[Asterisk-Users] Problems with incoming calls

[Rich Adamson]

> > This is a known issue with livevoip.com service. It's my opinion this
> > is really a design issue within asterisk, but Mark disagrees.
> 
> Your are correct - I do not agree with Mark but, he has never replied to
> any emails about this.
> 
> >
> > The problem is * must answer the incoming iax call from livevoip in
> > order to execute the IVR menues. When the caller then dials an extension
> > number, * responds to livevoip with "ringing" expecting livevoip to
> > provide the ringing to the caller. Since the call is in "answered"
> > mode, livevoip is simply ignoring the iax "ringing" command. Its my
> > opinion the livevoip is properly ignoring that iax function as the
> > call path has already been cut through, end-point to end-point.
> 
> Again, you are correct.
> 
> >
> > If you analyze this interaction in terms of real telephony standards,
> > iax should _not_ be issuing the "ringing" function back to livevoip,
> > but rather providing an inband audio ringback.
> 
> Once again you are correct.
> 
> 
> > So, your only choice is to live with it, or jump through hopps to
> > play an audio ringback within your extensions.conf context.
> 
> Or attempt to fix the code.
> 
> Currently, Asterisk is using the timing of the input stream to reproduce the 
> output stream. This means that when no RTP streams are being sent from the 
> peer Endpoint Gateway, Asterisk is unable to generate audio. This approach 
> or limitation leads to "one way speech" conditions. Plus - Some devices 
> don't generate audio until the answer supervision is received from the 
> called. For all these scenarios, no ringback can be presented to the calling 
> party. In cases where the endpoints are using silence compression, the audio 
> from asterisk is chopped.

Sure, and this last example is an issue mostly unrelated to whether
your livevoip server should/shouldn't respond to the iax "ringing".
Even if the code were "fixed", ringback is not going to be generated
by your user's asterisk system (my system in this example). Since 
both ends are asterisk based, Mark (as the sole architect of
iax protocol) can easily say iax isn't broken. But, that doesn't
hold any water when one vender responds one way and all the others
in a different way.