Library Projects: 1998: Building A Low Cost Library Catalog Searchable Via The World Wide Web
Description: The FMHI research library holds numerous special collections of books and articles which are not searchable via other avenues. Efforts to create databases in other popular programs (e.g., Oracle, DB2, etc) were stymied by the relatively high cost of implementation (approximately $6,000 of a single site licensed version of the Oracle program).
We decided on the following parameters:
The database contains over 5,000 records which are searchable across any of 8 different fields combined in groups of three. The system uses boolean logic and supports AND, OR, NOR, and NAND functions. The specific program used to generate the database was Pro-Cite, although any database manager that can export records in tab delimited ASCII will work. The record itself is based on the MARC standard (books format) used in many online bibliographic catalogues.
Colleges maintaining specialised collections of their own or customised databases might want to consider this approach for disseminating information over the World Wide Web.
The site for accessing this database is:
For more information, contact:
Platform: Sun Microsystems Sparc10/Model 30 running SunOS 4.1.3
After evaluating three different databases, Postgres95 was chosen as the database because it was relational, capable of handling large datasets, and had provisions for handling SQL queries. It also accepted data from tab delimited ASCII sources. Jolly Chen, one of the developers of Postgres95 provided specialized Perl source code capable of interfacing the database to HTML forms. With this final link in place, it became possible to use the world- wide- web as both a data gathering and data dissemination tool.
Ardis Hanson, Director, Institute Research Library
William D. Kearns, Ph.D., Director, Institute Computer Support