[MPlayer-dev-eng] LIBVO vs LIBVO2 or LIBVO must live!

[Nick Kurshev]

Hello!

I know that disscusing such subjects is not good thing here but I'll try ;)

Well - I guess that libvo2 will never be finished since it's not simplest job.
My issues:
1) interface of new libvo2 stuff seems not be able to do some right things like
encoding to file due it provides surfaces for decoding only.
In this connexion, libvo->draw_slice is more useful function.
2) Those goals which should be reached with libvo2 can be reached with libvo too!
3) It seems that libvo2 should decrease size of mplayer - it can be reached by
implementing of libvolib or similar library.
4) IMHO extending existing libvo stuff is more simplest task and will cause less number
of bugs than full rewriting it.
5) I don't know volunteers who want start libvo2 ;)

My suggestions:
I want extend current libvo stuff for following goals:
1) Direct rendering
(Unfortunately I don't know opensource codecs which provide this feature
but imho it would be easy to hack libavcodec and libmpeg2 for that. I prefer
start with libmpeg2 ;)I know that postprocessing will be unapplicable in this
case - but we can avoid this problem by providing "old classic" libvo interface
in case of PP ).
2) HW equalizing
(Since it expandable thing - it requries implementing something similar On Screen Menu.
Therefore it's not for immediate future - just a interface)
3) Notification of VO driver about movie characteristics before playback.
(Top easiest task - it's need only in case when direct rendering is unreachable
in case of commercail codecs for example)
4) Probably other stuff which will be useful in the future.
5) Implement libvolib at end.

I guess such extension will solve all needs in libvo2.

My initiative:
IF THERE ARE NO OBJECTIONS I'M READY TO START THIS JOB IN THE NEXT DAYS!!!

Best regards! Nick