more on dot "ex-ex-ex" domain boondoggle, err, proposal, back again

Begin forwarded message:

From: Dave Crocker <[email protected]>
Date: January 7, 2007 6:18:00 PM EST
To: [email protected]
Cc: [email protected], Bob Frankston  
<[email protected]>, Lauren Weinstein <[email protected]>
Subject: Re: [IP] dot "ex-ex-ex" domain boondoggle, err, proposal,  
back again


> Here we go again ... brace yourself for another round of knee-jerk  
> punditry
> about Sex And The Internet. I wish more writers would point out, as  
> I've
> written before, that dot-ex-ex-ex does not, in fact, actually *DO*  
> anything
> notable in terms of ratings.
...
> What ex-ex-ex *will* do, however, is produce an ongoing stream of  
> monopoly
> rents for the company proposing it (ICM Registry),


Dave,

The AP article looks pretty coherent on the topic, notwitstanding its  
reference to ICANN as "the Internet's key oversight agency."  
Certainly the ICM Registry proposal for .xxx membership looks like it  
has some interesting ideas, although it is not clear to me that they  
can be enforced for the long term.

With respect to ICANN's strategic role, here, it's nearly impossible  
to extricate underlying policy issues from the strong political and  
emotional distractions that are pornography's baggage. However a  
meaningful discussion of ICANN's role and choices, here, requires the  
separation, if the current choice is to be based on anything other  
than politics and emotion.

A decision was made, very early in the history of Domain Names, to  
have multiple
"templates" for the semantics of Top-Level Domain (TLD) labels.  DNS  
naming does
not require any particular scheme for choosing its labels, and a  
number of
different ones have developed.

In effect, we now live with 3 basic categories:  Geographic,  
Administrative,
Generic (GTLD).  There is nothing to prevent having more, but then,  
it probably
would not be any fun going through the hassle of gaining consensus on  
new ones...

The original GTLDs described a small set categories for registering  
institutions:

    MIL[itary]
    GOV[ernment]
    EDU[cational]
    [non-profit] ORG[anization]
    NET[working-specific]
    COM[mercial].

(It is also probably worth noting that these categories do not have  
to be
mutually exclusive.  For example, non-profit networking organizations  
and
for-profit educational companies exist, and there is no inherent  
reason they
should be limited to a particular GTLD.)

These categories served as registration criteria that were initially  
enforced.
Over time, the restrictions went away for .com, .net and .org. So we  
now have a
two-category policy framework for GTLD names.  Some GTLDs are merely  
descriptive
and others are highly prescriptive. That is, some represent a label  
for which
registration is open to anyone who considers it appropriate, while  
other labels
are restricted to those who satisfy specific criteria.  There are  
reasonable
arguments for and against both categories, but we are long past  
considering
whether to have either.

So the current controversy exists within the context of an  
established model
that provides for having some GTLDs restrict registration to  
organizations that
satisfy a clearly-documented set of criteria.  Not surprisingly,  
agreeing on the
labels and the criteria has not been easy, and there is no reason to  
expect
future choices to be different.

Examples of initial choices included:  Should .gov and .mil be  
restricted to government and military organizations only in the U.S.,  
given the history of DNS registrations?  Should .edu be restricted to  
4-year colleges and above, 2-year and above, any "regular" academic  
institution, or any organization that does teaching, or...? None of  
the previous restricted categories has been all that straightforward  
to decide, and some have undergone change, over the twenty-year  
history of DNS.

This sub-category is now called "Sponsored" because on-going  
responsibility for
administration of a qualifying GTLD is assigned to a sponsoring  
organization.
(This creates even more fun for the topic, because it requires  
deciding whether
the restrictive criteria are specified solely by ICANN or whether the  
sponsoring
organization can define the rules, as well as who to choose as sponsor)

One of the meta-policies for ICANN has been to decide whether it is  
going to have trivial rules for creating sponsored GTLDs or careful  
rules.

Trivial rules will mean a proliferation of Sponsored TLDs, including  
many that
are silly, wasteful, bogus, etc., along with some/many that are  
reasonable.
ICANN has chosen, instead, to add Sponsored TLDs slowly and only  
after extensive
consideration.  Of course, this does not really mitigate against  
having a new
one be silly, wasteful, or bogus.  But it does guarantee that the  
number of
potentially silly, wasteful, bogus Sponsored TLDs is small. (It also  
ensures a
small number that are useful.) Best of all is that it ensures there  
will be a maximum of controversy, for each one, since the Internet  
community gets to focus its attention on each one, over an extended  
period of time.

So, .xxx is merely an extreme example of the challenges associated  
with having
any Sponsored GTLDs at all.  Underlying the strong controversy about  
this
particular one are concerns about "legitimizing" pornography,  
misrepresentation
about the membership in .xxx or the implication of safety for the  
rest of the
domain space, and so on.  Then of course, there is the small matter  
of differing
norms and laws around the world, where the DNS must somehow survive  
the gauntlet
of possible criminal and civil interventions against a particular  
name and its
particular policies.

In other words, when there is a registration policy pertaining to the  
nature of
the organization, there are potentially major implications about  
membership and
non-membership. Unfortunately, ICANN's current approach places the  
responsibility for those implications on its own shoulders.

For me, the interesting question is whether there is a way it can  
delegate that
responsibility?

If ICANN were to choose a relatively open policy that permitted  
creating many
new GTLDs (sponsored or not) then it could probably stand back and  
let all sorts
of experimentation take place, including names of the type of .xxx.   
Certainly
there are many proponents of this path.  Howeever it would probably  
lead to a
very large number of TLDs; at a minimum there is some concern that  
having many
thousands of TLDs could hurt overall DNS performance.

More importantly, I think, is that that particular meta-policy  
decision ship
sailed a long time ago.  I'd be surprised to see ICANN bring it back  
home for
re-fitting.  However it might be worth considering, given that the  
current
approach implies that ICANN has the ability to navigate these  
difficult waters
for each "interesting" GTLD.  On the other hand, would a more open  
policy really
relieve it from the vagaries of legal intervention for GTLDs that are  
not only
controversial but might run smack into some jurisdiction's moral and  
criminal wall?

It is difficult to imagine a good outcome, either by creating .xxx or  
by refusing to create it, since there are such vigorous proponents  
and opponents in both directions, nevermind laws that support each side.

Given the lose-lose nature of the current candidate TLD, I find  
myself wondering
which choice leaves the loser with the weaker argument about damage.   
If .xxx
is not authorized, can its proponents claim that there is real damage to
potential registrants, their customers, or the Internet as a whole?   
If .xxx
does get authorized, can its detractors claim similar types of damage?

Given the range of global laws on pornography, my guess is that .xxx  
proponents
have the weaker claim for damages, since they are still free to  
register under
perfectly reasonable alternatives.

But then, IANL.

d/

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


-------------------------------------------