Before the
U.S. DEPARTMENT OF COMMERCE
National Telecommunications and Information Administration
Washington, DC 20230


In the Matter of


IMPROVEMENT OF TECHNICAL MANAGEMENT
OF INTERNET NAMES AND ADDRESSES

Docket No. 980212036-8036-01
Notice of Proposed Rulemaking


Comments of Craig Simon
March 19, 1998

I respectfully submit these comments in response to the "Green Paper"-hereafter called "the Proposal"-published in the Federal Register on February 20, 1998. I am presently engaged in dissertation research at the University of Miami's School of International Studies, and I have been studying the domain name controversy since March, 1997. I am grateful for this opportunity to share my opinions.


Craig Simon
PO Box 24-8911
Coral Gables, FL 33124
cls@flywheel.com

- - - - - - - - - - - - - - -
OUTLINE OF CONTENTS
Summary
Acronyms Used in this Discussion
I. Arguments
A The Proposal would impede growth of the gTLD space
B The CORE/SRS model would benefit individuals worldwide
C The CORE/POC process is open and legitimate
D The Proposal awards unreasonable commercial advantage to NSI
II. Recommendations for U.S. government action
A Endorse CORE/SRS, leveraging its economies of scale
B Endorse the SRS portability model
C Endorse the IETF open standards model over proprietary models
D Urge POC and CORE to enfranchise aRSC registrants
Closing
- - - - - - - - - - - - - - -

SUMMARY

1. The Proposal is likely to stifle growth in the use of new generic Top Level Domains (gTLDs). The alternative plan offered by the Council of Registrars (CORE) is better suited to accelerating such growth, especially among people in communities around the world where Internet access is still relatively expensive.

2. The Proposal would grant significant market privileges to the incumbent commercial registry. The Proposal's concept of enfranchising profit-oriented monopolistic TLD registries is fundamentally flawed. The model of non-profit portability embodied in CORE's Shared Registry System (SRS) offers an important safeguard against abuses and failures in the domain name registration market.

ACRONYMS USED IN THIS DISCUSSION


CORE Council of Registrars
DNS Domain Name System
FCFS First Come First Served
gTLD Generic Top Level Domain
IAB Internet Architecture Board
IAHC International Ad Hoc Committee
IANA Internet Assigned Numbers Authority
IETF Internet Engineering Task Force
IP Internet Protocol
ISOC Internet Society
ITU International Telecommunications Union
POC Policy Oversight Committee
aRSC Alternate Root Server Confederation
SLD Second Level Domain
SRS Shared Registry System
TLD Top Level Domain

I. ARGUMENTS

A THE PROPOSAL WOULD IMPEDE GROWTH OF THE GTLD SPACE

Appendix I of the Proposed Rule states "Each top-level domain (TLD) database will be maintained by only one registry and, at least initially, each new registry can host only one TLD." Similar language is expressed in section I/C.

I estimate that any party seeking to establish a new registry in compliance with the requirements further described by Appendix I would incur expenses of at least $200,000 annually. If the cost of being listed in a registry is set as high as $10, this means that a commercial registry would have to attract 20,000 registrants before it could make a profit. Using less conservative estimates, that number could exceed 50,000.

Either number may seem small when compared to 1.6 million names registered in .COM, or about 120,000 each in .NET and .ORG. Nevertheless, only 10 nations currently maintain national TLDs exceeding 20,000 registrants in Second Level Domains (SLDs). Thus, to justify investment in creating a new commercial TLD, investors would have to believe that their start-up TLD would quickly become at least as popular as the Italian, Argentinian and Swiss TLDs are today.

Implementing the Proposal would shut out many otherwise viable non-sovereign groups from participating as entities in the global domain space. The rule of market forces provides incentives only to the very largest TLDs serving the highest paying customers. Strong market disincentives inhibit the near term-creation of commercial TLDs serving smaller groups, such as .MUSEUM and .AIRPORT.

Moreover, this approach would unnecessarily add to the financial risk of creating language-based TLDs like .ESP (Spanish), .TRK (Turkish), .ARB (Arabic), and so on. The high mandated cost of doing business as a new TLD registry would raise legitimate concerns among savvy consumers about the long term viability of addresses in that zone. With no guarantees against the consequences of a registry's commercial failure, potential registrants will be deterred from using young, novel, gTLDs.

CORE has already invested several hundred thousand dollars building a high capacity registry server, and has worked with recognized leaders in the computer industry while assembling its system. CORE's non-profit SRS model offers the benefit of economies of scale, and is designed specifically for the purpose of broadening the TLD space. Yet the Proposal ignores the CORE system, which is practically ready to run. As a consequence, one may conclude that the effect of implementing the Proposal and blocking CORE would be to stifle growth in the diversity of populations that could organize under new TLDs.

Several individuals have demonstrated that a TLD can be assembled quite inexpensively if high speed Internet access and proper security features are already available in support of other commercial activities. In fact, there are working examples of alternative registries which already support multiple TLDs. However, the Proposal, like the CORE plan, for good reasons addressed elsewhere, favors separating registry and registrar functions. It is appropriate in this discussion, therefore, to factor registry expenses separately.

What follows is a very rough estimate of the expenses mandated by the "Registry Requirements" in the Proposal's Appendix I. Some sections are left blank because my guessing would be highly uncertain. A more thorough budget prepared by someone actively involved in these markets would reveal these are conservative estimates.

System
1.a 25,000 Hardware providing registrars with FCFS access.
1.b 24/7 and scalable
1.c 35,000 Two T1s and related connectivity (yearly)
1.d 5,000 Backup and archive
1.e 5,000 Transaction tracking
1.f Whois
1.g Registrar Port
1.h 50,000 Two zone servers (presuming T1s)
also 100,000 (staff, location, promotion, insurance, etc.)

Policies
2.a Dispute resolution forum
2.b Failsafe Plan
2.c 200 Staff training
2.d 800 Bulletproofing

Site requirements
3.a 2,000 Power backup
3.b 30,000 24/7 security
3.c 40,000 Redundant site with "hot switch over" capability

Yes, TLD registries should be robust and secure, but the Proposal stipulates this list of requirements without providing any estimate of real world costs. It makes no attempt to evaluate the impact of these expenses on the diversification of gTLDs. The Proposal hints that the prohibitive rule of one TLD per new registry might be amended in the future, but it is vague about when or how new TLDs would be distributed fairly among the registries.

B THE CORE SRS MODEL WOULD BENEFIT CONSUMERS WORLDWIDE

Domain name portability is one of the most innovative and attractive features of the SRS, offering a structure that works to eliminate the risk of unfair price manipulation by monopolistic TLD registries, while fostering competition among registrars. The Proposal includes language at the end of section I/B that recognizes these virtues (without acknowledging their source in the CORE plan), but then opts for the immediate creation of monopoly registries. The arguments given in the text of the Proposal are remarkably equivocal in that section, underscoring the weakness of its conclusion.

In my judgment, CORE's SRS model offers the prospect of an Internet that will be a more reliable and trustworthy venue for communication between people. On the other hand, adopting the Proposal might make the Internet more difficult to navigate, since the government's plan would permit entire TLDs to disappear from view as a result of market perturbations.

Over the next few decades, the Internet is projected to attract hundreds of millions and ultimately billions of participants. Given the contemporary surge of investments in Internet-related technologies, one may reasonably expect that individuals will increasingly use the Internet's addressing system to conduct financial and personal business. The rise of electronic commerce technologies which use Internet resource records for secure identity authentication is an example of this trend.

Nearly all parties to the DNS debate agree that the system should be stable and unified. I would add that any features which increase the possibility of disrupting someone's identity should be avoided. As it stands, the Proposal would allow market quirks to destroy entire gTLDs, thus wasting all of someone's efforts to establish a reliable presence at a particular domain name address. Commitment to the principles of competition and open markets should not be allowed to excuse the loss of someone's ability to represent herself or himself on the Internet as an individual or merchant with a stable, authentic identity.

Therefore, administering the Internet's Root Zone through a non-profit agency like CORE is a preferable alternative. If there are to be any secure locations in "cyberspace" where people can be spared from the complexity and fears of obsolescence associated with the Information Age, their representation within the address system should be that place. Otherwise, their virtual mailboxes would be rooted in quicksand, and web links to their home pages would mark a trail of dead ends and ruins.

The Proposal intends to introduce a new level of dynamism to the DNS, based largely on the stated desire to promote a higher pace of innovation. The logic of that premise should be reconsidered. It is a mistake to idealize innovation and "creative destruction" as an end in itself, especially when an individual's identity is put at risk.

The Proposal would brashly impose the American tradition of high tolerance (even desire) for quickening obsolescence on the rest of the world's population. Free enterprise and market competition are beneficial to society, but not within every aspect of it. The stable operation of essential facilities in the global community should not be abdicated to rampant, unchecked markets.

A reliable presence on the Internet (or whatever it transforms into) may eventually be as basic to a person's integration into society as facility with clocks, maps, money, and traditional forms of personal identification. Among the domain name plans currently under consideration, CORE's is best suited to provide a stable system that would be accessible, affordable, and durable.

C THE CORE/POC PROCESS IS OPEN AND LEGITIMATE

Those who have been following this dispute know that CORE was spawned through a long contentious process initiated by a group of veteran Internet engineers who wanted to introduce commercial competition to the domain name registration system. I have observed (and I think that many of those individuals recognize) that talent and skill at creating Internet technical standards does not translate easily into the political skills necessary for constituting new global institutions where so many novel and complex issues of commerce, trademarks, privacy, and jurisdiction intersect. Not all hackers can succeed as statesmen.

But shortcomings in tactical expression do not necessarily imply a corrupt political vision. During the course of my research I have been favorably impressed by many of the principles developed in the formal "Internet community" among leaders of the Internet Engineering Task Force (IETF), the Internet Architecture Board (IAB), and the Internet Society (ISOC). Open processes are highly valued among these groups, and I believe that the CORE plan does not stray from those precepts. Still, many critics have said those ideals have been sullied in this debate. This is an issue which must be addressed frankly.

In most decision making processes, various "insiders" and "experts" have a relatively strong influence over the outcome, simply by virtue of their access to the instruments of power. To outsiders who lack such privileges and who disagree with the fundamental precepts of the insiders, it often seems that "the fix is in," and undesirable outcomes are a foregone conclusion.

In other words, the responsiveness of the US government to outsiders is as open to question as CORE's responsiveness. Members of both sides have indeed made strong efforts to consider the arguments of all interested parties. They can also be faulted for failing to live up to a model of complete transparency and fairness. While CORE's most ardent critics call it an insidious Swiss cartel, critics of the Proposal have characterized its authors as puppets of the military industrial complex. In my view, neither attack has any substantive merit.

What is key is that many of the individuals who contributed to the discussions which created CORE have been engaged for quite some time in practical work and serious reflection about how to build an open and secure Internet. It would be a mistake to dismiss the fruit of their efforts so hastily.

Similarly, responsible individuals in the U.S. government have been making a forthright attempt to master an issue of stunning complexity and importance. The Internet is about to shed its remaining vestiges as Cold War era research project. If it continues to fulfill its promise as a critical global medium for education and commerce, decisions made now will have enduring historical significance.

The stewards of this puzzling transformation are all apt to make a few false starts and missteps along the way.

D THE PROPOSAL AWARDS UNREASONABLE ADVANTAGE TO NSI

Under the Proposal, Network Solutions, Inc. (NSI), would be allowed to keep exclusive control over the existing .COM .NET and .ORG registry, with the provision that other registrars would be allowed to "sell" into those TLD spaces. The plan thus grants exceptional advantages to NSI.

Only NSI is granted the privilege of hosting three commercial TLDS, while the five new registries called for in the Proposal would each initially be restricted to one. Thus, NSI is granted exclusive rights to leverage economies of scale and further consolidate its market advantages. This is strikingly inequitable, as many other respondents to the Proposal have also noted.

NSI would keep a long head start in numbers of registrants and brand identity. The company is already cash rich from its years as a monopoly provider, and well poised to initiate a series of mergers and acquisitions that will extend its advantages. Also, the company somehow anticipated the general outline of the U.S. plan. By creating WorldNIC, it has already restructured itself to comply with the Proposal's registrar requirements.

Perhaps these preferential arrangements were designed into the Proposal out of a sense of prudent caution and an honest wish to reduce the potential of destabilizing the smooth functioning of the incumbent registry. After all, NSI accounts for approximately two thirds of the domain names currently registered on the Internet. Yet, potential registrants in new TLDs are not accorded similar treatment or concern. This is consistent with my earlier conclusion that the Proposal would impede growth in the use of generic TLDs beyond those managed by NSI. If doing so was the government's intent, it should have been stated more clearly.

By mandating fairly high costs among the gTLDs which come online in the near term, the U.S. plan relieves NSI from the challenge of meeting any formidable threats in the registry market. Consequently, much of the pent up demand for domain registrations may move into legitimate national registries. By quirks of history, a few relatively small or poor nations have even handed over their privileges as sovereign states to commercial registry/registrars like .TO, .TM, .NU, and .IO. However, it is unlikely that these atypical oddities will be able to serve the public on a wide scale.

In sum, the U.S. plan virtually guarantees the perpetual "lock in" of a huge market segment in NSI's registry, while locking out the prospect that widespread use of new generic TLDs could proliferate.

II. RECOMMENDATIONS FOR U.S. GOVERNMENT ACTION

A ENDORSE CORE/SRS, LEVERAGING ITS ECONOMIES OF SCALE

If the U.S. government favors expansion in the use of generic TLDs, the Proposal under review is not the best path to that result. The Proposal's authors should clarify their goals and make a more consistent series of recommendations. If the authors ultimately decide in favor of accelerating gTLD expansion, they may find virtue in endorsing CORE's structure as a basis for moving forward.

Actions by the U.S. government which assist in the quick and confident implementation of CORE plan will further reduce the costs borne by users of the system. By leveraging its economies of scale, and by proper management of its funds, one can contemplate a day when the CORE registry performs its functions at nominal cost, perhaps charging under $2.00 per name.

Endorsing the CORE plan also embodies a safer strategy than the Proposal's. If CORE's plan proves unworkable for any reason, its organizational structure would be relatively easier to break up and transform. Moving in the other direction would be much more difficult, however. Consolidating several entrenched private entities into a single public trust would present far greater operational challenges. In other words, the CORE plan is more amenable to adjustment.

B ENDORSE SRS PORTABILITY MODEL

In light of the current administration's efforts on behalf of portable health insurance, it is unfortunate that the Proposal does not resonate more strongly with appreciation for the virtues of domain name portability. Reversal on this position would be a welcome move, reinforcing perceptions of consistency and coherence among the current Administration's policy initiatives.

C ENDORSE IETF OPEN STANDARDS MODEL OVER PROPRIETARY MODELS

Many of CORE's critics have said that CORE's promoters in the Internet standards community have been fooled or manipulated by ill-willed bureaucrats from within Geneva-based intergovernmental organizations such as the International Telecommunications Union (ITU). Given that the Internet community and the ITU have been adversaries in the past, this alliance is a remarkable occurrence, and deserves attention.

The U.S. should publicly welcome the Internet community's efforts to educate the ITU's leadership regarding the virtues of unencumbered, freely available technical standards. Moreover, if the United States government does reverse its position and endorse the CORE process, it should emphasize this issue, thereby reinforcing the open standards making processes of the IETF, while putting POC, CORE and the ITU on notice that those norms must be sustained. Participants in the IETF's technical work have undoubtedly made a beneficial contribution to society, and the laudable principles which mobilize that organization should be supported.

Normative issues are significant. For example, the word "netizen" has been used by many respondents to the Department of Commerce's Notice of Inquiry conducted in the summer of 1997, and to the Green Paper Proposal. This illustrates that the Internet has fostered lively interest in alternative definitions of affiliation and community. It remains to be seen whether simplistic concepts like "netizenship" will catch on, or fade away like fads and fashions. Nevertheless, there is value in reflecting on what it will mean if network-based patterns of social organization ever mature and take hold.

Just as the U.S. government provided initial stewardship of the Internet's exploding infrastructure, American social heritage will exert a lasting influence on the Internet's developing culture. It is no surprise that many of the Internet's "leading citizens" who are also Americans rank among the strongest advocates of First Amendment rights. Across the Internet one finds widespread insistence that a free press must be upheld as a bulwark against the potential abuses of central authority.

This does not imply that an Internet culture should be created entirely in the American image. Neither is this feasible. If and when a mature Internet culture emerges, it is likely to include elements that are quite distinguishable from the present international political order. For example, after several generations of using the Internet, people may declare they want to reassess the nation-state system. This may seem surprising and unrealistic today, but future "netizenship"--or some other amended loyalty--need not be a threat to Americans if it is nurtured with the best principles that the American legacy can offer.

D URGE POC AND CORE TO ENFRANCHISE ALTERNATE ARSC REGISTRANTS

This final suggestion is offered as a potential remedy to the sharp divisions that the DNS "wars" have created in the Internet community. The DNS controversy has gone far beyond technical or economic differences. An elegant solution needs to resolve political questions as well.

If the U.S. government were to endorse the CORE plan (a big if, of course), the government would then have leverage to push POC and CORE to enfranchise individuals and businesses who have registered names in the alternate .WEB and .ARTS domains before now. Since CORE also claims those TLDs, it is important to find ways that individuals with arguable (though legally weak) prior claims to relevant SLDs can finally be included in the Internet mainstream.

Such a compromise might not be entirely satisfactory to the "grass roots" entrepreneurs who want to continue operating the non-sanctioned .ARTS and .WEB registries, but it would eliminate the prospect of causing injury to their clients. Those domain name holders probably had little understanding of what a maelstrom they had fallen into when they submitted their registrations, and they would no doubt be quite pleased with a remedy that finally makes them visible across the entire Internet. Any "pre-registrant" in a CORE lottery queue hoping to register the same name in .WEB or .ARTS would be removed from the queue, and any pre-paid queue fees would have to be returned.

Implementation of this compromise would end the disputes over .ARTS and .WEB, and set a precedent for the migration of other alternate TLDs into a reunified root. It would also confirm CORE's commitment to the principle of stable, portable addressing, which should be the primary justification for U.S. endorsement of CORE's plan.

CORE's rules already permit operators of the alternate registries to sign up as registrars. After the aRSC registrants are moved to CORE's SRS, the U.S. government should encourage those alternate operators to join CORE, in the spirit of stabilizing and broadening CORE's constituency. WorldNIC should be encouraged to join as well.

In hindsight, it is clear that the IAHC made a serious tactical error by proposing new TLDs in conflict with the alternate unsanctioned registries. This only exacerbated tensions that were already frayed after months of harsh debate within the Internet community over how to add new TLDs. I'm aware that there were discussions between IAHC participants and at least one ARSC operator which tabled the prospect of "grandfathering" registrants from the alternate TLDs. Unfortunately, those negotiations terminated without agreement.

Therefore, it would be appropriate for the U.S. government to apply some diplomatic pressure in cyberspace. This conflict has been preoccupying far too many people for far too long. There is good reason to believe that the controversy will only grow more contentious if the parties are not reconciled. In the worst case, several more rounds of escalating strife and animosity could put the unity of the Internet's root at risk.

CLOSING

Thank you for your patient consideration of my comments, which, though lengthy, do not exhaust the other important topics raised in the Proposal, such as trademark protection, dispute resolution, the future of IANA, and IP number delegation. Over the last year I have had the special privilege of meeting, interviewing, and following the work of many talented and dedicated people who--by designing the "nuts and bolts" of Internet standards--have created the possibility for a new international reality and a more open global society. I hope that these suggestions can somehow advance their progress.

- - - - - - - - - - - - - - -
Craig Simon
- - - - - - - - - - - - - - -
cls@flywheel.com
- - - - - - - - - - - - - - -