On Apr 12, 2008, at 7:08 AM, Ben Campbell wrote:
>
> On Apr 12, 2008, at 4:16 AM, Dean Willis wrote:
>
>> I think at least one carrier has gotten a workaround in place to  
>> user caller-ID as an authenticator, but only when the call is  
>> originating from a mobile in their network and they've handled the  
>> GSM authentication themselves.
>
> Actually, I was contemplating an argument that we should just  
> establish a best current practice that says one should never trust a  
> phone number, even if it includes a 4474 signature, and that we  
> didn't need to put anything in the message to indicate that. But  
> you've just brought up an edge case where someone could really sign  
> a phone number and mean it, that is, if the gateway is controlled by  
> the authenticating carrier, it could have some out of band way of  
> knowing about said authentication.  and could in fact trust the  
> callerid with reasonable strength.
>
> Therefore, the fact that the From header contains a tel (or  
> user=phone) URL is not sufficient to for the receiver to infer the  
> authentication strength.
>
>

There's a much less "edge" case. It's entirely reasonable to have a  
phone identified by a phone number where that phone is attached to an  
IP network instead of the PSTN. ENUM could then be used to provide  
routing information to the phone. Such a phone could use strong  
authentication and encryption.

How, based on the From field (or any other information) can we  
differentiate such a phone from a PSTN gateway, consequently knowing  
when to trust and when not to trust the user part as signed by RFC  
4474? I've proposed in a separate message that "user=phone" be the  
differentiator -- in short, that PSTN gateways add "user=phone",  and  
that IP attached phones not add it.

The problem of course is backward compatibility. There are a few  
bajillion gateways out there that do not add user=phone. Consequently,  
were my proposal accepted, RFC 4474 could not be used with those  
gateways until they are repaired. I'm OK with that restriction.

--
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

Reply via email to