ogg dev - May 2008 - How Ogg mappings translate into the codecs parameter in Ogg media types

2008/5/28 ogg.k.ogg.k at googlemail.com <ogg.k.ogg.k at
googlemail.com>:
>>> FLAC and CMML for example have version fields.
>>
>> How many other "in-house" codecs have one?  This may be worth
knowing.
>
> ogginfo should tell you. With -v, I get "Version: 0" for Vorbis
streams.
> For Theora, I get 3.2.0.
>
> Which reminds me I've put a comment in liboggz's Theora BOS packet
decoder
> where there's a possible off-by-one that needs someone with knowledge
of the
> bitstream changes to ensure correct decoding. If anyone wants to have a go,
> look in oggz_auto.c, grep for 42.
This line?

  /* TODO: this should check against 42 for the relevant version numbers */
  if (op->bytes < 41) return 0;
> FWIW, though not in-house, Kate also has a version field.
Kate is in the house!

K.

On 28-May-08, at 6:20 PM, Conrad Parker wrote:
>   /* TODO: this should check against 42 for the relevant version  
> numbers */
>   if (op->bytes < 41) return 0;
I gather this means the USE_THEORA_PRE_ALPHA_3_FORMAT #if should  
instead be based on the length of the header packet. I'm not sure why  
you're bothering though, since you're not flagging the granulepos  
base counter change in beta 1, which applies to a lot more content.

  -r

2008/5/29 Ralph Giles <giles at xiph.org>:
> On 28-May-08, at 6:20 PM, Conrad Parker wrote:
>
>>  /* TODO: this should check against 42 for the relevant version numbers
*/
>>  if (op->bytes < 41) return 0;
>
> I gather this means the USE_THEORA_PRE_ALPHA_3_FORMAT #if should instead be
> based on the length of the header packet. I'm not sure why you're
bothering
> though, since you're not flagging the granulepos base counter change in
beta
> 1, which applies to a lot more content.
good point, could you fix that?

Conrad.