The MELVYL system in the larger context of the University of California's information technology planning

The MELVYL information system is an important component of the University of California's (UC) strategy to transfer from print to networked information. This paper traces the development of UC's library automation strategy over the past five years and the role of the MELVYL system within this strategy. This paper will highlight the need for systems, such as the MELVYL catalog, to function within an overall institutional strategy and the issues raised by such coordination efforts. (1) The evolution of the MELVYL system into a manageable production operation, and the development of an institutional library automation policy framework within which the MELVYL system could succeed, may be of interest to other information technology planners and managers in educational institutions charged with developing institutional initiatives in this vital area of information technology. ORIGINS OF THE MELVYL SYSTEM: A POLICY PERSPECTIVE The MELVYL catalog had its origins in the 1977 Plan for Development for the University of California library system. (2) Although the plan called for the development of a specific automated system to provide a unified view of the library collections maintained by University of California's nine campuses, which became the MELVYL catalog, the 1977 plan did not articulate an institutional library automation strategy. At this time, there was little coordination between the planning for a UC information technology infrastructure (both networking and computing) and library automation initiatives. In the late 1970s, library automation was still viewed as an in-library activity largely unrelated to broader universitywide information technology planning. This perspective preceded the age of ubiquitous networking. The lack of coordination between the MELVYL system and other universitywide computing and networking activities was acute by the mid-1980s. The MELVYL system operated from a computer center separate from other universitywide computing applications. As the first large-scale networked universitywide computing application, a universitywide computer network to support the MELVYL catalog had been developed in the early 1980s. By mid-decade, other multicampus applications were being deployed that also required computer networking support. Between the years 1980 and 1986 there was extensive planning for and subsequent implementation of an information technology infrastructure within UC. These networking efforts were not merged with those of the MELVYL system into an overall universitywide telecommunications system, however. In 1980, the work on the MELVYL system could proceed in isolation, as there were few projects under way on that scale. By 1986, however, a great deal of progress had been made on other fronts, and the lack of connection between MELVYL work and other universitywide efforts had become a major liability. Because the 1977 Plan for Development did not include a comprehensive plan that contemplated library automation requirements at the campus level, and because there was no subsequent funding strategy to address campus library automation needs, there was a natural tendency to assume the MELVYL system was the unarticulated strategy for universitywide library automation. The MELVYL system was not a strategy to meet campus needs, however, and was never intended to serve this role. Not surprisingly, the campuses were extremely unhappy with what they viewed as a failure to support local campus library automation requirements. Their needs had not been addressed due to a failure in the definition of roles and responsibilities, and there was no specific, programmatic funding for campus-based automation, By the mid-1980s, the MELVYL system itself faced terrific problems. It had outgrown the computer center that had been designed to house it, both in terms of physical space and computing resources. The resulting resource management approach of minimizing hardware and software investments was inconsistent with the needs of a production system destined to be the cornerstone for public-access information retrieval services within the university. …