[MPlayer-dev-eng] [PATCH] Enable gui slave commands

[Reimar Döffinger]

On Sun, Dec 11, 2011 at 02:33:40PM +0100, Ingo Brückl wrote:
> Reimar Döffinger wrote on Sun, 11 Dec 2011 14:11:25 +0100:
> 
> > On Sun, Dec 11, 2011 at 02:00:20PM +0100, Ingo Brückl wrote:
> >> Reimar Döffinger wrote on Sun, 11 Dec 2011 12:34:09 +0100:
> >>
> >> > On Sun, Dec 11, 2011 at 12:10:20PM +0100, Ingo Brückl wrote:
> >> >> Reimar Döffinger wrote on Sat, 10 Dec 2011 21:49:50 +0100:
> >> >>
> >> >> > On Fri, Dec 09, 2011 at 04:42:32PM +0100, Ingo Brückl wrote:
> >> >> >> Something like this has been remove in r31378, but in order for slackers to
> >> >> >> be able to operate the GUI with a remote control, we need it. I tried to
> >> >> >> minimize the impact on non-gui files and think I found a neat solution (first
> >> >> >> patch).
> >> >>
> >> >> > You shouldn't need the ifdefs in input.
> >> >>
> >> >> I assume you mean (only) input.h?
> >>
> >> > and input.c
> >>
> >> But wouldn't you get a wrong output (containing the gui command) for a
> >> non-GUI-MPlayer with -input cmdlist then?
> 
> > Not really.
> > We list fullscreen also for a MPlayer only compiled with -vo null etc.
> > Also, the command does not work with a GUI-MPlayer started in
> > command-line mode.
> > So IMO by putting it under an ifdef you just create 3 cases that people
> > who want to use it have to deal with instead of just 2: running as GUI
> > and not running as GUI.
> > What would be the advantage of having a third "not running as GUI and
> > compiled without GUI support"?
> 
> You don't see anything about a gui command if you've decided to not enable
> the GUI at all.

But if you send a gui command you get yet another error message.
And that something is in -input cmdlist doesn't say anything about
whether it will actually work, that depends on other things.

> But I don't mind removing the ifdef if you want to.

I see no point in keeping it, or rather more disadvantages than
advantages, so yes, I'd prefer that.