Comment 1 Egidijus Vaisnora 2010-10-01 04:05:39 EDT

Created attachment 180019 [details]
Patching 308895 test case to illustrate issue

Comment 2 Eike Stepper 2010-10-01 04:07:32 EDT

Please set the version where you want this be analyzed/fixed...

Comment 3 Egidijus Vaisnora 2010-10-01 04:15:45 EDT

Created attachment 180021 [details]
patch v2

Contains patch to test case + bug fix

Comment 4 Egidijus Vaisnora 2010-10-01 04:17:06 EDT

Ready for review

Comment 5 Eike Stepper 2010-10-01 04:39:16 EDT

Without a version this appears in no review queue here.

Comment 6 Eike Stepper 2010-10-01 05:41:56 EDT

Sorry, with Stefan's refactoring of the type mappings (committed just the other second) your patch does not match anymore ;-(

Comment 7 Egidijus Vaisnora 2010-10-01 06:42:18 EDT

Created attachment 180041 [details]
patch v3

Comment 8 Egidijus Vaisnora 2010-10-01 06:43:14 EDT

Again ready :)

Comment 9 Stefan Winkler 2010-10-01 07:45:20 EDT

Created attachment 180052 [details]
Patch-v4

Alternative solution.

- I copied the TestCase and removed all unneeded stuff (enhancing a Bugzilla-TestCase with a different number might create too much confusion ...)
- Actually, the bug was a problem in the framework. I should have marked setDefaultValue() in AbstractTypeMapping as final. Implementation of getDefaultValue() was a one-liner

TestCase passes now. Please check, if this solves your problem as well.

Comment 10 Stefan Winkler 2010-10-01 10:32:01 EDT

Created attachment 180058 [details]
patch-v5

- Resync with HEAD (conflict with AllConfigs in cdo.test project)
- Patch-v4 broke 308895 again. This patch now passes all tests.

Please review.

Comment 11 Stefan Winkler 2010-10-01 11:16:31 EDT

Committed to HEAD

Comment 12 Eike Stepper 2011-06-23 03:39:50 EDT

Available in R20110608-1407