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

[Gert Doering <[e-mail]>]

Hi,

On Fri, Apr 24, 2015 at 10:38:06AM -0400, Howard, Lee wrote:
> >[DJM]
> >I see one common theme in this discussion:  "IPv4 literals are harmful."
> 
> I think I read consensus on that statement.

+1  :)

> I'm not sure there is consensus that the IETF is the right place to make
> such a statement. There might be; I think we need a document to discuss to
> see if there is consensus.
> Somebody needs to write it for us to discuss it. Hint, hint.

I think the IETF is the right place to say so.  We document operational
problems, and we document how to move networks to dual-stack or IPv6-only,
and using IPv4 literal is a very clear operational problem.

As for the document: it could be a very short one ("do not use IPv4 literals
for services that are expected to work over the Internet, which contains
islands that use IPv4/IPv6 translation today, as it risks breaking your
service, or imposes operational costs elsewhere to work around it"), or a
longer one ("application programmers really should... IP agnostic API...
avoid this, avoid that... best practices like <see RFCxxx, RFCyyy, RFCzzz>").

I think both documents have merit, but the first one has a much higher chance
of actually getting anywhere in the near future.

Gert Doering
        -- NetMaster
-- 
have you enabled IPv6 on something today...?

SpaceNet AG                        Vorstand: Sebastian v. Bomhard
Joseph-Dollinger-Bogen 14          Aufsichtsratsvors.: A. Grundner-Culemann
D-80807 Muenchen                   HRB: 136055 (AG Muenchen)
Tel: [phone]           USt-IdNr.: DE813185279