[132] in North American Network Operators' Group
Re: US Domain -- County Delegations
daemon@ATHENA.MIT.EDU (Paul A Vixie)
Tue Aug 1 13:14:38 1995
To: nanog@merit.edu
In-Reply-To: Your message of "Tue, 01 Aug 1995 00:39:37 PDT."
<199508010739.AAA24971@puli.cisco.com>
Date: Tue, 01 Aug 1995 10:11:30 -0700
From: Paul A Vixie <paul@vix.com>
> We have certainly taken a solid whack at the root DNS server
> overload "problem". The disadvantage is we haven't solved the
> namespace collision problem, but I see that as one for the lawyers
> to solve, not us bit twiddlers.
we bit twiddlers have set up the current system in which for small values of
N, it is possible to use the naming service as a location service. as others
have pointed out, it used to be a fair bet that WWW.mumble.COM was Mumble's
WWW page -- but that's not true for a lot of domains (EXAMINER.COM among them)
and for increasing values of N it won't be true, statistically speaking, at
all.
i'm not concerned (here, anyway) about the legal aspects of name collision.
what i'm worried about is the "all the good names are taken" problem as well
as the "which EXAMINER do you mean?" problem.
i don't intend to solve the directory services problem off the cuff -- X.500
tried that and came up with something Truly Dismal. whois++ is an attempt and
i for one am satisfied to wait and see where that goes. the new invariant i
want to introduce is that "domain names don't have to be intuitive anymore" --
this will put the pressure on the directory services problem, where it belongs.
