[Asterisk-Dev] Bug report 4783 and RFC 3326: The Reason Header Field for SIP

[John Todd]

At 10:40 AM -0500 on 8/2/05, Tilghman Lesher wrote:
>On Tuesday 02 August 2005 02:42, Mikael Magnusson wrote:
>>  a new header, "X-asterisk-hangupcause", is introduced in bug report
>>  #4783 "Adding cleartext hangupcause to BYE and CANCEL" [1].
>>
>>  Wouldn't it make more sense to use the Reason header as defined in
>>  the standard tracks RFC 3326[2] instead? It explicitly states that
>>  the reason header is most useful in BYE and CANCEL requests.
>>
>>  Example:
>>       Reason: SIP ;cause=600 ;text="Busy Everywhere"
>
>That's great when you have a SIP-only system.  However, how do you
>translate between SIP codes and PRI codes?  IAX also used to send around
>text hangup causes, which are great when you're running only IAX
>clients, but aren't so great when you need to communicate the reason a
>call failed back across a PRI.  This is the reason we added PRI hangup
>cause codes to the IAX protocol, via an IE.  The same thing applies to
>SIP:  we need a way to transmit cause codes, regardless of protocol
>translations.
>
>--
>Tilghman


While there are certainly arguable points here, I'd suggest that this 
is relevant to the comparison:

http://www.zvon.org/tmRFC/RFC3398/Output/index.html

This defines a standard for ISUP-to-SIP mapping, which could possibly 
be used for PRI-to-SIP mapping, as many (all?) of the ISDN q.931 
messages are in there.  The issue is that q.931 codes are 
insufficient to describe some subset of SIP (or any IP protocol) 
failure messages.

JT