Oystein Grovlen - Sun Norway wrote:
Due to its high failure rate, ConcurrentImplicitCreateSchema was
removed from derbyall in February. The check-in of this patch has
added it back in the encodingTests suite. Why? It is not part of the
patch uploaded on JIRA, and I cannot see any relations between this
test and the topic of this patch nor to encodingTests.
This has resulted in failures in the nightly tests and the tinderbox
tests.
Thanks for pointing this out, Oystein.
The derby900 fix , also fixes the issue described in derby583. Both are
encoding related issues. The ConcurrentImplicitCreateSchema test was
used as the repro in derby583 and so I added this test to
encodingTests. But I dont need this entire test to reproduce the
problem, just needed an error message.
This test (as part of encodingTests suite) passed in my environment
using the 1.5 vm and I didnt realize that this test was disabled on
trunk (10.2 codeline). Sorry for the trouble. encodingTests suite is
only in trunk and hence this test was added to that suite only, as part
of patch (derby900.p1.diff.txt) for trunk, which is on jira. I assume
you looked at the 10.1 patch attached to the jira and hence didnt find
it there.
I have attached a patch (derby900.test.diff.txt) to derby900 to remove
ConcurrentImplicitCreateSchema test from encodingTests suite and instead
have a new simple test (TestEnc.java) for encoding issues to avoid any
other confusion.
Thanks,
Sunitha.
