client-server database query

We need to determine the appropriate database product for our needs. We would like to establish a bibliographic database of 1.5 million records, each record containing 7 fields, length of each record around 70 bytes. Using a client-server approach, we would like to be able to search on any of the fields and return results to a Gopher or WWW browser. One of the types of searches we would like to do would be one where, when the user supplied a data value to search the database on a given field, s/he would be presented with the "context" of that string in the set of indexed records on that key, and could then reformulate his/her search with one of the adjacent possibilities. (For example, a search across the database on
"electron" in the "Title" field would return the fact that there were 7
entries that met the criteria--and that there were 14 entries for
"electronic," and 23 more for "electronics"). We would like the product to
be truly client-server -- we would like to search it almost exclusively from other hosts over the Internet -- and not require a login or session to obtain results. Other issues include:

o reliability -- the database >will< be large o ease of maintenance
o accurate updating of data]
o ease with which client and client-handling routines can be written

Would an Oracle/Sybase type product be appropriate? Is there some other product out there which meets these specifications and would address our concerns?

Thanks very much,

Mike Kiley
The UCI Bookstore
University of California, Irvine
714 725 2430
mjkiley_at_uci.edu Received on Thu May 19 1994 - 00:41:04 CEST