Issue JBPM-462

[ejimenez]

Hey guys,

Respectfully, I think issue 462 (hence 468) can be fixed in jBPM 3.0.x. jBPM 3.1.x seems to be a pretty major upgrade, not everyone might be ready to upgrade immediately, including us as would have completed almost all work on our project when 3.1 is released for production use. Also we probably won't use many of its features as we've implemented them in our own ways and 3.1 would not replace those implementations.

The aforementioned bugs are trivial to fix in 3.0.2 (for release in 3.0.3) without any need of schema changes. In fact, I'm not sure schema changes solve the problem at all, as a schema change would mean that tokens would have a subprocess LIST. This is not exactly the bug. The bug is that there is no way for the subprocess-end event to know which subprocess ended.

The schema change fix would solve the problem of SAVING subprocess instances, but not the subprocess-end event issue.

The no-schema-change fix can be found in the issue comments:

http://jira.jboss.com/jira/browse/JBPM-462?page=all

Please comment on this. I would really like to use an unchanged jBPM in our environment to make minor version upgrades easy.

If you guys disagree, that's fine..we'll just have to maintain a list of things we have fixed to backport them to 3.0.3 and so on until we go to 3.1 (again, if that even happens)

[ejimenez]

Sorry, that's issue 438 instead of 468

[kukeltje]

Hey, it is opensource. If you can provide us with a fix and a testcase for it, we'd be happy to include it.

[kukeltje]

hmmm... sorry, the fix seems to be in there. Still, it would be nice if you could provide us with a testcase.

The issues is already reopened by Tom....

[ejimenez]

I would really love to, at this point I'm swamped with other work. I'll try to get something in next week. The fix has been tested in our performance testing environment so I'm pretty sure of it.

Do I attach the unit test to the issue description?