This is introduces a random hack, which is going to be reverted this fall.

extras/tools: use Mans Rullgard's gas preprocessor variant instead of David Conrad's since it is more up-to-date

Signed-off-by: Rafaël Carré <funman@videolan.org>

Otherwise, it might not have a window handle.

With Qt5, the window type must be checked at run-time: the windowing
system is always defined as Q_WS_QPA.

Conversion from FI32 to S16 is already supported by the Laurent's
generic audio format filter. Conversion to FI32 are useless for lack
of (functional) FI32 encoder or output.

None of the real audio outputs support FI32; they already typically
converted down to S16 internally. Moreover, mixing FI32 requires
64-bits (signed) multiplication, which non-FPU processors typically
do not support natively.

TODO:
- somehow implement mute callback
- initially report right volume to ui

VLC needs QtWidgets (which does not exist in Qt4).

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

This fixes a race between requesting mute and the aout actually muting.

There are still at least two known bugs with the mute widget:
 - Qt tries to sets the LibVLC status when LibVLC status changes
   (there is a hack around this for volume but not for mute).
 - Qt tracks the playlist "mute" variable instead of the aout one.

Volume and mute states are now independant. There is no need to
update them together in a single transaction.

Furthermore, other processes can change the volume and/or mute state
of VLC playback streams asynchronously. Thus volume-up/volume-down
and mute-toggle are not atomic operations even when protected by the
volume lock. We would need to have toggle and up/down provided by the
to the output plugins. That is probably impossible and overkill.
So accept the small race condition and simplify the code.