ICANN: Fisher and Zittrain

(entry last updated: 2003-07-01 20:50:01)

I’m *sure* Donna did better with this: yep – don’t miss her notes

Z: We are going to talk about internet governance today, in the effort to tie everything together.

ICANN – have your heard of it; are you on it? (Ray, next to me). Domain names, why do people care, and so what?

  1. How the mess came about: Start with Cerf, John Postel, et al. Dave Clark: “Well it started out with 12 people in a room…” The nerds having a good time, and trying to keep the rest of us away.

    Requests for Comments, within the IETF. Iteration through the comments until it converged to a protocol. RFC 2555 – “These notes were the beginning of a dialog, and not an assertion of control.”

    So we have an engineering problem: the internet is driven by numbers. Let’s put a mnemonic name that is associated with a number, and then we just appear to use easy to manage names. A so-called namespace; a lookup list.

    The original list didn’t scale too well, so the dot structure was added. And, at the same time, the list gets distributed according to the dot names. And the top level domains (edu, com, net) are found by asking the .root list. A heirarchy of distributed domain names dynamically resolved into numbers.

    Jon Postel maintained the .root at USC; mirrors were set up A, B, C, …. Internet Assigned Numbers/Name Authority.

    In 1993, Jon Postel decided the job was boring and hard. The NSF had been funding him, so they generated a RFP for someone to maintain the list of names. Jon might maintain .root, and the contractor (eventually Network Solutions Inc.) maintains the list. NSI figures out that a fee for renting the name from them would offset NSF fees.

    Problems emerge: NSI making a lot of money; why can’t others? We need more top level domains. And cybersquatting starts to take place, to the horrors of corporations.

    Josh Quittner, for example, found that he could get mcdonalds.com. Kaplan.com pointed to the Princeton Review, a competing group. Offered to give it up for a case of beer, but Kaplan elected to sue instead. “Kaplan has no sense of humor, no vision,and no beer” (alternative telling)

  2. Terry Fisher: Clean up #1 legal intervention – Lawsuits as a solution to the problem. The first pass at this was based on the use of trademark law (1993-1999).

    Four kinds of trademark infringement: (a) identical marks on competitive products; (b) similar marks on competitive product (areo instead of oreo) – will this confuse the consumer? becomes the operant question – measured against the marketing environment; (c) similar marks on non-competitive products – even more detailed showing required to consider consumer confusion – possible confusion includes confusion as to the source of the product, confusion as to endorsement (rolls royce radio tubes), confusion post sale (bolt on parts to make a car look like a Ferrari; prestige drops; thus confusing the consumer); (d) dilution – the defendant’s behavior dilutes the power of the trademark, either by blurring it or tarnishing it.

    Back to Terry Fisher: In the US, the dilution doctrine is coming into its own (originated in Germany). Trade Related Aspects of Intellectual Property agreement (TRIPS) leading to coordination across boundaries.

    Trademark law is not designed for domain names, so some manipulation required. The Toeppen cases suggest that squatting might confuse and certainly dilutes. Other countries move in this direction.

    Problems and limitations of trademark as applied to this problem. First, it is expensive to bring suit – thus settlements moved money; Second – jurisdictional variations make it hard; Third – you really have to strain to make it work. Dilution, in the US, applies only to famous marks, used in commerce, and requires a showing of blurring or tarnishing, which may be hard to accomplish. Problematic as this problem gets revved up.

  3. J. Zittrain: Clean up #2 political intervention – taking over the domain name system.

    Jon Postel still struggling with the problem, so he thought about changing the technology.

    • Jon could add some new names

    • Put out an RFC, add some new names

    • Forms a committee

    The International ad hoc committee – IAHC is formed. And they generate the gTLD-MOU – Generic Top Level Domain Memorandum of Understanding. A big boring document in search of consensus, but it did have some good ideas.

    A conflict emerged between NSI and Postel. Jon finds it’s time to stop running the A .root, so he gave it to NSI – immeditely regretting it. So, NSI was the keeper, but Jon was the administrator. And NSI made it clear to Jon that if the gTLD-MOU process was finalized, NSI would not abide by it.

    Jon got half the roots to conduct a test – use the B.root instead of the A.root. A huge brouhaha. The US Government moved in and Ira Magaziner was moved into Commerce, and moved into the problem. “The White Paper” a statement of policy that says it’s time to privatize domain names.

    Attempts within the internet community during the summer of 1998 to develop something that will work. The International Forum on the White Paper (IFWP) forms to discuss a solution to this initiative. First meeting held in Reston, VA – home of NSI. No real agenda, yet things start to percolate. Jon Postel, meanwhile, starts to think about a new IANA and the rules thereof.

    NSI again wants Postel out of the picture, so they also generate a set of rules. One suggestion to negotiate at harvard, canned by Magaziner, so the meeting was held in DC. Thus was formed ICANN (IANA was rejected for obscure reasons).

    At the same time, we have changes in the kind of disputes that arise. It’s not just simple cybersquatting anymore. So we get typosquatting – exploiting the misspelling of users (microsoft.com); conflicts between non-competitors; retailers exploiting name variant; commercial vs. non-commercial users (pokey.org); fan sites; parody and commentary (peta.org v peta.com – introducingmondy.co.uk).

    Two new dispute resolutions mechanisms contructed. One is the UDRP – uniform domain name resolution policy (?). ICANN recognizes certain registrars, and requires that its licensees sign a contract that includes the agreement to be boind by the UDRP – a contract of adhesion that is global in its reach.

    Ostensibly it governs abusive registrations and use of domain names – “legitimate interest” and “bad faith” as key terms of art that govern success in recovery (largely derived from trademark law). Imposes an expeditious process favoring plaintiffs; done by e-mail; 20 days to respond to a complaint; 14 days to decision. Limited remedies: no monies involved, but cancellation of registration or transfer of the name to the plaintiff. You have 10 days to take it to court, which can be difficult.

    See the Mueller Report (6/24/202) Success by Default:

    A New Profile of Domain Name Trademark

    Disputes under ICANN’s UDRP.

    In addition to the UDRP, we see an addition to trademark law, known as the Anticyberquatting Consumer Protection Act – the UDP on steroids, offering up a civil cause of action using much the same language as the UDRP. Similar terms of art, but also a safe harbor provision based on intent. The key difference is that the remedies are serious money – $100,000 per name, potentially.

    Courts seem to be willing to work this act in favor of plaintiffs.

  4. How well did it work? – the state of ICANN today. What did we learn, if anything.

    ICANN constructs a board, many committees, and some at large members. The at-large membership was called for by Magaziner – the Boston Working group.

    One worldwide election was held; a disaster according to the board; the public interest groups thought it went ok, or at least fixably. Adopted the results, and immediately went to work making sure that there would never be another election. “Succession by right of kings,” as it now stands – no real electoral process.

    When asked if they care about this disenfranchisement, the audience sits in (stunned) silence. Not many worred about this, although Ray London does care (note Barbara Roseman is here and refining some of the description of ICANN – ITU is acting to try to be a part of the process).

    Alternative roots/other competitive issues

    The changing organizations and the changing rules that are produced out of these groups. Against this a set of legal evolutions, moving us through a set of new approaches.


    1. First possession

    2. Avoid consumer confusion

    3. Provide incentives to establish good will

    4. Freedom of expression

    5. Identity/Community/Equality

    6. Efficient Web Navigation

    In a world of Google, do we really need domain names anymore? Two puzzles: (a) if you believe the architecture matters at a deep level, and that markets are a problem, yet we see that ICANN could, in fact, act like a government if they wanted to, and would have an effect on this intrinsically important architecture. (b) Or who’s in charge of this anyway? And how can we build a business upon something that is so oddly managed?

    A number of possible actions that affect you a great deal are listed: yet, governance is not generally viewed as the path to a resolution of conflict over these actions – other recourses are selected/applied. Why does the issue of how ICANN is governed supposed to be the answer to the problem of domain names? The second puzzle in Jonathan’s list.

    Sorry, I think I lost track of something at the close here. I’m sure to come back to this.