We are proud to announce the 1.0.4 release of OpenLexicon.

Our development team has been hard at work, responding to many requests. The result is a much richer and productive rules environment.  

Here are the release notes: 

• We have simplified the use of database tables! We added or improved process action support for Delete and Update functionality for both Oracle and MySQL databases - key columns are introspected from appropriate data dictionary elements.
  
• A simple way to export and control your projects! With our, 'Cache-to-Database' functionality - allows the user to load a cache file (serialized or XML) into a fresh metadata schema (directions for emptying a metadata schema for a cache load are also included)
• We have added manymetadata links added in Wizard so the user can review parameters or child objects contained by a parent business object, etc.
  
• Numerous cosmetic improvements to the Wizard - CSS cleanup, button appearance, and others
• We removed unnecessary tabs in the Wizard pertaining to Application object functionality
  
• For our java developers, we added a new class: com.hotapp.lexicon.test.RulesEngineTest, which guides new Lexicon users through using the Lexicon Business Rules Engine, helps diagnose initial problems, etc.
• If you use database persistenance, we have added a UI in Lexicon Wizard for generating and downloading MySQL table scripts, Oracle table scripts, and XSD scripts of Lexicon business objects
• We added new broker: LexiconFileProcessor, which reads data from a file and immediately executes a process list against it (LexiconProcessBroker uses a 2-step file load/process model) 
• We have expanded the testing ability of the business object. The business object testing screen can run arbitrary rules lists in the Wizard
• We have improved the doQuery method in JDBCModule to better process large result sets
• We now support TimesTen (Oracle's memory-resident database product) support announced. Please contact us!
• We added another new broker ProcessTableBroker, which reads all data from a table and processes it according to a process list; provides simpler inputs, and is more friendly to the BCP 
• Bug fix for misbehaving button on the process tab (Wizard) 
• We trimmed the code tree to remove unnecessary code elements and JAR files not needed for the deployment (resulting in a lighter WAR deployment)
• Web Services support, API, and methodology added - see documentation for details 
• Added path to LexiconUtils.castDatum for casting a double to a float value
• Improved LexiconUtils.castDatum for casting various types of data values to a Boolean (recognizes more variants of True/False)
• Improvements to LexiconMetaMethod (Lexicon Expression rule support) to better recognize JRE version and compilation JRE version target (e.g. a 1.5 JRE generating 1.4 compatible code).
• New LexiconUtils method called getWellTypedLexiconBean that performs a castDatum on all non-null data elements to the appropriate type based on the metadata definition of each attribute.
• Fixed a bug in LexiconFileProcessor broker where file-read data was not being cast to the appropriate type.
• Improved metadata object indexing in the Wizard - only display index elements for which there are existing metadata objects (e.g. do not display index "C" if there are no objects that start with "C")
• Improved behavior of the cache download link from the Wizard
• We have aAdded links to transformation rules for boolean expressions on the transformation business rules screen
• We now allow the user to choose between downloading a serialized cache file or an XML cache file