The World Wide Web and emerging Internet resource discovery standards for scholarly literature - Networked Scholarly Publishing

Library Trends, Spring, 1995 by Stuart L. Weibel

Document search and retrieval works better in a session-based model - the server should retain session context for the user (reusable result sets, for example). OCLC has developed a hybrid HTTP-Z39.50 server to bridge the stateless world of HTTP and the session-based Z39.50 world (Weibel et al., 1994). The WebZ server provides Web access to Electronic journals Online and soon will provide an entry point to the reference databases available under OCLC's FirstSearch system.

The WebZ hybrid server is acting as a Z39.50 client for the HTTP client, maintaining session information for any sequence of interactions from the same client. This is accomplished by putting a session ID in all URLs produced by the gateway and returned by the client. Authentication of the user is required only at the beginning of the session. Thereafter, the session ID in the URL is matched against active sessions maintained by the server, validating the request as legitimate. The sessions are aged by the server; if no subsequent requests are received within an arbitrary time interval, the session is closed and any subsequent request by the same client would require re-authentication.

Finding What You Want

No discussion of access to networked information can fail to note the major problem facing the net - i.e., finding something you want. Web browsers are just that - browsers. As browsers, they excel, and only the jaded (or perhaps a too-busy reference librarian) can resist the appeal of clicking from site to site around the world, stumbling upon gems and chestnuts that delight the inquisitive mind. But along with the gems are plenty of dead ends, and to find something from a dead start - either a known-item search or a keyword or fielded search - is quite another challenge, a challenge which the current information infrastructure does not adequately support.

The evolving information infrastructure that has served libraries for many decades has yet to be transferred to the digital world. Some aspects of the two worlds require different solutions, but the problems of networked resource discovery are more like those of the conventional library than they are different. The library community can contribute valuable experience to these solutions. The long-term investment of the library community in MARC records and the Anglo-American Cataloguing Rules (AACR2) (Gorman & Winkler, 1988) represents resents a working model of how object description can be formalized to support resource discovery and retrieval. Should MARC therefore be the basis for similar systems on the Internet?

MARC and the Internet

The MARC record must certainly be accorded a hallowed place in the history of library automation, and it is unlikely that it will be supplanted in the foreseeable future as the currency of bibliographic record exchange. It must be seen for what it is, however - a carrier syntax. MARC is a container with well specified capacities and a set of rules (AACR2) that have been developed over a long period of time to guide us in packing these containers.