[whatwg] Corrections for examples in section 3.14.10

[Ian Hickson]

On Sat, 29 Dec 2007, Ivo Emanuel Gonçalves wrote:
>
> On the examples:
> 
> "Vorbis audio alone in Ogg container
> 
>     <source src="audio.oga" type="audio/ogg; codecs=vorbis">
> 
> Speex audio alone in Ogg container
> 
>     <source src="audio.oga" type="audio/ogg; codecs=speex">"
> 
> This should be:
> 
> Vorbis audio alone in Ogg container
> 
>     <source src="audio.ogg" type="audio/ogg; codecs=vorbis">
> 
> Speex audio alone in Ogg container
> 
>     <source src="audio.spx" type="audio/ogg; codecs=speex">
> 
> Rationale: While Vorbis and Speex files (in Ogg) may use the .oga
> extension, it is far more common to see them using .ogg and .spx
> respectively.

On Sun, 30 Dec 2007, Silvia Pfeiffer wrote:
> 
> Xiph has decided to make .oga and .ogv the extensions of the future, to 
> avoid the current confusion between .ogg being in use for vorbis and 
> theora files. The use of extensions is mostly to select between 
> applications, so .oga for audio files and .ogv for video files make more 
> sense because they select between audio and video applications. The use 
> of .ogg for vorbis and .spx for speex files is still supported, but I 
> think we want to encourage people to use the new extension scheme. I 
> would thus support to keep the current specification, or at best provide 
> both but specify what is preferred.

On Sun, 30 Dec 2007, Ivo Emanuel Gonçalves wrote:
> On 12/30/07, Ralph Giles <giles at xiph.org> wrote:
> > I don't think it's a good idea to suggest people to start using .oga 
> > for Ogg Vorbis files. The plan where Ogg Theora -> .ogv, Ogg FLAC -> 
> > .oga, Ogg Ghost = .oga, but we keep Ogg Vorbis = .ogg as before makes 
> > more sense to me. I'm not sure what speex should do, and the presense 
> > of skeleton should be a major determiner between the old and new 
> > regimes.
> 
> Exactly.  The proposed corrections make sense.  They are also in line 
> with the I-D submited to the IETF regarding the issue.

On Sun, 30 Dec 2007, Ivo Emanuel Gonçalves wrote:
>
> For those not aware of what issue this thread is discussing, Xiph is 
> implementing a set of file extensions to make sure implementations work 
> as well as possible with different content encapsulated in Ogg.
> 
> Those are:
> .ogx for applications
> .ogv for video
> .oga for audio
> 
> Ogg files should use those file extensions with a notice that there is a 
> requirement (for the first) and a recommendation (for the other two) 
> that those files carry an Ogg Skeleton bitstream.  Ogg Skeleton is an 
> extension of Ogg to make it easier for implementations to demux content 
> inside Ogg containers.
> 
> However, as Ralph Giles mentioned, this is not entirely the case with 
> Vorbis and Speex.  If those codecs go alone in the container without the 
> Skeleton bitstream, which has been the behavior for the past years since 
> their conception they should use their legacy file extensions (.ogg and 
> .spx respectively) to make sure backwards-compatibility is not broken.  
> This is then the desired way to serve those files and why I proposed the 
> corrections in the original post of this thread.
> 
> Nothing prevents files or implementations dealing with Vorbis and Speex 
> to use .oga, as long as the Skeleton recommendation is noticed and .oga 
> is not used as the default extension of those two codecs.
> 
> Any other type of data encapsulated in Ogg, multiplexed or not, should 
> use the new set of extensions.

Ok, I've tried to update the specification to match this. Let me know if I 
got it right.

Incidentally the spec still doesn't have definitions for:

   Dirac video and Vorbis audio in Matroska container
   WMV9 video and WMA 2 audio in ASF container
   WMV8 video and WMA 2 audio in ASF container
   VC-1 video and WMA 10 Pro audio in ASF container
   XviD video and MP3 audio in AVI container
   Motion-JPEG video and uncompressed PCM audio in AVI container

...because I've no idea what the formal definitions for those are. Any 
help (and pointers to formal definitions) would be very welcome.


On Sat, 29 Dec 2007, Ivo Emanuel Gonçalves wrote:
> 
> Right below this section:
> 
> "Flac audio alone in Ogg container
> 
>     <source src="audio.oga" type="audio/ogg; codecs=flac">"
> 
> Should be:
> 
> FLAC audio alone in Ogg container
> 
>     <source src="audio.oga" type="audio/ogg; codecs=flac">
> 
> Rationale: The file extension is correct in the example.  The error is
> in how FLAC is written, as FLAC is an acronym.  May be worth fixing
> for correctness sake.

Fixed.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'