[SIPForum-techwg] SIPconnect 1.1 baseline - IPv6

[Elwell, John]

Spencer,

That was not how I interpreted your earlier message - I interpreted it
as MUST implement IPv4, MAY implement dual stack. I would have been
moderately happy with that, but it would still require additional
specification work (by us), e.g., to deal with negotiation of the IP
version between the two media endpoints. Making it mandatory to
implement raises the bar.

Or are you and/or Henning suggesting deployment only in environments in
which either all media endpoints are IPv6 or all are IPv4, thereby
avoiding the need for negotiation?

John

> -----Original Message-----
> From: techwg-bounces at sipforum.org 
> [mailto:techwg-bounces at sipforum.org] On Behalf Of Spencer Dawkins
> Sent: 14 August 2008 15:31
> To: Henning Schulzrinne
> Cc: techwg at sipforum.org
> Subject: Re: [SIPForum-techwg] SIPconnect 1.1 baseline - IPv6
> 
> Hi, Henning,
> 
> Yes, I agree ("distinguish" would be helpful). Let's assume 
> that I was 
> asking about "use" in my question below.
> 
> Thanks,
> 
> Spencer
> 
> 
> > It might help the discussion if we distinguished between 
> MUST  implement 
> > and MUST use, just as we do for security mechanisms, such 
> as  TLS, or 
> > transport protocols, such as TCP, for SIP. Given that all  
> mainstream 
> > server operating systems likely to be used for servers  
> support IPv6, I 
> > see no problem with mandating IPv6 implementation.  This 
> really is not 
> > hard - we've done this in our server years ago, so  I have some 
> > experience. Obviously, nobody is forced to run IPv6 or, as  
> a service 
> > provider, offer IPv6 services. We're the not doing the  
> industry a favor 
> > by pretending that we can just ignore the coming  address 
> crunch. The 
> > alternatives, such as yet more layers of NATing,  are far 
> worse than a few 
> > lines of socket code.
> >
> > Providing for both IPv6 and IPv4 does not prevent somebody, 
> e.g., in a 
> > closed environment (say, IMS or the military) from running 
> only IPv6  if 
> > they so choose.
> >
> > Henning
> >
> > On Aug 14, 2008, at 9:05 AM, Spencer Dawkins wrote:
> >
> >>>
> >>
> >> I agree that neither IPv4-only nor IPv6-only would be the 
> right  thing to 
> >> say
> >> in a 2009 specification, and remaining silent on IP versions isn't 
> >> helping
> >> anybody.
> >>
> >> So, we're going to have to have normative text at some 
> point on this
> >> topic...
> >>
> >> The issue with dual-stack is that you have to have an IPv4 
> address - 
> >> that's
> >> what's driving the IETF nuts now, because it's fairly obvious that
> >> dual-stack ALSO hits the wall when we reach the point where IPv4 
> >> addresses
> >> are either technically or economically unavailable. Dual stack is a
> >> transition mechanism, not a continued coexistence mechanism.
> >>
> >> But if this is the first time we say anything about IPv6, 
> and if we  have
> >> another opportunity to say something stronger after a 
> couple of  years, I
> >> would be OK with MUST IPv4, and MAY dual-stack, in a 
> specification  with 
> >> a
> >> two-year shelf life.
> >>
> >> Does anyone substantially disagree with this?
> >>
> >> Thanks,
> >>
> >> Spencer
> >>
> >>
> >> _______________________________________________
> >> techwg mailing list
> >> Send mail to: techwg at sipforum.org
> >> Unsubscribe or edit options at: 
> >> http://sipforum.org/mailman/listinfo/techwg
> >
> > 
> 
> 
> _______________________________________________
> techwg mailing list
> Send mail to: techwg at sipforum.org
> Unsubscribe or edit options at:  
> http://sipforum.org/mailman/listinfo/techwg
>