On Nov 24, 2008, at 6:52 PM, Paul Kyzivat wrote:
Hadriel Kaplan wrote:
-----Original Message-----
From: Christer Holmberg [mailto:[EMAIL PROTECTED]
Sent: Monday, November 24, 2008 4:24 PM
If we can't think of any legitimate use for an option-tag in
Require,
why should we allow it?
Because there may be a legitimate use for it tomorrow, or next
week, or
next year.
It occurs to me maybe we're talking past each other. When I think
of the *Require* header, I think of what does any random endpoint/
gateway getting this request have to support for this to succeed.
I can see no value in having that behavior, and plenty of harm in
doing so. I don't want a UAC maker to ever think it can require
UAS' to implement 199 in order for its request to succeed.
But maybe what you're talking about is *Proxy-Require*?
Well, we know tht Proxy-Require is way more evil.
Even so, it probably is the thing that a UAC might want to use if it
knew there were proxies doing the forking.
Trouble is - B2BUAs cause a lot of trouble with Require/Proxy-
Require. I suspect that Proxy-Require should have been MiddleBox-
Require, and so applied to B2BUAs.
So if you really need 199 responses any time a forked invite might
have been abandoned, then I think you must use *both* Require and
Proxy-Require. But that is pretty certain to guarantee that your
call will fail.
This has convinced me that there is no valid use of Require / Proxy-
Require.
Other than for testing for non-compliant proxies and UAs in your
signaling path, aka "diagnostics".
--
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
