[FFmpeg-devel] [PATCH 0/4] User controllable padding

[Michael Niedermayer]

On Wed, Jan 01, 2014 at 01:01:24PM +0100, James Darnley wrote:
> On 2013-12-30 20:16, James Almer wrote:
> > Bump lavf minor in the first patch.
> 
> Oh yes, done.
> 
> > -1 for default value, 0 for no padding. Muxers can and should clip the value if needed.
> > Ideally look up the corresponding limit for both flac and id3v2.
> 
> > You need to keep the default of 8192 if there's no user provided value.
> > Same with the value in the id3v2 patch.
> 

> Why should there be a "default value" value?  Why would different
> formats have different defaults for a feature intended to do the same
> thing on them all?

different formats have different requirements. different
recommanditions, differnt constraints due to compatibility with
various software.
Having per format defaults is the easy solution.
If you want to change the default padding used by some format then
theres the question if that breaks any software or hardware that reads
that format. This would have to be tested ...

but its a bit worse then that, even if for the current formats
theres a common default that would work for all, a format added
in the future might have different requirements, like the padding
being a multiple of 12345

also it might make sense to instead use private options for each
muxer that supports this, as that way it would be clear which
muxers support it and what is the valid range of padding for each
format, as well as the default, its not clear if this or a common
field in AVFormatContext is better

for AVFormatContext fields you need accessors though (for ABI compat),
see MAKE_ACCESSORS()


> 
> I do agree that a muxer will need to clip to a valid range.
> 
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel


-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Why not whip the teacher when the pupil misbehaves? -- Diogenes of Sinope
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20140101/672622b9/attachment.asc>