Re: report: URN Architecture Meeting at University of Tennessee, Oct 30-31

Keith Moore (moore@cs.utk.edu)
Thu, 09 Nov 1995 03:49:38 -0500


Message-Id: <199511090849.DAA23430@wilma.cs.utk.edu>
From: Keith Moore <moore@cs.utk.edu>
To: Renato Iannella <renato@dstc.edu.au>
Cc: moore@cs.utk.edu, urn@mordred.gatech.edu, uri@bunyip.com,
Subject: Re: report: URN Architecture Meeting at University of Tennessee, Oct 30-31 
In-Reply-To: Your message of "Thu, 09 Nov 1995 16:00:22 +1000."
             <199511090600.QAA02537@nova.dstc.edu.au> 
Date: Thu, 09 Nov 1995 03:49:38 -0500

> With a new TLD, are you saying that we (for example) would
> have a resolver at dstc.edu.au.urn  ??

Not exactly.  There would be NAPTR RRs in DNS for {something}.urn
(perhaps not dstc.edu.au.urn, but rather some less meaningful name --
semantic loading is bad if you're trying to provide stable names)
that pointed to one or more resolvers for URNs that you assign.
In this case your URNs would be of the form urn:/urn/{something}:{whatever}
(modulo a possible shorthand notation to get rid of the redundant "/urn")

Or you could define URNs of the form urn:/au/edu/dstc:{whatever} and 
add NAPTR records at dstc.edu.au that point to your resolvers.

You wouldn't have to use the .URN domain if you didn't want to; 
it is being proposed so that there will be a chunk of name space 
in which naming authority names can never be reassigned.  
(as well as having some other desirable properties)

Keith