[arin-ppml] Draft Policy 2009-1: Transfer Policy - Revised andforwarded to the Board

[michael.dillon at bt.com]

> Indeed.  Right now, the cost of inefficiency is zero and the 
> benefit to fixing it is zero, while the cost of fixing it is 
> usually nonzero.  It's not surprising that we've seen the 
> results we have.  Attach a monetary value to the resource, 
> though, and the entire picture changes.

Not unless the monetary value is above some threshold. For
instance, I've been working at companies paying ARIN's
highest fee level for some time. To us, it is effectively
zero. We know addresses are needed to make sales, and the
number is so low that nobody blinks an eye at it.

What threshold needs to be passed before the big companies, 
with the largest stocks of IPv4 are likely to take notice.
If these companies don't put addresses up for sale, there
will be nothing to buy. And if you don't get some million
dollar transactions to actually prove the monetary value,
then there is zero motive to act.

IPv4 runout raises a whole bunch of other issues that require
attention, such as IPv6 testing, systems support for IPv6,
migration off of some old devices that should have been gone
long ago, auditing IP addresses so that corporate officers
can attest to ARIN, chasing recalcitrant vendors who do not
yet have IPv6 roadmaps, arguments about whether NAT really
could mitigate any problems, and so on. With all of this
happening, I really think that it will be very hard to get
any attention for a plan to sell off some IPv4 address blocks.

Talk to Leo Vegoda about all the signoffs that he needed to
get to recover 14/8, digging through multiple layers of
mergers, acquisitions, spinoffs and demergers to find anyone
who considered themselves to be a stakeholder. I think that
roughly 100% of big ISPs will be in the same complex position
with regard to getting signoff for selling an IPv4 block.

This is just not a feasible direction to go in.

--Michael Dillon