[arin-ppml] Needs assessment

Owen DeLong owen at delong.com
Wed Jun 5 18:16:09 EDT 2013

> * people who attack the elimination of needs assessments on the grounds that "there are bad people out there who want to make money on addresses" seem to be missing the point. Nearly all of the organizations applying for IP addresses are trying to make money on them in one way or the other. Or did you think Google, Amazon and Verizon were charities? [1] Do you not understand that IP addresses are part of the service bundle offered by ISPs, cloud providers, hosting providers, etc., and that they are being "sold"? What is it, exactly, that makes selling a number block as an alienable commodity intrinsically evil but selling it bundled with internet service worthy and wonderful? 

That's not the basis I would use to attack removing needs basis, but, let me see if I can help you understand...

Many people feel that it's OK for you to make money selling software that you wrote. Many of those same people feel that it's perfectly fine if you write your software using libraries provided under the BSD or LGPL licenses. One could argue, as you do above, that this amounts to selling the libraries since they are inherently incorporated into your software (or at least parts of them are). However, even the BSD license (which is about as liberal as you will find on this subject in Open Source Software) does not permit you to sell the unmodified library as your own product.

It is my considered opinion that IP number resources are not property. Attempting to claim ownership over the number 3 makes about as much sense as attempting to claim ownership of an ocean. IP number resources are, instead, a community resource pool managed according to policies set by the community. ARIN (and the other RIRs and even ICANN to a certain extent) act as vehicles to facilitate the development of that policy and to apply that policy in the operation of the registries that track and support the effective utilization of the community resource pool.

If one were to look for an economic model that most closely represents this, I think a public library is closer than anything else in current practice. When you come to the registry, you can "check out" a set of resources providing that you can show appropriate policy conformance in your intended use. Your ability to use those resources is renewed on an annual basis so long as you continue to use the resources in accordance with that policy.

The policy at this time incorporates many things, including needs-basis assessment which you are arguing against. Some of us feel that needs-basis is an important tool in properly managing these community resources. We understand that you disagree. There have been many successes and many disasters throughout history using the model you propose as being ideal. I cannot point to a time in history where any library went to a policy of "if you have enough money, you can simultaneously borrow every book in the library as long as you want", so I cannot really evaluate how well such a policy would work out for libraries. Can you point to an example in history where such a policy has worked?

> In short, the removal of needs assessments from transfers after a free pool is depleted is a minor but important adjustment that we make in the dying days of IPv4. Time to relax about it. Dispense with the religion, and  focus laser-like on what makes for the most efficient methods of moving IPv4 numbers to their most highly valued use, quickly and with minimal friction. 

It's not a minor adjustment and, IMHO, it's not only unnecessary, but actually likely to be quite harmful. I'm really tired of your use of pejorative terms to describe any opposition. You're on the AC now. Please act like it. People who disagree with you are not inherently incorrect, inferior, or otherwise less important than your own opinion. I, for one, believe that determining the best use of IPv4 addresses strictly in terms of economics would be quite contrary to some of the most important uses of them today.


More information about the ARIN-PPML mailing list