Ruslan, This paragraph was key... On Mon, Nov 24, 2008 at 6:29 PM, ruslan <[EMAIL PROTECTED]> wrote:
> > All two issues go away if I specify PCEnhancer via -javaagent option. Which > makes me think that "Java 6 class retransformation" described in the > documentation doesn't work as intended (or I don't understand something) in > terms of control of "dirtiness" of objects. > > The "fallback" mechanism for doing enhancement via subclassing is not ready for prime time. Reference JIRA for several issues related to this processing [1] and it's been discussed on the dev forum several times. Unfortunately, since this feature already went out the door as enabled by default, it's difficult to turn it off by default. It was meant as an aid to demonstrate an easy "out of box" experience. It was never meant for production use. I would suggest turning this option off and using either static enhancement via the PCEnhancer tool or using the -javaagent option like you suggested. If both of these problems are resolved by using one of this preferred enhancement processes, then I would suggest that we open another JIRA with these additional short comings. Document for the RuntimeUnenhancedClasses property can be found here [2]. FYI, WebSphere Application Server runs with this option set to "warn". This way, a warning message gets issued and we still don't fallback to this subclassing mechanism. Let me know your results. Thanks, Kevin [1] http://issues.apache.org/jira/browse/OPENJPA-651, http://issues.apache.org/jira/browse/OPENJPA-650, https://issues.apache.org/jira/browse/OPENJPA-293 [2] http://openjpa.apache.org/builds/latest/docs/manual/manual.html#openjpa.RuntimeUnenhancedClasses
