[FFmpeg-devel] [PATCH 1/2] clarified the usage of ISO 639 language codes

[Stefano Sabatini]

On date Wednesday 2014-01-08 04:19:44 -0500, Helmut Tessarek encoded:
> On 08.01.14 4:02 , Stefano Sabatini wrote:
> > About the example comment, I'm not sure it's a good idea to add since
> > the "language" metadata is interpreted differently from each
> > container, and thus the use of ISO codes could be unrelevant.
> 
> I had a discussion with Michael about this yesterday on IRC.
> 

> I had the impression that I should clarify which ISO codes can actually be
> used, since the ISO 639 specification is quite extensive and it wasn't clear
> which ones to use.
> 

> It seemed to me that these 2 (639-2/T and 639-2/B) are the least common
> denominator that should work for all of them.

I don't think we can generalize how *all* containers interpret the
language metadata, so while this may be true for Matroska is probably
not true for many other containers, and thus the assumption should not
be applied to a generic example like the one in ffmpeg.texi.
   
> But maybe I misunderstood Michael, in which case the patch is irrelevant.
> 
> Can we get input from Michael on this?
-- 
FFmpeg = Fiendish Friendly Magic Puritan Elastic Governor