[ncdnhc-discuss] ICANN Reform: Role of ITU

[Michael Froomkin - U.Miami School of Law]

On Tue, 30 Apr 2002, Dave Crocker wrote:

> At 01:46 PM 4/30/2002 -0400, Michael Froomkin - U.Miami School of Law wrote:
> >We are not limited to those choices.  Having ICANN devolve and share power
> >regionally and functionally produces a better outcome to either of these
> >Hobson's choices.
> 
> This line of thinking continues to ignore two, fundamental constraints:
> 
> 1.  The hierarchical nature of the DNS requires a single, logical control 
> over allocation/assignment policy.  Any effort to "devolve and share power" 
> must be designed in a way that is compatible with this technical 
> constraint.  However no proposal for this has been put forward to permit 
> honest, diligent analysis of its operational feasibility.
> 

The above statement is a total falsehood and self-evidently wrong as it
is trivial to design systems that preserve the uniqueness of TLDs.
Preserving uniqueness is a bookkeeping job.  It's important, but need have
no other function attached to it - certainly not deciding on the semantic
contect of the TLD, or even who gets it, once basic assignment mechanisms
are in place.  The issue is not 'design' but political will, or lack there

Here are simplistic models to illustrate this point.  Moving from the
baby-talk to real life is left as an exercise for readers.

Model One:

Alice gets five TLDs a year that start with 'a' so long as they are not
yet in the root.

Bob gets five that start with 'b'.

Charles gets five that start with 'c'.
etc.

ICANN monitors to make sure no one tries to claim one that's been already
taken.

Model Two: 

N persons or institutions are given the right to choose one TLD a month. A
lottery is held to see in what order they pick in month one.  Each
subsequent month, the person who picked first goes to the end of the
queue.  All choices are first-come, first-serve, and if you miss your turn
you take what's left when you exercise your right to pick.

ICANN monitors to enforce uniqueness.

The key point here is what constitutes "choosing": a "chooser" picks not
only the TLD string but also the registry operator.  It can do it itself,
or it can delegate on any terms it wishes.

You can complicate this if you like by having ICANN set some minimum
standards for registries.  Myself, I'd only require two things: escrow of
data, and that the registry have a fixed, known, location so that anyone
who transacts with it knows what law applies.

My point is that it is trivial to come up with models that mean any
reasonable techincal criteria.  Designing to meet political needs could be
hard.  Once ICANN shows it has the will to move beyond the Lynn plan,
people will dedicate the energy to working to solve those issues.  But not
before.  ICANN has pulled the rug out from under outsiders who spent a lot
of time trying to solve its problems too often for that to happen without
a committment on its part.  (See,  e.g. shelving of ALSC, wanton amendment
of consensus .org report, and many prior similar acts...going back to the
MAC, at least.)


> 2.  The ability to reach consensus about things is not much better at the 
> regional level than it is at the global level.  For one thing, the 

We don't need to worry about this - local/regional/functional is easier
than global, and if power is devolved in sufficiently diverse ways, you
get competition to act sensibly, since not to act it to lose out to those
who are acting.  The current system creates no such incentive; the
incumbents have every incentive to block new entrants; the Board has every
incentive to maintain scarcity to maximize its power and revenue.

Plus, if you devlove things enough, it doesn't matter if some of the
beneficiaries have rough edges so long as things average out reasonably.
That isn't true of ICANN, where the frictions matter enormously because it
is a single point of failure.   This is the Internet lesson.

> construct of a "region" is often quite artificial.  For another, members of 

(Didn't stop ICANN from doing it for voting purposes.)  Any time you lump
people you are in the realm of the artificial.  Indeed, the Internet is,
the last time I checked, somewhat artificial.  That's life in modernity.
We do the best we can.  It is still better than a single point of
gridlock.

> a region often have widely disparate needs and goals.  Again, any proposal 
> for devolution and sharing of power needs to attend to this core fact of life.
> 

OK, so let's start by giving every country one new TLD of its choice, as a
short of hors derves to do whatever it wants with (including selling it).  
Subject only to FCFS.

And, meanwhile, let's auction a few TLDs to pay the bills.

And let's cut ICANN up:

-One body to do IP#

-One body to keep the master list of protocol numbers and TLDs ("ICAN")

-One body to do the UDRP (a wholly separate skill set from engineering -
get the American Law Institute to take it over, in partnership with
cognate non-US bodies)

-One body to run .arpa and maybe .int (or maybe 2 bodies)

-"ICAN" does no DNS policy at all except specifying minimum tech criteria
to be a registry, keeping the escrow.  Terms for who gets to be a delegate
TLD picker are worked out in advance.


PS. to Alejandro: beware people making comments to you in private they
don't want to see in private, especially if you agree with them. It's the
classic way for people to butter up the powerful in the hopes of future
favors, however subliminal the effect (and, yes, here you as the Board
Vice Chair are indeed the powerful in this case, since ICANN stands as the
chokepoint over a valuable resource).


Note: I had until recently killfiled Dave Crocker.  This post reminds me
why I did it.  And I think, after this, I'll put him back.  A response
below.  I won't respond to further messages from him, because I won't read
them.

-- 
		Please visit http://www.icannwatch.org
A. Michael Froomkin   |    Professor of Law    |   froomkin at law.tm
U. Miami School of Law, P.O. Box 248087, Coral Gables, FL 33124 USA
+1 (305) 284-4285  |  +1 (305) 284-6506 (fax)  |  http://www.law.tm
                        -->It's hot here.<--