Re: [emms-help] Re: make emms-player-start cycle players

[Jorgen Schaefer]

Andrea Russo <address@hidden> writes:

> Michael Olson <address@hidden> writes:
>
>> I've sent a patch for emms-player-mpd.el that makes its playable-p
>> function fail if it cannot connect to the daemon.  That should help a
>> little.
>
> Thanks!  Now I really got the solution proposed by Jorgen Schaefer.
> Of course this change would work also for other backends for which
> playable-p will check if the executable file exists.

I just applied a patch to `define-emms-simple-player' to check for
the existence of the player executable. This should solve your
problem. Please let us know when you continue to have problems :-)

> The only situation I can think of where may patch could help, is for
> container formats like .avi or .mkv in which the codecs used can vary
> and the best way to know if a player can play those media is by
> starting it and check its return code.

Indeed. The problem is that EMMS' "simple" players start the
process in an asynchronous process, thus the process wouldn't fail
to be created in such a case, as you noticed:

> Obviously my patch alone isn't enough without player return code
> checking.

But this won't work (easily). Currently, the simple players just
assume that the player process terminated when the process exited,
and play the next song. Guessing from the return value whether the
process terminated because it couldn't play the file, or because
there was an error while playing the file, or maybe even because
the player just felt like it, is rather impossible.

Hence, I don't think we should try other players for weird exit
codes.

Regards,
        -- Jorgen

PS. Thanks Michael for the _very_ quick patch! :-)

-- 
((email . "address@hidden") (www . "http://www.forcix.cx/";)
 (gpg   . "1024D/028AF63C")   (irc . "nick forcer on IRCnet"))