ICANN: It must take responsibility for its own failure

ICANN seems to be unwilling to appreciate how well the Internet has acted as a marketplace and is trying to fix what is not broken rather than providing simple and boring services. By confusing stable handles with commercial identifiers ICANN helps assure that the Internet cannot be stable since the handles all expire and, worse, can be reused. [Bob Frankston, 2002-02-26]

While this is a very harsh judgment on ICANN, I want to take pains to emphasize that I am criticizing ICANN itself and not the integrity of the individuals involved. I've known a number of the participants for many years and they are trying very hard to do what they see as best for the Internet. While I disagree with some of their decisions I want to be very explicit in praising their personal integrity and intent.

ICANN, "The Internet Corporation for Assigned Names and Numbers", recently issued a report entitled "The case for Reform". As with any such report the appropriate reaction should be a yawn if you react at all.

Unfortunately, ICANN is the single biggest threat to the Internet itself. This is not because the members of the board are evil or conspiratorial. I know a number of them personally and they have the best intentions and are trying very hard to serve the public good.

To quote from the ICANN Fact Sheet: "ICANN has been recognized by the U.S. and other governments as the global consensus entity to coordinate the technical management of the Internet's domain name system, the allocation of IP address space, the assignment of protocol parameters, and the management of the root server system".

Unfortunately, ICANN's mission has gone seriously astray because of a fundamental misunderstanding of how the Internet works and why it works. ICANN seems to be unwilling to appreciate how well the Internet has acted as a marketplace and is trying to fix what is not broken rather than providing simple and boring services. Assigning unique handles should be about as complicated and uninteresting as handing out the next available number.

Jon Postel did this part time with one assistant. One of his tasks was to steward the domain name system. The DNS was created to solve a simple problem. The Internet (IP) addresses kept being reassigned as the early Internet grew. The solution was to create a simple mechanism that mapped stable handles into addresses. This mapping would be managed as a local zone. Each zone is found by following a path from the root. Since each machine in a work group usually had a name, it made sense to use these names as locally unique handles. Pluto in workgroup A would be "Pluto.A" and the same name in another could be "Pluto.B". The root itself has an empty name so the full path is "Pluto.A.School.EDU." (though we don't usually type the final "."). This worked well for the prototype.

But it is dangerously naive to confuse the expedient use of simple names with commercial handles like trademarks. In the real world you are not allowed to use the name Rollecks for a watch since it is too easy to confuse it with Rolex. But you can have a House of Pizza in many towns and use the name Cadillac for a car company and a dog food company and a shoe polish within a single country and reuse names in different countries.

Such a naïveté would be humorous if didn't threaten the Internet itself!

I pointed out some of the issues in my comments in "The Tragedy of the .Coms" but I understated the problems. By confusing stable handles with commercial identifiers ICANN helps assure that the Internet cannot be stable since the handles all expire and, worse, can be reused.

ICANN's confusion is reflected in the larger world of governance. Since the DNS names are confused with commercial names, we shouldn't be surprised when governments impose commercial terms and conditions on their use. And we shouldn't be surprised when various groups try to use the control over the DNS to threaten any site whose DNS name happens to prefix a path to content that they consider illegal or just don't like.

Imagine if your right to use a number were limited by complex terms and conditions. Imagine trying to build new services, like the Web, on the Internet if you first had to justify yourself to organizations dedicated to maintaining the status quo.

ICANN is proud that its main accomplishment has been the creation of seven new global Top Level Domain names (gTLDs). And the marketplace respond with utter and profound indifference! Who knows what these new TLDs are and who cares? People rely on their bookmarks and search engines. They might try typing the company name and .COM but it seems silly to try all possible TLDs just in case. It's like trying to dial 1-800 (in the United States) followed by a company name and expecting it to work most the time and then trying 1-888 and then +41-800 (or whatever the standard prefixes are in each country).

Now that we have directories and the ability to save the results of searches, the DNS serves little purpose except to allow the kind of meddling that violates the basic end to end architecture of the Internet! Just like the First Amendment in the US constitution prohibited prior restraint and free speech and helped give the United States a vibrant economy, the ability to innovate on the Internet without having to first ask permission has reshaped the world for the better.

It is unfair to blame ICANN by itself for this dangerous and naive foolishness. It just reflects the popular confusion of .COM with trademarks and the DNS with the phone book. Pandering to this misunderstanding assures that ICANN must fail in its mission and that the failure will be ascribed to bad intent. The real reason it can't succeed is that it is being asked to deliver a technical solution to what are really complex social problems and, at the same time, entangle the technical mechanism with shortsighted social agendas.

It's like trying to "fix" math by creating procedures that give everyone a vote in the value of pi.

There is a need for some simple housekeeping. ICANN could have and should have fulfilled its mission by simply providing handles without commercial meaning.

ICANN must use this opportunity for real introspection. It must accept the judgment of the marketplace. If no one is buying your product, your first response shouldn't be to ask the government to fix the problem. ICANN has failed because it violated its basic mission and, far worse, because it violated the basic design principle of the Internet by confusing technology with social policy. Instead of assuring that the Internet continue to function, it has denied the users control over their own fate and has denied them the ability to keep the Internet a vital and stable source of innovation.

As I said, this was all done with the best of intentions. But good intentions are not enough. We must demand understanding and leadership and restraint. ICANN must:

Separate technology from social policy. An immediate step would be the creation of the last TLD, perhaps .DNS. Handles in this TLD would have no commercial value, no commercial terms, and would never be revoked nor reused.
Champion the adoption of IPV6 with encryption. This would put an end to the artificial scarcity of IPV4 addresses. Encryption is necessary to prevent meddling, even if done with the best of intentions since, as we've seen, intentions are no substitute for wisdom. And the best of intentions requires complete knowledge of the future.

It will take strong leadership to tell people who are very passionate in their beliefs that the DNS cannot serve their social agendas. I know how difficult it is to explain why the .XXX domain doesn't make sense because there is no perfect unvarying universal definition of pornography. And that is a very simple and obvious example. Even more difficult than disabusing people of their naive expectations is transitioning from being at the center of attention to just handing out numbers without glory.

Perhaps ICANN should take heart. George Washington's (the first President of the United States) greatest accomplishment was, perhaps, resigning after his term was over and going back to being a farmer. Ceding power can be the most noble act.

Change can be difficult. But in 1999 I wrote a proposal for how to introduce the new mechanisms without having to invalidate existing names. I'm not advocating forcing everyone to abandon .COM names, just asking for a path out of the quagmire.

© Copyright 2002, 2003 by Daniel Bricklin, Bob Frankston, and David P. Reed
All Rights Reserved.