[12395] in Commercialization & Privatization of the Internet
Re: About the creation of new 3-letter domain(s)
daemon@ATHENA.MIT.EDU (Noah Friedman)
Tue May 17 15:32:06 1994
Date: Tue, 17 May 94 11:06 EDT
From: friedman@gnu.ai.mit.edu (Noah Friedman)
To: dcrocker@mordor.stanford.edu
Cc: com-priv@psi.com, hostmaster@internic.net, hostmaster@nic.ddn.mil,
namedroppers@rs.internic.net
Reply-To: friedman@gnu.ai.mit.edu
In-Reply-To: <dcrocker@mordor.stanford.edu> Mon, 16 May 1994 13:47:26 -0700
>We have a dilemma. People give out their postal and phone contact
>information all the time. Why should email be different?
Because physical (geographical) reality and virtual (topological) reality
are two entirely different spaces. Why should email be forced to suffer
the same limitations as the postal system?
Let's be honest; geographics domain names are not the right thing in
principle, and not even in practice. The only reason to consider them is
as a means to balance the load on the DNS and as a feasible way to delegate
administration (essentially the same thing). At best they are a
compromise, and one that leans toward making life easier for computers, not
for humans.
A few months ago I suggested an alternative domain naming scheme for load
balancing without resorting to geographic names. The only question was how
to delegate administration of domains (since they would no longer be
geographically localized), for which the best solution I could think of
off-hand was to hold a lottery. People may think that's a strange way to
handle delegation, but with subdomains scattered all over the world, no POC
for any domain is likely to be better than any other.
