[MPlayer-dev-eng] [PATCH] Extra messages in slave mode

[Andras Mohari]

On Mon, Dec 05, 2005 at 10:29:46PM +0200, Oded Shimon wrote:
> The non-quiet thing is quite odd of it... requiring -identify is not a big 
> deal though...

Umm, not a big deal, but -identify was not meant to provide state
information about the player, or was it?

> Well, you could send it commands, it'll just end up in kernel buffer 
> until read..

Sure, but...well, the problem might be that I expect too much from GUI
frontends. I'm experimenting with a frontend of mine, and I'd like to
let it show what mplayer *is* doing instead of what it thinks mplayer
*should be* doing. So, I can send mplayer any command anytime *as long
as* I don't want the GUI to reflect mplayer's current state. If I *do*
want it, then I'm in trouble, because I need proper feedback. And
mplayer can spit out an awful lot of (hard to parse) messages, but most
of them have never been meant to be used by frontends.

So my dilemma is:
  A) Either forget about having a GUI that properly updates
     itself according to what mplayer is doing,
  B) or try change slave mode to provide more information to the master
     process.

After I saw Alexandre Conrad's post on mplayer-users, I chose B. :)
I thought that there were others too, who would benefit from a tiny bit
more complete slave mode.

> Could you wait a bit, I'm soon sending my patch. :) Then you can update it 
> there too.

OK, nobody is rushing me. :)

-- 
Andras Mohari