Hello Keith: >> > Well, it also matters that the set be constrained to some degree. >> > A large flat root would not be very managable, and caches wouldn't >> > be very effective with large numbers of TLDs. >> >> That's old fiction. If it works for .com it will work for ".". > >well, it's not clear that it works well for .com. try measuring >delay and reliability of queries for a large number of samples >sometime, and also cache effectiveness. > >let's put it another way. under the current organization if .com breaks >the other TLDs will still work. if we break the root, everything fails. Quick Question: Regarding Many TLDs vs. Fewer TLDs... If when .com breaks, the other TLDs still work... then, isn't that a good reason to have more TLDs? If you have millions of domains, across 1000s of TLDs, and, when one TLD goes down, then, doesn't it appear likely (statistically) that less domains would be effected in the event of such a problem? Regards, Mark