[FFmpeg-cvslog] r21731 - in branches/0.5: . Changelog doc/APIchanges libavcodec/avcodec.h libavcodec/utils.c

[Michael Niedermayer]

On Wed, Feb 10, 2010 at 06:35:20PM +0100, Reinhard Tartler wrote:
> On Mi, Feb 10, 2010 at 16:00:26 (CET), Ramiro Polla wrote:
> 
> >> Do you have one or more specific examples of applications that check the
> >> LIBAVCODEC_VERSION_MINOR setting?
> >
> > vlc.git$ grep -r LIBAVCODEC_VERSION_INT * | wc -l
> > 38
> 
> vlc.git$ grep -r LIBAVCODEC_VERSION_MINOR * | grep 21 | wc -l
> 0
> 
> >> As indicated in my previous post, when considering the relevant upgrade
> >> paths, I don't think this is currenly a problem.
> >>
> >> I agree that changes like this could open pandora's box if used
> >> carelessly, though.
> >
> > I don't think upgrade paths should have any effect on the issue here.
> > Uoti's example shows very well an use that may break.
> 
> Okay, so the argument is that someone might use LIBAVCODEC_VERSION_INT
> to check if the corresponding change in trunk is included or not:
> 
> ------------------------------------------------------------------------
> r17823 | schreter | 2009-03-05 08:35:06 +0100 | 3 lines
> 
> Add handling of frame position in the parser.
> 
> Luckily, this doesn't happen in vlc, but there is some risk that there
> might be other projects that rely on that. Personally, I think taking
> this risk has more benefit for the 0.5 branch than dropping this
> patch. Ideally, there is another way to express this addition and I'm
> open for suggestions.

I think the commits with version A.B.C in the main branch define API & ABI
any fork that uses x.y must be fully ABI&API compatible to main x.y
and this is not about "debian" upgrade paths. its simply like if you claim
html 2 support then you must support it.
You can always support more than x.y did but never less.

Anyway, personally i think the idea of doing any further work on a 1 year
outdated branch is wasted time. I wont stop you from wasting your time if
you like it but i am against setting the version to a value thats larger than
what is actually supported


>                                                                                                                          
> > Diego, how do you suggest we deal with version numbers in the releases?
> 
> CC'ing Diego, he mentioned that he doesn't follow the lists currently...

The ML seems to remove CCs in what gets posted to all
i think thats configureable by each subscriber but i might misremember
its a while since i read the docs

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Complexity theory is the science of finding the exact solution to an
approximation. Benchmarking OTOH is finding an approximation of the exact
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-cvslog/attachments/20100210/c28d62e9/attachment-0001.pgp>