On 7 nov 2008, at 19:18, EricLKlein@softhome.net wrote:This is why we have RFC 4864, and the comment that the IETF does not support NAT in IPv6. We need to find a way to make it crystal clear that NAT is not part of v6 and using it will be non-standard.And this is why I think this draft will cause more harm than good.I don't think RFC 4864 is clear enough. If the IETF is serious about avoiding IPv6 NAT, it needs to send out a warning to operators that they WILL have problems if they deploy it.
Agreed. Now the problem will be getting consensus across the various WGs that seem to have taken up beyond what v6OPS did and agree to make such a statement. I am sure that we now have Behave and Softwires DHCPv6 (and others?) looking into NAT as there is still a perception that NAT is needed even after Site Locals were depreciated in RFC 3879 which became an RFC back in September 2004. Given the fact that 3879 "formally deprecates them, but the deprecation does not prevent their continued use until a replacement has been standardized and implemented" - I am sure that they will be implemented by some networks anyway. (from the abstract of RFC 3879)
Eric