On date Monday 2011-04-18 13:06:45 +0200, Anton Khirnov wrote:
> On Mon, 18 Apr 2011 11:52:36 +0200, Stefano Sabatini
> <[email protected]> wrote:
> > On date Monday 2011-04-18 09:02:48 +0200, Anton Khirnov wrote:
> > > They've accumulated enough new APIs and corresponding deprecated cruft.
> > >
> > > This breaks API and ABI.
> >
> > Please wait so that we can fix the few remaining error code issues
> > (alternatively this can be delayed to the unstableness period right
> > after the change, but I prefer to do it before).
>
> I prefer bumping now, especially because of side data in AVPacket being
> ABI incompatible. We can declare the tree unstable for a few weeks, this
> should be enough to get rid of the error code issues
>
> >
> > BTW, libavdevice depends on libavformat, so it should be bumped as
> > well, same for libavfilter but being it (yet) unstable there is no
> > technical need for doing it (but may be a good idea anyway).
>
> I don't see anything deprecated in lavdevice, what would be the point of
> bumping it?
Error code changes, its functions will return different error codes
after the bump.
Also it depends on the libavformat ABI so a compiled application which
works with libavdevice won't work anymore after the bump.
--
Prototype designs always work.
-- Don Vonada
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
