[asterisk-dev] Dialstatus Oddity in 1.2

[John Todd]

At 11:49 PM -0500 1/22/06, Greg Boehnlein wrote:
>Guys,
>	I posted this to -Users, but got no response. I'm not sure how
>topically correct this is for -dev, but I tried looking at the code for
>IAX2 channel driver and couldn't find anywhere that this was documented.
>
>Anyone want to weigh in on this? Shouldn't an unavailable peer return a
>"CONGESTION" result rather than a "NO ANSWER"?
>
>--
>     Vice President of N2Net, a New Age Consulting Service, Inc. Company
>          http://www.n2net.net Where everything clicks into place!
>                              KP-216-121-ST
>
>[snip]

My mail is making it through to the list!  A miracle!

I would expect an unavailable IAX peer to return "CHANUNAVAIL", 
actually.  But this really does warrant some more discussion about 
what return codes are passed back to the dialplan, and why.

  - What should an improperly formatted peer name return?  In other 
words, an impossible destination or a destination that is not in 
<channel>.conf and not a fully-qualified user/host pair?

  - What should be accessible via the dialplan for SIP returncodes? 
IAX return codes? I would want to do something different for a SIP 
"404" error than for a repeated "407" error, as an example.

   I know that we would _like_ to have Asterisk treat all channels the 
same way, but sometimes that doesn't suit the needs of the particular 
task at hand.  While I am a firm believer in generalizing, it's 
always nice to be able to get under the hood and actually see what is 
going on if the dialplan author is sophisticated enough to use those 
tools.

   In short, I think that you've identified something that probably 
warrants a bug, but as usual the question to me at least seems larger 
than just this item.

JT