Agreed.

-d

> -----Original Message-----
> From: Hadriel Kaplan [mailto:[EMAIL PROTECTED] 
> Sent: Friday, June 27, 2008 8:06 AM
> To: Tschofenig, Hannes (NSN - FI/Espoo); ext Paul Kyzivat; 
> Hannes Tschofenig
> Cc: [email protected]; Elwell, John; Dan Wing
> Subject: RE: SIP Identity
> 
> 
> 
> > -----Original Message-----
> > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On 
> Behalf Of
> > Tschofenig, Hannes (NSN - FI/Espoo)
> >
> > I am trying to figure out where the problems are:
> >
> > Now, it would be interesting to hear whether these guys 
> have actually
> > ever turned it on and used it. What was their experience? Did they
> > indeed encounter problems with certain SBC configurations?
> 
> I am fairly confident in saying that there could not be an 
> SBC that wouldn't break 4474 signatures, since it's basically 
> inherent in the definition of an SBC to modify some of the 
> things that 4474 signs. (ie, if it's not modifying them, it's 
> not really acting in an SBC "mode" anymore)
> Except of course when the SBC is the signer/verifier.
> 
> 
> > Would more folks use it when some of the weaknesses on the 
> E.164 number
> > usage are documented?
> 
> I doubt it.  Plenty of weaknesses of RTP have been 
> "documented", which haven't driven SRTP's uptake.  It will 
> take actual real-world cases of abuse to drive it, IMHO.  But 
> by then it will be late.  First there's the problem of making 
> it work across already deployed infrastructure, which 4474 
> doesn't, and there can be no "flag day".  Given how long it 
> takes us to standardize things, get them into code/hardware, 
> and get people to upgrade, we're talking years.  That's why I 
> want to work on it now, just in case we need it in the next few years.
> 
> 
> > How would a change to the signature algorithm impact the situation?
> 
> By making it work across deployed infrastructure.  More and 
> more Enterprise-SP SIP Trunks are getting added every day, 
> and inter-SP peering is continually growing.  If an Identity 
> algorithm has a lot of false positives (signatures broken but 
> the request is really ok), it won't be much use and people 
> will ignore it.  No algorithm has a chance of use if legit calls fail.
> 
> -hadriel

_______________________________________________
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

Reply via email to