Re: draft-jcurran-v6transitionplan-02.txt

[John Curran <jcurran@mail.com>]

At 4:37 PM -0800 1/30/08, Fred Baker wrote:
>Thanks for this note, John. May I humbly request that, if you update it further, that you tag it draft-jcurran-v6ops-*? I tend to notice those a lot more easily than I do documents that don't reference the working group.

I'm more than willing to tag it in that manner.   I intentionally did
not originally, as it would have been presumptuous on my part to
assume that v6ops wanted to be associated with it in any manner...

>One question. I agree that the roll-out should be phased, and that providers world-wide need to take matters seriously, which I think is the high-order bit of your message. However, you provide very concrete dates on which cut-over events should take place, based on data which is the best available but is none-the-less somewhat approximate. We estimate that the IANA will hand out the last IPv4 prefix in the neighborhood of the end of 2009; one could easily imagine the actual event happening either a bit earlier or a bit later, or some RIR policy changing that makes it a lot earlier or a lot later. For example, Geoff suggests that the ISPs will simply start buying and selling addresses out of their allocations. If they are contemplating doing that, given that the RIRs allocate them based on a detailed assertion that the organization allocated to really needs more addresses to run its business, I should think that the RIRs could get pretty fussy about that analysis.
>
>So I wonder if the dates should be more approximate, or if operator fora such as NANOG or RIPE should be left room to diddle with them somehow.

Here's my take...   The exact dates that the IANA or the RIR's is
going to start turning down requests for lack of availability simply
does not matter.   We're not going to see an IPv4 "Free Pool"
countdown clock on the bottom of CNN for final 6 months, and
even now the vast majority of Internet connected organizations
have no idea what Geoff's numbers say today.

However, we're going to need to change global expectations on
what's required now, during transition, and post-transition to be
"connected" to the Internet.  Since the requirements change by
the particular phase, and we need to educating a global audience
about these phases, with materials in dozens of languages and in
nearly every fora on the planet, being able to declare the official
start and end of each phase (and not having these dates move
about without great fanfare) is actually a necessary prerequisite
for having a consistent global message and effective results.

So, the ID contains hard dates, since they maximize our chance
(however great or small) of overall success.

/John