I demand that Eduard Bloch may or may not have written...
> #include <hallo.h>
> * Darren Salt [Wed, Mar 29 2006, 05:32:47PM]:
>> I demand that Richard Thrippleton may or may not have written...
>>> I suspect it's a cross-wm issue, as there's a standard window property
>>> for 'occupy all virtual desktops', but if it helps I'm running icewm.
> It is not reproducible for me with default icewm settings (you can tune it
> to identify gxine and change the settings, of course). I just installed
> gxine, started, no such problems.
:-|
> I will attach my xprop output if you wish to compare.
Relevant differences follow.
With icewm:
initial state:
_NET_WM_STATE(ATOM) =
_WIN_STATE(CARDINAL) = 0, 63
after switching to full-screen:
_NET_WM_STATE(ATOM) = _NET_WM_STATE_ABOVE, _NET_WM_STATE_FULLSCREEN
_WIN_STATE(CARDINAL) = 1, 63
after switching back:
_NET_WM_STATE(ATOM) =
_WIN_STATE(CARDINAL) = 1, 63
With xfwm:
initial state:
_NET_WM_STATE(ATOM) =
_WIN_STATE(CARDINAL) = 0
after switching to full-screen:
_NET_WM_STATE(ATOM) = _NET_WM_STATE_STICKY, _NET_WM_STATE_FULLSCREEN
_WIN_STATE(CARDINAL) = 1
after switching back:
_NET_WM_STATE(ATOM) =
_WIN_STATE(CARDINAL) = 0
>> This is repeatable here. However, gxine is definitely calling
>> gtk_window_unstick() for the window (the same window is used for both
>> display modes).
> Ooookay... more details please. Icewm version, custom configuration, etc.
Current testing/unstable (1.2.25-1).
I've just made sure that I'm using the default configuration. It still
happens.
If you want to fiddle with gxine source, then app_state_cb() in
src/noskin_window.c is relevant. (I would consider a flag which is used to
record whether an event has been received for a stickiness change, but - as
things stand - stickiness would still have to be cleared on leaving
full-screen mode.)
> Btw, the first time I started it and began to play an AVI (freely
> available AVI trailer of Star Wreck, if you want to reproduce it).
[snip]
> The program 'gxine' received an X Window System error.
> This probably reflects a bug in the program.
> The error was 'BadMatch (invalid parameter attributes)'.
> (Details: serial 3092 error_code 8 request_code 138 minor_code 6)
[snip]
This doesn't happen here (0.5.5 or CVS HEAD).
--
| Darren Salt | linux or ds at | nr. Ashington, | Toon
| RISC OS, Linux | youmustbejoking,demon,co,uk | Northumberland | Army
| + Buy less and make it last longer. INDUSTRY CAUSES GLOBAL WARMING.
Mistakes are oft the stepping stones to failure.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
