Perhaps it would be better if outbound specified a special form of
contact address that was clearly bogus for routing in its own right, to
avoid any confusion that its value is significant when outbound is used.
Thanks,
Paul
Christer Holmberg wrote:
Hi,
Can we be careful about terminology here?
I don't think this has anything to do with third party registration.
Third party registration is when identity of the UA performing the
registration differs from the To-URI of the REGISTER request.
I think what this was intended to be about is when a Contact address
used in a REGISTER terminates at a different place from the device
sending the REGISTER request.
Now I guess the Contact address itself is irrelevant in outbound, to
the extent that its never resolved via 3263. We normally expect it to be
consistent with the source of the register request,
but does it really matter?
I don't think so. Incoming requests will be sent to where the REGISTER
came from, no matter what's in the Contact.
Regards,
Christer
Christer Holmberg wrote:
Hi Dean,
How would the keep-alives work for 3rd party registration? How would I
know that you have registered me, and that I now have to send
keep-alives (and where I would have to send the keep-alives)?
And, how would you know if the flow that you have registered for me
fails, in which case you should establish a new flow for me?
Regards,
Christer
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of
Dean Willis
Sent: 17. lokakuuta 2008 17:36
To: SIP IETF
Subject: [Sip] Question on non-outbound registrations in outbound
The current draft says:
4.2.3. Non Outbound Registrations
In an initial registration, a User Agent MUST NOT include a reg-id
header parameter in the Contact header field if the registering UA
is
not the same instance as the UA referred to by the target Contact
header field. (This practice is occasionally used to install
forwarding policy into registrars.)
A UAC also MUST NOT include an instance-id feature taf or reg-id
Contact header field parameter in a request to un-register all
Contacts (a single Contact header field value with the value of
"*").
This strikes me as odd in several ways.
The reg-id lets us know "which registration we are talking about". The
instance-id lets us know "which specific UA-appearance we are talking
about".
So for a 3rd-party registration, it might be useful to have reg-id or
instance-id values. Assume for example a voice-mail server that
registers a low q-value contact for an AOR, so that calls will
eventually reach it in a no-answer scenario. Withdrawing just those
specific registrations would seem to require a selection mechanism
from the set of registrations, and reg-id would appear to satisfy this
need.
In a different use case, assume a 3rd party registration is used to
install forwarding policy. The use of an instance-id in such a
registration would seem to allow the forwarding policy to target one
specific instance of the set of contacts registered to the target. In
other words, not "forward the call to sip:[EMAIL PROTECTED]", but
"forward the call to sip:[EMAIL PROTECTED]'s cell phone".
Now it's entirely possible that neither of these scenarios actually
work. But I don't see in the discussion of outbound anything about WHY
they wouldn't work, which means that the requirements language (the
MUST in outbound) is unsupported. I don't like unsupported
requirements in a specification.
--
Dean
_______________________________________________
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
_______________________________________________
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
