[arin-ppml] Prop-151: Limiting needs requirements for IPv4 transfers

[Owen DeLong]

On Jan 18, 2012, at 8:45 AM, William Herrin wrote:

> On Tue, Jan 17, 2012 at 9:22 AM, Owen DeLong <owen at delong.com> wrote:
>> Bill, you keep claiming that other RIRs have less strict policies.
>> Do you have any evidence or actual policy citations to back that up?
> 
> Owen,
> 
> You can find a comparison of RIR IPv4 policies at
> http://www.nro.net/rir-comparative-policy-overview/rir-comparative-policy-overview-2010-04#2
> 
> Differences in strictness are obvious even in the first section
> (section 2.1). ARIN requires you to demonstrate use within 3 months.
> All others allow 1 year. ARIN requires multihoming for requests
> smaller than a /20. LACNIC does not. And so on.
> 

ARIN only requires use within 3 months in a few select cases. Usually it
is actually 1 year, so, I'd say that's generally comparable.

I would support relaxing the multihoming requirement on smaller allocations
within the ARIN region, but, again, I don't see that as being a significant
disadvantage to ARIN members vs. inter-regional transfers.

> In a few details one or two of the other RIRs come out more strict.
> But, on average ARIN's policies are the strictest.
> 

Looking at the exact same comparison and having worked with 4 of
the 5 RIRs on actual allocations/assignments, I would say that they
are roughly comparably strict overall. Yes, there are differences, but,
I don't agree with your contention that those differences would create
a substantial disadvantage for ARIN constituents in an environment
where inter-regional transfers are allowed.

> And here's a another difference described in that document: APNIC
> allocates a lot of addresses to National Internet Registries. ARIN
> does not; registrants hold addresses directly. What isn't clear (and
> the AC has NOT researched despite the problem being pointed out) is
> that an APNIC transfer policy would require the NIRs to permit
> out-region transfers the way our draft policy binds ARIN. We've
> already determined that transfer reciprocity is NOT required for
> another RIR's policies to be found compatible for the purposes of
> allowing out-region transfer of ARIN addresses.  We know this because
> the President of ARIN was asked the question and he said so.

I believe you misunderstand and/or misrepresent the functioning of
NIRs in the APNIC region. I think it would be more accurate to say
that APNIC issues a lot of addresses through NIRs rather than to
the NIRs. My understanding is that NIR policies cannot and do not
supersede APNIC policy and that a registrant who obtained APNIC
addresses through an NIR still has all the same rights and abilities
with regard to those registrations as if they had obtained them through
APNIC directly. As such, I believe this to be a non-issue.

> 
> 
> On Tue, Jan 17, 2012 at 1:50 PM, Owen DeLong <owen at delong.com> wrote:
>> I don't believe for one moment that APNIC's policies or their application of
>> those policies is significantly less strict than ARIN"s. If it were, APNIC would
>> have consumed a much greater quantity of IPv4 address space.
> 
> It is, as a matter of simple fact, less strict. How much significance
> you find in the difference is a matter of opinion. I respectfully
> disagree with your opinion that the difference is not important.
> 

I disagree.

First, the page you referenced is out of date and ARIN's actual policies have
been significantly relaxed since then.

In actual fact, there are areas of policy where ARIN is more strict than APNIC
and areas of policy where APNIC is more strict than ARIN. Overall, I think it
is roughly a wash. Admittedly, this is a subjective opinion and there is no
way to make such an evaluation objective, so, likely we will continue to
disagree on the matter.

> 
> On Wed, Jan 18, 2012 at 5:36 AM, Chris Grundemann <cgrundemann at gmail.com> wrote:
>> "evidence" and "facts" are pesky little buggers indeed...
> 
> They sure are.
> 

Yes, but, to be useful, they must be current.

Unfortunately, there really isn't an objective way to compare the
relative strictness of policies. That's one of the reasons I find your
proposed language to be extraordinarily problematic from a policy
perspective.


Owen