Thanks, Alan. My problem was slightly similar.
I found that the generated classes and methods for the returnValueEvaluators
(and other generated classes/methods) were being overwritten for every
process in the same package. Thus, the last compiled process in my
changeset for a given package worked fine, and the preceding processes were
exhibiting the strange behavior. The reason it "worked" on my local machine
was due to the fact that I was only debugging with one process in my
changeset.
A workaround for this is to have every process in a different package. But
this makes me believe there is something still very wrong, here. I have yet
to determine if this is an issue in the trunk or 5.1, or if it's simply
something in my local configurations and/or loaded classes. I may drill
further to determine that.
I would highly recommend these debugging tools and parameters to anyone
trying to debug a server side issue with these generated classes:
drools.dump.dir
http://blog.athico.com/2008/02/looking-under-drools-skirt.html
Eclipse Remote Debugging
http://www.eclipsezone.com/eclipse/forums/t53459.html
Cheers.
--
View this message in context:
http://drools-java-rules-engine.46999.n3.nabble.com/Drools-Flow-inconsist...
Sent from the Drools - User mailing list archive at
Nabble.com.