Re: [Uri-review] [sipcore] Proposal: sip6 URI scheme

["Dan Wing" <[e-mail]>]

> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On
> Behalf Of Rick van Rein
> Sent: Thursday, April 26, 2012 2:27 AM
> To: [e-mail]; [e-mail]
> Subject: [sipcore] Proposal: sip6 URI scheme
> 
> Hello,
> 
> I am developing to get better support for SIP over IPv6.  One of the
> problems that I have encountered is the interoperability between
> endpoints that run only IPv4 or IPv6.  To solve this situations prior
> to call setup, I am proposing a sip6 URI scheme.  Using this, end users
> and their tools should have an easier time deciding whether or not to
> use the URI.
> 
> A similar thing applies to enforced ZRTP --being sure before calling
> that the call will be private-- and I combined that with the URI scheme
> proposal, so as not to register more URI schemes than required.
> 
> Your feedback is kindly welcomed.

It's a layering violation, and -- worse -- will cause far more harm
than it could possibly help.  Please don't do this.

http://tools.ietf.org/html/rfc6157 describes how SIP is expected to
support IPv6, and transition from IPv4-only to dual-stack to IPv6-only
with the least amount of harm.  It works over NAT64 and NAT46, too
(SIP6, if deployed, can't), and NAT64 will be necessary for IPv6-only
endpoints to access IPv4-only endpoints, such as being deployed by
T-Mobile USA's IPv6-only service, Swisscom's IPv6-only trial, and
probably other mobile operators and, some day, eventually wireline
operators.

If RFC6157 fails, which is why you are proposing SIP6, you need
to back up and explain to everybody the failure of RFC6157.

-d