Letters to the Editor - The Internet Protocol Journal - Volume 6, Number 1

Ruling the Root


As a matter of principle, I don't mind having my book
Ruling the Root
reviewed by David Crocker. Mr. Crocker was a significant figure in some of the key events covered in the book. His assessment and opinion of the book had the potential to be quite interesting.

One can only be disappointed with the results, however. The review reveals an inability to rise above partisan sniping and engage rationally with an different view. That, as a matter of policy, is why serious journals don't publish unsolicited reviews of books. Unsolicited reviewers tend to fall into one of two types: unabashed promoters with a personal interest in the success of the book, or people with an axe to grind trying to shoot down a perceived enemy.

I offer a rebuttal only because I think it is vital that the Internet technical community, the presumed readers of
The Internet Protocol Journal
, achieve a higher standard in their discussion of Internet-related policy issues.

Ruling the Root
is a serious attempt to analyze the intersection of technology and policy. It offers a way of understanding that intersection based on theories of institutions and property rights. I know that this intersection irritates many engineers, who often harbor a wish that it would go away. By now we should know that it won't. Technical systems raise political issues. Technical people, economists, lawyers, and policy analysts, therefore, must be able to engage in rational dialogue about institutional issues, even when the discussion comes uncomfortably close to home. If we can't, the world is in big trouble.

The review completely misses this big picture. It begins with an attempt to belittle the policy significance of domain name management by inventing a mythical decree that all street names have to be in an obscure language. Crocker's attempt at humor falls flat, given today's headlines. Virtually the same day his review was published a German registrar was ordered to take a domain name away from a Web site with objectionable content. Not too long after, an ICANN Task Force published a WHOIS policy proposal that allows domain names to be shut down after 15 days if someone challenges the accuracy of the contact information, raising issues of privacy and harassment. ICANN regulates the prices of registries and entry into the market for domain name services. No one, not even ICANN itself these days, pretends that domain name administration is an exclusively technical matter.

Instead of engaging on those terms, the review concentrated on factual nitpicking. Take this one: "...the book does not consider NSI's role in ICANN-related political processes." This is an astoundingly inaccurate statement. The index of the book under "Network Solutions" contains 33 listings under 5 separate headings.

The book analyzes at length NSI's origins and ownership changes, its opposition to the IAHC and gTLD-MoU, its implied threat to establish a new root, and its policy conflicts with ICANN and the U.S. Department of Commerce.

Crocker claims that I "[characterize] the pre-ICANN
International Forum for the White Paper
(IFWP) as 'the real arena for arriving at a decision [about the details of the new organization].'" His use of a sentence fragment covers up what appears to be a deliberate distortion. I really wrote that some people viewed the IFWP in that way, while others, notably Joe Sims, Jon Postel, and the Information Technology Association of America, did not; see pages 176-178. I wrote at length about how that basic lack of agreement between adherents of IFWP and followers of IANA over legitimacy led to lasting conflict over ICANN's formation.

Crocker was one of Jon Postel's appointees to the
International Ad Hoc Committee
(IAHC). The review takes issue with my characterization of the IAHC, but unfortunately only to maintain Crocker's fictional self-conceptions. He denies that the IAHC ever claimed that "the root was theirs to dispose of." He also denies that IAHC was intended to be the seed of an alternative DNS governance structure. He's wrong on both counts. There is a voluminous record on this question, comprising contemporary news accounts, e-mail list archives, and my own recorded interviews with principal figures such as Don Heath.

Crocker's assertion that IAHC was "explicitly subordinate to IANA" is rather disingenuous, because IANA's U.S. government funding was ending and IAHC was explicitly perceived by Postel and ISOC as a mechanism for continuing its funding. So IAHC was intended to be the governance and support structure for IANA, just as ICANN now is. Indeed, today's ICANN has many features in common with the IAHC proposal, such as the shared registry concept, the slant toward intellectual property interests, the treatment of TLDs as "public resources," and a compulsory and uniform dispute resolution procedure.

What is really at issue here? It is this: Crocker cannot accept the simple fact that a political battle was under way for control of the root, and Postel/IAHC, as well as NSI and the U.S. government, were contenders for that control. Crocker's review challenges the claim in the book that Postel's root redirection exercise in January 1998 was "apparently" based on "concerns about the direction U.S. policy was taking." This judgment was based on interviews with people who were involved with Postel's effort. Of course I cannot read Postel's mind, but neither can David Crocker. My interpretation of why Postel acted is based on the timing and on evidence drawn from first-hand participants. Crocker offers an alternative interpretation, plausible but based on nothing but his own assertion. There is plenty of room for legitimate debate about historical interpretations. Such debate is useful, however, only if it is aimed at discovering the truth.

Regarding the status of IANA, I am sure we will never agree. I see it fundamentally as a DARPA contractor subject to U.S. governmental authority; Crocker views it in almost mystical terms as the embodiment of the Internet community. He says nothing about who paid the bills. Yet, we are not as far apart on the facts as he wants to make it seem. Contrary to the review, the book does document in great detail how a new community for Internet standards development grew up around the old DARPA-funded cadre of Postel, Cerf, and the IAB, and created its own standards of legitimacy and process. My book doesn't dispute Postel's tremendous respect and legitimacy among the technical community. But when it comes to institutionalizing control and ownership of the name and address roots of the Internet, whoever pays the piper calls the tune. And Postel's ability to perform the IANA functions was supported by U.S. government money from day one.

Hence, it was unrealistic to expect Postel to be exempt from governmental authority after domain names became resources of economic value and produced legal and political conflict over that value. Nor is it correct to imply, as Crocker does, that knowledge of the operational details of a technology automatically confers wisdom as to the correct public policies that should be adopted when that happens. Of course, policy decisions must respect technical facts and technical constraints. It is this relationship between technical system, technical community, and the worlds of business, law, and government that is cntral to the story told by
Ruling the Root.

Crocker's final stab at discrediting the book involves some rather spurious charges of ethics problems. "In his criticism of dispute-resolution activities, he neglects to mention that he is a paid arbitration panelist," he writes. Crocker here refers to the fact that I was one of the few nonlawyers allowed by WIPO to serve as one of three judges in domain name—trademark disputes brought under ICANN's UDRP. The "pay" he refers to is a $500 or $750 honorarium for each case. I do about ten cases a year. I fail to see any conflict of interest or ethical problem here. Crocker implies that my meager remuneration for assuring that justice is done in UDRP cases somehow corrupts me, but he knows perfectly well that I am an opponent of the UDRP and would happily stop receiving those honoraria if the darn thing went away. Besides, no one is in a better position to understand what is right and what is wrong with UDRP than someone who is involved in the actual cases. I do not even understand what his concern is about the noncommercial DNSO constituency. I deal with it in one sentence in the book, and most of my activity in a "management capacity" (i.e., as an elective representative) came after the book manuscript was written.

—Milton Mueller
Syracuse University

The author of the book review responds:

Professor Mueller's response discusses his goals of the book and his opinions of my review, to which he is, of course, entitled. He characterizes
Ruling the Root
as an academic consideration of the policy issues pertaining to the Domain Name Service, which he casts as global Internet administrative services. Note that the tag line to the title of his book, however, casts it even more generally as "Internet governance." Academic and policy work need to be conducted carefully. Unfortunately, Professor Mueller confuses the issues, rather than elucidating them.

The opening, mythical decree of the review was carefully constructed to make the perspective of the book on communication system administrative policy clear: Professor Mueller confuses an administrative agency, such as ICANN or its telephonic equivalent, with a national government such as Germany. He also confuses control over administrative information, such as names and addresses associated with registrations, with primary content, such as a Web page.

Professor Mueller defends his writing about the IFWP as merely reporting the view of others, rather than being his own advocacy. However, his reporting is highly selective and results in his confusing the difference between tension that was
the IFWP process, versus
IFWP and IANA. His casting the issue as being with IANA is contrary to the formal documentation of IFWP, and contrary to the style and content of its process. IFWP was not designed, nor was it conducted, as a decision-making body.

Professor Mueller confuses the actions and intent of the IAHC with those of IANA (and ISOC). He claims to have extensive substantiation for his assessment of the IAHC. Yet none that is relevant to this confusion appears in his book or his letter. This omission is in spite of the fact that his view is at odds with the formal charter for the IAHC, the group's published report, and the direct record of the group's actions.

The review cites IANA's community-based authority. Professor Mueller confuses this with a rejection of the importance of funding, which it was not. He further confuses the IETF technical standards specification process with the operations administrative work of IANA. He continues to misunderstand the role of operational expertise in policy planning for critical infrastructure services, and he ignores the particular 15-year history of successful administrative policy activities provided by operations geeks, for DNS and IP addresses.

Lastly, given the minor points that Professor Mueller chose to address in his response, it is curious that he fails to respond to the primary ethics point raised in the review, namely his pattern of erroneous or absent citations that substantially undermine many of the assertions of his book.

—Dave Crocker
Brandenburg InternetWorking

Zero Configuration


I recently read Edgar Danielyan's article on Zero Configuration Networking in the December 2002 issue of IPJ. As is always the case, as a journalist Edgar is entitled to hold and express his own opinions, so as I began the article I didn't know whether to expect glowing praise of Zeroconf, or a savage attack. Thankfully I needn't have worried. I found an excellent and well-balanced article.

I have two brief comments to make.

1. Since Edgar wrote his article, the old expired Internet Drafts have been updated. The drafts Edgar worked from discussed names ending in The actual shipping version of Mac OS X 10.2 ("Jaguar") uses names ending in just local. to designate link-local names, (link-local names are locally assigned, unique only within the local link, not required to be globally unique).
2. Edgar expressed the opinion that Zeroconf is only useful on small networks, not large networks.

While Edgar is correct that Zeroconf per se is aimed at solving the "small network" problem, discovering your local peers is useful no matter how big the network. At the recent IETF meeting in San Francisco, there was a large network with full connectivity to the Internet, including IPv6, yet the printers were still advertised using Rendezvous, and for Mac users those printers showed up automatically in the "Printer" popup menu in the print dialogs, with zero configuration.

There is also the issue that Rendezvous (the Apple product) will go beyond just what is required for Zeroconf (the IETF Working Group). Service Discovery, on which Rendezvous is based, doesn't have to be used only with link-local multicast DNS. It can also be used with conventional unicast DNS. For a preview of what the future might hold, you can browse to find an example list of printers at my house. Type:
nslookup -q=ptr

Thanks for publishing a great article.

—Stuart Cheshire
Apple Computer, Inc.

List of Acronyms

DARPA Defense Advanced Projects Agency
DNS Domain Name System
DNSO Domain Name Supporting Organization
gTLD-MoU generic Top Level Domain-Memorandum of Understanding
IAHC International Ad Hoc Committee
IANA Internet Assigned Numbers Authority
ICANN Internet Corporation for Assigned Names and Numbers
IETF Internet Engineering Task Force
IFWP International Forum for the White Paper
ISOC Internet Society
UDRP Uniform Domain Name Dispute Resolution Policy
WIPO World Intellectual Property Organization