How difficult is that?
Fix crash introduced by 1a4bb059

(Feel free to cleanup)

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>

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>

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

Signed-off-by: Laurent Aimar <fenrir@videolan.org>

Those cards are normally accessed with the standard FrameBuffer,
through X11 or svgalib instead. The mga_vid driver is not even in the
kernel (or not anymore).

This would enable third party vout display plugins, and remove the
maintainance hurdle that the wrapper was. As a side effect, legacy
video output plugins will always be probed after new vout display
plugins even if they have a higher priority.

This fixes the plugin names in the user interface. On the other hand,
legacy video output plugins cannot be selected explicitly anymore
(though they still take part in automatic probing).

Close #3105

This should simplify the mms-bitfield issue and avoid the compilation of the dvdread plugin on Windows...