Memory associated with a FcPattern is freed when the pattern is
destroyed, so copying the font filename needs to be done before
the call to FcPatternDestroy.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Modified-by: Rafaël Carré <funman@videolan.org> (simplify return)
Signed-off-by: Rafaël Carré <funman@videolan.org>

This reverts commit 5fc0ccda.

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

Headers normally present under normal Linux are missing in bionic.
Signed-off-by: Rafaël Carré <funman@videolan.org>

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

The library introduces noise in unclear circumstances. There does not
seem to be any problem on VLC side though. I do not have time and
expertise to troubleshoot this.

The time report callback could only be called from pf_play() so this is
cleaner and less prone to coding mistake.

Also pass the relative drift instead of the absolute play time stamp.
The drift is less subject to scheduling noise, i.e. drift evolves much
slower than current time advances.

Now, the detached window position is restored on relaunch. Furthermore,
this fixes some positioning issue when switching from macosx-background to
detached video view.

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

(regression from 5eeeb50d)

(regression from c9c770e7)

close #5174

refs #5174

refs #5174

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

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

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

Far too many clueless users push the volume far too high[1]. Then they
complain that VLC distorts the sound... Really anything above 100% (0dB)
is prone to saturation. In my experience, loss of quality is noticeable
at +6dB and unbearable at +10dB.

For the references:
In VLC 1.1, the UI was restricted to 200% (+6dB) by default,
while the audio output core allowed up to 400% (+12dB).
PulseAudio recommends +11dB.
Microsoft Windows limits to +0dB.

[1] I actually compared VLC 1.1, VLC 2.0 and WMP with an external
decibel metter. They give the same output at 0dB. At 6dB, both VLC
versions give the same decibel readings and slight distortion. So the
complaints are purely caused by confusion and cluelessness from users.