[arin-ppml] Draft Policy ARIN-2014-20: Transfer Policy Slow Start and Simplified Needs Verification

[David Huberman]

John wrote:

> A transfer policy mechanism which allows receipt up to a limit based on 
> current holdings provides far more certainty for those who wish to plan
> for the future, as they can go to market knowing precisely that limit. 

What is the virtue of a limit?  

It's not the prevention of speculation and hoarding.  Those will always 
happen outside the view of ARIN policy.  Speculators and hoarders will 
buy blocks on the open market and simply not engage ARIN because
there's no reason to.  Contract law makes it trivial to ignore ARIN.

It's not conservation - there is no such thing as conservation in IPv4.
85% of the address space ARIN issued over the last 10 years went
to less than 20 companies.  (At a significant penalty, I might add, to
the little guys and especially new entrants, who got screwed in ARIN 
policy for 17 years.)

Before anyone answers this, please ensure you're knowledgeable about
the IPv4 market today. I am. I characterize it as VERY robust.  Tons of 
supply, with new suppliers showing up every month.  Outside of China,
prices are low; it's a buyer's market.  There's no speculation that I can 
find, short of a one-off speculator who is a well-known fraudster. There
is certainly hoarding by the larger companies, but ARIN policy today
isn't stopping that, and no policy passed here can stop that. Think about
that last sentence carefully.  ARIN policy is powerless to stop hoarding.

So how do we write policy that helps the non-big guys?  By removing
artificial policy barriers that require lots of paperwork with ARIN beyond
simply, "I bought this /20 from this company, please update Whois".

ARIN's job should simply be to verify the seller is the bona fide registrant,
and that the seller agrees to the transfer, and that the buyer signs an
RSA and pays whatever fees are necessary to cover the costs of the 
transaction processing.

Let's simplify ARIN processes, make ARIN policy fit the REALITY of 
network operations in a post-exhaustion world, and move on with 
talking about RPKI, DNSSEC, IPv6 and other actually important things 
that will shape our future.

David