🚀 go-pugleaf

kre@cs.mu.oz.au (Robert Elz) writes:

] In <1992Aug31.234012.17114@bby.com.au> djk@bby.com.au (David Keegel) writes:

] | mainly because of the fact that AARNet charges (directly or
] | indirectly) a right-to-use the network fee

] Which is totally irrelevant,

Not totally -- no-one (sane) is going to register a name for a machine
which is not connected, and in a large percentage of cases this means
paying some money (possibly indirectly) to/for AARNet.

] no-one has ever claimed that sites with names in AU have to be AARNet users

Of course not.

] A better question might be "Can you guarantee that there will never be
] more than a few hundred individual's systems registered for mail?"

How about "Do you want to put random meaningless strings into people's
e-mail addresses?"  Or kre's question with "individual" replaced by
"company".

If there was some logical "ident" which could be applied, then by all
means, use it.  But no one has found it yet, and I don't see the point
in using contorted e-mail addresses just because somebody can't guarantee
that the number of possible future nodes is less than N.

[ When I say "logical", what I am hoping for is that someone with an idea
of the local subsection of the namespace could make a guess at a site's
E-mail address, given Real World information.  For instance, I work at
a company called BBY, so bby.com.au is the logical domain for my work.
Someone who has seen e-mail from other Aus companies could guess that
postmaster@bby.com.au is worth e-mailling to about my address. ]

I consider using the state in priv.au to be a half-logical ident.  Is
this better or worse than random?  Good question.
-- 
+----------------------------------------------------------------------------+
| David Keegel   <djk@bby.com.au>   Ph: +61 3 614-8922   Fax: +61 3 614-8742 |
+----------------------------------------------------------------------------+