[MPlayer-users] The future of mplayer

[emaillist+mpu at dogphilosophy.net]

On Wednesday 29 January 2003 08:21 pm, Arpi wrote:
> > On Wed, Jan 29, 2003 at 06:34:07PM -0700, emaillist+mpu at dogphilosophy.net
> > wrote:
> > > P.S. MPlayer MUST quit wasting all that time on silly GUI's and work on
> > > implementing support for the Theora codec!  :-)
> >
> > Does Theora even exist? I thought it was still vaporware... *ducks*
>
> since Theora == VP31 in OGG container, it's already supported (in theory,
> i've never seen any theora files, but both the planned codec & container of
> it is supported so it should work)

Alpha 1 of the Ogg Theora release appeared mysteriously back on September
25th.  There hasn't been any VISIBLE development on it since then, but
there is evidently a fair amount of effort being put into it from behind
the scenes.  Alpha 1 was pretty limited in what the example programs could do,
but what they did they seemed to do well, especially for first-working-release
as-yet-unoptimized code.

The release seems to have included "native" source code for dealing with
the initial "Theora" (VP31, or was it 32?  The Theora-related CVS module
at Xiph was called "vp32" before they created a "theora" directory...) video
codec which, while not yet optimized, did seem to work fine (the "native" 
code, I think, adds a little bit of features beyond the original vp31 -
support for more resolutions and such.)

Though obviously the "must" in my post above was a joke, it would be spiffy
if there were support for encoding video with a native Theora/vp31/32 codec
(in .avi's until enough of a formal specification is available to generate
"Ogg Theora" files directly...)  (For that matter, the ability to encode
from the original vp31 windows .dll wouldn't be too bad, either, but obviously
native would be better...)  I have no idea how difficult incorporating this
code into mencoder (or libavcodec/ffmpeg?) is, so I'm in a perfect position to
insist that it ought to be done and should be very easy. :-)

The FAQ at www.theora.org implies that there will be updates that add
capability to the original VP codec to make "Theora", but that the original
VP31 (/32) would then be a compatible "subset", so in theory one could
take an .avi with vp32 codec video and later re-mux it into a compliant
Ogg Theora file.  I'm under the impression that other than the changes 
existent in the Alpha 1 release code to support more resolutions, etc., there
isn't going to be much of anything else done to it, so I would GUESS 
(possibly erroneously) that if it actually were ported now, that the
final video codec for Ogg Theora would PROBABLY still be compatible.

(And, yes, I AM the same hyperventilating pest who keeps posting to Theora-dev
asking if anything's actually happening on it yet... :-) )