[wg-c-1] Compromise

[Craig Simon <cls@flywheel.com>]

                    A Plan to End the DNS War

1. Scope

     The proposals in this document are offered for consideration
     by the Internet community and particularly by ICANN's DNSO
     Working Group C, Drafting Committee 1. This committee has
     been tasked to make recommendations concerning new gTLD
     deployment in the IANA-approved root zone.


2. Concepts

2a. Compromise

     It is time to reunify and strengthen the Internet's name
     space. Rapid and broad expansion of the number of IANA-
     approved gTLDs should be pursued. I propose starting with a
     compromise aimed at integrating certain non-IANA zones into
     the legacy Internet name space.

     The plan to increase the number of gTLDs may be summarized
     as follows: Settle disputes with operators of zones which
     have been the subject of divisive contests. Begin gTLD
     service for those zones first. Follow by adding the TLD
     candidates suggested to the IANA in 1995 and 1996.[1] Then
     continue to augment the IANA root with suggestions offered
     by a broadly-based and informed Internet community.

2b. Goals

     Such expansion is intended to accelerate the commodification
     of domain names in ways that would preclude any artificial
     scarcity of names. This approach should expedite the reduction
     in price of domain name registration for the consumer. It is
     safe to expect that price competition and market
     differentiation will become more pronounced at the registrar
     level as the number of commodity gTLDs proliferates. 

     To reduce the likelihood that any future regime of Internet
     administration might seek to impose speech or content
     control through DNS management, and to guarantee the
     abundant availability of names in all gTLDs, ICANN should
     avoid models of gTLD differentiation which restrict second
     level domain name registrations within particular gTLD
     suffixes. This does not preclude the construction of
     specialized zones elsewhere in the Internet's name space,
     under ccTLDs or under SLDs at the 3LD level and below.


2c. Strategy

     The first gTLD added to the listing of the IANA-defined
     legacy root should be .web. 

     I propose that verifiably current SLD registrations now
     listed in any non-IANA .web zones will be "grandfathered"--
     enfranchised in the IANA-approved .web gTLD. The .web zones
     operated by IODesign and Name.Space are apparently the only
     such zones currently offering commoditized service. I
     further propose that incentives be provided to encourage the
     non-IANA zone registry operators to join the ICANN-
     administered system as registrars.

     This model should be followed in good faith for domain name
     holders in other publicly-offered non-IANA zones so that the
     Internet's name space will be as unified as possible before
     augmenting the legacy root with more gTLDs.

3. Commentary

3a. Coherence and Commodification in the Internet Name Space

     The thrust of this proposal is to reunify the Internet's
     public name space by settling outstanding claims to gTLD
     zone authority. 

     Resolution of the controversy surrounding .web must be given
     priority.  The issue has been a recurring source of
     political tension because the zone is unusually attractive,
     offering a very high potential for popularity and market
     success as a gTLD. The .web suffix was among the very first
     and most frequent new gTLD requests submitted to Jon Postel
     after NSI commodified domain name registrations in September
     1995 (These requests came separately from, in order, Scott
     Adams, Mike Lester, Chris Ambler, and Craig Harkins[1]). It
     was one of the seven new gTLDs designated by the IAHC in
     early 1997. And it has been the subject of an unsuccessful
     law suit by IODesign and a long public campaign waged by
     Chris Ambler, an officer of that corporation. Name.space and
     Paul Garrin have also undertaken litigation demanding rights
     to sell registrations in .web and other zones.

     My judgment as a long standing observer of these events[2]
     is that Mr. Ambler's claims to have been granted authority
     over .web by the IANA in July 1996 do not hold up.
     Nevertheless, it is possible that the officers of IODesign
     may reinitiate litigation if a .web zone registry other than
     theirs is added to the IANA-approved root. ICANN must be
     prepared for such a contingency in the event this proposed
     compromise is not accepted.

     A suitable compromise must be both practical and principled.
     ICANN must confront this issue and seek an amicable
     resolution that is mutually satisfactory to all parties for
     the following reason: A coherent name space is essential to
     the integrity and stable operation of Internet-based
     communication. 

     Responsible administration of the Internet therefore
     requires integration of .web and other existing publicly-
     offered non-IANA zones into the legacy root. The zone is
     strategically valuable because of its ready memorability as
     a public Internet identifier. Nothing is to be gained if it
     were to be ignored or left idle, while countless more months
     or even years of dispute over its authority would be costly
     and unnecessary. Settlement of these disputes will fortify
     confidence in the strength and stability of the IANA system
     and will punctuate the transition to a new era of reliable
     and responsive Internet management.

     Below I offer two scenarios for compromise. Two are
     necessary because another outstanding policy matter must be
     settled first: ICANN's endorsement of a proprietary or a
     shared registry model.

3b. Option 1: Implementing .web as a Proprietary TLD

     Reunification of the Internet's name space deserves higher
     priority than preference for the non-profit SRS business
     model over the proprietary gTLD model. It is better to see a
     monopolized for-profit .web in the legacy root rather than
     no .web at all.

     If a properly constituted ICANN board deems a proprietary
     gTLD model to be acceptable, there would be no principled or
     practical reason to deny existing non-IANA zone
     proprietorships admission into the legacy root.  Moreover,
     it would be inappropriate and unfair to focus on creating
     new gTLD proprietorships in a market where existing
     proprietorships are still being denied visibility in the
     legacy roots. The proper course would be to resolve disputes
     for primacy between proprietorships which claim authority
     over identically named zones. If competing parties can not
     be reconciled, ICANN should establish the sort of Fair
     Hearing Councils which have been suggested elsewhere.

3c. Option 2: Engineering the Transition to Shared gTLDs:

     I have stated some arguments in support of shared not-for-
     profit TLDs elsewhere.[4] Regardless of my own preference
     for it, I do not believe a broadly-based and informed Internet
     community has yet endorsed or rejected this model. Nor would
     I argue that the White Paper mandated one model over the
     other. 

     Yet it is clear that if a properly constituted ICANN board
     adopts the CORE type of non-profit registry / distributed
     registrars model, the transition period is bound to be
     politically more complex and technically less familiar than
     simply expanding the number of monopoly proprietorships.
     NSI's legacy of rapid growth and high profitability will
     continue to tempt other businesses into running their own
     zone registries, even if this means offering service outside
     of the legacy root and undermining the coherence and
     operational stability of the Internet's name space.

     If ICANN commits to the shared registry model, I propose
     that ICANN should offer what might be called a double
     enfranchisement to the operators of contended zones: 1)
     Offer the zone proprietor an expedited application to become
     an ICANN-approved registrar, a generous discount on any fees
     required to attain registrar status, and a temporary (30-60
     day) opportunity to function as the exclusive initial
     registrar for that gTLD, and; 2) Offer holders of domain
     names within the zones at issue a free year of inclusion the
     IANA-approved shared registry.

     The domain name holder's enfranchisement can be referred to
     as "grandfathering." In practice, this funds the amount any
     registrar would have to redeem to the IANA-approved shared
     registry for each domain name record. This covers the
     standard cost of maintaining the domain name record as well
     as any additional fees that might be required by ICANN.
     Grandfathering can play out in one of two ways: 1) The
     credit would be used to offset the ex-proprietor's (and
     newly constituted registrar's) cost of listing the domain
     name holder in the IANA-approved registry, or; 2) In case
     the proprietor is not willing to participate in the ICANN
     system, the credit would be used to enlist an existing
     ICANN-approved registrar to provide free service for one
     year to the domain holder. This encourages the alternate
     registrar to build a relationship with the domain name
     holder by effectively subsidizing the fees the registrar
     would be obliged to pay to list the record in an IANA-
     approved registry.

     In either circumstance the existing domain name holder will
     be granted visibility to hundreds of millions of users in
     the legacy root as soon as possible. 

     The new gTLD would be introduced into the IANA root in a way
     that enhances coherence and operational stability. Care must
     be taken to avoid penalizing the holder whose domain name is
     being copied into the IANA-approved TLD. In the best case,
     if the existing zone proprietor becomes an ICANN registrar,
     the domain name holder would not be inconvenienced in any
     way. The holder can simply maintain its relationship with
     the registrar, or can choose a new registrar if desired. 

     In the worst case, if the proprietor balks, there might be
     some confusion experienced by the domain name holder as the
     new circumstances become clear. In such a case it would be
     necessary to add the domain name record and any available
     contact information to the list of one of the ICANN-approved
     registrars. Then the domain name holder would be notified
     that one year of free service within the IANA-approved gTLD
     is being provided by way of that registrar. The domain name
     holder would then face the choice of sustaining the
     relationship with one or both of the zone registrars. It is
     safe to assume that the registration which guarantees
     visibility in the legacy root would be more attractive.

     This grandfathering strategy is designed to smooth the
     transition to a unified root. If successful it will help
     migrate as many as 40,000 registrants from non-IANA zones
     into gTLDs within the legacy root.  It intends no prejudice
     or penalty against domain holders whose zone proprietors in
     the worst case may refuse to join the legacy root, despite
     being encouraged to join as registrars under generous
     circumstances. In other words, the compromise is structured
     to forge coherence in the Internet's domain name space, so
     that new gTLDs can be added under the most auspicious
     possible circumstances

     Since the number of domain name holders in non-IANA zones is
     small, this registrant-subsidization and registrar-discount
     strategy will have a negligible impact on the shared
     registry's and ICANN's revenue streams.  The greatest cost
     to all participants involves learning the art of compromise.
     It is a cost that must be widely dispersed.  Nevertheless, it
     is clear that the potential benefits of unifying the Internet's
     name space far outweigh the costs of sustaining the present
     conflict.


4. Contingencies:

     If any of the suggestions introduced here are ultimately
     incorporated in the final product of this drafting group,
     their implementation should be deferred pending ratification
     by a fully constituted ICANN board. I believe it is
     necessary to first decide the question of proprietary vs.
     shared gTLD registries before determining which compromise
     to pursue, and that ICANN's interim/initial board is
     not yet sufficiently representative of the Internet
     community to set this policy.


Craig Simon
14 July, 1999
cls@flywheel.com


Notes

[1] Jon Postel. "The IANA's File of iTLD Requests."
http://www.gtld-mou.org/gtld-discuss/mail-archive/00990.html 

[2] Craig Simon, Ph.D. Candidate at the University of Miami's
School of International Studies. A page relevant of links and
writings is available at http://flywheel.com/ircw/overview.html .

[3] Pending the satisfaction of technical criteria pertaining to
the connectivity, interoperability, and scalability of the
registry, and perhaps social criteria pertaining to security and
inclusivity. These parameters are fundamental to the scope of
this proposal to Discussion Group C.

[4] Craig Simon. "My Response to the Green Paper."
http://www.flywheel.com/ircw/gpsubmit.html