Date: Thu, 24 Mar 1994 00:58:23 +0100
From: Titia.vanderWerf@konbib.nl
Message-Id: <2D90BB7F.F4401*/G=Titia/S=vanderWerf/UA-ID=000000001/ADMD=400NET/C=NL/@MHS>
To: URI@BUNYIP.COM (Receipt Notification Requested) (Non Receipt Notification Requested)
Subject: comment on draft specs: URN, URC
Karen wrote:
> If we could agree on consensus before Seattle that would be
> great, but if not, then, please, at Seattle. It's less than
> 4 pages - shouldn't be hard to agree on. I had a certain
> number of murmurs after the last round, so let's hear it!
Alan answered:
> Seconded. If people could enumerate (briefly) their concerns
> with the current draft and post them to this list by Wednesday,
> I will incorporate such points into the agenda for specific
> discussion. I'm not talking about the character set question
> which can (I hope) be pretty quickly resolved but more about
> any philosophical underpinnings described in the document.
> If nothing comes up we'll use the time to review the draft
> URN specification (Peter Deutsch and Chris Weider) which will
> probably be have to be revised in light of Karen and Larry's
> document.
My main concern at this point, after having read the Specificati-
on of URN is the question: Why not consider using the existing
international naming schemes for identifying resources and
services in a unique fashion: ISBN (International Standard Book
Number) and ISSN (International Standard Serial Number)?
ISBN meets the requirements for URN's functional capabilities
(except the grandfathering of existing systems).
ISBN also meets the requirements for URNs as they are encoded.
The naming authorities for ISBN are organized in national ISBN
offices by country and the central registry is the International
ISBN office.
ISBN has been designed for automated book ordering systems and
is therefore appropriate for computer based systems.
The structure of ISBN is very straightforward:
It contains 3 series of numbers, each of which may vary in
length, enabling extensibility.
ex. 0-672-22588-3
1. country code : ex. 0 = USA
90 = Netherlands
2. publisher code: ex. 672 = Macmillan inc.
3. serial number given by the publisher
The last number is a check digit.
ISBN is not solely reserved for book material. On the contrary
it is defined very broadly, and includes all type of material,
also electronic media, online resources and databases...
I will send the definition tomorow, when I am back at the office!
ISBN is basically based on an identification system of publishers
- Internet resources are no different then books in this respect:
resources are also published on the network and there is always
a publishing organisation/individual behind the resource.
Probably some of you have already studied the applicability
of ISBN for Internet resources, but I have never encountered
arguments against using ISBN for URNs...
so please enlighten me!
As for the Specification of URCs, this is a very complex case.
Indeed libraries have been making citations (catalogue cards)
for ages, and most of the catalogue functions correspond to
those of the URCs, but encapsulation... at best that would be
the CIP -description made by libraries for publications in print
and printed by the publisher on the inside of the cover page
of a book.
I find the characteristic "Time to Live" of URCs as "no greater
than 0" a bit in contradiction to the lifetime characteristic
(persistence) of URNs they are supposed to encapsulate.
I wish you a lot of succes reviewing the draft specifications
at Seattle and good luck!
kind regards,
Titia van der Werf
Editorial Board SURFnet InfoServices
Royal Library
Prins Willem Alexanderhof 5
2595 BE THE HAGUE
THE NETHERLANDS
tel: +31-70-3140615
fax: +31-70-3140501
e-mail: Titia.vanderwerf@konbib.nl