On 07/14/2015 06:18 PM, Owen DeLong wrote:
On Jul 14, 2015, at 13:00, Adam Thompson <[email protected] <mailto:[email protected]>> wrote:
[...] but although there's a good technical and contractual justification for giving those entities their own distinct subnets, they have nothing more than a contractual relationship with me. Yet I'm still not an ISP as far as I'm concerned.
By policy definition, you are, actually.
Owen

Yes... in the same way that crossing the street in the middle of the block at lunch yesterday, and speeding slightly on the way to work this morning, made me a criminal. Yet, strangely, I don't think of myself as a criminal.

One of the things every so-called 'leadership' school teaches is "Never give an order you know will not be obeyed".

While it's origins appear to be military, a corollary has been stated[1]:
“Never pass a law that huge numbers of people will break”. Passing such laws does little or nothing to change human behavior, but does a great deal to undermine the rule of law.
Policymakers are subject to the same forces - if a policy is ignored or deliberately violated by a sufficiently large group of affected entities, it ceases to be usefully enforceable. "Sufficiently large" is directly relevant to both the enforcement resources available and human group dynamics.

As I stated in a previous email, I am personally aware of dozens of "enterprises" who have partially-ceded control of their IP address assignment to 3rd-parties for entirely valid, normal, non-ISP-like reasons. Off the top of my head, extranets, HVAC systems, and security systems all come to mind. (For clarity, yes, I'm still talking about the public portable address space.) IMHO, it doesn't make sense to suddenly call them ISPs unless renting IP address space is a noticeable part of their business, or perhaps if a substantial part of their allocation is partially under someone else's control.

If you think it does make sense to call them an ISP, then I think that's a signal that it's high time to discard the distinction between ISP and Enterprise, and move to a fee-for-service model instead.

My understanding - and this could easily be wrong - is that the distinction originally came into being because of the different amounts of effort required on ARIN's part to service the customer; an ISP was expected to cause significantly more work over a year than an enterprise.

I get that as an enterprise, paying the lower fee structure, I shouldn't have access to all the "features" an ISP - paying the higher fees - does. That's fairly straightforward. But on the flip side, some ISPs are SWIPing and requesting new allocations all day long, but some only do it once a year. Lumping them together isn't fair, either... if an ISP never delegated address blocks and had relatively static requirements, why on earth would they self-declare as an ISP? I'm not *saying* I know anyone who's doing that, but...

(I also favour a so-called "flat tax", could you tell? <grin>)

Reducing the distinctions between ISP and Enterprise makes sense to me.
Unfortunately, I don't think the financial discussion can possibly happen only *after* the theoretical discussion, as originally postulated. This is an area where IMO the usually-clear(ish) distinction between Community, AC, Board and Staff won't entirely work.

-Adam


[1] I've heard it before, but the only reference I could find to it today was http://www.twopotscreamer.com/never-give-an-order-that-you-know-will-not-be-obeyed/. FWIW, Gen. D. MacArthur also said "Never give an order you know can't be obeyed", which isn't quite the same thing.
_______________________________________________
PPML
You are receiving this message because you are subscribed to
the ARIN Public Policy Mailing List ([email protected]).
Unsubscribe or manage your mailing list subscription at:
http://lists.arin.net/mailman/listinfo/arin-ppml
Please contact [email protected] if you experience any issues.

Reply via email to