Re: [v6ops] The need for local-ipv4 socket transition solutions -- NAT64/DNS64 remains insufficient

[Tore Anderson <[e-mail]>]

* Mikael Abrahamsson <[e-mail]>

> Until then, the operating system will have to implement functions
> like 464XLAT to work around this situation. It's not good, but I'd say
> this is pretty typical transition problem.

Yep.

Justine Vick from Microsoft held a very interesting presentation at the
V6 World Congress on Wednesday about their plans to make one of their
buildings IPv6-only. One of the things she had identified a need for
was to have a CLAT in the desktop version of Windows. She had already
established contact with the relevant teams to try to make that happen
(I asked her to add the Windows Server folks to her CC list). Fingers
crossed!

Once we can translate IPv4 on our network borders/edges and keep the
network itself IPv6-only we have way to proceed with the transition.
Hopefully it will eventually dawn on the apps and web sites that use of
literals or depending on AF_INET sockets isn't particularly smart if
they want the optimal performance and functionality. (But if they don't
it's no big deal either, they can get the best-effort service level.)

Tore