I would argue that DERBY-880 and DERBY-881 are not regression test
failures since they're not currently part of the standard derbyall
suite. I have also raised some concerns about what we mean by
"compatibility" -- what interfaces must remain compatible? -- which have
never really been resolved.
David
Mike Matrigali wrote:
I don't see JIRA entries for the encryption failures in the regression
test suite component, maybe they were filed before it existed? There
are 3 issues currently unowned:
DERBY-880 jdbcapi/connectionJdbc20.java fails for server (10.1.2.2)
and client (10.2.0.0) compatibility testing with > SQLSTATE(null):
java.sql.SQLException: Stream of column value in result cannot be
retrieved twice Unassigned Ramandeep Kaur Major Open
Open UNRESOLVED 27/Jan/06 03/Feb/06 Test
DERBY-881 jdbcapi/LOBTest.java fails for server (10.1.2.2) and
client (10.2.0.0) compatibility testing Unassigned Ramandeep
Kaur Major Open Open UNRESOLVED 27/Jan/06
03/Feb/06 Test
DERBY-654 unit/T_RawStoreFactory.unit fails with an assert failure
in J2ME/CDC/FP Unassigned Deepa Remesh Minor Open
Open UNRESOLVED 27/Oct/05 23/Jan/06
I admit I haven't being paying close attention to the tinderbox
reports, but I am impressed by the following reports - for those
who have not taken a look:
http://www.multinet.no/~solberg/public/Apache/TinderBox_Derby/Limited/index.html
I probably could have easily narrowed down some questions I sent out
earlier this week to a specific change from looking at this. Nice job.
It would be REALLY nice if these reports had ZERO errors normally -
as you point out the encryption ones, and a few others are consistently
failing. It
would be nice to get owners for the issues and then get them out of the
suite so that this tool could be used to highlight quickly NEW errors.
David W. Van Couvering wrote:
+1 on taking this one out, for same reasons.
I *am* worried about the encryption test failures, is anyone looking
at these?
David
Mike Matrigali wrote:
I have seen this fail 5 or 6 times in the last 2 weeks. I haven't
seen any new requests for more information, from the discussion on
the jira item it seems like the issue is understood (though not the
solution).
My quesion is since no new info seems to be gained by having it
fail in other environments, and since there is an assigned person
working on the problem, is it time to remove this test from nightly
runs (or maybe from specific environments)?
It seems like a number of items are lingering on the regression
test suite list. I believe from the discussion a few weeks ago
the goal was to work toward shrinking this list to 0, unless there
was a good reason to keep a known failing test running in the
nightly suite. Especially ones where there is someone assigned, it
seems reasonable to take the test out unless the assigned person
wants it run.
/mikem
begin:vcard
fn:David W Van Couvering
n:Van Couvering;David W
org:Sun Microsystems, Inc.;Database Technology Group
email;internet:[EMAIL PROTECTED]
title:Senior Staff Software Engineer
tel;work:510-550-6819
tel;cell:510-684-7281
x-mozilla-html:TRUE
version:2.1
end:vcard
