[FFmpeg-devel] [PATCH] Add WebM to the Matroska demuxer name

Reimar Döffinger Reimar.Doeffinger
Thu Jul 15 23:37:15 CEST 2010


On Thu, Jul 15, 2010 at 11:12:40PM +0200, Reimar D?ffinger wrote:
> On Thu, Jul 15, 2010 at 04:57:40PM -0400, Alex Converse wrote:
> > On Thu, Jul 15, 2010 at 4:30 PM, Reimar D?ffinger
> > <Reimar.Doeffinger at gmx.de> wrote:
> > > On Thu, Jul 15, 2010 at 04:10:29PM -0400, Alex Converse wrote:
> > >> So am I going to wind up burying this patch because we can't agree on
> > >> what the ideal libavformat API should be?
> > >>
> > >> What's making me irritated here is the patch is not being held up
> > >> because it is broken but because people dislike the established
> > >> interfaces.
> > >
> > > It is "broken" in so far as it _will_ break MPlayer, MPlayer
> > > uses the short name to decide whether to accept or reject an
> > > autodetection (since for some formats the lavf demuxers
> > > are not working well enough).
> > > This may not have been the intended use, but it seems likely
> > > to me other applications may have misused it in similar ways,
> > > since I am not aware of any other way to uniquely identify a
> > > _given_ AVInputFormat.
> > 
> > Under the current ABI it is a comma separated list that is allowed to
> > grow (though It grows quite infrequently). It seems like doing it the
> > right way is possible under the current ABI (all you would have to do
> > is compare the first name up until the first comma or NUL). Perhaps
> > you can enlighten us on why mplayer feels it necessary to do it the
> > way it does.
> 
> Probably for the same reason that av_guess_format does that?

Btw. even av_find_input_format is rather questionable in my opinion.
av_find_input_format(format->name) will return nothing for all formats
that contain a ',' in name.



More information about the ffmpeg-devel mailing list