The Safety Information Center at Triodyne, Inc.. provides library services to the firm's engineers and scientists as well as fee-based research services for external clients. The library's diverse collection supports mechanical engineering, environmental engineering, human factors, vehicle operation and design, product liability, and safety engineering.
We estimate that there are about 160,000 potential machine-readable records in our collection, of which about 3% (the book collection) is catalogued through OCLC. The remaining resources include current and retrospective government reports, standards, patents, product literature, an extensive vertical file collection, vehicle manuals, a special collection on fire, and a collection of deposition transcripts. This proportion illustrates a problem many science-based libraries have: much of the working collection is nonjournal and nonbook material. Yet this is precisely the material that benefits most from fast access through standard bibliographic practices. Of the library staff of sixteen, only four catalog and maintain the files. Therefore, any file management activities have to be simple, fast, and easy to learn and implement.
Begun in 1979, the library became automated in the early '80s with the purchase and installation of digital-based (VAX) WPS-Plus/VMS. This software became the basis for eight separate collections. WPS-Plus/VMS provided the opportunity to create bibliographic records and list process them; subject searches and quick sorts were impossible. When other departments at Triodyne bought computers, they purchased IBM PCs and Macintoshes, thus establishing an environment under which noncompatible systems existed. The lack of standardization became a drain on productivity. After an executive-level evaluation, the company decided to develop a PC local area network (LAN), retain the Macs (essential for our graphic department), and concentrate on changing the library VAX to PCs.
To prepare for the switch to PCs, the library developed a plan that included the following: a description of current applications, a list of current library software, a "wish list" of future library applications, a list of desirable features (for instance, variable length fields, Boolean search features, networkable, import/export, and so forth), and a list of the equipment that would be needed. Our future applications list was a jumble of bibliographic, invoicing, administrative, and word processing requirements. We realized we needed a general-purpose software program that could be used to develop different kinds of databases, allowed searches, was user-friendly and easy to manage. Although the switch from VAX to PC was "mandated" from the top, we were optimistic that this was an opportunity to improve our day-to-day operations and speed up our year-end reporting process.
The library began by evaluating PC software (hardware requirements were handled later by our technical staff). One of our most immediate needs was to track the library's acquisitions activities and we kept that foremost in mind during the evaluation process.
The library was already using PARADOX, a relational database manager, and INMAGIC, a text manager, for special one-time-only applications. Because they were already in place, we considered them first. The librarians most closely involved with both strongly believed that PARADOX was too difficult for the existing staff and INMAGIC was not multifaceted enough for our needs.
After a review of the literature, it became apparent that flat-file database software had progressed to the point where it would meet our three basic criteria, that is, enable searches, be user friendly, and be easy to manage. We identified several candidates, obtained and studied review articles, ordered demo disks, set up a test database based on the acquisition files we wanted to track, and tested each demo. The test database included as many different kinds of acquisitions records as we could think of. …