AustLit software is a custom built implementation of a design based on the FRBR data model and strongly influenced by the work of the ABC Harmony, INDECS and Topic Map groups.

The system was built using:

• Java 1.3 programming language and runtime
• Tomcat (current production release) open source servlet environment
• Apache (current production release) open source web server
• Xerces-J 1.3 open source XML parser
• Xalan-J 2.1 open source XSL processor
• Oracle 8i database
• OCLC Z39.50 Java library for interrogating holdings information from National Library of Australia's Kinetica service
• Microsoft Internet Explorer 5.5 web client used for data maintenance

The system currently runs on a Sun Microsystems Blade-1000 workstation (750Mhz) with 1GB of memory.

The AustLit database implements the FRBR data model in a very highly normalised form based on the topics and associations of the Topic Map paradigm.

The server software provides the mapping from the database structures to FRBR style objects and from these to a common XML format. Information is output by passing the XML representations of the FRBR object through XSL (eXtensible Stylesheet Language) stylesheets. Different stylesheets are used to generate extracts and encodings such as HTML, EAD fragments and, when required, MARC and RDF (Resource Description Framework).

The Oracle interMedia software is used for free-text searching of titles and names.

The data maintenance facility passes XML objects representing the FRBR view of the data being maintained and instructions for generating the maintainer's user interface to a web browser. Javascript in the web browser using the XML DOM (Document Object Model) maps the data to dynamically constructed HTML forms. The Javascript constructs selection lists and popups, and validates data entry using XML tunnelled over background HTTP connections with the web server. (Screenshot)