Bob Penfield wrote:
> 
>> -----Original Message-----
>> From: Paul Kyzivat [mailto:[EMAIL PROTECTED]
>> Sent: Tuesday, June 17, 2008 11:44 PM
>> To: Bob Penfield
>> Cc: [email protected]
>> Subject: Re: [Sip] I-D ACTION:draft-ietf-sip-199-00.txt
>>
>>
>> Bob Penfield wrote:
>>
>>> On the topic of UAS's sending 199, I actually think things will work
>>> better if they did. I know there has been lots of discussion about
>>> B2BUAs doing it, but consider a UAS that is aware that it has multiple
>>> registered contacts (via presence or registration event package or
>>> other means), it might want to send the 199 to be sure that it does
>>> reach the UAC (in case the proxy does not support it).
>> Bob, I don't understand the point you are making. Can you explain further?
>>
>> Are you talking about a case where there is a forking proxy in front of
>> the UAS, and the proxy doesn't know about 199, and the UAS wants to send
>> the 199 to give the UAC a warning in case the final response is delayed
>> by the proxy?
> 
> Yes, that is pretty much the case I was considering.
> 
>> While there may be some cases where that would be advantageous, in many
>> others it would just increase the message traffic for every failing
>> call. And I don't see how the UAS would be able to discern the cases
>> where it would be advantageous. I guess it could be configured in those
>> cases where all inbound calls are forked, but how many such cases are there?
>>
> 
> If the UAS knows that the AOR has multiple registered contacts, it knows that 
> inbound calls will be forked. In that case, it could send the 199. As long as 
> proxies don't generate 199s in addition to ones they forward, the increased 
> traffic could be limited.

How often does a UAS know that there are multiple contacts for this AOR?

Even if it does, it presumably can't know if it is the last one tried, 
the first, or whatever.

> Having proxies generate responses on behalf of a UAS is a bit of a violation 
> of the end-to-end principle from a purist point of view. By certainly having 
> only proxies generate 199s would limit the increase in messages.

Yes, it is.

> I was just trying to point out that there might be cases where real User 
> Agent would send 199 and not just B2BUAs.

I do appreciate the point that there *can* be cases where the UAS would 
have enough info to justify sending a 199 response. But the cases where 
a UAS is likely to know that in practice seem limited, with the 
exception of B2BUAs. I would just like to avoid giving UA developers the 
idea that they *ought* to be generating 199 responses on a regular 
basis. Doing so "to be helpful" is likely to be sub-optimal in many cases.

>>        Paul
>>
> _______________________________________________
> Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
> This list is for NEW development of the core SIP Protocol
> Use [EMAIL PROTECTED] for questions on current sip
> Use [EMAIL PROTECTED] for new developments on the application of sip
> 
_______________________________________________
Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [EMAIL PROTECTED] for questions on current sip
Use [EMAIL PROTECTED] for new developments on the application of sip

Reply via email to