[MPlayer-dev-eng] Reverted patch of time-based PTS locking

[Reimar Döffinger]

On Mon, Aug 20, 2012 at 11:17:51PM +0200, Pásztor Szilárd wrote:
> BTW I tried your suggestion of the mpi || drop_frame thing and it
> didn't
> work as causes a wrong overwrite of sh_video->pts and a possible bogus
> limitation of the pts buffer (resulting in pts drop) by the delay var.

Right, there's even a comment about that.
But what about leaving the !mpi and only moving drop_frame?
It's still not quite right I believe (at least for hard frame drop),
the thing we'd really need to know is "would we get a frame if we had
not set drop_frame".
There is also the issue that we'd really want to fetch the PTS value
in case of a decode error.
So I believe that would make the "right" solution to make the change
you tested, but also have a if (!got_frame) where we had a if (!mpi
...).
Though I might be missing even more things.