New top level domains
“Top level domains” are things like .com, .net, .de. Historically, we haven't had very many of them, and the process for creating new non-country-code domains (.de and .fr are country codes) has been slow and often inexplicable from the outside.
The policy advice arm of the generic top level domain world within ICANN is called the Generic Names Supporting Organization. (Generics are domains other than country codes, in general. Generically.) That group has a leadership council called the GNSO Council.
The GNSO Council is right now working on a set of policy recommendations about how to create new top level domains in a standard way. As you might imagine, there are many questions being raised about how to resolve conflicts among applicants and what strings (sets of letters and numbers) should be considered acceptable — also which strings should be “reserved” (made unavailable for various reasons). ICANN's current draft operating plan puts aside more than a million dollars for this process.
The same communications principles/decisions keep coming up as part of this discussion. What is ICANN's role? Are new TLDs (that's the lingo) like spectrum allocation decisions (traditionally viewed as public resource management, in which the licensee takes on public trustee obligations) or like potential new newspapers (traditionally viewed as private actors subject to no particular licensing regime other than compliance with applicable law)?
My view is somewhere in the middle. I think that the scarcity we now have in gTLDs has been entirely artificially created, and that it should be possible to have many more of them — subject to the resource constraints of making sure that the applicants are technically and financially viable. These resource constraints might require creation of a limited number of TLDs each year. It's fine with me to have some kind of minimal “objections” regime to show whether there is some global consensus against the creation of a new string. But in the absence of such a consensus, I'm not sure that ICANN has any particular legitimacy to serve as a gatekeeper for the content of particular strings.
I'm also not sure that the demand to create these things is limitless, and I think given the important roles played by search engines people will run out of energy to invest in a new TLD. People “find” things online without searching for particular strings in their browsers — instead, they search for names inside search engines. But for the moment, there are applicants who want to run new TLDs and they should have a predictable process to go through. Once they're created, they'll need to be subject to global consensus policies about stability etc.
If you have any interest in this subject, follow the links above and you'll see where the GNSO is.
==update: and there's a public forum on this topic on Monday here in Lisbon, from 2pm to 4pm (five hours ahead of NYC). It will be webcast and you'll be able to send in comments before and during the forum.