Re: [Sipping] query on draft-ietf-sipping-dialogusage-06

You need to read section 5.2 more carefully - a 408 affects only the  
transaction by itself.

Some usages will be terminated if a transaction inside them receives  
a 408 (3261 says
to send a BYE in an INVITE created dialog).

And remember that the far UA being completely gone is only one way  
you could end up with a 408.
You could have a proxy that has (for instance) a hard time dealing  
with some particular SUBSCRIBE
refresh request but would pass an re-INVITE on that same dialog along  
just fine. The UA itself may
be failing in a similar fashion.

So, when we discussed this, we determined that the safest thing to  
specify was that the usages came
down independently.

RjS

On Jun 19, 2007, at 2:01 AM, Malleswara Rao Ankem wrote:

> Section 5.2) of draft-ietf-sipping-dialogusage-06 indicates that at  
> 408 response (or a time out) to a transaction within a usage of  
> dialog should affect only that specific usage of the dialog and  
> should not affect the other usages as such.
> But is n't the 408 response  indicating that the request has timed  
> out ? which basically should mean that the other UA has either died  
> or is in some infinite looping condition?
> Why should not the entire dialog state change to terminated ? why  
> should a UA wait for timeouts on all the other usages ?
>
> -Mallesh
>
>
> _______________________________________________
> Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
> This list is for NEW development of the application of SIP
> Use sip-implementors@cs.columbia.edu for questions on current sip
> Use sip@ietf.org for new developments of core SIP

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors@cs.columbia.edu for questions on current sip
Use sip@ietf.org for new developments of core SIP