[FFmpeg-devel] [RFC] libvorbis decoder wraper

[Clément Bœsch]

On Sat, May 05, 2012 at 12:26:03PM +0200, Michael Niedermayer wrote:
> On Sat, May 05, 2012 at 12:21:37PM +0200, Clément Bœsch wrote:
> > On Sat, May 05, 2012 at 10:57:25AM +0200, Michael Niedermayer wrote:
> > > This puts the libvorbis wraper back (updated to todays API)
> > > I needed this for the ogg/vorbis timestmp/seeking fixes. Not sure if it
> > > should be applied or left for reference in the ML / a branch
> > > 
> > > comments welcome
> > 
> > What was the reason of dropping it in the first place? Even if the
> > internal decoder is better, I think it makes sense for a lot of people
> > (users and dev) to be able to do some comparison in the implementations in
> > order to track bugs, or compare performances and efficiency. It can also
> > be used as for demonstration purpose on how to use the API (FFmpeg one,
> > and the library one).
> 
> The commit that removed it said this:
> 
> commit f9def9ccc6ecfe1778d4daa62a7ada27b5f79bfc
> Author: Diego Biurrun <diego at biurrun.de>
> Date:   Sun Nov 4 12:55:32 2007 +0000
> 
>     Remove libvorbis Vorbis decoding support.  Our native decoder is complete
>     and has no known bugs, any remaining issues will hopefully be uncovered now.
> 

Yes, but that looks to me like a good reason to keep ours, not drop the
alternative.

The issue might be to reconsider/clarify the purpose of the project; are
the lib* wrapper a workaround, or extra features?

While I don't think we should add more glue to extra libs all over the
project, I still think it's good to have reference implementations
available (see the points in my previous mail) as long as it's not much
code and that they are not the default when our implementation is good
enough.

Anyway, just my opinion/comments, any decision is fine with me.

-- 
Clément B.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 490 bytes
Desc: not available
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20120505/82963695/attachment.asc>