[vorbis] Why is Vorbis development slow?

[John Ripley]

> From: Sebastian Gesemann [mailto:sgeseman at upb.de]
> Sent: 30 September 2003 11:53
> To: vorbis at xiph.org
> Subject: Re: [vorbis] Why is Vorbis development slow?
> 
> On Tue, 30 Sep 2003, engdev wrote:
> 
> > OK, OK, I get the message. I was only 'following up' with 
> my thoughts which
> > were provoked by a user wishing to write his own 
> implementation, only to be
> > told to refer to the code. This is not what he wanted, he 
> wanted to create
> > from 'scratch' if you like, from the documentation.
> > 
> > This was several months back. I never saw him on this list 
> again - I assume
> > he gave up and "copied" the code.
> 
> Might be me...
> 
> Well. It wasn't easy, but I succeded in overcoming the difficulties.
> I had to fill the information gaps in the spec with my common sense.
> (It didn't work for the floor-0 stuff, anyway)
> 
> Wow, and i did'n look at the reference code once.
> So, I'm really kinda proud. :-)
> 
> The specs aren't perfect, but with enough background knowledge about
> other source coding techniques, it's possible to build a vorbis
> decoder without the reference code.
> 
> @ John Diplay :

I assume that's supposed to be me :)

> You didn't implement Floor-0 either, did you ?

It didn't seem worth the effort, seeing as I don't have any floor-0 files
anyway. I notice that Tremor has trouble running real-time on embedded
(portable) players when it encounters floor-0 files. I should think it's
possible to get it much faster, but the algorithm itself doesn't exactly
lend itself to being efficient.

I had to look at the reference code for the residue packet decode. The
pseudocode was far too mangled to make head or tail of back then, and
there's a lot of "ok" edge cases to handle (premature end-of-packet being
"ok" really complicates matters). I think it has been corrected now.

I'm now quite tempted to properly finish off my decoder and use it instead
of Tremor in the portables I work on, if I could only work out the rights
nightmare that would ensue...

- John Ripley.
--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.