Message-Id: <9309182224.AA06832@mocha.bunyip.com>
To: Richard Wiggins <WIGGINS@msu.edu>
Subject: Re: URN Usage
In-Reply-To: Your message of "Fri, 17 Sep 1993 00:07:19 EDT." <9309170427.AA05008@mocha.bunyip.com>
Date: Sun, 19 Sep 1993 00:24:08 +0200
From: "Harald T. Alvestrand" <Harald.T.Alvestrand@uninett.no>
I don't see what the problem is.....
it seems rather clear to me that we want an URN that names intellectual
content - one which can be used in reference lists in journals and so on.
It seems also clear to me that we want URNs for any item that has unique
properties - such as the ASCII version of a PostScript text, or the
JPEG 20 Kbyte image of the Mona Lisa; again, because we need to refer
to them.
It should be painfully obvious that in some cases we need the ability
to walk back the chain of descent: ASCII descended from GIF descended
from PostScript != ASCII descended from French translation of PostScript.
But I believe this is part of URC, not URN; any protocol *really* using
URNs should have the ability to say "this object is created from URN
xyzzy".
Harald A