[MPlayer-users] BUG: "mplayer -identify" now plays

[D Richard Felker III]

On Wed, May 14, 2003 at 04:21:24PM -0800, Sycotic Smith wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> > > D Richard Felker III wrote:
> > > > couldn't both identify and show the movie at once (might not matter
> > > > for local files, but consider a network stream.
> > > I don't quite see the logic.. afair -identify printed script-edible output,
> > > why would anyone need that, standard MPlayer output fits more to humans.
> > 
> > Consider a web interface to remotely start mplayer on a headless box
> > connected to tvout... It might be nice for the web interface to parse
> > and show you info about the file you're playing.
> > 
> > Rich
> 
> Just my $0.02 US on this topic... What was the FIRST reason this
> option was created? IIRC, it was to show you what was in the
> container file and exit. Am I right, or was it just exiting due to
> the break; or exit; within the parser for that option? I would say
> if the addition was to have a way to get the details *without*
> having to give any other options, then it *should* exit after giving
> the details. If it is there to parse the info and pass it along to
> say a GUI or frontend, then it is correct the way it is *now*.
> 
> If correct in its current incarnation, maybe there should be a
> different option that just gives this output? As it stands - I am
> sure there are people out there that will disagree - this is idiotic
> to even have it included, since anyone can write a script to capture
> the same info from the normal output. *Forcing* the user to type an
> elongated cmdline just to find out this info is IMHO not the right
> idea.
> 
> Rich does have a point up there, but I just felt from a *USER*
> stand-point this topic needs to be clarified just a *tad* more. So
> all-in-all, I wouldn't say this is a BUG, but rather a 'feature
> change request' by the original poster.

I agree that changing the behavior was probably a bad idea. It was
good to remove all the stupid hacks (e.g. if (not identify mode) then
(don't initialize vo/ao), etc.), but the old behavior could have been
kept by modifying -identify to set vo and ao to null, and then a new
option could have been added to identify without exiting. Maybe
someone could still do this... Or would it annoy everyone even more to
change the behavior again??

Rich