[ppml] Policy Proposal: Expand timeframe of Additional Requests

[Ted Mittelstaedt]

>-----Original Message-----
>From: David Conrad [mailto:drc at virtualized.org]
>Sent: Thursday, August 16, 2007 4:24 PM
>To: Ted Mittelstaedt
>Cc: Public Policy Mailing List
>Subject: Re: [ppml] Policy Proposal: Expand timeframe of Additional
>Requests
>
>
>Ted,
>
>On Aug 16, 2007, at 2:52 PM, Ted Mittelstaedt wrote:
>> The first hoarder to sell IP's will no doubt try what your saying -
>> the next one will undercut, the thrid one will undercut more, and  
>> so on.
>
>This is, of course, assumes a market in IP addresses will behave like  
>any other market.  However, there are some who do not believe this  
>would apply to IP addresses and that speculators would engage in  
>"Enron-like" behavior, driving the price of IPv4 addresses to ruinous  
>heights. 

Fine with me since that would impact the large networks ability to
complete if they chose to not push IPv6 - and it is only the large
networks pushing IPv6 that will get the IPv6 Spruce Goose off the
ground.

>I suspect reality is somewhere in between and largely  
>depends on the number of sellers.
>

And whether the community carves out exemptions to allow it.

>> Keep in mind what the hoarders are doing is borderline criminal - they
>> lied on the RSA to get the stock, if ARIN catches up with them ARIN
>> could revoke the entire assignment.
>
>As I mentioned, there are around 100 legacy /8s.  By definition,  
>these addresses were not assigned under an RSA.

Under a current RSA.  I had thought there WERE criteria for the
original assignment.  ARIN also is not under obligation to manage
WHOIS records for the legacy holders, and is doing it for the good
of the NON-legacy holders, mainly.  If the legacy holders get
heavily into IPv4 sales it will destroy the usefulness of the
whois records that cover the legacy holders, thus there would no
longer be an incentive to the community to have ARIN keep track
of it. At that point the community could have ARIN drop all whois
records for the legacy holders and it would essentially throw those
blocks to the dogs.

You do realize ARIN won't allow a transfer of a legacy block to a
"purchaser" without a signed RSA and billing commencement, regardless of
whether a legacy holder has made a side deal to "sell" the block to
someone.

There's a lot of ways I can think of to make hoarding and sale of
legacy blocks an extremely uncomfortable thing for the legacy holders
to engage in.  I'm sure the community can think of many more.

Furthermore since the legacy holders have held those blocks for so
long I think it's not reasonable to assume they aren't already using
quite a large amount of them - when they know they have them of course.
I suspect a lot of legacy blocks are tied up with entities that have
no legal existance any longer and are kind of in limbo.

> One may assert that  
>those addresses implicitly fall under the RIR policy regime by  
>tradition or some such, but I honestly can't imagine such assertions  
>would go unchallenged.
>
>> Another analogy is Freon, AKA Refrigerant 12.
>> ...
>> What happened is
>> that within 3 years the price of Freon on the market crashed, and  
>> today
>> there's tons of it out there, even though it hasn't been  
>> manufactured for
>> years.
>
>An interesting analogy.
>
>> If IPv4 "sales" are allowed widely it will drive a stake in the  
>> heart of
>> IPv6 deployment and billions of dollars that the large networks  
>> have already
>> invested in getting ready for IPv6 will have been wasted.
>
>An interesting assertion that might need a bit of explanation.

OK well here's the idea on that.  Organizations are naturally
going to do the cheapest thing.  If ARIN and the community makes
IPv4 "sales" easy, pleasant, and sanctioned things, and there's
a lot of IPv4 rolling around that is unused and easy to untie,
then hoarders are going to flourish, the price of IPv4 will be
low, and most ISP's will simply buy IPv4 from a broker rather than
bothering with waiting in line for a IPv4 block to become freed
up in one of the RIRs.

But I just cannnot see how you could make such a thing happen for
IPv4 sales without running afoul of a lot of policy and assumptions
that people have made so far.  The utilization requirements for
one thing - how can a hoarder operate without having a large stockpile
of unused numbers, and how can they get this without lying - unless
they were a legacy holder - in which case the legacy holders sell
off their holdings and ARIN forces the "buyers" to sign an RSA and
meat utilization requirements, or
the legacy holders attempt to "rent" the IP's out while not changing
the whois data - which ruins whois, and thus provides incentives for
the community to just drop recording of legacy IPv4 blocks in the
public records.

You cannot have it both ways as the saying goes.

Worse, what would you do if you were a
legitimate org that had been paying your money for the last 6 years for
your /21 or whatever, and then found out you could turn it back in
to ARIN and go to a broker and rent IPv4 from the broker for a quarter
of the cost of what you had been paying ARIN?  I would think a lot of
people would go to the broker and sing a contract with them - that
reduces control over the blocks that the RIRs have, the RIR's wouldn't
stand for it.  Then all the IPv6 incentive policy we have been talking
about here becomes useless because the people paying for the IPv4
aren't paying the RIR now they are paying the broker and the broker
is going to have an incentive to make them NOT update to IPv6 (since
that will put him out of business) by making it cheap to retain the
IPv4.  And how does the RIR you apply incentive policy to a broker
if they are going to allow brokering/hoarding?  The whole thing
becomes a house of cards with contradictory and hypocritical policy
that is just going to invite lawsuits.

>An  
>alternative view is that the unpredictable (other than being higher)  
>cost of obtaining IPv4 will encourage ISPs to migrate to IPv6 more  
>quickly in order to regain some level of predictability in the costs  
>of addressing.


That too.

Ted