[FFmpeg-cvslog] r21850 - in trunk/libavformat: avi.c avienc.c metadata.c metadata.h mp3.c utils.c

Reimar Döffinger Reimar.Doeffinger
Thu Feb 25 07:53:32 CET 2010

On Wed, Feb 24, 2010 at 09:58:45PM -0800, Mike Melanson wrote:
> M?ns Rullg?rd wrote:
> >Justin Ruggles <justin.ruggles at gmail.com> writes:
> >
> >>James Darnley wrote:
> >>
> >>>On 16 February 2010 17:32, rbultje <subversion at mplayerhq.hu> wrote:
> >>>>Log:
> >>>>Set lavf identification string globally in av_write_header(), rather
> >>>>than inside the muxers. Remove special handling of "encoder" tags from
> >>>>AVI and MP3 muxers.
> >>>What are the chances of getting this reverted?  I ask because my "flac
> >>>tags" patch already writes the LIBAVFORMAT_IDENT string in the
> >>>VorbisComment as the vendor_string.  This means that the encoder tag
> >>>is redundant.  I would also argue that it is wrong because libavcodec
> >>>is the encoder.
> >>From Xiph.org VorbisComment recommendations:
> >>http://wiki.xiph.org/index.php/VorbisComment#ENCODER
> >>
> >>- The vendor string is usually used to store the name and version of the
> >>underlying codec library
> >>- The idea of ENCODER is to store the name of the user-visible
> >>application, for example ffmpeg2theora.
> >
> >I wouldn't trust an organisation that can't distinguish between
> >containers and codecs on this matter.
> >
> >>They also recommend using a URL in the ENCODER tag.
> >
> >Which their own software doesn't do.
> We're all just a bunch of Ogg-hating wankers, says Monty of Xiph:
> http://blog.gingertech.net/2010/02/20/googles-challenges-of-freeing-vp8/#comments

He's definitely right on the Ogg-hating part.
However that a software developer refuses to see that needless dependencies
between already large software (demuxer and decoder) is a rather bad thing
is quite scary to me.

More information about the ffmpeg-cvslog mailing list