Questions and Answers

How things might work in a world were TLDs are brand names run by directory companies.

Q: What if two people want the same name?

In most cases they can have it. With lots of TLDs, Apple computer might register apple.yahoo, while Apple records registers apple.lycos, and the Washington Apple growers registers apple.google.

What about a trademark conflict?

Of course, registering a domain is just that. Using a domain might involve violating a trademark. For example, say you register kodak.yahoo before the film company does. Kodak has the power to sue you in trademark court to give it up.

That's the way trademark law works and is supposed to work. There are tons of directories out there, entirely independent. I can place an ad for "Kodak Services" in the Yellow Pages, and probably it will go in. Kodak will sue me, and make me pull it and give up the phone number perhaps, but they don't shut down the directory or get it to recall all the issued Yellow Pages.

Kodak has what's called a famous mark, and it's also a made up word that never existed before the company made it up. It will have an easier time in trademark court than most companies. If Apple computer tried to sue over apple.yahoo, normal interpretation would have them lose unless you're using it to sell computers.

Q: Is this enough protection for trademark holders?

Trademark holders don't get any special protection in most media. Anybody can print books and flyers, advertise on TV, or list in most directories under a trademarked name. Nobody pre-checks in most cases that the user is the owner of the mark. The law leaves enforcement to the holder, using the courts as a weapon.

Sure, perhaps a name like Kodak would be recognized and the newspaper might not let you run a big ad pretending to be Kodak. But that's more the exception than the rule.

Would we want internet naming to be more regulated than the print it is replacing?

Q: Won't trademark holders try to sue TLDs? They try to sue Network Solutions

Indeed, it was suits against NSI that led to some of the bad policies that were put in place which heavily favour trademark holders. And indeed they will try. It is vital that these cases be watched closely and the courts encouraged to rule for the internet as they would in the paper world. The publisher of the directory is not liable for the actions of users of the directory that violate a trademark. Sure, if a company sues about trademark infringement, and the court order from a court of valid jurisdiction says the domain should be changed, they will obey the court order.

(Ideally however, the order is on the original registrant to make the changes. The TLD owner would normally only be brought in if the registrant refused to obey the court's order.)

Q: Won't big holders just register "their" name in every TLD?

That does happen when there are just a few TLDs, and will happen until the number grows very large. However, it is assured that in an open, branded system, this will not happen for many names.

Imagine there are 10,000 companies who feel the need to register "their" name in every TLD to be sure they own it.

This means there is a strong incentive to create a new TLD. If you do (and that's easy if you already manage one TLD) you are assured 10,000 registrations and some moderate price like $30. In other words, a quick $300,000 per year. Plus any other registrations you might get.

Thus people will keep creating TLDs until the number of companies that want to register in all of them drops low enough that it's no longer worth making a new TLD. And so, except for a very few names from sucker companies, there will always be a TLD you can register a name in. And for sure a TLD where you can register a very close version of a name even for those few monopolized names.

Q: What stops IBM from registering .IBM for their own sites?

The rule that any TLD must be mostly for resale would forbid this. As enforcement, if somebody brought a complaint that the domain was being used as a vanity TLD instead of for resale, the TLD could be lost and possibly a penalty paid. If your TLD has a risk of being lost, you won't base your valuable business names on it, so generally nobody will get a TLD with any significant risk of vanishing. Nor will anybody want to use one.

This also applies to the rule forbidding generic use. If a TLD doesn't delete domains that end up having a generic use, the whole TLD could be lost, which nobody will risk.

Q: Won't WIPO and powerful trademark holders fight to stop this system?

They might. I think giving them their own space in which to set their own rules might satisfy some of their needs. But they have no more right to control other spaces than they have to right to place a priori restrictions on publishing itself. The internet is the future of publishing, and the names are how people get around.

Q: With vast numbers of TLDs, won't people get confused?

We seem to be able to handle thousands of street names, town names, state names and country names in remembering the addresses of things in the everyday world. And we remember vast numbers of brands. Some TLDs will be obscure. Some well known. It will be a competitive market, and the better known ones will be able to charge more for registration.

Q: Aren't brand-name TLDs an very corporate answer?

Parties could create a TLD for any purpose -- not simply to meet market needs and make money. For example, a non-profit organization devoted to certain principles might create a TLD to reflect these principles, rather than the need for cash. While branding is normally thought about as a tool of business, it really is just a system to differentiate otherewise similar things. In the beginning, it was simply there to tell which ranch a cow was from.

Q: Aren't most terms generic in some language of the world?

A term isn'th inhernetly generic on its own, it is the use of the term in a generic fashion that makes it a generic term. Thus the generic term "apple" is OK as a trademark for computers, since it has no generic meaning when applied to computers -- there is no such thing as an "apple" computer, other than those given that name by that computer company.

Thus imagine that "Apple Directory Company" decided to create a .apple TLD. They could do so, but they could not sell domains in it to people who are selling actual apples, the fruit of the tree. This may seem bizarre at first, but it prevents their TLD from have a value as a generic as the place to get a domain if you sell apples. If somebody did register a domain like "washington.apple" for the sale of apples, others could file a complaint, and the domain would be stripped. That risk would pretty much stop people from wanting to get generic domains, without any need for pre-enforcement.

Trademark law acts the same way. If you create a directory of cars and call it "the apple directory," that can be a protected name, nobody else can do a directory of cars under that name, or perhaps (almost) any other sort of directory. The one exception is a directory of fruit related products. You can't stop anybody else from having a directory of apple growers called "The Apple Directory" -- no one party is allowed to own that. And so it should be in internet naming.

Q: What about cybersquatting?

With a lot of TLDs, cybersquatting is effectively impossible. Since each TLD is branded, no single TLD is the "one right domain" for any party, and if you have a name you want for a 2nd level domain, you can pretty much always find a TLD where you can register it.

It is probable, even likely, that some brand-name TLDs will do a very good job at marketing and gain cachet as "the" place for certain types of domains, simply by pleasing the market. However, dispute policy will be entirely up to that TLD. If they have a dispute policy that doesn't please registrants, then fewer people will register in such a domain and they won't keep the cachet.

Q: Why not just switch all applications to use search engines to look up names?

The use of search engines should be encouraged in all places where people might be trying to find a site without being sure of its name. Many web browsers can already be configured to do that. If a name is ambiguous or generic, a search engine is the right tool to find the actual site you're looking for.

However, people don't want a search engine's multiple answers when typing in an e-mail address off a business card, or following a web link. Both the provider of the name and the user want the name to resolve reliably and uniquely to the party which provided it. If you mail steve@apple.yahoo, you don't want your mail program to pop back, "Did you mean Apple Computer or Apple Records?"

In addition, mail programs currently don't do this, and it would take years for them to adapt to doing so. As such the ability to search when you're unsure on a name is a great feature to be encouraged, but it can't be the only option.