Over on Twitter I said something related to new TLDs. I have some views on what I said that are for another entry, but for now I want to sort out some views on new TLDs in general.

New top level DNS domains come in two flavours; there are country TLDs (and country-like ones) and what are officially called generic TLDs, things like ".online" and so on. Adding new country TLDs is obviously required and is sensible, and I feel this is also the case for new TLDs for country-like entities that aren't strictly countries, such as the European Union (which is not in ISO 3166-1, although 'EU' is reserved for identifying it and it has .eu as a TLD). I think this is most people's view on country TLDs, since it's hard to argue that new countries should be frozen out of having their own TLD when existing, older ones get TLDs.

(I'm neutral on the use of country TLDs to create attractive looking URLs, for example various domains using .ly and .io.)

What I have mixed feelings about are all of the generic, non-country TLDs (gTLDs), which are sometimes called vanity TLDs. There are a lot of these gTLDs (also) and some number of them wind up getting shut down after a while, among other issues. People can also criticize a number of gTLDs for being either unnecessary or not having a good general use, and some of them have clashed with common unofficial internal use (for example, .dev, which also forces HTTPS for all websites in it). All of this makes it sound like there should be either no new gTLDs at all or at least a lot fewer of them.

However, the choices for how many new gTLDs to have are none, some, or as many as people want. If the choice is "some", IANA is put in the position of deciding which gTLDs and which gTLD purposes are worthy enough to be approved, and which aren't. I doubt there's any broad agreement on what the criteria for this should be, which means that IANA would basically be making arbitrary decisions (either for gTLDs or for the gTLD criteria). The choice of "none" feels unsatisfactory and excessively conservative; it's saying that despite the Internet's evolution since the original gTLDs were picked (and despite the wart of .mil), nothing needs to change. That leaves the choice of allowing as many gTLDs as people want to pay for, which is the current situation.

So on the one hand I don't really like the explosion of gTLDs and the various consequences of that, but on the other hand I'm not sure I see a better practical choice. You could tinker around the edges, for example by saying that all gTLDs have to have a public purpose and not be restricted to a single company (so no .google or .microsoft), but I don't think that would be a big change. And I'm not sure that those company domains are the worst gTLD annoyances.