See http://forum.videolan.org/viewtopic.php?f=14&t=90169

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

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

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

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

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

It makes sense, especially since the libav/FFmpeg split

Also, __USE_UNIX98 is a libc-internal definition. The build system is
not supposed to define that directly (we have _GNU_SOURCE).

(Note __BUILTIN__ was not used anymore as we have !__PLUGIN__)

I don't this abomination to leak through the stable LibVLC interface.
To link statically, modules should be linked into LibVLC instead.

And then, there is no need to reintroduce the old "builtin" vlc-config
type if all modules are built the same. In prehistoric times, the
distinction was needed because some modules were static and some not.

GCC takes the last -O* option, which defaults to -O0 anyway. With
./configure forcing -O0, there was no way to select a custom
optimization level via external CFLAGS.

Optimizing VLC code for size is probably not such a great idea,
considering that media playback is performance critical. There are
better ways to reduce code size than -Os, e.g. Thumb(2) code on ARM.

And of course, removing unneeded plugins and feature is the best way
to reduce code size.

In many cases, OSS was used when the builder forgot to install the
alsa-lib development headers. This should not be an issue anymore, as
ALSA is now required on Linux by default.

I believe the fallback to OSS from ALSA and PulseAudio is causing more
confusion than it solves problem. Besides, the OSS emulation by ALSA
cannot mix, so OSS is almost completely useless on Linux, unless OSSv4
was specifically installed.

Moreover, Linux distributions have started dropping OSS completely.

This ensures that at least either ALSA or PulseAudio will be compiled
by default (on Linux).

This will be needed if/when we get a PulseAudio input and SD.

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

This restores commit 45b7ce9e.

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

This reverts commit 45b7ce9e.
Nothing wrong with the commit, but I need to revert it to refactor the
deinterlacer.

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

Currently the V4L2 device to be used can only be set via the media URL.
This patch allows the :v4l2-dev option to set the device as well.
Setting the device via the media URL still takes precedence.

This is useful if a program auto-detects the V4L2 device so that the
media URL only needs to contain e.g. the tuner settings.
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>