Are Multiple Roots Outside the Scope of ICANN?
Date: Thursday June 07 2001, @03:15PM
Topic: Alternate Roots

Mueller writes "In the debate over multiple roots, it is common to see assertions like this: "Alt.roots have no rights inside the ICANN controlled namespace, and ICANN has no rights inside the alt.root namespace. They are mutually exclusive."

It would be nice if this were true, but it isn't."

Sooner or later, we will have to stop pretending that multiple DNS roots are "mutually exclusive." If they really were, no one would care about the issue. Everyone could do their own thing, and no one's thing would affect anyone else's.

Alternate roots are not, for the most part, "private" name spaces. They are intended to be substitutes for the ICANN root. They are open to public use. They are almost always designed to be compatible with the existing TLD assignments in the legacy root. This in itself ought to be enough to refute the myth of mutual exclusivity.

However, when conflicting name assignments are made, as in the .biz case, it becomes quite clear that one root must drive out another, or lead to some kind of higher-level compatibility arrangement to avoid widespread conflict. This is because the DNS root is characterized by very strong network effects. If one must choose between a .biz supported by a small root system with say 10,000 users, and a .biz supported by a large root with 50 million users, compatibility concerns will push most users into the one with the widest scope of compatibility.

We can ignore alt.roots now because they are so much smaller than ICANN's root as to be insignificant. But what if an alternate root becomes as large as or larger than ICANN's? This possibility has been demonstrated by

Prior to Stockholm, I tried to initiate a DISCUSSION of this issue in DNSO, without presupposing a result. At the VERY LEAST, it would have been nice to come up with a consistent policy regarding TLD assignments that are already being used by other roots.

It should not be that difficult to define criteria to guide ICANN as to when it should avoid making conflicting assignments, and when it should not bother to avoid conflict. E.g., if I have jumped on a TLD name two weeks ago, have no customers, and no real operational capacity, there's no reason ICANN should respect such a claim. On the other hand, it seems self-destructive and destabilizing for ICANN to make TLD assignments that conflict with legitimate, established companies. A company like Ambler's IOD, for example, was using .web and claiming it for many years. has many customers both among ISPs and the public. Conflicting assignments in those cases could be construed as nothing more than a monopoly trying to crush its competitors.

In short, ICANN's relationship to alternate roots is an issue very much like Microsoft's relationship to other operating systems or the old AT&T's interconnection with alternate long distance networks. In fact, AT&T DID say, in the 1960s, that alternative long distance networks, if they were to be allowed at all, had to remain non-connected to the public switched network. It was the interconnection of private microwave networks to the public system (by MCI, no less) in 1978 that ushered in the legal and regulatory revolution that redefined the nature of the telephone system to make it more competitive.

If, as Esther Dyson now claims, ICANN is an "antitrust authority," (and I'll have fun with that assertion in a later post) what are we to make of a competition-enhancing agency that not only refuses to recognize that it is in competition, but actively asserts its right to create conflicts that can snuff out competitors, or lead to extended battles over supremacy that will create problems for users?

In Stockholm, that simple policy issue got mixed up - deliberately by some people but due to ignorance in other cases - with the silly notion that there should be no coordination at the root level.

Whatever one's position on how alternate roots ought to be treated, the real tragedy of Stockholm is that open, bottom-up discussion of this issue was deliberately pre-empted by the ICANN management. ICANN CEO Stuart Lynn unleashed a PR barrage a few days before the meeting purporting to describe an already existing policy, and the Board and Names Council let him get away with it. The so-called discussion draft was nothing more than an another assertion that ICANN is the "authoritative root" and a rather scurrilous set of attacks on the legitimacy of alternate roots; i.e., on ICANN's competitors. And of course, as we learned from the Names Council vote, the business interests that dominate the DNSO did not want to discuss it at all.

It is useful to analyze what the myth of mutual exclusivity accomplishes in this regard. For ICANN, the myth gives it a green light to create conflicts with any alternate roots that might threaten to achieve the critical mass needed to challenge ISPs' adherence to the ICANN root. If our name spaces are "mutually exclusive" after all, it doesn't matter if I come crashing into yours with twenty tons of network effects.

Actually, the "mutual exclusivity" myth would not be so bad if the ICANN supporters would simply be consistent about it. If alternate roots are really outside its scope, then shut up about them! As everyone knows, however, ICANN management and certain Names Councillors recognize that the success of alternate roots really does threaten their monopoly control over DNS administration. If alternatives do achieve widespread use, ICANN will have a tough choice: either fight them and create widespread compatibility problems, or coordinate with them and give up a lot of its ability to impose onerous policies on Internet users, registries and registrars.

This is the corner that ICANN has painted itself into. It must now pretend that other DNS roots are outside its scope and utterly irrelevant to its decisions and processes, while at the same time scurrying about Washington and the IETF lobbying hard to convince everyone that they are destructive and dangerous. Hypocrisy, thy name is spelled I-C-A-N-N!

This discussion has been archived. No new comments can be posted.
Are Multiple Roots Outside the Scope of ICANN? | Log in/Create an Account | Top | 20 comments | Search Discussion
Click this button to post a comment to this story
The options below will change how the comments display
Check box to change your default comment view
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Re: Are Multiple Roots Outside the Scope of ICANN?
by tlr (reversethis-{gro.tsixe-ton-seod} {ta} {relsseor}) on Thursday June 07 2001, @11:08PM (#766)
User #34 Info |

If ICANN would perform to the satisfaction of market participants, there would be no point in taking on the cost of creating alternative ways of resolving new TLDs - like does. Maybe, instead of trying to get rid of alternative roots the hard way, ICANN should try to get the current mess cleaned up gratuitously, and to perform better in the future.

[ Reply to This | Parent ]
Re: Are Multiple Roots Outside the Scope of ICANN?
by Wireless on Friday June 08 2001, @04:56AM (#768)
User #2902 Info
To obtain a detailed perspective on's proposal for a market-based TLD creation and allocation structure -- reducing ICANN's role to a merely technical one -- read the white paper found here:
[ Reply to This | Parent ]
  • 1 reply beneath your current threshold.
Re: Are Multiple Roots Outside the Scope of ICANN?
by fnord ( on Friday June 08 2001, @06:08AM (#771)
User #2810 Info
M. Stuart Lynn states in his draft that Over the past several years, some groups have established alternative root nameservers on the public Internet. I found this a rather curious statement. What is the 'public' internet? I am presumably free to walk down the street and enter a house of ill repute. Does that make it a 'public' place?

I previously did a bit of content analysis regarding the repetition of the word 'community' in Lynn's draft. The 46 times it appears there is surpassed by the 48 times 'public' is used. ICANN seems to be floating a meme here, that to it has fallen the sacred trust of not only numbering each house, but specifying which should be considered 'on the other side of the tracks' for the greater good.

In actuality I may choose to use or ORSC or Walid or an increasing host of others to act as a taxi service and neither I nor the cab driver pay much attention to such abstractions. ICANN may or may not be a single point of failure but it is a convenient single point of control. It is apparently scaring the pants off the control freaks that this single point is no more than a shared hallucination, only agreed to when it was consensual. ICANN's unilateral and repeated claim that the illusion is still real because they say so just doesn't wash anymore. Taxi! -g

[ Reply to This | Parent ]
Re: Are Multiple Roots Outside the Scope of ICANN?
by SimonHiggs on Friday June 08 2001, @04:08PM (#775)
User #2898 Info
Please see the following internet drafts for the technical details of the alternate roots. Maybe you won't then jump to such false conclusions:

Alternative Roots and the Virtual Inclusive Root

Root Zone Definitions
[ Reply to This | Parent ]
  • 1 reply beneath your current threshold.
Re: Are Multiple Roots Outside the Scope of ICANN?
by 300baud (service at evesnetwork dot com) on Saturday June 09 2001, @02:13AM (#778)
User #2869 Info |
Mutliple roots are mutually exclusive. Period

My testbed DNS server queries multiple roots and works quite well. Each TLD belongs to a single root and zero collisions.

The "it can't be done" argument washes almost as well as Mr Lynn's "single point of control" premise.

ICANN does have authority for the USG A root, but to try and exclude all other roots from existense is nothing more than Microsoft style monopolism.

Idealism and freedom are not dead on the Internet.

[ Reply to This | Parent ]
Re: Are Multiple Roots Outside the Scope of ICANN?
by Mueller ({mueller} {at} {}) on Monday June 18 2001, @11:44AM (#896)
User #2901 Info |
If multiple roots are mutually exclusive, and you know what you are talking about technically (which I doubt) then please reply to this message by identifying yourself promsing to pay me $50 if you ever claim that other roots will lead to technical instability or otherwise lead to compatibility problems on the Internet.

Ya can't have it both ways. Either they are "mutually exclusive" and therefore don't effect the ICANN root, or they cause instability. Either/or. Let me know which side you pick as soon as you make up your mind.

And oh, as for my "filthy economical interests" I'm a university professor.
[ Reply to This | Parent ]
  • 5 replies beneath your current threshold.

  • This article comes from ICANNWatch

    The URL for this story is: