On 11/23/11 23:08, Nov 23, Paul Kyzivat wrote:
> Lets ask Adam to comment on this.
>
> Adam?

Yeah, I saw this on the resip list but didn't have time to dig into it. 
My recollection is that you're not supposed to send a NOTIFY on a dialog 
while another NOTIFY is still pending on the same dialog. I distinctly 
recall having discussions around this point in the 1999-2000 timeframe, 
and I think we elected not to prohibit it (although I don't remember 
why). It's water under the bridge at this point.


...
>> so I don't understand how in RFC 5057, section 5.1, 500 (or unknown 5xx)
>> show up in "Transaction Only" and not in "Destroys Usage" - what am I
>> missing?

RFC 5057 reflects more recent operational experience than RFC 3265. Even 
though it is only informational, I would take its notion of what 
destroys a dialog versus just failing a transaction to be more informed 
than RFC 3265's. In fact, in the revision to RFC 3265 currently 
underway, we specifically use the RFC 5057 rules for destruction of 
dialogs, usages, and transactions.

So, in this case, I believe that the B2BUA stack has it wrong, and 
shouldn't be terminating the dialog. This will be much clearer when the 
next version of SIP Events is published.

/a
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to