Recycling database records

"Our users are used to searching and don't care for A-Z lists. We don't want to maintain a separate database of databases. Let's catalog the database record once, recycle it and use the discovery API to build a database search feature. " That was our thoughts when introducing our new web site. When filtering on databases Summon API was called and a relevancy ranked list was presented. But immediately voices were raised from researchers and post-graduates that they had difficulties using the tool. So, we decided to build a more traditional database list yet keeping the main principles: •To search databases from the general library search box. •To maintain in one place only. •To retrieve the records through several search services. To build a tool that facilitates discovery and provides additional features we had to use a source with more stringent metadata. Thus we dropped the Summon API and instead used the API from the original source, the national catalog of Sweden, Libris. A team of librarians and IT-developers developed a database search feature and list that better met the needs of both students, faculty and librarians. We, the librarians, got an understanding about APIs. We also learned by painful experiences that to make MARC records at least a bit machine readable we need to catalog with thorough control. The IT-developers learned about the MARC reality we still live in