FW: infoURI standard officially blessed

From: Matthew Cockerill <matt_at_biomedcentral.com>
Date: Tue, 15 Nov 2005 18:42:44 -0000

PURL http://purl.oclc.org/ISBN
is defined, and resolves to:
URL http://names.oclc.org/
which givens (me) a server not found error.

I agree, the choice of 'registration authority' for a namespace is a social issue.
But a large part of that social issue is a sense of 'which authority has a policy on registration which will lead to practical applicability.

The info URI folks policies seem to make registration of a URI scheme something that strikes the right balance between being incredibly difficult/baroque - and therefore impractical for routine use (top level URI schemes fall into this category),
versus being almost totally free-for-all (e.g. wikipedia), and therefore likely to exhibit data quality issues problems.

As ever, if you go for the middle ground, you get attacked from both sides.

Note that as well as introducing lots of practical issues relating to domain name ownership, and physical server organisation, tieing identity to resolution ends up affecting how people use things - e.g. because purls act as redirectors to particular "web resources", that's how they are being used. Any given purl privileges some web resource as *the* target.

Info tries hard to opt out of this, and says 'there is no privileged target', and there is no implied need to dereference in order to make use of this information.

I'd say that less is more here, in exactly the same way that XML was more than SGML by being less.


PS In terms of avoiding infinite regress, I have a feeling that this debate about whether all URI should be dereferencable is a bit analagous to debating whether a computer language should have primitive types or not (and equally unlikely to reach an agreed resolution)

-----Original Message-----
From: Stefano Mazzocchi [mailto:stefanom_at_mit.edu]
Sent: 15 November 2005 13:51
To: general_at_simile.mit.edu
Subject: Re: infoURI standard officially blessed

Alf Eaton wrote:
> Stefano Mazzocchi wrote:
>> I'd much rather invest on automatically dereferenceable URIs now and
>> avoid paying the discoverability price (not only the technical one but
>> also the social one!) later on.
> Stefano,
> Just as example, which dereferenceable URI would you use for the book
> "The History of Love", ISBN 0393060349, so that everybody would be able
> to use that URI as an identifier for that book?



I perfectly understand why


is more appealing than the PURL above: the info URI doesn't feel "owned"
as much as the PURL. But this is a mere illusion and it's due to the
fact that you are giving dereferencing away for just identification.

If you later want to dereference the info URI, you might need to hit a
lookup mechanism. In the best possible scenario, this lookup mechanism
is an extension of the DNS, but it's unlikely since DNSs are so hard to
change and deal with and sysadmins don't like that, so it's probably
going to be yet another web-redirection service, so you automatically
translate the above into, say:


and tell me, how is this any different than the above that we can
*already* have today?

Ah, well, yes, purl.org is owned by OCLC and info-uri.info is owned by
Herbert Van de Sompel.

At the end of the day, it's all about who you trust the most, which is
fine and human and *has* to be taken into account.

So, using a URN scheme allows us to move forward without having to fix
both indentification and localization at the same time, *but* don't get
fooled thinking that localization will be more natural with the 'info'
URI and that the 'resolver owenership' problem is solved, it's only
pushed under the carpet, probably hoping that nobody will notice while
diverse powers take ownership of those hubs and become yet another
"network solutions".

Stefano Mazzocchi
Research Scientist                 Digital Libraries Research Group
Massachusetts Institute of Technology            location: E25-131C
77 Massachusetts Ave                   telephone: +1 (617) 253-1096
Cambridge, MA  02139-4307              email: stefanom at mit . edu
This email has been scanned by Postini.
For more information please visit http://www.postini.com

Received on Tue Nov 15 2005 - 18:40:15 EST

This archive was generated by hypermail 2.3.0 : Thu Aug 09 2012 - 16:39:18 EDT