[Agreed that it's a no brainer.](posted 8904 days ago)Originally there was a concern that if the TLD space was totally unlimited, it would result in a huge, flat file with millions of addresses in it. That defeats the purpose and efficiencies of the hierarchical system that was designed precisely to avoid that. If the entire file has to be updated continuously every time there is a single TLD change, then the system becomes unwieldy.
I'm not sure whether this is still the case and whether router efficiency is still threatened or how large a list would push the envelope. Fact check needed.
> Namespace is effectively unlimited, > and the market is the best -- indeed, the only -- viable means of > allocating this resource. The value of existing second-level names in > existing TLDs will, of course, decline -- but the policy should not be > chosen on the basis of protecting existing nameholders. TLDs can be > allocated on a first-come first-served basis -- if someone wants to > start a .biz, or .stuff, or .computer, or .store, or .law, or . . . . . > registry, why should we stop them from doing that?
Stability, financial security, competence. This is a technical service after all although it's a pretty simple one. One doesn't want commercial or personal entities investing good money in their websites only to discover that the registrar has gone bankrupt. What happens to those TLDs? It seems reasonable to set some criteria, the same way we do for state corporations as a protection to consumers. In this case, the agency responsible for administration simply couldn't handle even a thousand applications at once. If the gTLD is issued to an entity for its own use, without 2LD resale to the public (e.g., www.ibm) then this aspect may be of less concern. But we do bump all the present registrar problems up to another level (trademark disputes, equal access problems, etc.).
> The tools to > navigate through this more complex space are available, or will be > developed in response to consumer demand if those TLDs are successful > in attracting registrants.
Yes. But this may also reduce the need to issue more gTLDs. The main reason to add more (in my view) is simply to ensure that there will be competitive pricing and service in the gTLD registry function. That doesn't exist at present. Of course, that requires setting up new registries to compete with NSI.