> Thanks for the bug fixes. Nothing like a new Release Candidate to
> bring the bugs out of the woodwork! (Sorry, JC, I guess we should have
> waited until Friday like the schedule said.) I'll be putting them into CVS
Sorry for confusion. I'm revisiting the freeglut code and found lots of strangeness.
I'm working on these now but I don't care these are fixed in freeglut-2.4.0.
1. Reshape callback called twice on X11.
Regression of freeglut_main.c 1.129 by Sven Panne.
also affects freeglut_gamemode.c 1.33
1.129(glutEnterGameMode() patch) should be solved by a patch of Bugs#1007994.
2. [Windows] Position and size of subwindow are incorrect.
window->State.Width = fgState.Size.X;
window->State.Height = fgState.Size.Y;
if( ! fgState.Position.Use )
x = CW_USEDEFAULT;
y = CW_USEDEFAULT;
3. Full screen position is wrong on X11/KDE.
Because adjustment of window decoration
if (x || y)
is done before the window manager changes its position.
x. Menu is accidentally obscured by windows on Windows.
> I was able to reproduce your first problem and have no trouble
> believing that the other two also occur. Since GLUT doesn't allow menus
> under game mode at all, anything "freeglut" allows is a bonus. I would
> classify these problems under the "doc, it hurts when I do _this_" heading.
Sorry. I forgot I had changed One demo program by commenting out
This is not game mode-related.
x. Multiple windows with a menu
You should force "at most one menu on the screen" policy.
Done by John?
x. A bugfix to prevent from invoking motion/passive-motion callbacks when some menu on
any window is active.
derived from Bugs#1051977
Same as x.
x. Callback of glutMenuStatusFunc never called.
There isn't any code for this purpose ever.
I'm not certain of correct patch on Windows.
Done by John?
x. Deactivate menus when non-freeglut windows pop up.
I should port the patch to current CVS.
I found more simple way (using WM_ACTIVE) on Windows.
x. Is menu action on pressing button necessary?
After selecting the entry on the menu, button release event is passed to underlying window.
If you ignore button press event and make only release event a trigger, things will go well.
x. Arguments x/y of fgCheckActiveMenu() ambiguous.
Absolute coordinates on X11 and relative coordinates on Windows.
Though this causes no problem currently, the future is unknown.