[arin-ppml] IPv4 Transfer Policy Change to Keep Whois Accurate

Owen DeLong owen at delong.com
Sat May 21 20:31:46 EDT 2011

On May 21, 2011, at 6:45 AM, Mike Burns wrote:

> Hi McTim,
> I don't see the slippery slope.

I do.

> IPv4 addresses have *already* been monetized.

There is a difference between monetization and commoditization.

> In the near future, once the free pool has exhausted, the only way to get IPv4 addresses will be to pay for them.

> We can bemoan the fact, but we can't change that fact in any way I can see or that you have proposed.

I don't see how that is particularly relevant to his point.

> Let's not setup a false dichotomy, as in pass my proposal and face the monetized IPv4 market, don't  pass my proposal and somehow the deep pockets don't win.

Pass your proposal and the deep pockets get an overwhelming victory handed to them on a silver platter.
Don't pass you proposal and the deep pockets at least have to engage in some level of fraudulent manipulation to achieve that result.

> With ARIN's existing policy, the deep pockets will still beat the shallow pockets.

Perhaps, but, perhaps not to the extent or as easily or as readily or as quickly as under your proposal. Any one of those is
an overall win for the community.

> The fundamental problem I have with your post is that you fail to recognize the significant game-changer that is exhaust.

I disagree.

> If you have read what I have written, you will find no mention of monetizing IPv6, and really there was no movement towards monetizing IPv4 until the brick wall of exhaust nearly hit us in the nose.

While you have not mentioned it, you have expressed support for competing privatized registries. The obvious and inevitable result of such a move
is to dilute the public policy process with the obvious conclusion being the commoditization of all IP number resources, including IPv6 addresses
and autonomous system numbers.

Just because you have not mentioned something does not mean that it is not a probable future outcome of the continued evolution of policy along
the lines that you have expressed support for.

> So that's where we are, like it or not, we are at the exhaust point, and moving forward in IPv4 will require the purchase of IPv4 addresses.

I don't think anyone has denied this, but, the question is whether the purchase of IP addresses is best managed with or without the inclusion
of the existing justified need policy. McTim and I (and many others) believe that the community is best served by policy that includes the
justified need. You believe that it is better to commoditize IPv4 addresses rather than merely monetize them.

> I argue that the needs basis is the appropriate model for handing out free addresses, but the purpose of the needs requirement, that is distribution to users who will put the addresses to productive use, has been superceded by the actions of the market, which I have argued will have the same overall effect of driving these assets to their most productive use.

Which argument has been refuted multiple times. You have yet to provide any proof whatsoever that your argument actually
holds water. You have presented nothing to indicate that the monetization/actions of the market will somehow direct addresses
to efficient use rather than to profitable acquisition. There is nothing to couple efficient use to profitable acquisition, and, in fact,
many reasons to believe that the two are utterly unrelated.

> As far as arresting what you consider a slide towards private registries, what could be better than arming our registries with the optimal tools to encourage registry at ARIN in order to forestall any flight to a private registry, which presumably would have some advantage in the customer's eyes?

If you eliminate policy and stewardship  in favor of commoditizing address registrations by the registry, then,  there is little difference between
a single registry with a commoditized policy and competing registries other than a possible issue of price drivers.

> If we insist on maintaining the needs analysis as a roadblock to (many? some? a few?) transactions, we are driving that registration out of the system you are praising, and to what end? I say we run the risk of actually incentivizing the switch to private registries, and lest we forget, there is nothing in the world to prevent private registries from popping up and succeeding if they provide some utility to customers and to network operators.

What you call a roadblock, many of us call proper stewardship. You are insisting that there are a significant number of transactions
occurring outside of policy and that would somehow magically start being registered with ARIN if we only deleted this needs basis
requirement. Further, you assert that a growing number of these transactions will occur without this policy and that the elimination
of needs basis will somehow magically eliminate them.

Unfortunately, you have failed to provide any evidence of any of these four assertions:
	1.	That such transfers are occurring.
			I believe that they are, but, we have no proof and there have not been
			enough of them to even be quantifiable. I am inclined to believe that they
			are occurring in such small quantities as to be insignificant. Certainly not
			in such quantities as to discredit the ARIN registry.

	2.	That such transfers will increase after runout.
			There's really nothing behind this assertion and no proof or strong reason
			to believe runout will affect the number of such transfers positively or negatively.

	3.	That removal of needs basis will cause such transfers that have happened to be
		registered with ARIN.
			Yes, it will remove one disincentive from registration, but, in most cases, this
			 is not the most significant disincentive and is unlikely to make a difference to
			most parties unwilling to work within the ARIN policy framework.

	4.	That removal of needs basis will prevent such transfers from occurring in the future
		outside of ARIN policy.
			As in (3) above, there is no evidence to support this assertion.

> We are stewards not just of IPv4 addresses, but AS numbers, Ipv6 addresses, AND Whois. At this point let us not make policy designed for the stewardship of free resources that operates at the risk of Whois accuracy and usefulness.

Indeed, we should not adopt this policy for exactly that reason. Not only is it poor stewardship for IPv4,
but, it also sets bad precedents for the future stewardship of IPv6 and AS Number resources as well
as being likely to further degrade the accuracy of whois as the IPv4 landscape degrades into a
free-for-all of random trading both with and without ARIN registration.

> The Internet succeeded, in your eyes, because it was open, transparent, and collaborative. Why then, are we engineering policy which is less transparent and less open? Remember, my policy does away with NDAs and allows for transparency in transactions by removing ARIN's transactional impediments and faciliating the registration on the public Whois registry.  Maintaining needs drives transactions into the darkness and threatens the functionality of Whois, of which we are stewards.

The current policy is quite open and transparent. I don't see any greater openness or transparency
contained in your policy. It does not do away with NDAs, nor could any policy do away with them.
I'm not sure why you think it would.

You again repeat the assertion that this requirement is driving transfers away from ARIN process
in spite of recent evidence of an uptick in registrations of past transfers in the ARIN process with
needs basis preserved.

How do you explain this dichotomy between provable fact and your assertions?

> Do we want transparent transactions, at least as far as knowing who controls what netblock? Then let's end the policies which prevent ARIN from registering the transfer of otherwise legal transactions.

Of course we do. However, not at the cost of abandoning any proper stewardship of the
resources in the process. Referring to transactions that don't meet needs basis as
"otherwise legal transactions" is, IMHO, a lot like suggesting that we legalize theft
so as to remove policies which prevent this "otherwise legal graft".

Transactions which do not meet needs basis are harmful to the community in general
and the community has long maintained that they should not be allowed. Just as theft
is harmful to the community and therefore illegal in almost any society run by law.

Yes, there are a certain number of these transactions that will probably manage to stay
somewhat under the RADAR and survive in spite of their non-compliance with ARIN
policy. In reality, this has probably been true for a very long time. However, ARIN policy
like any community standard for the decent conduct of the citizens of the community
in the community interest is generally dependent on voluntary compliance. Enforcement
to any extent is an absolute last resort in any just structure of law. The average speed
in 70 MPH zones on freeways in the US is 77.8 MPH. Something on the order of 80%
of drivers operate their vehicles in excess of posted speed limits.

Strangely, few seem to be calling such speed limits "unenforceable" or suggesting
that they should be repealed on that basis.

OTOH, I'd be surprised if there are more than 1% of resource holders who have
made such transfers outside of ARIN policy and I doubt that number is likely to
get above 5% even with current needs-basis preserved. As such, I think these
transfers are outliers and not a sufficient body to represent any sort of consensus
to change the policy.
> Let me just end with the crucial factor- we don't have any more IPv4 addresses left to steward, we have already moved into the commoditization of these addresses, but as long as the free pool of AS numbers and IPv6 addresses exists, stewardship requires some constraint on their allocation, and I have seen no mention anywhere, by anyone, that these things should be commoditized. Thus the fear of lifting needs requirements for addresses already allocated should not be transmogrified into some fear of ravenous capitalists set to devour the Internet model.

Just because they no longer sit in a free pool does not mean that our
responsibility for stewardship has ended. We still have a responsibility to
administer even a fully allocated address space. In some ways, the question
of how we fulfill that role may be more defining than the much simpler
questions of how to administer the allocation/assignment of addresses from
an existing free pool.

The monetization of addresses is not necessarily the same as the
commoditization of the addresses. I agree that the monetization is
a necessary and inevitable result of the current situation. I do not
agree that commoditization is inevitable or even a proper or useful
outcome for the community.

You speak of fear, yet, this is not about fear from my perspective. I
realize that by playing the fear card, you hope to discredit my
arguments. I think that is rather transparent and useless in the
debate, however.

The reality is that commoditization rather than mere monetization
of addresses will lead to ravenous capitalists devouring at least
one aspect of the internet model. To me, the most crucial aspect
of the internet is it's ability to democratize communications.

If you allow ravenous capitalists to consume addresses without
restriction, then, you allow that democratization to be terminated
by said capitalists.


More information about the ARIN-PPML mailing list