[Asterisk-Dev] Configurable "hold" extension

John Todd jtodd at loligo.com
Fri Sep 19 11:47:27 MST 2003


[Note: new thread created since this is a separate concept - moved to -dev]
>[previous ideas and thread about discmans, copyrights, etc. deleted]
>
>Hmm, what do you think about about creating a fake extension (like s,
>t, fax etc.) called, say, "hold" that would be called every time moh
>is played now?
>
>to get the old behaviour you'd do:
>exten=>hold,1,MusicOnHold
>(or sth)
>
>and you'd get the required flexibility for just about anything.
>examples off the top of my head follow:
>* dial a sip extension which streams an .asf using some
>proprietary/windows/etc. software
>* some agi plays you nice music while mixing in some real time
>generated info ("you've been on hold for $time. if you're pissed off
>already, dial $phone and complain" ;))
>* well, the top of my head seems to end here but i'm sure you'll find
>more creative uses :)
>
>cheers,
>  grzegorz nosek

Grzegorz -
   This is a very interesting idea.  However, I don't understand 
enough about how Asterisk handles recursive calls to know how complex 
this would be to implement.  It sounds reasonable enough, though. 
The only problem might be if "Dial" applications are called out of a 
"hold" context - where does call control go?  Can the original 2nd 
leg of the call pick up the user that was put on hold?  Does that 
interrupt the existing call that the 1st leg may now be involved with?

   I like this idea the more I think about it.  With this method, it 
may be possible to dump users into a queue if they are put on hold 
for too long, even if the original leg was not from a queue.  I 
suppose a Dial statement in the middle of the context could be 
handled as a "Transfer" if it were to occur, and the first leg would 
be disconnected...

   In any case, this sounds like a good project.  I don't think Mark 
or anyone at Digium right now has the time to implement; do you have 
time to work on development of this potential addition?

   Mark, does this present any obvious no-start issues?

JT



More information about the asterisk-dev mailing list