On Mon, 3 Dec 2007, Tom Horsley wrote:
On Mon, 3 Dec 2007 17:06:16 +0100
Jindrich Makovicka <[EMAIL PROTECTED]> wrote:
I once wrote even simpler example, using only plain X libraries, no GTK:
http://www.mail-archive.com/fvwm@lists.math.uh.edu/msg14479.html
I forget to include a build command in the original post. Something like
gcc -o test -O2 -L/usr/X11R6/lib -lX11 -lXpm -lXext test.c
should do the job. Could you test this code if fvwm still fails? If the
window shows up for one second, it is wrong, if it shows for 6 seconds
(all the time the program runs) it is ok.
It doesn't fail for me, the window stays up the whole time. If I
restore the sleep(1) call, then the window disappears for a second,
then comes back. Whatever is going on with GTK is apparently
somewhat different.
I remember fixing that bug. the issue with gtk windows can however be seen
by removing the first sleep, and replacing XUnmapWindow with
XWithdrawWindow(dpy, win, scr).
That is: it happens when a program requests to be moved to the withdrawn
state before it has been reparented by fvwm.
/Viktor