Web Magazine for Information Professionals

EDINA: WWW, Z39.50 and All That!

Peter Burnhill gives a briefing note on what EDINA and the Data Library are doing about the World Wide Web (W) and the Z39.50 Protocol (Z).

Why W? Well, a lot has been said and done about W. W clients are cheap (free?) and easy to come by. End users like them, they have point-and-click graphical interfaces. But the problem for the national bibliographic data services is how to provide access to the databases held in text management software, like BasisPlus used by both BIDS and EDINA, in a way that allows users to build queries interactively, across several physical databases.

Why Z? Well, because this provides, or will provide, an economic and versatile way to search complex databases and retrieve records without having to use the search commands particular to the database software in question. Inter-operability, as they say, in a client/server environment. Quite a lot is being said about Z, and you can have a look at Ariadne or Dlib, although I have yet to find a 'Noddy Goes Z' explanation. To jump in at the 'official' deep-end, try http://lcweb.loc.gov/z3950/agency/ which gets you to the Library of Congress.

Two problems: end-users want to use the Web browser on their desktop (eg Netscape), not a special Z39.50 client; the protocol spoken by W (http) needs to be translated back and forth into the Z protocol.

So what have we been doing? First, as part of a collaborative project with BIDS and COPAC, we have been looking at OCLC SiteSearch, which contains a product called WebZ and a search engine called Newton with a Z39.50 server. WebZ acts as a W to Z gateway. The result of that work is experimental Web access to PCI, currently being field-tested by a group of academic support staff in a number of universities.

Last month, we received the source code for an alpha-level Z39.50 server for BasisPlus from IDI, to port onto Solaris. If that can be made to work, then we should be able to use WebZ to connect to PCI in BasisPlus, and therefore be on the way to providing W searching for PCI, and subsequently for Biosis Previews.

We are also building upon the advances made in SALSER, a 'virtual' union catalogue of serials (http://edina.ed.ac.uk/salser/). This was launched in 1994 with concurrent searching of WAIS-indexed files, Version 1 of the Z environment. Over the course of the next quarter, SALSER is to be ported into SiteSearch, into Version 3 of Z. This would offer the possibility of allowing an end user to discover an information object (article) of interest in PCI and then be able to locate the carrier (journal) via SALSER.

All this is to state that we do have a real interest in both W and Z: we rate both as important in the UK electronic library.