On Tue, Sep 08, 2015 at 04:29:48PM +0200, Clément Bœsch wrote: > On Tue, Sep 08, 2015 at 03:00:25PM +0100, Derek Buitenhuis wrote: > > On 9/8/2015 11:52 AM, Alex M wrote: > > > Could somebody please confirm that this issue is still considered > > > outstanding and that a working patch has not been submitted? In which > > > case > > > I’ll start work on this asap. > > > > Working this cleanly into the API is no small feat. Clément was working > > on this recently, and had even done some patches. I don't know what happened > > to them (they're somewhere in the mailing list archives). There was a lot > > of discussion on them. > > > > Yes, here is the last one IIRC: > > https://ffmpeg.org/pipermail/ffmpeg-devel/2015-January/167317.html > > > Hopefully Clément will respond to this thread to clarify. > > > > .o/ > > The state: > > - I don't need it myself anymore (not using the Apple frameworks to avoid > generating such file was an easier and more reliable solution with > regards to everything in the world not supporting this feature) > > - the uncertainty about FATE changes (see thread) still hold > > - Lacking more test samples > > - There might be some pts/dts adjustment in lavfi that might still > conflict with this edit list support (see previous point for > confirmation) > > - Still not exactly a satisfying solution (typically, the export of the > timeline is not generic, it's a dump of the atom) > > I can rebase my work and let you play with it if you are interested. >
Note: if you think you can design a generic data structure for the timeline, it might be interesting: this might be useful for matroska as well for example. Not knowing the latter, I didn't take the risk of doing so and only exported the mov atom. -- Clément B.
pgpAQpGVPQbPd.pgp
Description: PGP signature
_______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
