[ppml] Proposed Policy: IPv4 Countdown

>-----Original Message-----
>From: Stephen Sprunk [mailto:stephen at sprunk.org]
>Sent: Friday, March 16, 2007 7:47 PM
>To: Ted Mittelstaedt; Owen DeLong
>Cc: ARIN PPML
>Subject: Re: [ppml] Proposed Policy: IPv4 Countdown
>
>
>Thus spake "Ted Mittelstaedt" <tedm at ipinc.net>
>>>Nothing is simple when lawyers get involved, and this sort of
>>>contract is novel in a variety of ways, so until it's been tested in
>>>court, we really don't know how enforceable it is.
>>
>> OK so according to this logic since no contract is really valid until
>> it's tested in a court, there is no point for businesses to bother to
>> sign contracts with each other because none of those contracts
>> have been tested in court and so are bogus anyway.
>
>Ah, good ol' reductio ad absurdum...
>

Hey, you started it.

>New contracts are generally based on prior contracts and case law
>so one has
>a basis to determine how courts will view them.  While I'm sure ARIN's
>counsel has done their best with the RSA, it is unique enough we're really
>not sure what the courts will think of it.  ARIN is still in court over
>their first lawsuit (well, I haven't seen any discussion of a
>conclusion to
>the case yet) regarding the RSA, so we don't really know what its
>status is.
>The result could be anything from 100% enforceability to anti-trust or
>restraint-of-trade actions against ARIN.
>

Interesting hand waving.  So I guess then that nobody can write a
contract over any new idea, product, or service until it is court-tested.

Yah, right.

Well, the lawsuit your referring to - Kremen VS Arin, available
here: www.internetgovernance.org/pdf/kremen.pdf has enough huge flaws
o drive a truck through.

For starters Kremen claims to have a judgement against all assets of
Cohen.  Kremen claims Cohen has assigned numbers out of his block to
Mexican ISPs and is getting money from this.  However if Kremen has an
actual judgement against Cohen that imposes a trust on Cohen's assets,
(as the lawsuit claims) then any money that Cohen would get from this
alleged unnamed Mexican ISP would go immediately into Kremem's bank account.
So I do not see how exactly that Kremen can claim any harm here since
by court order Cohen's income is his, and further, Cohen is in a US
prison.

Secondly, Kremen claims that not having IP numbers is detrimental to
his business.  However, if he wants IP addresses all he needs to do is
go to ARIN and request assignment through the usual means.  Cohen's
addresses - that Kremen owns the revenue from - cannot be used to deduct
from the total number of addresses that Kremen is entitled to, assuming
he proves usage.

What the entire lawsuit boils down to is Kremen is suing ARIN because
as part of ARIN's requirements Kremen must agree to disclose what he
is doing with the IP addresses and who is using them, and Kremen does
not want to do this.  That is unsurprising since Kremen is a spammer.

I fail to see how Kremen can prove in a court how disclosure of what
he is doing with the IP addresses is harmful to his business - unless
that is he intends on doing something illegal like violating the federal
CAN-SPAM act - and that is the central basis that Kremen is resisting
simply requesting the allocation from ARIN.  ARIN has in fact said
they would give Kremen the allocation - assuming he signs the appropriate
paperwork which requires him to disclose what he's doing with the IP
addresses and demonstrates appropriate need.

The other big flaw is the complaint assumes ARIN is a monopoly and
thus illegal under Sherman Anti-trust act.  What this conveniently
ignores is the fact that just because an organization is declared a
legal monopoly does not mean that it's actions are illegal.  If ARIN is
in fact found by the court to be a legal monopoly then it is a huge
stretch to assume it will have to be broken up.  What has happened
recently with technology companies that have been shown to be monopolies
is the courts have favored regulation over divestiture, as a remedy.
That is what happened with Microsoft, and Intel, recently.  (in fact
the Intel complaint didn't even make it to a court, it stayed with
the FTC)  You see, what the Sherman Anti-trust act assumes is that
monopolies always harm the consumer, thus it is best to break them up.
What was shown with the breakup of the Bell Telephone company in the
80's is that divestiture only helped long distance customers, it was
in fact harmful to local dialtone consumers and did not lower prices
of local dialtone.  That is why courts are very leery of breaking up
technology monopolies these days, and are favoring regulatory relief
instead.

Since ARIN is already subject to
regulation by ICANN I do not see how much would change if in fact the
court did declare ARIN a monopoly.  In fact all that would likely happen
is the court would sit down and review the rules regulating ARIN and would
almost certainly uphold all of them, as a result Kremen would be in
worse shape because now he would have a court upholding ARIN's right to
make him sign a contract to get his netblock.

>> Or better yet, businesses should sign contracts then immediately
>> break them so they can go to court and have the court rule if the
>> contract is valid or not so they can know if it's OK to break it.
>>
>> You have been watching too much LA Law on television.
>
>Never saw it.  This is based on discussions with my own attorney regarding
>drawing up personal contracts, negotiating contracts with partners and
>customers at my employer, discussions on this very list regarding the RSA,
>etc.
>

My attorney says your attorney is full of baloney.

>>>All of the address space is either reserved or assigned to an
>>>RIR.  The issue is that legacy assignees/allocees(?) are not
>>>bound by any contract with their respective RIR that dictates
>>>the terms of that relationship.
>>>
>>>In theory, since there's no contract ARIN has no legal obligation
>>>to maintain those registrations, but the community has, to date,
>>>felt that there is a moral obligation to maintain them at no cost.
>>
>> What exactly do you think "maintaining" something means?
>> OK, I'll capitulate.  I agree with you that ARIN should initiate
>> processes to reclaim IP addresses from non-contracted address
>> space because to not do this would mean that they are failing
>> in their duties to maintain such space.
>>
>> You cannot have your cake and eat it too, I am sorry to say.
>
>You can't agree with me, because I said nothing of the sort.  ARIN is
>maintaining registrations for legacy space (and for most IPv6 space, for
>that matter) without fees.  They're actively providing a service
>(if nothing
>more than rDNS in many cases) at no charge because that's what the
>community
>has told it to do.  Part of the reason the community has made that
>decision
>is because we're unsure if we have an alternative, but that's not the only
>reason.
>
>I would support a policy proposal that directed ARIN to actively try to
>reclaim address space that was no longer in use, regardless of
>what terms it
>was assigned/allocated under.

Well, good because that is what I am in favor of as well.

> I would _not_ at this time support
>a proposal
>to revoke registrations for space that _was_ being used simply
>because there
>is currently no fee attached, or to impose a fee on registrations that
>currently do not require one.
>

Neither would I.  Rather, I would base revocation on a number of
other factors.  For example, is the netblock in question being advertised
via BGP anywhere on the public Internet?  Can I sit at my Internet-connected
PC and run a traceroute command with a destination of an IP in a netblock
and have it actually go somewhere other than to my default gateway where
it then fails due to network unreachable?

>Side note: if ARIN _were_ to start charging for services that are
>currently
>free (including, but not limited to, maintenance of legacy
>blocks), we would
>need to either lower fees across the board, accelerating consumption*, or
>find new expenses (that didn't violate ARIN's charter!) to offset the
>additional revenue.  Both of these results seem to be of dubious value to
>the community.
>
>>>Then quit arguing about it and submit a policy proposal.
>>
>> Someone was complaining a few days ago about policies written
>> wth no input.  Now your wanting policies written with no input.
>> I guess there is no satisfying people.
>
>Float a draft of _actual changes_ to the NRPM and see what people want you
>to modify in order to gain their support.  Waving your hands
>saying "the end
>is near!  we need a policy!" is not constructive.
>

Now see here, -I- didn't start doing that.  I merely pointed out to
the originator of this discussion that attempting to artifically end IPv4
assignments early was an unworkable idea for a variety of political
reasons.

>>>Do note that the projections for how long address-reclamation
>>>efforts will extend the exhaustion are on the order of six months.
>>>That means it'll take us longer to reach consensus and
>>>implement the changes than the period of time we'll buy by
>>>doing so, meaning we're better off _not_ doing it and instead
>>>spending our time figuring out how to get people on IPv6.
>>
>> Ah, now the truth comes out.  You want IPv6 and are happy to see
>> reclamation efforts on IPv4 fail so it hastens the day for IPv6.
>
>We do not have a choice.  The IPv4 address space _will_ be exhausted, and
>it'll happen in about four years if we do nothing.  The best
>projections, by
>people who are quite authoritative on the matter, is that reclamation will
>buy us six more months.  If you want to claim, as you did in another
>message, that it'll really buy us 5-10 years,

I claimed the possibility existed that it might buy another 5-10 years.

> you need to come up with
>better studies and data than we already have.

I don't see how anyone including myself can come up with any studies worth
a damn that will predict how reclamation turns out.  Nobody really knows
how much of the allocated address space out there is really being used
according to the justifications submitted when it was allocated (years
earlier in many cases) and a lot of assumptions are sheer speculation.  Such
as the assumption by another poster on this list that growth of nodes on
the Internet automatically means consumption of IP addresses rises.
Technologies such as NAT and virtual hosting throw a wrench in that kind
of assumption.

You have to simply make a decision, are we going to try reclamation or not?
If we are, then the next decision is are we going to try reclamation based
on monetary adjustments, (changing fees) or by setting policy?

if we are not going to try reclamation then the best thing would be, now, to
set a date for converting the Internet backbone to IPv6 that is well within
the most pessimistic assumptions out there for the end of IPv4 allocations.
That would completely avoid all "run up" of IPv4 by speculators, it would
also pull the rug out from under ill-conceived lawsuits like this Kremen vs
ARIN thing, it would give all the ISPs a clear date for budgeting hardware
upgrades, and it would make the transition as smooth as possible.  It would
also prevent ISP customers from endless chasing after "the lowest bidder"
ISP, it would discourage hoarding and other contactural violations and in
general be kinder to animals and plants.  Naturally you would assume that
a lot of people might simply buy proxy servers instead of renumbering.

This is how the Great Renaming worked for DNS.

>And we'll be in the
>same boat
>again after that much time anyways, so we might as well save the
>effort and
>convert now before we buy/deploy _another_ hundred million routers and PCs.
>

Well, that's the argument against reclamation.  But, what are YOUR
arguements
FOR any alternative?  I have read a lot of posts on the lists bitching
against
reclamation efforts, but none of the bellyachers have submitted any kind of
alternative other than "let's just do nothing until the last minute then -I-
will switch over to IPv6 and laugh at the rest of the dummies who are
runnning
around trying to preserve their IPv4 networks"

>I don't buy the "all the studies show I'm wrong, but let's try it anyways
>and find out while wasting millions of dollars and putting off our only
>remaining viable option" argument.  But that's just me; float an _actual
>proposal_ and see what others think.
>

Your building a house of assumptions there, not the least of which is that
reclamation efforts will cost millions of dollars.

>S
>
>* If one asserts that higher fees would discourage consumption, one must
>also accept that lower fees would encourage consumption.  Either
>consumption
>is linked to price or it isn't.
>

Of course it is but the problem is that it is not directly linked, there is
some elasticity there.  A 10% jump in fees will probably not result in a 10%
increase in IPv4 being returned.  If anything, it will probably cause people
to dig in their heels and hold on to what they have and start viewing it as
an "investment" with the expectation that the price will be run up by the
market and they will be able to make a killing.

I do not favor setting fees as a tool for changing IPv4 allocations.  I am
much more in favor of setting policy then if people violate the policy,
suing
the pants off of them.  This does have a requirement that the organization
be run by someone with balls.  Unfortunately, I am in a minority because
most
people subscribe to the belief that all we have to do is adjust pricing and
people will do what we want - like magic! As a result if a reclamation
policy
does get passed, it most likely will be a limp-handed politically correct
poofball that will make all kinds of naieve assumptions that organizations
will
fall all over themselves to return unused space just because they think it
might cost a bit more money to hold on to it.

Ted