DNS articles
Brad Templeton Home
Brad Ideas
(My Blog)



Jokes / RHF

Photo Pages

Panoramic Photos

SF Publishing








RHF Home

Copyright Myths

Emily Postnews


Burning Man

Alice Pascal

The Rules for Guys

Bill Gates

ICANN readies to auction off the English language

ICANN readies to auction off the English language

ICANN is moving swiftly to expand the top level domains (TLDs) in the domain name system. While nobody has been happy with the limited set, including the virtual monopoly on valuable names in Verisign's ".com" domain, I believe ICANN is about to compound the mistake and make it bigger.

The ICANN gTLD plan will add, they estimate, 400 to 1,000 new TLDs per year. They call them generic TLDs, though there are really (at least two) types of TLDs that are being applied for. Applicants must pay a $185,000 application fee, most of which is not refundable, to go through the approval process.

Competition with .com and the rest is sorely needed. The method for this, which is new "generic" TLDs in which 2nd level domains are resold, raises the greatest concern. But there is also concern about the non-resale corporate TLDs that are expected to be sought, with ".ibm" and ".microsoft" given as examples. Those domains just push all the problems of .com up to the root, where they can't be pushed any higher -- more on that can be found here.

Trademark law figured out centuries ago that nobody should get monopoly ownership rights in generic terms, such as words in English (or other languages.) Instead, trademarks must be adjectives, variations of a generic term. Famously, there is a trademark for "Apple Computers" and another for "Apple Records" but none for just "Apple" or "Apple Orchards." Ownership rights are granted in the specific industries but never in general.

Since then, domain names have become by far the most important thing in the naming of companies and products. Traditional branding is still very valuable, but few create a new brand or company today without first assuring they can get a good domain name to match it -- one that will be memorable and typable. They often end up buying the domain name from somebody else who happened to have registered it first.

The .com domain became the USA's, and to some extent the world's de-facto place for branding. Ownership of a domain like "drugstore.com" or "business.com" has become like ownership of those English words when it comes to naming. People know it, and have paid millions in transactions to purchase such domains. There have also been big and expensive battles over names containing brands, even though trademarks are not supposed to exist outside of their contexts. Most such battles bypassed the courts to use the UDRP system, and most were won by the party with the most money. (Though there were occasional exceptions like nissan.com.)

To allow competition with .com, ICANN plans to welcome new TLDs with names like .sport, .music, .green (adding to a small set granted in the prior decade like .aero, .travel and .museum.

As a result, some domain company will own .music, giving them effective monopoly ownership of that word in internet naming, and thus global branding. (Very similar TLDs like .musical would be forbidden by the new rules, though synomyms like .songs in theory will be allowed and soften the monopoly.)

The owners of .music will want it because they hope it will become the only real choice in a domain related to music. It will have that status not because of their hard work at building a brand, but because it is a generic term with inherent value. The very thing trademark law forbids giving to somebody.

In addition, they can sell portions of their new feudal feifdom. Somebody will pay dearly for rock.music or pop.music or classical.music, again gaining a monopoly over a generic term. This may go to the party who claims it first or the party which pays the most at auction, and this may seem fair, but the reality is that nobody should own these terms.

We hope for competition, but in reality, .music won't be competing with .sport for customers. We'll see some competition in fields where there are truly equal synonyms, such as the controversial .xxx vs .porn or .eco vs .green. But it will usually be duopoly competition at best. In places where one name is the obvious winner, it will probably come down to an auction -- a bidding war for monopoly rights to generic terms and words.

Competitors for .com & the failure of the early gTLDs

It's possible that .music will provide serious competition for .com within music companies and brands. In particular, new bands may go there just because all the names in .com seem taken.

It's actually less certain that the more generic commercial domains will do so well. Already there have been several attempts to unseat .com including ICANN's early experiments with .biz and .info along with efforts by many small countries to promote their two-letter country domains as global commercial registries, such as .to and now .co. So far, these experiments have fared poorly. People have taken domains in them, but it's generally clear that something like a .biz domain (as well as a .net) were poor second choices by people who could not get the .com.

Even more surprisingly, and perhaps as counter-evidence to my thesis above, the few new generic TLDs like .travel and .museum have also been failures. They exist, and sell domains (usually at nice monopoly rent prices even higher than Verisign) but it's pretty rare to see a site advertise a domain in .travel or the others. In spite of the value of the generic terms, the lack of public awareness has kept the huge strength of .com as the only real choice. Anything else is second place. ICANN hopes that this will change as new gTLDs become well known and common. On the one hand, I suspect they are right. On the other hand, might I might hope they are wrong, which would mean the gTLD plan causes less damage, it is still a problem if Verisign's .com remains the global monopoly for everything. Once this is done, however, there is no going back. There is no higher level to punt the problem up to. When problems arise, it is likely that the "fix" will be to put tons of regulations on the TLD owners, reducing competition.

If the generics do not get customers, fewer people will be able to pay $185,000 just to apply for one. That price, which is claimed as a cost-recovery price by ICANN, seems crazy-high to me, and I would expect it to drop.

Real competition & regulations

I believe for real competition, we should have new TLDs, but they should be subject to very few regulations by ICANN. Unfortunately, powerful forces, which have managed to get themselves declared "constituencies" and "stakeholders" in the ICANN world, think differently. (Personally I do not understand why there is any other constituency than the users who look-up and register domains, and possibly the courts or national laws of the places this happens.)

Strangely, it is the trademark "constituency" which is holding up the new gTLDs. They know that if there are hundreds of them, they can no longer just register theirname.tld in every TLD. They have demanded the power to block the use of "their" names in every TLD, through a Uniform Rapid Suspension system.

They also plan regulation of certain types of TLDs. Anything that's a place name like .paris or .canada they will limit to the actual governments of those places. I actually feel that nobody should own these generic terms. In many places, the governments of places do not own their names. The U.S. government does not own the names "United States of America" or "USA" and people are free to use them in all sorts of ways. The City of San Francisco does not own the name "San Francisco." (ICANN does realize that it gets even worse with a name like .paris as there are many cities of that name, though one particularly large and famous one.

Finally ICANN has a special process for what it calls "community" domains, where the operator has to show they are the recognized leader or master of that community. This is the area where we will actually see the corporate vanity TLDs.

I believe for proper competition you want a decent number of TLDs which all started on an even footing. Neither should be inherently better or worse than another for those seeking a 2nd level domain. Instead, those domains might become more valuable because of the efforts of their owners. This is how brands work. A brand like "Kodak" had no meaning when it began, but through lots of hard work and marketing it became one of the most valuable names in photography. This is how domains should work as well -- no inherent value, but through hard work and money their owners should be able to make them valued places to register domains.

The curse of dot-com

Unfortunately, all approaches, including ICANN's, face the tremendous strength and public awareness of .com. It is uncertain if the new gTLDs will remove the cachet of .com, and if they can't, they will become small niches and the problems of DNS caused by it will continue.

I believe the only answer is harsh. Strong efforts must be made to devalue .com. While Verisign will fight hard to prevent that, the truth is they should not have been granted that monopoly and it is appopriate to rewrite the terms of it. However, the registrants of .com domains would also fight any effort that makes their high-cachet domain less valuable or doomed.

In particular, the phase-out of .com for generic domains should be announced. Sites which have a generic name in {d .com) or any other name which might have multiple parties with a right to use it would be forced to migrate to a new domain. This means that apple.com would have to migrate out (as nobody should own the word "Apple") though Microsoft.com and other coined term domains could stay.

To migrate out would not mean that apple.com goes away. Rather, at a cutoff date perhaps 2 years into things, all web requests to apple.com would do a Redirect Permanent (301) to Apple Computer's new domain, and Apple Computer would have some time before ceased new advertising of that domain. All web requests and E-mails and everything else would still work using the apple.com domain, but browser users would see the new domain and e-mail from Apple staff would use the new domain as Reply-to and eventually From header.

It could be that two options are offered: Migrate sooner and it's cheap. Migrate with more time and it's expensive, eventually ridiculously expensive. Either way, though, migration would be with enough time to wind out any advertising campaigns and logo material. (Old cards and printed material would still work as long as the site wanted to pay, people would just not be making more of it.)

The opposition from existing holders would be very high, and it may be the case that the domain remains and the problem is never fixed.

Alternatives to the present course

  • Forbid generic TLDs. Instead allow coined term TLDs with no inherent value.
  • Require TLDs be exclusively for resale of 2nd level domains and lower. The company owning the TLD may not use it, except for their NIC and domain-sale operations. This blocks most vanity TLDs.

Other Issues

It should be noted that there are other problems with some of the proposals. It is almost certain that the gTLDs will be censored, so you may not be able to get .fucked because of vetos by censors, raising all sorts of questions. (Mostly people wll be unwilling to spend $185K on dirty words.)

At the other end of the spectrum, while competition is hot for domains like .xxx or .porn there is a lot of opposition in the porn community to these, because of proposals for laws that will force adult sites into this TLD. Aside from the monopoly issues, the plan is that this would be used for filtering and regulation of adult sites. I wrote a satire of issues around this some years ago.

The Uniform Rapid Suspension system is another dangerous idea, with many censorship risks. Already there are laws being proposed to use the takedown of domains to censor or block copyright infringers, "terrorists" or other people who face the ire of governments or big corporations.