Why not to restrict first-level domains to companies which can demonstrate that they have 1000+ hosts?
I have an idea :-)
Why not create more top level domains to reduce the congestion in .com? Then let other companies run some of them so a free market can decide what policies they like best.
Because there's already a lot of them (i didn't count but it appears to be close to a hundred) and you'll have real hard time convincing root NS operators (who, FYI, provide that service free of charge) to add any more. And then, if you stopped to think for a minute before writing, it won't fix anything. O(N exp(M)) = O(exp(M)). In other words, if you throw in more TLDs you may (in the absolutely ideal case, when registration suddenly spread equally in all TLDs) it will delay the namespace collapse by 6 months for every doubling of number of TLDs. Since the real-world scenario would be far from the ideal case what you offered would cause a lot more problems and fix nothing. The only way to fix domain name service is to stop the insane flat-tree registration and start growing tree in depth. Which is -- if you want to start a registry ask NIC to allocate you something like MYREG.COM and register whoever you want *under it*. No negotiations with root NS operators. No additional load on root NSes. But, i guess, that solution does not appear to be as "sexy" as screaming bloody murder about InterNIC policies. The only problem i see with what InterNIC is doing now is that they encourage proliferation of the "first-level domains are sexy" hysteria. John et al -- how about stopping to register names under .com *NOW* and start registering them under .AA.COM (with understanding that later there will be .AB.COM etc, some of them delegated to other registries)? Name collisions could be resolved in that way, too, rendering the trademarked-domain wars unnecessary. (I won't say it'll make all lawyers to jump from the bridge, but still). It remains a mystery to me why so many supposedly intelligent people keep running in circles trying to argue against the trivial mathematics. --vadim
On Mon, 22 Apr 1996, Vadim Antonov wrote:
And then, if you stopped to think for a minute before writing, it won't fix anything. O(N exp(M)) = O(exp(M)).
Assuming that a .ALT domain is created in which *ONLY* 3rd level registrations take place (i.e. kirk.tlhIngan.alt) but the second levels (i.e. tlhIngan.alt) are all automatically created and served from the same set of nameservers, will this buy any more time? In order to get people to register at the third level you need some sizzle to go with the steak. I think .ALT is sizzle and automatically created 2nd levels are also sizzle that will get people to actually use it. I would expect that such a .ALT domain will quickly grow to approximately half the size of .COM. This is still an unbalanced tree but I don't understand where your math originates. What is the critical component that is O(exp(M))? Michael Dillon Voice: +1-604-546-8022 Memra Software Inc. Fax: +1-604-546-3049 http://www.memra.com E-mail: michael@memra.com
Vadim Antonov writes:
Which is -- if you want to start a registry ask NIC to allocate you something like MYREG.COM and register whoever you want *under it*. No negotiations with root NS operators. No additional load on root NSes.
Be it known then that I will accept applications for registration under commercial.com. Email for details (hostmaster@commercial.com), which should give me time to make some up. Why is this thread on nanog? Most everything that's been said here has been said (several times) on com-priv (along with a lot of other things making even less sense, of course). -- Dick St.Peters, Gatekeeper, Pearly Gateway, Ballston Spa, NY stpeters@NetHeaven.com Owner/operator, NetHeaven 518-885-1295/1-800-910-6671 Internet for Albany/Saratoga, Glens Falls, North Creek, & Lake Placid First Internet service based in the 518 area code
participants (3)
-
avg＠postman.ncube.com -
Dick St.Peters -
Michael Dillon