W3C home > Mailing lists > Public > www-zig@w3.org > September 2001

Re: What libraries need -- Why ZNG

From: Sebastian Hammer <quinn@indexdata.dk>
Date: Sat, 29 Sep 2001 10:07:01 +0200
Message-Id: <4.2.0.58.20010929094437.017b7cb8@bagel.indexdata.dk>
To: "Stevens, Pat" <stevens@oclc.org>, www-zig@w3.org
At 22:50 28-09-2001 -0400, Stevens, Pat wrote:

>2)  As Thomas Place of Tilburg commented, libraries also want 'to give
>integrated access to scientific information sources' and other network
>resources not under their direct control.  An increasing amount of library
>budgets are being spent on licensing electronic resources which reside
>somewhere else on the network.  This is BtoB with the library as the service
>manager, and Information Provider as supplier.
>
>The providers do not have the same interests as librarians and they don't
>have government funding as an inducement. It would seem that since libraries
>are the buyers in this relationship that they would be able to set the
>standards for access. For complex reasons, that doesn't seen likely to
>happen.  Integration approaches that are easier to implement could help make
>this integration possible.

It is a possibility, but as you suggest, the reasons why the providers are 
unwilling to provide open, interface-independent access to their data are 
complex, and just handing them a simpler standard may not address their 
core concerns. Looking at some of the big, commercial suppliers of 
full-text journals, for instance, are there any that have already, 
voluntarily implemented interface-neutral web-services (like the NLM)? 
After all, the requirements for such an interface are fairly 
straightforward and they don't really need library people to tell them how 
to do it if they feel it's a good business move.

I find it hard to imagine an organisation like Elsevier not implementing 
Z39.50 just because they get stuck reading the standard or because they 
can't figure out the ASN.1.    :-)

>3) Libraries would like to help users seeking information elsewhere on the
>Web aware of the print and electronic resources available through the
>library.  They would like to help users accessing search engines and other
>online resources like encyclopedias to find related information that their
>library can provide. In a similar fashion, university libraries want to
>integrate library resources into online curriculum management.

This seems a distinct possibility -- perhaps especially within specific 
subject areas where, for instance, scientific libraries may fit very well 
into a portal environment. This is an argument for ZNG that I can readily 
support. However, in the case of even moderately successful portals there 
is a real danger that the library system behind the scene will be exposed 
to user loads beyond anything it has been designed to handle. That's a real 
engineering problem we've observed in concrete experiments doing just what 
you suggest above.

Just to poke fun at poor Ray, consider the LOC -- a sexy resource if you 
like books in general -- which paradoxically may have one of the slowest, 
most unreliable Z39.50 servers on the face of the planet. What would happen 
if Google decided it would be neat to provide parallel searching in the LOC 
as a service to their users? Very little. The LOC would have to do some 
*significant* re-scaling to support even a moderately successful internet 
service.

Given the current state of technology, and the development I can foresee 
over the next few years, it may well be that an OAI-like approach (ie. 
replication of database contents) would work better for anything more than 
a very local, low-key sharing of data, say, within a specific field of 
research.

That's not an argument against ZNG or Z39.50, but a warning, again, that 
the concrete protocol may well turn out to be the least of your problems.

--Sebastian
Received on Saturday, 29 September 2001 04:08:06 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:13:27 UTC