Re: [dispatch] New Version Notification for draft-kyzivat-dispatch-trs-call-info-purpose-00.txt

[Brian Rosen <[e-mail]>]

>> If we do proceed with this, I don't think the call-info is the appropriate place to put it. I think a new header would be the right thing.
> 
> We considered this, and thought that Call-Info was a more appropriate choice. But either is acceptable to us.
It is, exactly, information about the call, but I agree that a new header is an acceptable solution.


>> If the goal is purely to check the user is in the US, then having the VRS check that they were sending media to an IP address inside the US seems like it would be a better solution.
> 
> Perhaps, if the FCC agreed. Maybe it could be done as an interpretation.
> But it may also not be visible due to media relaying by the default provider.
> 
Yeah, ubiquitous use of SBCs pretty much makes any addresses seen by the recipient a non-starter.
Please also note that the providers want the check to be the responsibility of the provider who is actually providing the interpreter service, and they explicitly support the “dial around” case that requires this work.

While a VPN would fool the check, any form of NAT on an ISP or home network is fine.  Like anything else, this check merely increases the cost of fraud, it doesn’t stop it.

I also point out that use can be restricted to a private networks (the one that interconnects providers).  It’s pretty much in everyone’s interests to strip the header if the call goes outside the domain of interest.  We can make that text as strict as we want.