[MPlayer-dev-eng] [RFC][PATCH]removing gmplayer

[Diego Biurrun]

On Fri, Jun 11, 2010 at 07:10:14PM +0200, Reimar Döffinger wrote:
> On Fri, Jun 11, 2010 at 11:17:24AM +0200, Diego Biurrun wrote:
> > On Fri, Jun 11, 2010 at 12:17:43AM +0200, Reimar Döffinger wrote:
> > > Right now I don't see a real cost in keeping it, slowly squezzing out remaining
> > > users via rot.
> > 
> > The costs are very real code complexity and ugliness.  The GUI is badly
> > architected.  It should not spread its fingers all over the core code.
> 
> Any particularly bad issue?
> E.g. I think it's okay to just remove the GUI code from mp_msg, but I wasn't
> able to find any other significant GUI code under CONFIG_GUI _except_ a huge
> load in vo.
> 
> > What made you change your mind?  Last time you said that if no maintainer
> > is found, we should drop it.  After 1.0rc4 is the best moment for that.
> 
> Did I? I remember it more like it being acceptable to drop it.

We talked about this at froscon I think.  You said that we might make
an announcement together with 1.0rc4 that the GUI will go the way of
the dodo unless some person steps up to maintain it.

Nobody will step up to maintain it, it's a complete dead end.

> However the "complexity and ugliness" right now don't feel like
> an issue to me _right now_.

It seems you haven't looked at the code in a while ;)

> > > If you want to raise awareness, you might change the --enable-gui to
> > > --enable-deprecated-gui or such?
> > 
> > That might be a start, but do you really expect somebody to stand up and
> > maintain the GUI?  It would have to be rewritten completely anyway.
> 
> Maintain so to make it competitive? No. Slowly build it back, all the while
> making sure other GUIs gain the features it has but aren't exposed (e.g. via
> slave mode), maybe.

What sort of features are missing?

I don't see a slow death as being better than a quick kill.

Diego