Re: draft-kaplan-sip-session-id-00.txt

Dale Worley wrote:
> I'm getting a bit confused by this discussion, and I think part of the
> problem is terminology:
> 
> Let's avoid using the word "match" unless we mean a process which has
> the usual properties of "match":  (1) If A matches B, then B matches A.
> (2) If A matches B, then A and B are considered to some degree to be the
> same thing.  Thus, if two requests have the same Via branch parameters,
> they "match" because they are (should be) identical.  And if two
> requests have the same call-id and tags, then they "match" because they
> are part of the same dialog.  (And that relationship is symmetric.)
> 
> I think what people want is the concept "references".  If A references
> B, then B does not necessarily reference A.  Thus, a dialog event
> package references a dialog, but the dialog does not reference the
> dialog event package.
> 
> I believe that we intend for two message to have the same Session-Id if
> and only if they are legs of the same dialog (in the sense that each leg
> was generated by a B2BUA as a continuation of another leg).  (Actually,
> that should be "dialog set", as created by one out-of-dialog request.)

This is still very fuzzy to me.

If a B2BUA (B) is doing 3pcc:

	         |------ C
	         |
	A ------ B
	         |
	         |------ D

and it first connects A to C, and later transfers the call to D, will A 
and C and D all have the same Session-Id?

Then take same picture, but make B be a conference focus, to which A, C 
and D are all connected. Will they ever have the same Session-Id?

Also, how are dialog event packages supposed to work? Is it expected 
that they will be extended to contain Session-Ids?

	Thanks,
	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