Doing the same thing over and over again ...

[Ken Ryan <wkryan@xxxxxxxxx>]

and expecting different results?

The 1998 White Paper proposed a number of coordinated functions for
the organization that became ICANN, including: oversee policy for
determining the circumstances under which new top level domains would
be added to the root system, and coordinate the assignment of other
Internet technical parameters.

The 1998 MoU states under Principles: This Agreement promotes the
management of the DNS in a manner that will permit market mechanisms
to support competition and consumer choice in the technical management
of the DNS.  This competition will lower costs, promote innovation,
and enhance user choice and satisfaction.

A lot has changed in the ensuing decade - approximately 960 registrars
are now accredited and new gTLDs have been added in two rounds - but
we are still 'discussing the policy for determining the
circumstances'.

Has anyone kept a scorecard?  Registration statistics, comments from
the US Department of Justice and the distribution of UDRP cases all
demonstrate user preference for legacy gTLDs.  New TLDs are sub-prime.

Trademark protection has been a recurrent theme in gTLD policy.   As
long as one gTLD is predominant (which it undeniably is) you cannot
protect trademarks.  Domain names (SLD plus TLD) must be unique while
trademarks generally are not.

More than 3,000 US trademarks include the word 'united', but there can
be only one united.com which leaves 3,000 disenfranchised trademark
owners.  A simple search of the US trademark database generates 10,000
records for the term 'American'.  The problem is not limited to gTLDs,
Denmark (pop. 5.5 million) has at least two trademarks registered on
the family name Hansen, but there can be only one hansen.dk.

We cannot fine-tune the registration policy to generate trademark
equity, and we haven't even considered the name recognition
investments made by companies that don't own trademarks.

New gTLDs have been added successfully to the root, but they have
neither increased user satisfaction nor solved identified problems.
Trademark owners have felt the pressure to make defensive
registrations.  Now we are poised to do the same thing over again.

What is our objective: to safely increase the name space in agreement
with principles outlined in the MoU, or to repeat the past
experiments, layering the process with more bureaucracy, and expect
different results?

We've seen what users prefer - consumers are not driving the demand
for more gTLDs.   We claim to respect intellectual property but ignore
the gross inequity of protection the DNS can afford.  We want to
promote innovation, but what is our track record supporting 'new and
different'?   What alternatives have we examined?

A belief that technical solutions are impossible generates only a
self-fulfilling prophesy; start instead by assuming that a technical
solution can be found, and then ask how.  My own suggestion is
multiplexed domain names (tested, it works), but the field is open for
innovation.

Since the GNSO hasn't addressed any of these issues, may I suggest
that we postpone the introduction of new gTLDs until alternative
solutions have been solicited, tested and evaluated from vantage
points that include technical, intellectual property, consumer
acceptance, and the impact on cost for, and competition among, name
consumers -  i.e. businesses in general – not just presumptive
registries.