Re: [wg-c] Eureka?

[Jonathan Weinberg <weinberg@mail.msen.com>]

At 09:45 AM 8/6/99 -0700, Kent Crispin wrote:

>On Thu, Aug 05, 1999 at 11:47:19AM -0400, Jonathan Weinberg wrote:
>> 	As for structured name space: I've indicated in this list that I think
>> lots of new gTLDs would be a good idea.  I've indicated that competitive
>> registries choosing their own TLD strings may be a good idea.  I've
>> indicated that a centrally planned name space may be a bad idea.  That
>> doesn't make the issues inextricable.
>
>What makes them inextricable is the fact that there is only one name
>space.  Either it is planned, or it is not.  It is simply incoherent
>to say that it is planned, but that we allow free choice of names to
>some set of participants -- that free choice undermines any
>possibility of planning.
>I don't like to use analogies, but perhaps one will help: You have a
>tract of land, and you say part of it will be planned development,
>and part of it will be given to a developer to develop, free of any
>plan.  The planned part is to be for residential units.  The 
>developer decides to build a toxic waste dump.  The two uses don't 
>work together.  In fact, the use of the land as a toxic waste dump 
>prempts the use of any of the nearby plots for most purposes.
>Of course it is not possible to plan everything in advance, and any
>use will preclude some other choices -- that's a given.  But one can
>be prudent in ones choices, and pick things with wide acceptance, or
>that have relatively less impact.
>
	I think we must be talking past one another.  I agree that allowing a set
of TLD operators free choice of names can disrupt a centrally planned name
space.  When I said the issues weren't inextricable, my thought was that
whether there should be a lot of gTLDs or only a few, and whether the
namespace should be planned or unplanned, are separate issues.  One can
believe in having a lot of gTLDs and a planned namespace; one can believe
in having a lot of gTLDs and an unplanned namespace; one can believe in
having only a few gTLDs and a planned namespace.  (I don't know whether I
know anybody who believes in having only a few gTLDs and an unplanned
namespace.)  The point I made earlier was that *if* one believes in a
planned namespace (as you say you do), then it doesn't seem to me to make a
lot of sense to say "well, let's pick five gTLDs and add them now, and
later on we'll figure out how -- and whether -- we can make them fit into a
planned structure."  Indeed, I think the sort of ad hoc approval that would
result from the "series of working groups" approach you've advocated in
other posts would give us neither the benefits of planning nor those of
market processes.

>I believe you have indulged in a rather serious implicit
>mischaracterization what I have proposed, as well.  You speak of
>"planned namespaces" with echos of faceless gray secret soviet style
>planned economies. 
>But the planning I refer to is planning through public open processes
>in a manner common in civilized democratic societies.  If we may
>pursue the real estate analogy just a little further: land use
>planning is a ubiquitous feature of modern urban society.  Many
>people would like to treat the DNS as a wide-open frontier landrush,
>where we need to support homesteading capitalists, and give them the
>widest possible leeway to develop the wilderness.  But many others
>believe that we are well past that stage, and the shop keepers and
>farmers need a more settled and controlled environment.
>
>[snip]

	Central planning isn't necessarily bad; as you point out, zoning is
central planning, and it usually works well (and better than the
alternatives).  When I first started out thinking about these issues, it
seemed to me that NewCo should structure the DNS based on the Yellow Pages
model.  But later on I became more skeptical.  For central decisionmaking
to make sense, one's got to believe that the structure the planners will
come up with will be better than the one that would grow organically out of
market processes.  While I'm open to argument, I'm not sure that's true in
this case.  I've worked in the government, and I've studied communications
regulation for a long time.  Both of those experiences tend to teach one
that even really well-meaning central planners (and even ones who engage in
"public open processes") often do a really bad job.  If conditions are in
place that would allow meaningful market selection of TLD strings, I'm not
at all sure that wouldn't be a better approach.
>
>
>To extract the meat from your last sentence above, you basically say:
>"The TM community is unreasonable."
>I presume you base this position on statements to this list that TM
>folks have blocked new gTLDs in the past -- that would be
>understandable, given some of the previous posts.  
>The similar condensation of your more complete position, as I
>understand it, goes like this:
>"The TM community is unreasonable, so we must structure things so 
>that they will be forced to allow more gTLDs, because I believe that 
>'more gTLDs' is an intrinsic good.  To make them feel better about being 
>forced, we will do it over a year or two."
>However, I think your position (as I understand it) is based on at
>least three bad assumptions:
>1) TM people are unreasonable;
>2) their actual (as opposed to strawman) position is unreasonable;
>3) they are the only ones who favor a cautious approach.
>I realize I have characterized your position in a rather bald way, 
>but I think it is important that you understand how it looks to me.
>Is there any fundamental error in the way I have characterized
>it?

	I base this position on my observations and experience — like you, I've
been involved in this process for a while.  (No, I haven't been involved as
long as you have.  You've been involved since God was knee-high to a
grasshopper.  I've only been involved for two years.  But it's enough.)
When I was working for the government, I talked to TM folk (and to others
as well) about DNS issues.  I don't want to knock the TM community.  I like
and respect a number of TM lawyers.  (And in general I'm not a big fan of
saying critical things about other people.)  But I think it's the case that
an awful lot of the TM community are, well, single-minded about their
concerns.

	I'm not sure I understand your position that adding a lot of new gTLDs is
"forcing" TM folk to do anything.  Adding new gTLDs quickly doesn't "force"
opponents to "allow" that rollout any more than adding them slowly "forces"
advocates of the opposite position to "allow" that.  I do think that an
approach under which ICANN plans out a several-year phased rollout, and
places the burden on opponents, if evidence comes in demonstrating that
additional new gTLDs are a bad idea, to bring that evidence to ICANN's
attention and call for a halt or a slowdown, is the approach best
calculated to reach good results.

Jon


Jonathan Weinberg
weinberg@msen.com