AW: [Sipping-tispan] Parallel Ringing

Hello Jeroen

Thanks for the comments, my comments inline

greetings,
Denis Alexeitsev

-----Ursprüngliche Nachricht-----
Von: Bemmel, Jeroen van (Jeroen) [mailto:jbemmel@lucent.com] 
Gesendet: Freitag, 17. Februar 2006 15:31
An: Alexeitsev, Denis; sipping-tispan@ietf.org; TISPAN_WG3@LIST.ETSI.ORG
Betreff: RE: [Sipping-tispan] Parallel Ringing

Denis,

My comment would be: this feature sounds straightforward enough, but I expect it is in fact quite tricky to realize, and may even introduce more issues than it solves. 
>I hope that the issues will not overweigt the simplicity, but we will see 

When you start forking, an issue that arises is HERFP. See http://www.ietf.org/internet-drafts/draft-jbemmel-sipping-herfp-solution-00.txt <http://www.ietf.org/internet-drafts/draft-jbemmel-sipping-herfp-solution-00.txt>  for details; basically a quick error response from one of the targets could block succesful accepts from others, causing the call to fail

>Surely the PRNG will inherit some of  the forking problems and we will need to assert what troubles they can cause.

Second, what about currently deployed voice mail servers? They would not understand the 'parallel ringing' flag and respond immediately anyway. A 'Require: parallel-ringing' might avoid that (should trigger an error response), but proxies aren't currently allowed to insert a Require (and the error response would cause HERFP). The proxy might instead be adapted to delay processing of a voicemail server response, but then you have the issue how to recognise a response from a voicemail server.  

> processing of the flag as it is also in the Call-Forwarding case is not mandatory. If a voicemail provider does not care about it, it should not worry, but a 
> more user friendly one will get a possibility to react. 

And then there are probably more issues. In other words, I expect that you would probably need additional SIP procedures to make this work (in particular for error scenarios) 

>I will be happy to discuss them