Release Notes ============= Jena2-pre3 (May 2003) Third preview release of Jena2. It is a preview release, not intended for serious development. Developers are encouraged to try it out and send feedback to jena-dev@yahoogroups.com. This release does adds the first release of the relational database-backed persistence for Jena2. Jena1 database models are not compatible with Jena2 database models. See /doc/DB/ for further details. Database: See the directory /doc/DB/ for preliminary documentation. + Bulk update interface supported (performance impact not yet evaluated). + Reification storage optimized to avoid quad bloat. + Table layout changed to use a "denormalized" schema for faster response time (but uses more space). + Configuration options specified as RDF statements; config files (e.g., Mysql.config) no longer used. + MySQL supported. Oracle, Postgresql, BDB coming soon. Known Issues: - Database consistency for reification triples: Jena applications that add and delete reification triples explicitly should do this within a transaction. If a transaction is not used and a failure occurs during the Jena update, the database may be left in an inconsistent state. ModelFactory + The ModelFactory has been improved to make it easier to build ontology models with different reasonless. It is also now possible to use the ontology model API with OWL or DAML triples stored in the new persistent storage models. Reification + reification Model API unchanged since preview 2 + base Models can be created with any of three reification styles + the default style is Reifier.Convenient, which hides reification triples + all three styles support the ReifiedStatement API family + reification over inference models may have surprising behaviour OWL Syntax Checker: pre-alpha + The OWL Syntax Checker is a 95% complete implementation as defined in the OWL Tests WD. There is currently very little documentation, and there is a 3-5 second start-up delay. + To run the syntax checker, try: java -cp *jena-jars* com.hp.hpl.jena.ontology.tidy.Checker URL [Level] where *jena-jars* is all (some of) the jars in the lib directory in the distribution and URL is the data to be checked, and [Level] can be used to get an explanation as to why [Level] is incorrect. Level is Lite DL or Full If [Level] is correct then this option has no effect. Tools + The schemagen tool (jena.scehamgen) has been extensively updated. It now RDFS, OWL and DAML+OIL schemas into Java vocabulary classes. New capabilities are documented in a howto in doc/tools/. The old schemagen has been renamed schemagen_orig. Vocabularies: + Dublin Core vocabulary updated to current version + Added vocabulary classes for DC terms and DC types + Class DC is the current (1.1) Dublin Core - Old DC available via class DC_10 Misc + One JAR file has changed name: it is now called concurrent.jar