[wg-c] OK, *really* final version

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

	Immediately after sending out the "frozen" version of the report, I got a
message from Bob Broxton with additional suggestions.  I've added in a few
of his changes, but this is it -- this is the really and truly final version.

	Again, votes are due by  3 a.m. UTC (10  p.m. US EST) on Monday night /
Tuesday morning.

	Thanks.

Jon

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

Report (Part One) of
Working Group C of the Domain Name Supporting Organization
Internet Corporation for Assigned Names and Numbers


	This document is Part One of the Report of Working Group C.  It sets out
the rough consensus of  the group regarding whether there should be new
generic top-level domains (gTLDs), and if so, how quickly they should be
added to the root as an initial matter.

Introduction and summary

	Working Group C has reached rough consensus on two issues. The first is
that ICANN should add new gTLDs to the root.  The second is that ICANN
should begin the deployment of new gTLDs with an initial rollout of six to
ten new gTLDs, followed by an evaluation period.  This report will address
each of these issues separately.  For each of the issues, it will summarize
the discussions within the working group, arguments pro and con, and
comments received from the public.  It will then briefly summarize the
ongoing work of the group.

Procedural and outreach history

	The Names Council approved the charter of Working Group C on June 25,
1999, and named Javier Sola (Business constituency) as its chair.  On July
29, the working group members elected Jonathan Weinberg co-chair.  The
working group includes extensive representation from each of the
constituencies.  It is open to anyone who wishes to join, and currently has
about 140 members, many of whom are inactive. (For most of the life of the
working group, no NSI representative participated.  When WG-C's co-chair
solicited greater participation from the Registry constituency, Don Telage
explained that NSI had chosen not to involve itself in the WG-C process.
That representational gap has been filled now that Roger Cochetti and Tony
Rutkowski, WG-C members from the start, have joined NSI in senior
policymaking capacities.)

	On October 23, 1999, the Working Group released its Interim Report.  That
report described the issues on which the Working Group had reached rough
consensus to date.  It also included seven "position papers," setting out
alternative scenarios for the introduction of new gTLDs.  Those position
papers usefully illustrate alternate approaches to expanding the name
space, and address a broader range of issues than does this Report; they
are available at <http://www.dnso.org/dnso/notes/19991023.NCwgc-report.html>.

	On November 23, 1999, the Names Council formally requested public comment
on the Interim Report.  This call for comments was publicized on a variety
of mailing lists maintained by the DNSO, including ga-announce, ga, and
liaison7c (which includes the constituency secretariats).  In addition,
WG-C's co-chair spoke at the meetings of most of the constituencies at the
Los Angeles ICANN meeting, and urged constituency members to file comments.
Nearly 300 comments were filed in response to the interim report.  They
included responses from leading members of all of the constituencies but
two - the record does not include comments from the ccTLD or Registry
constituencies (although ccTLD members participated in the discussions that
led to the Interim Report, and WG-C's co-chair expressly solicited the
comments of both of those groups).

	The initial draft of this report was circulated to the working group on
March 2, 2000, and the report was presented to the Names Council on March
8.  The working group approved this revised version of the report in a vote
that closed on March 20.


Issue One - Should There Be New gTLDs?

Discussions within the working group

	The working group quickly -- by mid-July, 1999 -- reached consensus that
there should be new global top-level domains.  There was very little
dissent from this position.

Arguments supporting the consensus position

	Expanding the number of TLDs will increase consumer choice, and create
opportunities for entities that have been shut out under the current name
structure.  Today, .com stands astride the name space: it has more
registrations than all other top-level domain names combined, and is ten
times the size of the largest ccTLD.  Yet it has become nearly impossible
to register a new simple domain name there: Almost a year ago, in April
1999, a survey found that of 25,500 standard English-language dictionary
words, only 1,760 were free in the .com domain.

	This situation is undesirable.  It requires companies to register
increasingly unwieldy domain names for themselves, and is inflating the
value of the secondary (speculators') market in .com domain names. Existing
second-level domain names under the .com TLD routinely change hands for
enormously inflated prices. These are legitimate trades of ordinary,
untrademarked words; their high prices reflect the artificial scarcity of
common names in existing gTLDs, and the premium on .com names in
particular. The inflated value of the speculators' market imposes
additional costs on businesses making defensive registrations of domain names.

	Companies that currently have a domain name in the form of
<www.companyname.com> have an extremely important marketing and
name-recognition tool.  They have an advantage over all other companies
that do not have addresses in that form, because the companyname.com firms
are the ones that consumers, surfing the Net, will be able to find most
easily.  If the name space is expanded, companies will be able to get
easy-to-remember domain names more easily, and the entry barriers to
successful participation in electronic commerce will be lowered.  Addition
of new gTLDs will allow different companies to have the same second-level
domain name in different TLDs.  Those businesses will have to compete based
on price, quality and service, rather than on the happenstance of which
company locked up the most desirable domain name first.
 
	Similarly, addition of new gTLDs could enlarge noncommercial name space,
and allow the creation of top-level domains designed to serve noncommercial
goals.  One proposal made in WG-C, widely applauded in the public comments,
advocated the creation of a new top-level domain to be operated by North
American indigenous peoples.  Other examples are easy to imagine.

	Creation of new generic top-level domains can be beneficial in other
respects.
 One proposal before WG-C, with significant support, urges the creation of
multiple registries, each capable of managing registrations for multiple
TLDs, so as to eliminate the single point of failure for the registration
process.  Under this view, multiple new gTLDs are necessary to support the
multiple registries needed for stability.

	Adding new gTLDs to the root, finally, is an important part of ICANN's
mandate.  ICANN was created because the institutions that preceded it were
unable to resolve the intense political and economic conflicts created by
demand for new top-level domain names. The U.S. Department of Commerce's
White Paper saw the establishment of policy "for determining the
circumstances under which new TLDs are added to the root system" as one of
ICANN's fundamental goals.

Arguments opposing the consensus position

	Three arguments were made in WG-C that cut against the addition of new
gTLDs.  First, some working group members suggested that the perceived need
for new gTLDs was illusory.  Public commenters raising this issue included
Bell Atlantic and Marilyn Cade.

	Second, some working group members suggested that an increase in the
number of top-level domains could confuse consumers, because it would be
harder for consumers to keep in mind and remember a larger set of top-level
domains.  Accordingly, any increase in the number of new gTLDs should be
cautious.  Notwithstanding requests, though, no working group member
offered studies or other evidence backing up this view.

	Finally, some working group members raised trademark policing concerns:
Expansion of the domain space will create additional opportunities for the
registration of domain names that are confusingly similar to existing
trademarks.  It will present a risk that bad actors will seek to confuse
consumers by registering SLD strings identical to those registered by
others in other TLDs.  It will likely increase trademark owners' policing
costs and the costs of defensive registrations.

	The relationship between domain names and trademark rights presents an
important and difficult issue, and is appropriately addressed by registry
data maintenance requirements, dispute resolution mechanisms such as the
UDRP, and any other device that ICANN may choose to adopt, as well as by
national legislation. Trademark owners' concerns in this regard are
important ones, and not to be overlooked.  In public comments on the
Interim Report, a substantial number of commenters urged that deployment
should be delayed until after implementation of the uniform dispute
resolution procedure, improved domain name registration procedures, and
adoption of a system for protecting famous marks.  They included, among
others, Jonathan Cohen (then an NC member, IPC), Dr. Victoria Carrington,
AOL, British Telecom, Disney, INTA, Nintendo of America and Time Warner.
Steven Metalitz expressed a similar view: "New gTLD's should be inaugurated
only when, and to the extent that, established and proven procedures are in
place in the existing gTLD's to improve the quality and accessibility  of
registrant contact data, as well as satisfactory dispute resolution
procedures."  The comments of the WG-C Rapporteur of the Business &
Commercial constituency urged, on behalf of the constituency, that
"business requirements such as the effective implementation of the UDRP and
international business practices such as jurisdictional domains"should be
addressed satisfactorily before new gTLDs are deployed.  The Software and
Information Industry Association noted its support for adding new gTLDs,
but only after the creation of a robust, responsive whois system.

	Other commenters, by contrast, do not believe that trademark-related
concerns justify delay in the introduction of new gTLDs.  These included
Hirofumi Hotta (NC member, ISPCPC) (emphasizing that discussion of
famous-mark protection should not delay the gTLD rollout), Kathryn Kleiman
(NC member, NCDNHC), Michael Schneider (NC member, ISPCPC), Computer
Professionals for Social Responsibility, Melbourne IT, AXISNET (Peruvian
Association of Users and ISPs), the United States Small Business
Administration's Office of Advocacy, Register.com, InterWorking Labs,
Tucows.com, InterAccess Company and PSI-Japan.    Raul Echeberria (then an
NC member, NCDNHC) filed comments urging that the establishment of new
gTLDs was important and positive, but that rules should be devised to avoid
massive speculative purchases of domains in the new TLDs, or trademark
holders simply duplicating their existing domains.

	Within the working group, the argument that ICANN should impose
substantial delays on the initial deployment of new gTLDs in the interest
of adopting or perfecting trademark- protective mechanisms won little
support except from Intellectual Property constituency members. 

Public comments

	The discussion above canvasses many of the public comments received.  By
far the largest set of comments, however, addressed a specific
implementation of the principles discussed above.  Nearly 180 commenters (a
majority of the comments filed) supported the creation of a particular
proposed new domain: .NAA, proposed as a new gTLD to be run by North
American indigenous peoples.


Issue Two - What Should be the Nature of the Initial Rollout?

Discussions within the working group

	In working group discussions, members of the working group initially
expressed sharply varying positions on the nature of the initial rollout.
Some working group members urged that ICANN should immediately announce its
intention to authorize hundreds of new gTLDs over the course of the next
few years.  While ICANN might interrupt that process if it observed serious
problems with the rollout, the presumption would be in favor of deployment
to the limits of the technically feasible and operationally stable.  If
ICANN simply deployed a small number of new gTLDs with no commitment to add
more, they argued, the public would have to make registration decisions
based on the possibility that the small number of new gTLDs would be the
only options.  This would give the new registries oligopoly power and the
ability to earn greater-than-competitive profits; it would encourage
pre-emptive and speculative registrations based on the possibility of
continued artificial scarcity.  By contrast, they urged, an ICANN decision
to deploy a large number of gTLDs would enable competition and a level
playing field: If ICANN announced an intention to add hundreds of new gTLDs
over a three-year period, no new registry could exercise market power based
on the prospect of a continued artificial scarcity of names.

	Other working group members took the opposite approach.  New gTLDs, they
urged, could seriously aggravate the problems facing trademark
rightsholders in the existing domain name space.  Accordingly, they urged,
new gTLDs should be introduced only slowly and in a controlled manner, and
only after effective trademark protection mechanisms had been implemented
and shown to be effective.

	A third set of working group members took still another approach.  In the
long term, they stated, it would be desirable for ICANN to allow the
deployment of new gTLDs to the limits of the technically feasible and
operationally stable.  As a short-term matter, however, the immediate
deployment of hundreds of new TLDs would not be prudent.  The operationally
safer course, rather, should be to deploy a smaller number, and to follow
that deployment with an evaluation period during which the Internet
community could assess the initial deployment.  ICANN would go on to deploy
additional TLDs if no serious problems arose in the initial rollout.

	The proposal that ICANN start by deploying six to ten new TLDs, followed
by an evaluation period, was crafted as a compromise position to bridge the
gap separating the three groups, and to enable a rough consensus to form in
the middle ground.

	In September 1999, the WG-C co-chairs made the determination that the
working group had reached rough consensus supporting the compromise
position.  Because there had been no formal consensus call, though, the
working group held a vote in December 1999 to reaffirm that consensus.
Following the lead of Working Group B, the working group determined in
advance that a two-thirds margin would constitute adequate evidence of
rough consensus.  The vote reaffirmed the "six to ten, followed by an
evaluation period" compromise position as the rough consensus of the
working group, by a margin of 44 to 20.  (A substantial number of working
group members did not cast votes.  In addition, some working group members,
having been solicited to vote, sent messages to the list explaining that
they were declining to take a position at that time, and listed themselves
as consequently abstaining.  Neither the non-voters nor the abstainers were
counted in figuring the two-thirds majority.)

Arguments supporting the consensus position

	The "six to ten, followed by an evaluation period" consensus position has
the advantage of being a compromise proposal supported by a wide range of
working group members.  In a bottom-up, consensus-driven organization,
broad agreement on a policy path is valuable for its own sake.  The sense
of the bulk of the working group is that this proposal strikes an
appropriate balance between slower, contingent deployment of new gTLDs and
faster, more nearly certain, deployment.

Arguments opposing the consensus position 

	Three arguments were made in the working group against the proposal.  The
first was that the contemplated initial deployment was too large; rather,
some WG members urged, it would be appropriate, following the
implementation of effective intellectual property protections, for ICANN to
roll out no more than two or three new gTLDs.  The second argument was that
the contemplated initial deployment was too *small*: that, as detailed
above, a deployment of only six to ten, without an upfront commitment to
roll out many more, will be a half-measure that would grant oligopoly power
to the lucky registries selected for the initial rollout.

	Commenters expressed agreement with each of these positions:  Bell
Atlantic and Marilyn Cade supported the introduction of just a single new
gTLD at the outset; British Telecom and Time Warner urged the initial
rollout of only a few.  The submission of the WG-C Rapporteur of the
Business & Commercial constituency, on behalf of that constituency, urged
that ICANN should start with a "very small number" of new gTLDs.  Other
commenters, including Jonathan Cohen (then an NC member, IPC), Dr. Victoria
Carrington, AOL, Disney and Nintendo of America, generally endorsed the
statement that the introduction of new gTLDs should be slow and controlled,
and should incorporate an evaluation period.

	By contrast, Hirofumi Hotta (NC member, ISPCPC), Kathryn Kleiman (NC
member, NCDNHC), Michael Schneider (NC member, ISPCPC), Computer
Professionals for Social Responsibility, AXISNET, InterWorking Labs,
Tucows.com and InterAccess Company supported the  position that ICANN
should, at the outset, announce a schedule for introducing hundreds of new
TLDs.  The Office of Advocacy, U.S. Small Business Administration concluded
that ICANN should start with a limited introduction of new TLDs followed by
an evaluation period, but that ICANN should announce in advance that it
would continue with a steady introduction of additional TLDs so long as
pre-announced technical criteria were met.    Raul Echeberria (then an NC
member, NCDNHC) stated that ICANN should evaluate the operation and market
acceptance of the TLDs added in the initial rollout before creating or
announcing more.  Melbourne IT, PSI-Japan and Register.com all supported
the compromise position of an initial rollout of six to ten new gTLDs
followed by an evaluation period.

	Most WG members concluded that a deployment of fewer than 6-10 would not
give ICANN the information that it would need to make sensible later
decisions, and was smaller than caution dictated.  At the same time, most
WG-C members felt that an initial commitment to many more than 6-10 would
not be operationally sound.  Until we see the consequences for the domain
name space of adding new gTLDs, there are advantages to a more circumspect
path.

	The final objection raised was that the consensus agreement answered the
wrong question: The working group, said some, should not be addressing the
number of new gTLDs at all before resolving such issues as whether the new
top-level domains should be general-purpose (like .com), special-purpose,
or some combination of the two.  These issues are discussed in this report
under the heading of "ongoing work," and certainly it would not have been
inappropriate for the WG to have sought to reach conclusions on those
matters before discussing Issue Two.  But most members of the working group
concluded that the size of the initial rollout could and should be
addressed first, before resolving less tractable issues.

Ongoing work

        Remaining questions before the working group include how the new
gTLDs deployed in the initial rollout, and their associated registries,
should be selected.  In initial discussion and straw polls on this issue,
working group members fell into several camps.  One group urged that ICANN
should first select new gTLD strings, and only then call for applications
from registries wishing to operate those TLDs.  A second group urged that
ICANN should select new gTLD registries on the basis of objective criteria,
and allow the registries to choose their own gTLDs in response to market
considerations.  A third group suggested that registries should apply
describing their proposed gTLDs, and that an ICANN body or process would
then make selections taking into account the characteristics of both the
registry and its proposed gTLD. The working group considered the third
option, viewed as a possible middle ground, as a consensus call, relating
only to the initial rollout of six to ten new gTLDs.

	Thirteen "yes" votes were cast in that consensus call, and five "no"
votes.  While the votes cast were markedly in favor, it's the view of the
co-chair that a finding of rough consensus, at this date, would be
premature.  Only a small number of people voted:  In contrast to the 64
votes cast on the consensus call relating to the size of the initial
deployment (well over half of the membership of the WG at the time), only
eighteen people chose to cast a vote on this matter.  Even some active
participants in the discussion of the consensus call did not cast votes.
This makes the vote less reliable as a gauge of the views of the working
group as a whole.  Other factors making it difficult to draw an unambiguous
consensus from the vote include the facts that some of those who voted
"yes" added additional caveats conditioning their support, and that voters
may have had varying understandings as to how the term "registry" in the
consensus call should be understood, and what an application would entail.
("No" voters urged both that the consensus proposal would give too much
discretionary authority to ICANN, and that it would preclude ICANN from
considering gTLD proposals that came from entities other than would-be
registries.)

	It appears to be the sense of the working group, among both supporters and
opponents of the consensus call, that ICANN's selection process should be
procedurally regular and guided by pre-announced selection criteria.
Further, it appears to be the sense of the working group that the namespace
should have room for both limited-purpose gTLDs (which have a charter that
substantially limits who can register there) and open, general-purpose
gTLDs.  The working group extensively discussed a set of eight principles,
drafted by Philip Sheppard (NC member, Business) and Kathryn Kleiman (NC
member, NCDNHC), against which applications for new TLDs might be judged.
The proposed principles, in their current iteration, incorporate the
keywords Certainty, Honesty, Differentiation, Competition, Diversity,
Semantics, Multiplicity and Simplicity.  However, the working group has not
so far achieved a consensus on the content or usefulness of the principles.

Conclusion

	In summary, Working Group C has reached rough consensus on two issues. The
first is that ICANN should add new gTLDs to the root.  The second is that
ICANN should begin the deployment of new gTLDs with an initial rollout of
six to ten new gTLDs, followed by an evaluation period.  The working group
is continuing to address other issues, including the mechanism through
which new gTLDs and registries should be selected.  While there is
sentiment within the working group for the compromise position that
registries should apply describing their proposed gTLDs, and that an ICANN
body or process should make selections taking into account the
characteristics of both the registries and their proposed gTLDs, a finding
of rough consensus on this point would be premature.

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

A detailed summary of the public comments on the working group's Oct. 23,
1999 interim report is available at
<http://www.dnso.org/wgroups/wg-c/Arc01/msg00490.html>.