Re: [Sip] SIP INFO

[Brett Tate <brett at broadsoft.com>]

Because of race/glare conditions, I prefer INFO 489 not MUST or SHOULD trigger the release of the call.  Concerning calls, I see little difference between impacts of 489 and 405/501/420.  Thus I prefer option 2.

Option 2 appears reasonably consistent with notes 3 and 12 within RFC 5057 section 5.1.

> -----Original Message-----
> From: Eric Burger [mailto:eburger at standardstrack.com]
> Sent: Monday, March 23, 2009 9:29 AM
> To: Brett Tate
> Cc: SIP IETF
> Subject: Re: [Sip] SIP INFO
> 
> On Mar 23, 2009, at 6:19 AM, Brett Tate wrote:
> > What was the outcome of INFO 489 response MUST trigger release of
> > call discussion?
> 
> Whoops - that was slide 2...
> 
> *Get a 489, can all still be fine?
> 	*Option 1: Tear down dialog
> 		Something is hopelessly broken, give up
> 
> 	*Option 2: ignore it
> 		-We say INFO does not change dialog state,
>                    and a 489 tearing down dialog changes state
> 		-But, handling 489 differently for INFO makes INFO "special"
> 		-But what if I said "package X is OK" in an INVITE, but there
> is
>                    glare between you sending an INFO from package X
> and my
>                   dropping X in my ACK?
> 
> 
> Thoughts?