[12533] in Commercialization & Privatization of the Internet
Domain names, company names, and trademarks
daemon@ATHENA.MIT.EDU (Barry Shein)
Fri May 20 11:54:07 1994
Date: Fri, 20 May 1994 00:38:46 -0400
From: bzs@world.std.com (Barry Shein)
To: draperK@io.com
Cc: com-priv@psi.com
In-Reply-To: Draper Kauffman's message of Wed, 18 May 1994 18:08:42 -0700 <199405181104.GAA28328@illuminati.io.com>
>From: draperK@io.com (Draper Kauffman)
>* Domain names are intended to be mnemonics for IP numbers. When they get
>longer and more complicated than the numbers, people will revert to using
>IP numbers and handing them out on business cards, putting them on
>letterhead, and building databases of them, just as we do with phone
>numbers and fax numbers already.
No, that won't fly. Small changes in LAN configurations will change IP
addresses. Split an office ethernet into two and you generally have to
change the IP numbers on half of them (or put another way trying to
keep them the same would be a disasterous constraint on the physical
network design.)
No system adminstrator with any sense would tolerate the prospect of
people needing a a stable IP number. Where "stable" might be only
weeks or months. It's not comparable to the stability of some other
schemes (like phone numbers.) Shuffling around IP numbers is very
common. Just being moved down the hall often necessitates changing
your machine's IP number.
>* Using phone numbers for domain names? First, they aren't portable.
Neither is apple.com if you stop working for Apple.
No scheme other than random digits or similar is very portable. Even
people's names aren't all that stable, a lot of them change their name
when they marry. Company names change all the time (Standard Oil
becomes Esso becomes Exxon, more or less, that's about the largest
corp. in the world or top three or so anyhow.)
>Second, the phone company can already change your phone number on a whim to
>suit its own convenience--do you really want them doing that to your
>Internet address as well? Instead of enslaving the net to the
>inconveniences of the phone system, we can use the net to overcome those
>inconveniences. Lose track of a friend when he moves? No problem. Send
>him email and ask for his new phone number.
Well, as has been pointed out before, these are small objections and
just odd cases that no doubt happen but don't really quite prove that
therefore the phone systems of the world are unworkable.
No scheme will be 100% free of all possible problems.
The required action is to propose a better solution, not pick at odd
cases in others' proposals unless you truly believe you have something
that simply proves a scheme cannot work, or can be argued to be a more
serious objection than any with some other scheme you are championing.
The point being it's much stronger to argue *for* a scheme than
against one in these sorts of discussions. Others have pointed this
out here before, it's a common unwritten rule of this sort of
standards derivation. I hope you can see the logic behind it: We have
to come out with something in the end, even keeping the old way is a
decision, objections or no. Rarely is any scheme absolutely perfect,
but the final choice will be one in the list (including doing
nothing.)
-Barry Shein
Software Tool & Die | bzs@world.std.com | uunet!world!bzs
Purveyors to the Trade | Voice: 617-739-0202 | Login: 617-739-WRLD
