[arin-ppml] Q1 - ARIN address transfer policy: why the triggerdate?

Paul Vixie paul at vix.com
Tue Jun 24 11:30:08 EDT 2008


> From: Tom Vest <tvest at pch.net>
> 
> Hi Paul,
> 
> How do you think the "dribs and drabs" are going to make it to the  
> little guys?

some will just sit in the ARIN free pool, unusable by larger entities since
they will be too small to be worth the paperwork.  some will be returned and
handed back out.  some will be illicitly transferred via the same mechanisms
that appear to have aided geoff mulligan lately.  or, 2008-2 or something like
it may come into existence.  or your proposal to reserve some part of the pool
for IPv6/IPv4 gateway networks may gain traction.

> What would cause the owners of something uniquely valuable -- even  
> something that the owner couldn't use directly -- to part with it on  
> "reasonable" terms?

that depends on the coming battle between the regulators and the speculators.

if the buy-price is set by those who need ipv4 space to expand their networks,
then large networks won't bid on small blocks, and the sell-price will be by
definition something small/medium networks can afford to pay (and amortize).

if the buy-price is set by those who want to sell later, as in the current
world oil market, then all bets are off.  the invisible hand of a free market
automatically puts fungible commodities to their "highest and best use" which
in the case of electric power in california in 2001 was to sideline capacity
so as to push up the price.

this buy-price delta is the reason i support the must-qualify provisions in
2008-2, even though i'm still on the fence as to whether i like 2008-2 at all.

this delta is also why i'm particularly concerned about the geoff mulligan
story, and the other anecdotes told here recently, maybe indicating a culture
not unlike that during america's prohibition experiment in the 1920's, where
there's a lot of winking and nodding going on and most transactions are off
the books.  the community has to decide whether this is what we want, since
ARIN could ratchet up enforcement, but if the current system is seen by many
as working, then perhaps the community should investigate a more APNIC-like
transfer policy without any must-qualify provision.

> Assuming they are willing to part with it at some price that's less than the
> absolute max that "the market will bear", what will prevent other players
> who are willing to pay more than little guy "reasonable terms" from buying
> it up and internalizing the difference?

speculation can be a very profitable activity for those who engage in it, but
the resulting volatility is bad for consumers.  i do not particularly want my
actions and values to be monetized in this way, not for ipv4 space, not for
oil, not for electric power in california in 2001.  if we here are amateurs
as randy bush has often said, then the speculators will win no matter what
and we should plan for that future (which is the reasoning that led me to
once call APNIC's transfer policy the "Lay Down and Die Act of 2008").

there are a couple ways to avoid volatility in the upcoming endgame.  one is
to find ways to accelerate ipv6 deployment, such as gating ipv4 allocations
so that they are only available for dual-stack networks, starting well before
runout, perhaps even starting as soon as $NOW.  another is to find ways to
structually drive out speculation, like offering the LRSA, pushing for the
must-qualify provisions in 2008-2, ratcheting up ARIN's enforcement, asking
USG to formally authorize ARIN as its successor in interest for legacy space
and investigating policies like APNIC's "sign an LRSA or lose your in-addr
and whois".

i'd prefer that the ARIN community seriously pursue anti-volatility actions.
(note, as before, this is a personal position, not nec'ily that of the BoT.)

paul



More information about the ARIN-PPML mailing list