Re: draft-ietf-ipv6-unicast-aggr-v2-02.txt

[Erik Nordmark <Erik.Nordmark@sun.com>]

> I'm concerned about the way in which Section 2.0 discusses the
> role of the RIRs.  Given the ugliness of the bugs under these rocks,
> I wonder if the best thing might not be to justify the change
> solely based on the technical argument and cut the rest of the text.
> This would reduce 2.0 a good bit, to something like:

[Co-author hat on - I put those words in the document and I'm sure
they can be improved.]

Ted,

As I recall the history a non-trivial part of the motivation for
killing the RFC 2374 approach was that the IETF shouldn't be setting policy
and that document contained policy language around who could get a TLA etc.

So if for some reason we don't want to mention the term RIR
in the document, how about rephrasing it along the lines of
"the IETF shouldn't be setting policy for address allocation" without
explicit reference to who should set the policy?

Also I think, but I'm not certain, that there were folks in the WG
that wanted to point out that the IETF have provided technical
input to the RIRs process. That was part of the arrangement that was
made to get the allocation policy discussion out of the hands of
the IPv6 WG, thus I don't think it hurts.

---

Re-reading the whole section it seems to be that most of the mentioning
of the RIRs is to document the histroy of how we got here, including
the handing over of IPv6 allocation policy. Thus I think that can't easily be
construed as taking sides in some current debate between the RIRs and ICANN.
The only problematic future looking thing is 
   It is likely that the
   RIRs' policy will evolve as IPv6 deployment proceeds.
since it seems to say that the RIRs will always set the policy.
So at a minimum we should change that, perhaps to use passive voice, to:
   It is likely that the IPv6 address allocation policy
   will evolve as IPv6 deployment proceeds.

But I'm a bit concerned that different folks might interpret such a change
as either
 - the possibility that the IETF can take back the policy control
 - that we are pre-judging that the RIRs will not continue to 
   be in control of the policy. (The text doesn't do that itself, but
   the discuss comment to the WG and/or Bob explaining why this change
   is necessary might have this effect.)
Perhaps I'm overly concerned?

  Erik