Keywords: Z39.50, Java, OPAC, Library Automation
Access to online library catalogs (OPACS) and other information resources using the Z39.50 protocol is finally becoming widely operational. The question now is how the client software should be designed to access the Z39.50 compliant bibliographic databases. This paper explores the current methods for designing clients, forms based Web interfaces, gateways, and implementations of JAVA based Web search software. This paper advocates the use of the JAVA programming language to implement the ideal Z39.50 client.
In the 1992 presidential campaign, Bill Clinton's campaign headquarters had a sign prominently displayed that said, "It's the economy, stupid." It was James Carville's attempt to focus all the energy of the campaign on what he believed was the main issue. All of us in the information field need a similar sign that says, "It's the Internet" (We don't need the "stupid" part.) Indeed, it is the Internet: not just the Internet in it's present incarnation, but the general idea of true client-server technology. The days of the stand-alone application and even the proprietary client are numbered. Internet technology is just too powerful and too convenient. The only excuse for not accessing online catalogs via the World Wide Web is that we can provide greater functionality, but functionality loses to convenience every time. The more convenient interface will be preferred over the more functional interface.
Consider just the convenience of not having to distribute, maintain, and update custom client software in an organization. We can leave that to Netscape or Microsoft. We can assume that our users have a minimum level of technology if they are accessing the Internet. With the latest browsers we can assume that users have Java as part of their arsenal. Java is a programming language that allows the person navigating the Web to download an application, sometimes called and applet, with the Web page. This capability makes the custom Internet client obsolete. The Web page carries its own software, in effect customizing the generic Web browser.
In the library context, we want access to library catalogs wherever they are. Eventually, we want the document to be right "behind" the bibliographic record. This is the vision of the digital library. Prototypes of digital libraries now give us a taste of what is to come. The digital library experiments already have formidable collections of research materials on the Web in image, text, or SGML formats. The documents, illustrations and all, can be downloaded from within the bibliographic record.
Today, the most common access method on the Web is the use of forms based interfaces embedded in the Web browser. But the limitations of HTML and the current generation of browsers make it difficult, if not impossible, to do sophisticated information retrieval from complex library systems. The server tools needed to publish information on the Web this way are easily and cheaply available, but they cannot produce large result sets, except by providing a page at a time.
The current state-of-the-art in Web browser interfaces to library catalogs are systems like SIRSI's WebCat. These interfaces are far superior to the command line interfaces of the previous generation that many libraries still use. The interface allows the user to search the catalog by the usual author, title, subject categories. But more impressively, it allows the user to link easily to tracings of subject headings, or to link to all the works by the same author. Perhaps most impressively WebCat allows the user to link to the call number of the book, effectively allowing the browsing of the shelf to the left and to the right of the retrieved work.
There are serious limitations to the approach of using the generic browser as an interface to the online catalog and other information retrieval systems. "The individual WWW client has no knowledge of the application domain in which it operates. It receives a stream of graphical user interface primitives (such as buttons, text-entry fields, and formatted response data) from the server, and naively displays these to the user. The WWW inherits a problem that has haunted users since the first information systems went on-line using simple character terminals: no two information systems share the same interface characteristics. Each new system requires the user to master a new interface structure, and, with the advent of graphical interfaces such as the WWW, a new set of custom-designed icons and symbols." (Hammer and Favaro, p. 1)
"The Z39.50 standard was developed to overcome the problems associated with multiple database searching such as having to know the unique menus, command language, and search procedures of each system accessed. Z39.50 simplifies the search process by making it possible for a searcher to use the familiar user interface of the local system to search both the local library catalogue as well as any remote database system that supports the standard." (Turner) The work within the Z39 community has concentrated on the development and adoption of the protocols, so that the information gets to the user. " In the ten years since work on Z39.50 began, little attention has been given to the end user, the one who is supposed to ultimately benefit from the implementation of the standard. 'To put it simply: we [library professionals] get so hung up about information standards that we forget why we are using the information.' (Anderson, 46 quoted in Iltis)"
The lack of attention paid to actual users of library systems has characterized much of the technological development in the field. All too often systems are designed and built to make the lives of librarians easier, even at the expense of the user. Since librarians purchase library systems, they are the target market of the developers of library systems. Users are seldom involved in the purchase of library systems. This is not to say that librarians are not concerned with user satisfaction, but that the user is not the major factor in the design of systems. Experience with users suggests that they are very satisfied with the services they get, but also suggests that few users have any idea of what they can realistically expect. It is almost impossible for the library user to know what they are missing. If users were considered, all systems would be accessible from a common standard interface within the most widely used Web browsers. Clearly, this makes the most sense.
As Web browsers become an integral part of the computer operating system and the applications merge with the Web browser, we can expect users to want to access their libraries from their computer, rather than access a computer from within the library. This is a major shift of emphasis that will have profound effects on all libraries.
According to Baldacci and Favaro, the Z39.50 protocol was developed to avoid the problems associated with network connectivity and with varying interfaces. Search failure is often the result of seemingly similar interfaces that are actually very different. (Baldacci and Favaro) The new versions of the standard should directly address the problems of diverse interfaces and allow user-system dialog from a single interface.
The more powerful Z39.50 Internet clients are independent clients that allow searching of any system that is Z39.50 compliant. These clients, including SIRSI's VIZION allow the user to configure the client to search multiple information systems and to organize the results. By customizing the client, the user can organize Z39.50 servers as a personal set of online resources. Making sense of all the information resources now available online is a major task in itself.
SIRSI's VIZION goes a long way toward helping the user to organize information resources available on the Internet. It provides a consistent interface to Z39.50 resources and also allows the inclusion of other resources in "folders" that are accessible through easy to use icons.
The goal of the Z39.50 enterprise is to make many diverse information retrieval systems look and feel the same to the user. The same instruction set will do the same tasks in the same way.
A variation of the Z39.50 approach is to use the technology to provide locations of information as well as bibliographic data. An example of this is the effort to provide a locator to many available services and resources is the GILS (Government Information Locator Service) effort by the U.S. Government. If successful, the locator will provide an easily accessible way to find government information. A second example of the application of Z39.50 for providing a locator service is a global locator service for environmental information.
The solution to the problem of providing greater retrieval capability while maintaining the simplicity of the standard Web browser is to create the Z39.50 client as a Java application. This will provide the power and versatility of the proprietary Web client, but with the convenience of using the standard Web browser technology. By the end of 1997 most major application software suites will be Web based and Web browsers will be integrated into operating systems.
A Java based client will be essential for the user to be able to take advantage of this integration. Java promises to dramatically extend and expand the power of conventional Web browsers.
The Java solution will create the bridge from library catalogs to most major Windows and Macintosh applications. The goal should be to have the ability to search several library catalogs from within a word processor, downloading the bibliographic references to a personal database and then accessing the text and graphics of the sources while writing the document, copying quotes as needed.
This utopia can be realized either by the development of a stand-alone proprietary Web client or by the creation of a sophisticated Java application that will allow a generic Web browser to have the same power as the proprietary client. Since users are already familiar with their Web browser, the Java based solution is preferable.
Mark Kelly states that "Lowest common denominator metadata schemes yield lowest common denominator results for searchers. Java and other developments will enable stateful communication between client and server by providing clients the ability to download software which speaks something besides HTTP. The Web (which I define as information linked by HTTP) may become only a skeletal infrastructure to get to the interesting things on the Internet/Intranets. In this way the Web may be the vehicle of its own replacement". This optimistic view suggests that Java will provide the means to make the Web a truly powerful tool. The work on integrating the Web browser into application software also suggests that this is on the horizon.
Some efforts to create a Java solution are already underway. At Stanford University, Harold Finkbeiner is developing code for a Java client. Art Rhyno at the University of Windsor is also developing a Java Z39.50 client. These are research projects and at this writing are not yet available for general use.
The Java solution is clearly indicated, because the complexity of the interface between the Web client and Z39.50 servers demands the capabilities that only a Java application can provide. The Java client will be able to negotiate the search session and be able to handle large result sets coming from the server, displaying the results in "bite sized" pieces for the user.
___________________________
The research for this paper was supported by SIRSI Corporation.
WebCat is available as an interface to any Z39.50 compliant library catalog, but works best with SIRSI's Unicorn system. This interface is available to the catalogs of Rice, Carnegie Mellon, and Stanford Universities and many others.
VIZION is a product of SIRSI Corporation (www.sirsi.com)
The locator can be found at: http://enrm.ceo.org
Kelly, Mark. Three Modest Suggestions for Distributed Indexing/Searching Development. 9 April, 1996 Paper presented at the Distributed Indexing/Searching Workshop, MIT, May 28-29, 1996 Sponsored by the World Wide Web Consortium
___________________________
1. Hammer, Sebastian, and John Favaro. "Z39.50 and the World Wide Web." D-Lib Magazine (March, 1996): URL: http://www.ukoln.ac.uk/dlib/dlib/march96/briefings/03indexdata.html.
2. The 1996 GILS Conference. College Park, MD: U.S. National Archives, November, 1996. URL: http://www/dtic.mil/gils/.
3. Kunze, John A., and R. P. C. Rodgers. Z39.50 in a Nutshell (An Introduction to Z39.50). Bethesda, MD: Lister Hill National Center for Biomedical Communications, National Library of Medicine, July, 1995. URL: http://www.nlm.nih.gov/publications/staff_publications/rodgers/z39.50/z39.50.html.
4. Lynch, C. Using the Z39.50 Information Retrieval Protocol in the Internet Environment. Berkeley, CA: University of California, Office of the President, December, 1994. URL: http://ds.internic.net/rfc/rfc1729.txt.
5. Addyman, A. M. Facilitating the Creation of Z39.50 Origins in the UK. England: University of Salford, January, 1994. URL: gopher://ukoln.bath.ac.uk/00/Publications/Z39.50/zcapir.txt.
6. St.Pierre, M., et al. WAIS over Z39.50-1988. : IIIR Working Group, June 1994. URL: ftp://ds.internic.net/rfc/rfc1625.txt.
7. Buckley, C., and P. Ryall. Position Paper: Z39.50 & Ranked Searching. URL: http://lcweb.loc.gov/z39.50/agency/indexing/peter.html.
8. St. Pierre, M. Z39.50 and Semantic Interoperability. : Blue Angel Technologies, URL: http://lcweb.loc.gov/z3950/agency/indexing/margaret.html.
9. Christian, E. Proposal for an Information Locator Service. : MIT Distributed Indexing/Searching Workshop, May, 1996. URL: http://www.usgs.gov/gils/w3c.html.
10. Kelly, Mark. Three Modest Suggestions for Distributed Indexing/Searching Development. Cambridge, MA: MIT Indexing/Searching Workshop, May, 1996. URL: http://pages.prodigy.com/ZUPN84A/www.htm.
11. Baldacci, M. B., and John Favaro. Online Public Access Catalogs and the World Wide Web. URL: http://lcweb.loc.gov/z3950/agency/indexing/baldacci.html.
12. Moen, W. The ANSI/NISO Z39.50 Protocol: Information Retrieval in the Information Infrastructure. URL: http://www.cni.org/pub/NISO/docs/Z39.50-1992/50.brochure.toc.html.
13. Z39.50 Bibliography. Washington, DC: Library of Congress, URL: http://lcweb.loc.gov/z3950/agency/biblio.html.
14. Denenberg, R. Z39.50 Recent Developments and Future Prospects (Presented at the September 30, 1996 Z39.50 Seminar at the Royal Library of Belgium. Washington, DC: Library of Congress, Sept, 1996. URL: http://lcweb.loc.gov/z3950/agency/brussels/kbr.html.
15. Dempsey, L., R. Russell, and J. Kirriemuir. "Towards distributed library systems: Z39.50 in a European context." Program 30.1 (January 1996): 1-2. URL: http://www.aslib.co.uk/program/article2.html.
16. Wood, A. Z39.50 Client Reviews. : Distributed systems Technology Centre, Resource Discovery Unit, 1996. URL: http://www.dstc.edu.au/RDU/reports/zclients.html.
17. Turner, Fay. An Overview of the Z39.50 Information Retrieval Standard. Ottawa, Canada: National Library of Canada, June 1996. URL: http://www.nlc-bnc.ca/ifla/VI/5/op/udtop3.htm.
18. Nebert, D. D., and J. Fullton. Use of the ISite Z39.50 software to search and retrieve spatially-referenced data. Reston, VA: Digital Libraries 95, 1996. URL: http://csdl.tamu.edu/DL95/papers/nebert/nebert.html.
19. Iltis, S. Z39.50. Seattle, WA: University of Washington, March, 1995. URL: http://www.cqs.washington.edu/~camel/z/z.html.
20. Freedman, M. White Paper on Javafied Willow. Seattle, WA: University of Washington, Oct., 1996. URL: http://www.washington.edu/willow/java.html.
21. Funfrocken, S. Distributed systems - practical course SS96 task 4: Z39.50JavaClient - Applet. 1996. URL: http://pages.prodigy.com/ZUPN84A/zjavaeng.htm.
22. Rhyno, A. Welcome to Zjava. Windsor, Ontario, Canada: University of Windsor, July, 1996. URL: http://janus.lamf.uwindsor.ca/units/library/zjava.html.
23. Hinnebusch, M. "Z39.50: Where is It and Who Cares?" The Public Access Computer Systems Review 1.2 62-66. URL: http://info.lib.uh.edu/pr/v1/n2/hinnebus.ln2.
24. Gravano, L., et al. STARTS Stanford Protocol Proposal for Internet Retrieval and Search. Stanford, CA: Stanford University, October, 1996.
25. Lougee, W. P. Strategies for Digital Library Development: The University of Michigan Case Study. Ann Arbor, MI: University of Michigan, August, 1996.
Prepared for TICER, First International Summer School, Tilburg, The Netherlands.
26. ---. Agent Architecture for Digital Libraries: The University of Michigan Framework. Ann Arbor, MI: University of Michigan, August, 1996. Prepared for TICER, First International Summer School, Tilburg, The Netherlands.