[ARIN-consult] discounting registration fees for IPv6 assignments

[Jimmy Hess]

On 10/28/12, Bill Woodcock <woody at pch.net> wrote:
> On Oct 28, 2012, at 11:25 AM, Jo Rhett <jrhett at netconsonance.com> wrote:
> …except that we tried that experiment for, what, eight or ten years?  So if

There may have been an experiment tried for some number of years, but
there were no conclusive results.

> people weren't motivated by the discount-all-the-way-to-free, nor by the
> phasing-out of the discount over several years, do you really think they'd
> be motivated by it being discounted _again_, when they weren't the first

Discount, especially discount to 0  of  new IPv6 requests,  is not
motivation per se.
It is a removal of one barrier.  The removal of a barrier makes other
motivations more effective.

If nobody in an organization finds a motivation,  the org is not
likely to waste the time in even applying for IPv6.

If the organization, or some people in it _DO_ have a motivation,
they may start the steps towards IP deployment,  but when they are
researching what is required for them to obtain address resources,
they find there are these costs,   that the organization cannot
justify.

Therefore, although there is motivation,  the deployment does not move
ahead... it gets delayed for consideration in  2014, 2015, 2016, ....

> It doesn't fix the ridiculously-large volume discounts, but it makes a step
> in the right direction.

IP resources are _not_  assets that ARIN sells to applicants.  ARIN
does not charge a fee for a certain number of IP address seats,  so
there is nothing the least bit wrong  that larger allocations have a
lower average cost per IP.

Now they do have these allocation-size based pricing,  which are
obviously meant to distribute ARIN costs  based on the relative sizes
of networks.

But it's not reasonable to say that the region's most massive IP
networks with the /12s  should pay almost all ARIN's costs,  via
per-IP pricing.

That makes ARIN too reliant on a small number of large organizations
for its survival, which creates extra risks,  those organizations
might object, and just collude and refuse to pay, it   makes entry
costs too low,   and  eventually means that smaller orgs are deprived
of meaningful participation.

> It doesn't get rid of the ISP/end-user distinction, but it makes a step in
> the right direction.

ISPs and End users at not the same.    ISPs  are allocated resources
to reassign to end users  connected to them;  a distinction is
appropriate.


>                                 -Bill
--
-JH