[ARIN-consult] PDP Consultation Reminder

Owen DeLong owen at delong.com
Wed Jun 27 00:58:35 EDT 2012

On Jun 26, 2012, at 8:04 PM, William Herrin wrote:

> On Tue, Jun 26, 2012 at 10:20 PM, Owen DeLong <owen at delong.com> wrote:
>>>> while some of our worst have enjoyed strong
>>>> supermajorities.
>>> This is the more interesting statement. I'm actually very curious
>>> which policies you would identify here.
>> 2007-23 (I'll note that at the time it seemed like a good idea and I supported this one).
> Really? There was nothing even vaguely close to agreement on any other
> approach. Still isn't. Runout wasn't going to wait for us. And if I
> recall the dissent, it was more in preference to some particular
> alternative rather than objections to this approach per se.

The belief when this passed was that it would not significantly skew
the runout dates of the big-three RIRs to be very far from each other.

Unfortunately, it put APNIC way first, RIPE a distant second and ARIN
a very distant third. That might seem great for ARIN at the moment, but
it's really bad for the internet over-all.

The goal of the policy was actually to provide a relatively even runout
date across the board.

We failed to achieve the policy goal and I suspect we have only begun
to see the true consequences of this.

> This policy brought the IANA free pool to an orderly end where nobody
> accused anybody else of gaming the final allocation. Imagine if APNIC
> had requested the last two /8's and they genuinely were the *last* two
> /8's! I'm not sure how 2007-23 could be viewed as bad policy, even
> with 20/20 hindsight.

We can agree to disagree. I know you seem to believe that ARIN running
out last is perfectly fine and that you think APNIC squandered its resources.
You're entitled to your opinion, but I don't agree with you.

>> 2009-8 (same situation as above).
> I continue to think this is wise policy. It *should* be harder to get
> addresses from the remaining free pool than through transfer. That has
> encouraged the transfer process to develop before we hit rock bottom
> emergency all gone.

We pulled the trigger too early and that's exacerbating the problems with 2007-23.

We should have set the trigger when ARIN got down to the last /8, not when ARIN
received it.

The policy itself isn't bad, but the timing as we implemented it was terrible.

>> 2012-3 (maybe a controversial selection, but there
>> was little opposition in the AC in spite of significant
>> opposition on PPML)
> I'll give you this one. I think the policy was reasonable but the
> community consent wasn't there. It should have gone back in the drawer
> for a cycle or two and then come back out when the idea was no longer
> startling. It shouldn't have passed by a majority let alone a
> supermajority.


More information about the ARIN-consult mailing list