Did you use the correct hibernate dialect? For Oracle this is :
hibernate.dialect=org.hibernate.dialect.Oracle9Dialect //Oracle 9 hibernate.dialect=org.hibernate.dialect.OracleDialect //Oracle 8
There is also something with a mapping of 0=false, 1=true or something. I think it is documented in the wiki and in the SF forums. Don't know, but hope this has already made it into the oracle hibernate config files
Yes, I have the correct dialect. I've determined that I can solve the problem by including the following in hibernate's configuration (jbpm.hibernate.config.xml in my case)
hibernate.query.substitutions=true '1', false '0'
I suppose the above config line is the correct solution and just hasn't made it into the public jbpm-db build. I just wanted to be sure I hadn't made a common build mistake or something.
The correct hibernate mapping is actually
hibernate.query.substitutions=true 1, false 0
since the boolean field is using int (not char) as far as I recall.
I commited the entry in hibernate.properties in CVS.
Oh .. and btw. it would be great if you could run the whole test suite with Oracle using the jbpm.db project and let us know where things go wrong.