[MPlayer-users] ANNOUNCE: X11 fullscreen fixed ?

[Nilmoni Deb]

Arpi,
	U r right about the swscale assert() bug. I tried very hard but
failed to reproduce it. So thats closed now.

But the KDE WM detection and borderless window problem is still there. Its
definitely reproducible (about once in every 6 mplayer runs).

Here's a portion of the output (don't consider this a bug report but just
a small sample) from a good run:

Start playing...
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes
[x11] Detected wm is KDE.
[xv] dx: 0 dy: 0 dw: 1024 dh: 768
[xv-fs] dx: 0 dy: 0 dw: 1024 dh: 768
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes%
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes%
[x11] NET style stay on top ( layer 0 ).   15/ 15   0%  0%  0.0% 0 0 0%


Here's the corresponding output from a bad run:

Start playing...
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes%
*** [vo] Allocating mp_image_t, 352x240x12bpp YUV planar, 126720 bytes%
[x11] Detected wm is KDE.
[x11] Detected wm is KDE.
[xv] dx: 0 dy: 0 dw: 352 dh: 264
[xv-fs] dx: 0 dy: 0 dw: 1024 dh: 768
[xv] dx: 0 dy: 0 dw: 1024 dh: 768 -0.002    5/  5   0%  0%  0.0% 0 0 0%
[xv-fs] dx: 0 dy: 0 dw: 1024 dh: 768
[x11] NET style stay on top ( layer 0 ).   20/ 20   9% 15%  2.4% 0 0 0%


Notice that there all three mp_image_t allocations are _before_ KDE 
detection in the bad case. I am wondering if thats the real issue.
Or is it the value returned by vo_x11_check_events ?

thanks
- Nil