> > Well same for divx. It's just MPEG4. I don't think this argument is
> > totally valid. Supporting different codec backends is useful if one,
> > say, has better error checking than the other, or has better precision.
> That's correct. The issue with divx is that both libavcodec and libdivx4
> seems to have different problems decoding some divx streams and the M$
> codec is only available on intel. If libavcodec would do flawless divx
> decoding for all streams with the same or better performance as
> libdivx4, libdivx4 should, IMO, be ditched too.
libavcodec has no known div3 decoding bugs now.
Maybe xine has old version in its CVS, or you found some buggy streams not
yet known by libavcodec developers. Report them at ffmpeg-devel list,
i'm sure they will be fixed ASAP.
I've a big collection of strange div3 files, most of them was buggy with
libavcodec or divx4, but now thay all play fine with libavcodec.
Bugs are never fixed unless being reported... you should know that.
A'rpi / Astral & ESP-team
"I don't RTFM? Wow. What's the meaning of this? It's new for me."
-- Martin Baum, a tipical MPlayer user...