output will be attached in file autoconf_U32.txt. There are some warnings: ": integer expression expected 2" and (4 times) ": integer expression expected 4". I upgraded my cygwin version recently, and I think that the problem arised after upgrading, even without changing the fltk sources. Autoconf version is now 2.61, but I don't know the previous version that worked. IIRC, 2.60 had the problem, too. My linux box with autoconf 2.59d does define U16 and U32: #define U16 unsigned short #define U32 unsigned /* #undef U64 */ but this might also be a system related issue. If someone can help and needs more info, please ask for it. I can't see anything strange in config.log. Albrecht | ||
matt 02:18 Feb 19, 2007 | I updated my Cygwin to check this and now I can't compile anything anymore. Cygwin is driving me nuts and I suggest dropping Cygwin support alltogether! OK, I am kidding, but how in the world do I uninstall and reinstall Cygwin? I could not find an uninstall script. | ||
matt 02:18 Feb 19, 2007 | I updated my Cygwin to check this and now I can't compile anything anymore. Cygwin is driving me nuts and I suggest dropping Cygwin support alltogether! OK, I am kidding, but how in the world do I uninstall and reinstall Cygwin? I could not find an uninstall script. | ||
wavexx 06:45 Feb 19, 2007 | Looks like $ac_cv_sizeof_* cache vars of autoconf might not be set, resulting in a bad test expression. This may be an autoconf regression (like many in the past) or a cygwin issue. Try attaching the result of "sh -x ./configure ..." and config.status if possible. | ||
cable_guy_67 06:53 Feb 19, 2007 | Hey Matt, have you gotten this worked out? The way I uninstall is to right click on c:/cygwin and hit delete. :) The only thing I can think of is to make sure you add cygwin/bin to your path... I just updated my cygwin to a new machine a few days ago and have not had any troubles. Albert is this building with --enable-cygwin or --disable? My Box: AMD Turion64 X2 XP current $ ./fltk-config --version 1.1.8 $ autoconf --version autoconf (GNU Autoconf) 2.61 Copyright (C) 2006 Free Software Foundation, Inc. This is free software. You may redistribute copies of it under the terms of the GNU General Public License <http://www.gnu.org/licenses/gpl.html>. There is NO WARRANTY, to the extent permitted by law. Written by David J. MacKenzie and Akim Demaille. $ gcc --version gcc (GCC) 3.4.4 (cygming special, gdc 0.12, using dmd 0.125) Copyright (C) 2004 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. $ ld --version GNU ld version 2.17.50 20060817 Copyright 2005 Free Software Foundation, Inc. This program is free software; you may redistribute it under the terms of the GNU General Public License. This program has absolutely no warranty. $ svn info Path: . URL: http://svn.easysw.com/public/fltk/fltk/branches/branch-1.1 Repository Root: http://svn.easysw.com/public/fltk/fltk Repository UUID: ea41ed52-d2ee-0310-a9c1-e6b18d33e121 Revision: 5698 Node Kind: directory Schedule: normal Last Changed Author: matt Last Changed Rev: 5697 Last Changed Date: 2007-02-13 09:38:43 -0500 (Tue, 13 Feb 2007) Hope this is of some help. Mark | ||
cable_guy_67 07:03 Feb 19, 2007 | Ahhh, its with --disable-cygwin Albrecht. I see what you are talking about now. :) Compiling fl_draw_image.cxx... In file included from fl_draw_image.cxx:39: fl_draw_image_win32.cxx: In function `void innards(const uchar*, int, int, int, int, int, int, int, void (*)(void*, int, int, int, uchar*), void*)': fl_draw_image_win32.cxx:133: error: `U32' does not name a type fl_draw_image_win32.cxx:134: error: `bmibuffer' undeclared (first use this function) fl_draw_image_win32.cxx:134: error: (Each undeclared identifier is reported only once for each funct ion it appears in.) fl_draw_image_win32.cxx:169: error: expected primary-expression before "static" fl_draw_image_win32.cxx:169: error: expected `;' before "static" fl_draw_image_win32.cxx:178: error: `buffer' undeclared (first use this function) fl_draw_image_win32.cxx:180: error: `U32' has not been declared fl_draw_image_win32.cxx:183: error: expected primary-expression before "static" fl_draw_image_win32.cxx:183: error: expected `;' before "static" fl_draw_image_win32.cxx:188: error: `line_buffer' undeclared (first use this function) fl_draw_image_win32.cxx:190: error: `U32' has not been declared make[1]: *** [fl_draw_image.o] Error 1 make: *** [all] Error 1 Guess I haven't built that way in a while... Mark | ||
Albrecht.Schlosser 07:22 Feb 19, 2007 | matt wrote: > I updated my Cygwin to check this and now I can't compile anything anymore. Cygwin is driving me nuts ... Sorry, if this happened to you. What changes did you make, I might try to check it here (post a diff, please). - wavexx wrote: > Try attaching the result of "sh -x ./configure ..." and config.status if possible. Please have a look at my attached file autoconf_U32.txt. This is not the full output, though. I cut as much as seemed appropriate to show the problem, but of course I can also attach the full output and config.status, if this would help. Right now, I would prefer to wait, if Matt can find something with what he has done so far ... - cable_guy_67 wrote: > is this building with --enable-cygwin or --disable? It's all the default (i.e. --disable-cygwin). I just tried it with --enable-cygwin, and that seems to "do the right thing": #define U16 unsigned short #define U32 unsigned But I need native windows exe files, using cygwin (the dll) is not what I want and can do. And, BTW, it's _only_ the (annoying) configure problem. After adding U32 by hand, everything works fine. Thanks to all who replied so far ... Albrecht | ||
matt 07:24 Feb 21, 2007 | Note to self: if you expect Cygwin to work, do not install the Atmel Microcontroller developer environment, or Cygwin will try to translate FLTK for Atmel. Sigh. | ||
matt 08:20 Feb 21, 2007 | OK, I got Cygwin running. The expression: if test $ac_cv_sizeof_int -eq 4; then fails with: : integer expression expected 4 This is because $ac_cv_sizeof_int is "4\r", not just "4". I don't know enough about autoconf to know how to fix this. | ||
mike 10:04 Feb 21, 2007 | Please report this bug to the Cygwin folks - "\r" is a CR, so they aren't treating end-of-lines properly in their tools... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 4 - High, e.g. key functionality not working |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | incomprehensible text strings with kdesu |
Version: | 1.1.7 |
Created By: | blackmoon |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
blackmoon 15:23 Feb 15, 2007 | If i run a fltk app with "kdesu" (command for run graphical app with root privileges under linux with kde), i'm unable to see te text string in the button label, i see only °°°°°°°°. It work fine if i run it with "sudo" command or "gksu" command on the same sysyem. | ||
greg.ercolano 08:41 Feb 16, 2007 | This might not be an FLTK issue; check to see if kdesu is setting some environment variable that is somehow messing up fltk's ability to access fonts. You might try adding a line like: system("printenv > /var/tmp/foo"); in your FLTK program, and compare the contents of /var/tmp/foo when invoked from kdesu vs. sudo, to see if there's any critical variables being set (LOCALE, etc) that is causing the trouble, and if so, possibly modify your app to look for this problem and prevent it. I'm not sure, just chiming in.. | ||
blackmoon 05:19 Feb 18, 2007 | Hi, i've add the system("printenv > /tmp/env.txt"); line to my app and this is the differences that i've got if i run it with sudo or with kdesu. I don't know if this can help... ------------ sudo printenv ------------ KDE_FULL_SESSION=true HOME=/home/testing _=/usr/bin/sudo SUDO_COMMAND=./test_app ---------------------------------------- ------------ kdesu printenv ------------ KDESU_USER=testing HOME=/root _=/usr/bin/kdesu SUDO_COMMAND=/usr/bin/kdesu_stub - ---------------------------------------- | ||
mike 05:35 Feb 18, 2007 | kdesu resets the home directory to ~root, which might be picking up different fonts, etc. | ||
greg.ercolano 08:37 Feb 18, 2007 | Follow through by doing a test, where you make a script that sets those SUDO variables before invoking your app. Then tell kdesu to invoke the script instead of your app; that way the variables will be set before your app starts running. See if it makes a difference. An example script to do this might look like: ---- snip #!/bin/csh -f setenv KDE_FULL_SESSION true setenv HOME /home/testing setenv _ /usr/bin/sudo setenv SUDO_COMMAND $0 /path/to/test_app $*:q ---- snip | ||
matt 02:16 Feb 19, 2007 | Blackmoon, please verify if setting the correct environment variables for X11 to find the fonts will fix this problem. Thanks. | ||
blackmoon 06:27 Feb 19, 2007 | I've verified the font paths in xorg and seem to be ok. I've also tried with a live distro, but with the same result. | ||
mike 03:58 Feb 22, 2007 | I'm not seeing a problem with kdesu on Fedora Core 6 (which has KDE 3.5.6) - what distro and version of KDE are you using? (FWIW, this is almost certainly a kdesu bug...) | ||
blackmoon 01:52 Feb 23, 2007 | I'm using kubuntu edgy (6.10), which has KDE 3.5.5, on amd64 machine. | ||
matt 05:07 Feb 25, 2007 | Googleing kubuntu, I seem hundreds of messages about kdesu problems. Did you apply the required post-install patches that fix the "sudo"vs. "su" issue? Did you try to launch FLTK apps directly from the root account? Did they have the correct fonts? Can you start the kcontrol as root and set fonts for the root account correctly there? | ||
blackmoon 08:53 Feb 26, 2007 | The post-install patch should be already installad in this version of kubuntu. If i run the application from the root shell with "kdesu" (sudo -s; kdesu ./my-app), i can see the correct fonts in the button label. How can i fix this issue, for run the application without use the root shell, but only the kdesu command from normal user shell? This is a kubuntu bug? | ||
matt 13:24 May 01, 2007 | Please contact your Linux distributor for this issue. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | X11 |
Summary: | Unexpected FL_KEYUP events appear when a key is pressed for a long time |
Version: | 1.1.7 |
Created By: | 142857 |
Assigned To: | matt |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
142857 09:26 Feb 25, 2007 | keyup.cc 1k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
142857 09:26 Feb 25, 2007 | When I hold down a key for some seconds, a series of FL_KEYDOWN events are produced. The bug is that among all those FL_KEYDOWN, at some time appears an isolated FL_KEYUP event or more. The time and rate of FL_KEYUPs is not constant, but after ~ 300 events, there has been normally one. Try the example. You can see the problem if it appears something like this: ... 191: FL_KEYDOWN 192: FL_KEYDOWN 193: FL_KEYDOWN 194: FL_KEYDOWN 195: FL_KEYDOWN 196: FL_KEYDOWN 197: FL_KEYUP 198: FL_KEYDOWN 199: FL_KEYDOWN 200: FL_KEYDOWN 201: FL_KEYDOWN 202: FL_KEYDOWN 203: FL_KEYDOWN 204: FL_KEYDOWN 205: FL_KEYDOWN ... To have a bit more of fun while you test the bug, try the synthesizer "zynaddsubfx": select an instrument which sounds interesting, and hold a key (with the keyboard on your computer) for some seconds. It restarts randomly. However, with a mouse click on the key, the sound is stable. fltk 1.1.7 on Debian; g++ 4.1.2 | ||
matt 13:30 Feb 27, 2007 | To quote the source code of the keyboard event handling system: // Stupid X sends fake key-up events when a repeating key is held // down, probably due to some back compatability problem. Fortunately // we can detect this because the repeating KeyPress event is in // the queue, get it and execute it instead: Well, in some cases, the KeyPress event is not in the queue yet, so a few KeyUp's slip through the net. This is a problem of the implementation of X Servers. I have to suggestions for fixes on your side. They are to specific to fix in FLTK: 1: try calling Fl::key_state() whenever you get a FL_KEY_UP event. The state may still show the key as pressed or 2: whenever you receive a FL_KEY_UP, set a very short timeout, then wait if you get another FL_KEY_DOWN on the same key. It should come pretty much immediatly, so calling Fl::check() may already be enough. Matthias | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | X11 |
Summary: | x11: fullscreen() ignores panels positioned at the upper screen border |
Version: | 1.1-current |
Created By: | fsausm |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
Name/Time/Date | Filename/Size | ||
---|---|---|---|
fsausm 10:10 Jun 06, 2007 | Fl_Window_fullscreen.patch 0k | ||
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
fsausm 10:10 Jun 06, 2007 | Fix attached. | ||
spitzak 11:55 Jun 06, 2007 | Rats we may have to add two methods. fullscreen actually was intended to go over panels and everything and give you all the pixels on the screen. Another function that does this may be needed. | ||
fsausm 03:52 Jun 07, 2007 | fullscreen() should do what it's name suggests: simply bring a window to fullscreen using a ICCCM-compliant way (like e.g. GTK does). This indeed may break older window managers. I also really don't see a point in having two fs methods (like in 2.0) as there is only one way to do it right and the current behaviour can be obtained by a combination of calls to resize() and border() anyway. Maybe this should be discussed on the lists (I know it has been from time to time). For the current implementation, there are some more issues, at least in my environment (XFCE-4.4): - calling fullscreen on a maximized window does nothing but remove the window's border - fullscreen_off() leaves the window without border (so I have to enclose the fullscreen calls with hide() and show()) I didn't file STRs for these as they are somehow a result of the current design. Enough ranting for now as it doesn't belong here anyway :-) Regards, Freddy | ||
matt 04:31 Jun 07, 2007 | Um, yeah, I think there are a few possible interpretations of fullscreen. MSWindows can make windows fullscreen, but they keep the title bar and don't overlap system toolboxes. Mac OS X has a similar mode which changes the graphics coordinates, but not the decoration at all. And then there is what I would call the gamescreen, a window that absolutely covers evrything on a single screen, even deactivates a dynamic system menu. But if that is the case, there is no reason to stick with the system screen resolution either, especilly for games that are not vector graphics oriented. Either way, I see this as a feature and would prefer to push this to 1.2. Just think of the risk of going "game screen" with an application that is not prepared to let the user change to other apps without using the system menu... . | ||
matt 04:33 Jun 07, 2007 | PS: I have no problem whatsoever with the first patch of course. An evn better patch would use the current screen and maximize to that screen instead of the main screen. PPS: Maximized, btw., is what I called fullscreen in the posting above. | ||
fsausm 05:54 Jun 07, 2007 | By "first patch" do you mean the one I attached to the STR? If so then this one is absolutely needed (at least on my environment, because I have a panel at the top of the screen which covers my app when going fullscreen and leaving space of the height of the panel at the bottom). As for fullscreen() I don't think it should mimic a particular behaviour (as it does now) instead it should rely on a platform's native way of doing it, which (almost?) every system today provides. And if there is more than 'one true way' one could provide flags to fullscreen() to adjust it's behaviour similar to wxWidgets' version: <http://www.wxwidgets.org/manuals/stable/wx_wxtoplevelwindow.html#wxtoplevelwindowshowfullscreen> (although they use it mainly to show or hide widgets) One thing I forgot to mention in my last post: is there really a need for setting the window parameters when leaving fullscreen mode? Shouldn't the window's position and size be restored by the window manager? | ||
mike 07:09 Jun 07, 2007 | That's not how fullscreen() works. We actually resize the window, as there is generally no way to ask the window manager to maximize the window (which is a completely different semantic than "fullscreen"...) Given that fullscreen() is working as advertised, we will not be "fixing" this in FLTK. If you want to maximize the window, use: window->resize(Fl::x(), Fl::y(), Fl::w(), Fl::h()); | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | FLUID |
Summary: | FLUID sets group attributes for no reason |
Version: | 1.1-current |
Created By: | StanS |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
StanS 11:00 Jul 25, 2007 | If one creates a Group in FLUID, the generated code looks like this: { Fl_Group* o = new Fl_Group(25, 25, 15, 15); o->end(); } // Fl_Group* o But if the group is asserted to be an instance of a class, it looks like this instead: { MyGroup* o = new MyGroup(25, 25, 15, 15); o->box(FL_NO_BOX); o->color(FL_BACKGROUND_COLOR); o->selection_color(FL_BACKGROUND_COLOR); o->labeltype(FL_NORMAL_LABEL); o->labelfont(0); o->labelsize(14); o->labelcolor(FL_FOREGROUND_COLOR); o->align(FL_ALIGN_TOP); o->when(FL_WHEN_RELEASE); o->end(); } // MyGroup* o This is exactly backwards. It is when one creates a custom widget derived (as is often the case) from Fl_Group, that one is more likely to want to take control of the default values of these attributes. | ||
matt 03:25 Oct 07, 2007 | Well, we are in trouble here. FLUID sets attributes only when they differ from the default value. For custom groups, we don't know what the default values are, so we set all of them, just to be sure. This decision was made quite a while ago to fix related issues. But I can see how this is in the way when your derived widget's purpose is to *override* exactly those defaults. Do you have a suggestion how to fix this without changing existing projects? It seems we need yet another button in Fluid (which I would like to avoid, but that is probably the only solution). | ||
StanS 05:26 Oct 09, 2007 | I wish I know what the "related issues" are. It's hard to see how an existing project would suffer if the code to set attributes to the same values they would have without any code at all were eliminated. However, taking what you say at face value, I suppose one could supply fluid with the desired default values for a custom group. Via a data file perhaps. | ||
mike 08:24 Nov 20, 2007 | There is no way for us to know what the subclass assumes, so we have to set everything - otherwise you'll get undefined behavior. Will not fix. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | menu::add while in use |
Version: | 1.1-current |
Created By: | danh |
Assigned To: | matt |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
danh 11:27 Aug 15, 2007 | In my project, i'm adding a bunch of items to the menu. Sometimes, if items are added while the menu is open, the menu items start redrawing incorrectly, and some menu items become unusable. This usually leads to a crash. This may be related to the internal array that gets reallocated on demand, but I dont know. Here's a stack trace (from 1.1.7, VS2005): >fl_measure(const char * str=0xfeeefeee, int & w=0, int & h=0, int draw_symbols=1) Line 275 + 0x9 bytes >fl_normal_measure(const Fl_Label * o=0x0012f570, int & W=0, int & H=0) Line 52 + 0x15 bytes >Fl_Label::measure(int & W=0, int & H=0) Line 97 + 0xf bytes >Fl_Menu_Item::measure(int * hp=0x00000000, const Fl_Menu_ * m=0x003f3398) Line 129 >menuwindow::find_selected(int mx=360, int my=8) Line 442 + 0x10 bytes >menuwindow::handle(int e=3) Line 609 + 0x17 bytes >fl_fix_focus() Line 597 >Fl::grab(Fl_Window * win=0x00000000) Line 102 >Fl::release() Line 255 + 0xa bytes >Fl_Menu_Item::pulldown(int X=286, int Y=536, int W=560, int H=25, const Fl_Menu_Item * initial_item=0x00000000, const Fl_Menu_ * pbutton=0x003f3398, const Fl_Menu_Item * t=0x00000000, int menubar=1) Line 795 >Fl_Menu_Bar::handle(int event=1) Line 63 + 0x3f bytes | ||
danh 12:40 Aug 16, 2007 | Ok, I did a bit more testing. It doesn't have anything to do with the dynamic memory [de]allocation for the menu items. The problem seems to be that the popup menu code can't handle a changing menu. | ||
matt 10:36 Oct 06, 2007 | Changing the menu array of a menu that is currently open is not allowed. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | FLTK, CMake, and VC8 |
Version: | 1.1.7 |
Created By: | mengda |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mengda 20:59 Aug 31, 2007 | Build FLTK with CMake and VC8. When BUILD_SHARED_LIBS is disabled, everythinkg works fine. But if I enable BUILD_SHARED_LIBS, there is linking error when generating fltk.dll. The error message is: 1>------ Build started: Project: fltk, Configuration: Debug Win32 ------ 1>Linking... 1>LINK : warning LNK4031: no subsystem specified; CONSOLE assumed 1>fl_call_main.obj : error LNK2019: unresolved external symbol _main referenced in function _WinMain@16 1>F:\Develop\FLTK\fltk-1.1.7\Build\bin\Debug\fltk.dll : fatal error LNK1120: 1 unresolved externals 1>Build log was saved at "file://f:\Develop\FLTK\fltk-1.1.7\Build\src\fltk.dir\Debug\BuildLog.htm" 1>fltk - 2 error(s), 1 warning(s) | ||
robert_a 05:23 Sep 18, 2007 | I have the same problem. Also subsequently the linker appears to search for fltk.lib rather than fltk.dll ? is this just normal if the latter is not found? I believe this is related: it seemed there was some difficulty in getting the included png library to build if 'BUILD_SHARED_LIB' option was selected, since the shared included zlib could not be found, for some reason it appears to look for the static lib even though the dll has been built. Is this a problem within VS 8 ? or in the CmakeLists.txt? 1>pngrio.c 1>pngget.c 1>Generating Code... 1>Linking... 1>LINK : fatal error LNK1181: cannot open input file 'fltk_zlib.lib' ------------------------------------------------------^^^^^^^^^^^^^^ Shouldn't this use the fltk_zlib.dll? $ find . -name fltk_zlib\* ./bin/debug/fltk_zlib.idb ./bin/debug/fltk_zlib.lib <===== this gets build if BUILD_SHARED_LIB OFF ./bin/debug/fltk_zlib.pdb ./bin/release/fltk_zlib.dll <===== The .dll has been built ./bin/release/fltk_zlib.idb Subsequent errors similar to the mentioned by Original Poster: 4>fl_call_main.c 4>Generating Code... 4>Linking... 4>LINK : warning LNK4031: no subsystem specified; CONSOLE assumed 4>fl_call_main.obj : error LNK2019: unresolved external symbol _main referenced in function _WinMain@16 4>D:\utemp\fltk_ms_build\bin\Release\fltk.dll : fatal error LNK1120: 1 unresolved externals 4>Build log was saved at "file://d:\utemp\fltk_ms_build\src\fltk.dir\Release\BuildLog.htm" 4>fltk - 2 error(s), 112 warning(s) ..... 9>CodeEditor.cxx 9>Generating Code... 9>Linking... 9>LINK : fatal error LNK1181: cannot open input file 'fltk.lib' 9>Build log was saved at "file://d:\utemp\fltk_ms_build\fluid\fluid.dir\Release\BuildLog.htm" 9>fluid - 1 error(s), 110 warning(s) | ||
gga 16:08 Sep 18, 2007 | Last time I tried it, the current CMakeLists.txt in fltk1.1 is indeed incomplete for some build type (don't recall which one - i think shared). That being said, if you build a shared lib on windows (DLL), you end up with both a .DLL and a .LIB file. Albeit .LIB is confusingly ALSO used on windows for static libraries, the content of the files are different when using it for a shared library. When linking your executable, you always link against the .LIB file regardless. When using a DLL, the .lib file which is really just a stub for the location of the DLL functions. The .DLL is only used at runtime, never during linking. | ||
mike 08:29 Nov 20, 2007 | Not to be fixed unless we get an update from the CMake contributors... | ||
mike 09:55 Dec 15, 2007 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Build Files |
Summary: | Makefile generated by ccmake and ./configure are inconsistent |
Version: | 1.1.7 |
Created By: | haxu |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
haxu 09:21 Sep 03, 2007 | I have built FLTK with two different approaches: 1. ./configure and make 2. ccmake and make Using the first approach, it generates a library named 'libfltk_z.a'. However, the second approach will generate a library named 'libfltk_zlib.a'. I think this is due to the Makefile generated by ./configure and ccmake does not agree on the filename: libfltk_z VERSUS libfltk_zlib. | ||
mike 08:22 Nov 20, 2007 | This will not be fixed unless the cmake contributors (Andy Cedilnik and Julien Jomier) provide an update. | ||
mike 08:14 Dec 15, 2007 | This STR has not been updated by the submitter for two or more weeks and has been closed as required by the FLTK Configuration Management Plan. If the issue still requires resolution, please re-submit a new STR. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Build Files |
Summary: | fltk apps on 64 bit Fedora6 crash immediately |
Version: | 1.1-current |
Created By: | joss |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
joss 01:53 Nov 26, 2007 | Actually, fc6 comes with 1.1.8 as default. I encountered exactly same error in my own fltk app, but I then noticed that the system fltk apps had same problem, ie window gets displayed and then application immediately crashes [top of stack trace shows: #0 0x00002b67b986da5f in XftCharIndex () from /usr/lib64/libXft.so.] yum install fluid gdb fluid Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 47724522584512 (LWP 19740)] 0x00002b67b986da5f in XftCharIndex () from /usr/lib64/libXft.so.2 (gdb) (gdb) (gdb) where #0 0x00002b67b986da5f in XftCharIndex () from /usr/lib64/libXft.so.2 #1 0x00002b67b986abc5 in XftDrawString8 () from /usr/lib64/libXft.so.2 #2 0x00002b67b940fe62 in fl_draw () from /usr/lib64/libfltk.so.1.1 #3 0x00002b67b940cdea in fl_draw () from /usr/lib64/libfltk.so.1.1 #4 0x00002b67b940d4f2 in fl_draw () from /usr/lib64/libfltk.so.1.1 #5 0x00002b67b941181e in fl_normal_label () from /usr/lib64/libfltk.so.1.1 #6 0x00002b67b93e3f0c in Fl_Menu_Item::draw () from /usr/lib64/libfltk.so.1.1 #7 0x00002b67b93e6a4c in Fl_Menu_Bar::draw () from /usr/lib64/libfltk.so.1.1 #8 0x00002b67b93d4e78 in Fl_Group::draw_child () from /usr/lib64/libfltk.so.1.1 #9 0x00002b67b93d4f77 in Fl_Group::draw_children () from /usr/lib64/libfltk.so.1.1 #10 0x00002b67b9400ff4 in Fl_Window::draw () from /usr/lib64/libfltk.so.1.1 #11 0x00002b67b93ce8a1 in Fl_Double_Window::flush () from /usr/lib64/libfltk.so.1.1 #12 0x00002b67b93c12b9 in Fl::flush () from /usr/lib64/libfltk.so.1.1 #13 0x00002b67b93c150e in Fl::wait () from /usr/lib64/libfltk.so.1.1 #14 0x00002b67b93c15ed in Fl::wait () from /usr/lib64/libfltk.so.1.1 #15 0x000000000043c295 in toggle_sourceview_cb () #16 0x0000003c4a21da44 in __libc_start_main () from /lib64/libc.so.6 #17 0x0000000000410549 in Fl_Browser_::draw () #18 0x00007ffff1b373c8 in ?? () #19 0x0000000000000000 in ?? () I can provide access to the relevent server if wanted [contact joss@softwarethatworks.co.uk] | ||
mike 07:51 Nov 26, 2007 | 1.1.8 has not been released. Please report this to the Fedora folks, or retry with the current 1.1.x developer snapshot. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Build Files |
Summary: | test/Makefile failure |
Version: | 1.1-current |
Created By: | ajapted |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
ajapted 04:22 Nov 27, 2007 | The blocks, checkers and sudoku targets in the test/Makefile have an erroneous $(INSTALL_BIN) command in their build command list. Perhaps it works for MacOS X, but if fails for other operating systems. This $(INSTALL_BIN) was added sometime between r5479 and r5940. Cheers -- Andrew Apted | ||
mike 07:27 Dec 15, 2007 | It is not an error, and it will work on all operating systems that provide the POSIX shell - basically all build environments using make. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 3 - Moderate, e.g. unable to compile the software |
Scope: | 2 - Specific to an operating system |
Subsystem: | Core Library |
Summary: | segfault at Fl_Widget::type at ../FL/Fl_Widget.H:110 |
Version: | 1.1.7 |
Created By: | mmokrejs |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
mmokrejs 05:01 Feb 05, 2008 | When I execute seaview application build on fltk on amd64 machine without any arguments, its GUI comes up. I use mouse to go for File/Quit to quit the application and the thing dumps core due to the following bug in fltk: (gdb) where #0 0x00002ba73aebe03e in Fl_Widget::type (this=0x3a1) at ../FL/Fl_Widget.H:110 #1 0x00002ba73aebc2ef in Fl_Widget::damage (this=0x6b0bb0, fl=1 '\001', X=6, Y=5, W=46, H=25) at Fl.cxx:1092 #2 0x00002ba73aebc502 in Fl_Widget::damage (this=0x6b0bb0, fl=128 '\200') at Fl.cxx:1078 #3 0x00002ba73aebc582 in Fl_Widget::redraw (this=0x6b0bb0) at Fl.cxx:1035 #4 0x00002ba73aeef4aa in Fl_Menu_Button::popup (this=0x6b0bb0) at Fl_Menu_Button.cxx:61 #5 0x00002ba73aeef5fc in Fl_Menu_Button::handle (this=0x6b0bb0, e=1) at Fl_Menu_Button.cxx:78 #6 0x00002ba73aed9c99 in send (o=0x6b0bb0, event=1) at Fl_Group.cxx:67 #7 0x00002ba73aeda283 in Fl_Group::handle (this=0x68e460, event=1) at Fl_Group.cxx:195 #8 0x00002ba73aed9c99 in send (o=0x68e460, event=1) at Fl_Group.cxx:67 #9 0x00002ba73aeda283 in Fl_Group::handle (this=0x6875b0, event=1) at Fl_Group.cxx:195 #10 0x00002ba73aebca48 in Fl_Window::handle (this=0x6875b0, ev=1) at Fl.cxx:1013 #11 0x00002ba73aebc08c in send (event=1, to=0x6875b0, window=0x6875b0) at Fl.cxx:662 #12 0x00002ba73aebcf8f in Fl::handle (e=1, window=0x6875b0) at Fl.cxx:700 #13 0x00002ba73af1947a in fl_handle (thisevent=@0x7fff6fe44b30) at Fl_x.cxx:972 #14 0x00002ba73af1974c in do_queued_events () at Fl_x.cxx:175 #15 0x00002ba73af197d4 in fd_callback () at Fl_x.cxx:293 #16 0x00002ba73af19c51 in fl_wait (time_to_wait=0.26623200000000002) at Fl_x.cxx:241 #17 0x00002ba73aebde9d in Fl::wait (time_to_wait=0.26623200000000002) at Fl.cxx:349 #18 0x00002ba73aebdf4e in Fl::run () at Fl.cxx:357 #19 0x0000000000418ef3 in main (argc=1, argv=0x7fff6fe45028) at seaview.cxx:378 (gdb) More details are at http://bugs.gentoo.org/show_bug.cgi?id=207107 | ||
mike 06:41 Feb 05, 2008 | General support is not available via the STR form. Please post to the FLTK forums and/or mailing lists for general support. Note that the "this" pointer is bad in your backtrace - more than likely the application is corrupting the heap. | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | end() in Fl_Scroll missing!? |
Version: | 1.1-current |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | None |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 00:21 Feb 14, 2008 | By accident I've found that end() seems to be missing in Fl_Scroll. So far I haven't had any misbehaviour in my applications but if it's a bug then it is probably a serious one. Fl_Scroll::Fl_Scroll(int X,int Y,int W,int H,const char* L) : Fl_Group(X,Y,W,H,L), scrollbar(X+W-Fl::scrollbar_size(),Y, Fl::scrollbar_size(),H-Fl::scrollbar_size()), hscrollbar(X,Y+H-Fl::scrollbar_size(), W-Fl::scrollbar_size(),Fl::scrollbar_size()) { type(BOTH); xposition_ = 0; yposition_ = 0; hscrollbar.type(FL_HORIZONTAL); hscrollbar.callback(hscrollbar_cb); scrollbar.callback(scrollbar_cb); } | ||
mike 05:20 Feb 14, 2008 | It is not a bug - Fl_Scroll is a container widget like Fl_Group, so we don't call end(). If it was a standalone widget then we'd call end()... | ||
Application: | FLTK Library |
---|---|
Status: | 2 - Closed w/o Resolution |
Priority: | 2 - Low, e.g. a documentation error or undocumented side-effect |
Scope: | 3 - Applies to all machines and operating systems |
Subsystem: | Core Library |
Summary: | inconsistency in Fl_Slider event handling |
Version: | 1.1-current |
Created By: | kontakt.andreas-schoemann |
Assigned To: | mike |
Fix Version: | Will Not Fix |
Update Notification: |
Trouble Report Files:
No files
Trouble Report Dialog:
Name/Time/Date | Text | ||
---|---|---|---|
kontakt.andreas-schoemann 01:21 Feb 14, 2008 | I have an Fl_Scroll widget embedded in a custom widget. When I push one of the sliders of one of the scrollbars of the Fl_Scroll then the FL_PUSH is not handled by the Fl_Slider if the slider has 'full size'. Hmmmh, hard to describe, better you look at the code to see what I mean: int Fl_Slider::handle(int event, int X, int Y, int W, int H) { switch (event) { case FL_PUSH: ... S = int(slider_size_*ww+.5); if (S >= ww) return 0; ... } } Here 0 is returned if the slider size is greater or equal to the width/height of the slider. This behaviour is not consistent with the scrollbar knobs which always return 1 when they get pushed. Imho the slider should behave identical. As my custom widget handles Fl_RELEASE events too it is irritating to see the custom widget action triggered when I push the slider of the Fl_Scroll. Looking through the code I think it is save to remove if (S >= ww) return 0; so that the my code now looks like this: int Fl_Slider::handle(int event, int X, int Y, int W, int H) { switch (event) { case FL_PUSH: ... S = int(slider_size_*ww+.5); ... } } | ||
mike 05:21 Feb 14, 2008 | This is completely consistent - the slider isn't using the event, so it returns 0 as required. | ||
END OF ASSIGNMENT |