Hmmm... I have seen SBCs out there that remove the authentication header
as well and not all SBCs generate authenticate challenges (for example,
they rely on a softswitch later on to do it). Also, there is the issue
of a change in Nonce in the middle of the call when a new request gets a
challenge with "stale=true".

I'm not sure we can refrain from introducing a new value generated by
the SBC.

> -----Original Message-----
> From: Dan Wing [mailto:[EMAIL PROTECTED]
> Sent: Friday, July 13, 2007 3:45 AM
> To: Gilad Shaham; 'Hans Persson'
> Cc: [email protected]; [EMAIL PROTECTED]
> Subject: RE: [Sip] SIP Identity using Media Path
> 
> > Dan Wing Wrote:
> > > > > > Section 4.1, item about Contact:
> > > > > >
> > > > > > How can the Contact header usefully be used in the
> > > > > > signing process? An
> > > > > > SBC along the message path will happily replace it.
> > > > >
> > > > > I have removed that in -01 (which isn't yet published, of
> > > > > course).  There
> > > > > was some thought it was necessary, but I agree it should be
> > > > > removed from the signature.
> > > > >
> > > > > On a similar note, I am considering removing CallId from
> > > > > the signature.
> > > > > Oftentimes the Call Id value contains an IP address (in
> > > > > dotted decimal
> > > > > or hex), and an SBC or B2BUA may also want to rewrite such
> > > > > a CallId.  I
> > > > > have made a note of that in -01 so this can be discussed.
> > > >
> > > > I thought about that as well, but apparently I forgot about
> > > > adding it to
> > > > my mail. I think it would be a good idea to take Call-ID out as
> > > > well.
> > >
> > > Ok, thanks.
> > >
> >
> > I agree with this, but keep in mind that the call-id serves as a
nonce
> > in RFC 4474. It might require introducing a new header/parameter to
be
> > the nonce.
> 
> Thanks for that reminder.  You're right, and I agree that having the
> authentication service add its own nonce is the best way to do it;
such
> a nonce need only be a random number, as it's already qualified by the
> authentication service's domain (and certificate).  Thus, SBCs won't
> need to also destroy it.  I'll include that in my working revision
> to sip-identity-media.
> 
> -d


_______________________________________________
Sip mailing list  https://www1.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