[MPlayer-cvslog] r27556 - trunk/libmpdemux/demux_lavf.c

Reimar Doeffinger Reimar.Doeffinger at stud.uni-karlsruhe.de
Wed Sep 10 09:13:45 CEST 2008


On Tue, Sep 09, 2008 at 11:50:48PM +0300, Uoti Urpala wrote:
> On Tue, 2008-09-09 at 19:58 +0200, Reimar Döffinger wrote:
> > On Tue, Sep 09, 2008 at 04:45:50PM +0200, uau wrote:
> > > The internal demuxer must remain the default at least
> > > until the subtitle issues are resolved.
> > 
> > Let me make this clear, and I intend to say it only once: unless you
> > intend to maintain both the subtitle code and the internal matroska
> > demuxer, this is not for you to decide, at least certainly not you alone
> 
> SSA/ASS subtitle functionality is an important feature. It's not OK to
> just leave that broken in svn. And the breakage did already lead to user
> complaints. I can keep the subtitle code working (and have just had to
> fix the breakage Aurelien caused...).

Well, as far as I saw it, it was rather a minor breakage before you changed it
back.

> If you mean that Aurelien should
> be allowed to leave MPlayer Matroska handling broken in any way he wants
> if he's maintaining the lavf Matroska demuxer I do not agree.

I think it is reasonable to accept some minor breakage in the process of getting
a unmaintained demuxer replaced by a lavf demuxer that in several cases is already
better, and several bugs that only exist because it has not been tested enough.
I see certainly no reason for unilateral action, esp. since the "fix it or I will revert
in some days" might have worked very well without disturbing the process of
improving lavf stuff.

> > and not by overriding the decision of the matroska maintainer without
> > any input from the person maintaining the file you change.
> 
> Aurelien has not maintained Matroska in MPlayer for years.

He is still the official maintainer and to my knowledge he has always promptly
reacted to questions and patches (though sometimes only to personal mails, but still).

> The main point is: the svn version should stay usable. The current state
> of the lavf Matroska demuxer causes practical problems for a lot of
> people. Using the internal demuxer causes problems for far fewer.

Given I have not seen any on MPlayer-users I just can not image it is that many.
At least not enough to skip the "ask official maintainer, currently even active again,
to find a better solution within a few days" methodology.
I really can't understand why you seem unable to see how to solve problems by working
with people more instead of just overriding them and their opinons.

> There
> is no development reason to make the lavf demuxer the default until the
> known problems are fixed.

There is: testing and immediate testing of improvements. And yes, we are
"misusing" our users as testers. We have been for a long time, and I do not
really think it works that much worse for our users on average.



More information about the MPlayer-cvslog mailing list