Pekka Pessi wrote:
> "Vijay K. Gurbani" <[EMAIL PROTECTED]> writes:
> 
>>Folks: Before the Dallas IETF, I had sent an email to gauge
>>consensus on whether the "received" parameter should have the
>>delimiters '[' and ']' when its content is an IPv6 address.
> 
> It should have. Beer should be served in bigger glasses, too.

Yes, I don't think you will have problems establishing consensus
around the beer-and-the-glass part ;-)

> Accept both, generate format according to the RFC 3261 grammar (no
> brackets). 

That is probably what should be the case if we decide to leave
the ABNF as-is.

> The problem is that you are a bit late. It is written that way in
> RFC 3261. Every RFC-3261-compliant device, supporting IPv6 or not,
> must be able to handle Via lines containing the RFC3261-style
> received parameter with naked IPv6address.

RFC 3261 has ambiguities - this is one of them, probably the
most benign, but an ambiguity nonetheless.  With a little work,
this can be easily sorted out in one of the following ways:

   (1) Have the sipping IPv6 torture test document update the
   IPv6 ABNF (although, this will require it to be a standards-
   track document, and not an informational).

   (2) Open and maintain a bug in bugs.sipit.net, pointing new
   developers to the appropriate behavior.

   (3) Write a quick draft updating the IPv6 ABNF in RFC 3261.

   (4) Agree that we will leave the representation of the
   received parameter as-is in RFC 3261, but put leniency torture
   tests around this in the sipping IPv6 torture test document.

Any preferences?

Thanks,

- vijay
_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to