[Ffmpeg-devel] [PATCH] simplify libavcodec/allcodecs.c

Diego Biurrun diego
Tue Nov 28 17:11:05 CET 2006


On Tue, Nov 28, 2006 at 05:00:14PM +0100, Dominik 'Rathann' Mierzejewski wrote:
> On Tuesday, 28 November 2006 at 16:10, Diego Biurrun wrote:
> > On Tue, Nov 28, 2006 at 03:07:12PM +0100, Dominik 'Rathann' Mierzejewski wrote:
> > > On Tuesday, 28 November 2006 at 12:37, Diego Biurrun wrote:
> > > > On Tue, Nov 28, 2006 at 10:14:35AM +0100, Benjamin Zores wrote:
> > > > > >
> > > > > > Same as for allformats, but applied to allcodecs.
> > > > > > Line count divided by 3.
> > > > > > I will apply soon.
> > > > > 
> > > > > +#define REGISTER_ENCDEC(X,x)  REGISTER_ENCODER(X,x); REGISTER_DECODER(X,x)
> > > > > 
> > > > > The name sounds pretty bad at hearing.
> > > > 
> > > > It's not that bad and only used internally.
> > > > 
> > > > > Why not simply REGISTER_CODEC instead (which should mean coder-decoder) ?
> > > > 
> > > > Codec is much too ambiguous IMO.
> > > 
> > > It is not ambiguous. It means exactly what Benjamin said.
> > 
> > I'm fully aware of the textbook definition.  Yet you would say things
> > like "ffplay supports the H.264 codec" without a second thought.  What
> > does codec mean in the context of this example?  Hardly encoder.  That's
> > what I call ambiguous.
> 
> Not in my book. The sentence in question could be rephrased as: "ffplay
> supports video data encoded with the H.264 codec". At least that's the
> meaning I derive from it.

There you have it: Now codec means a compression format, not a coder or
decoder implementation.  I maintain that the term codec is ambiguous.

Diego




More information about the ffmpeg-devel mailing list