Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

SUBPICTURE_ALIGN_* must be used instead.

Previous commits showed it wasn't necessary.

Close #3262

German l10n: changed the proposed translation to a more formal one, which is more appropriate in this context.

even when in fullscreen. Close #3602

(Containers with >10 entries will no longer be ignored.)
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

--00148536e60e0cebb80485af170d
Content-Type: text/plain; charset=ISO-8859-1
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

The function used by the Win32 implementation is not available, so just
leave it unimplemented.

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@videolan.org>

It is not well tested yet.

It is implemented for win32 and when sched_getaffinity() is available.

It is cleaner. Also, it seems that ffmpeg-mt correctly handles it
and so will allow using it.

aout_Restart() is quite buggy (IMHO) and restarts the audio output
from the interface thread. Then Qt4 external dialogs are buggy as
they deadlock if called from the interface thread (I cannot find a
direct way to use BlockingQueuedConnection for foreign threads and
DirectConnection for the interface thread).

Until either or both of those bugs are fixed, we cannot use modal
dialog from the audio output.

This reverts commit 110693d5.
This fixes #3599.

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

code inspired from xcb.c

And don't unlockgl if lockgl wasn't successful.