Split out from Issue2550805:

Bernhard saw a test fail with the patch from issue2250805 running
the testsuite close to mightnight. And the test was okay five minutes
after mightnight.

this should show the issue
  python run_tests.py .  testFilteringRangeGeekInterval

John also reported seeing something like this.
And Bernhard saw this in 2010 at one testrun with anydbm as well.

Hypothesis: There is a defect that is related to the time when the test
runs or other factors that makes it fail only sometimes.


Details:
Test run with postgresql 9.2 on Opensuse 12.3.
FAIL: testFilteringRangeGeekInterval
(test.test_postgresql.postgresqlDBTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib64/python2.7/unittest/case.py", line 327, in run
    testMethod()
  File "/home/bernhard/hacking/roundup/roundup-
hg/./test/db_test_base.py", line 
1351, in testFilteringRangeGeekInterval
    ae(filt(None, {'deadline': '-2d;'}), ['5', '6'])
  File "/usr/lib64/python2.7/unittest/case.py", line 511, in
assertEqual
    assertion_func(first, second, msg=msg)
[...]
AssertionError: Lists differ: ['6'] != ['5', '6']

First differing element 0:
6
5

Second list contains 1 additional elements.
First extra element 1:
6

- ['6']
+ ['5', '6']


And old failure I have in my personal developer logs:
svn up and then
        Revision: 4482
   python run_tests.py

testFilteringRangeGeekInterval (test.test_anydbm.anydbmDBTest) ...

Failure in test testFilteringRangeGeekInterval
(test.test_anydbm.anydbmDBTest)
Traceback (most recent call last):
  File "/usr/lib/python2.5/unittest.py", line 260, in run
    testMethod()
  File
"/mobilehome/bernhard/hacking/roundup/svn/trunk/./test/db_test_base.py"
, line 1251, in testFilteringRangeGeekInterval
    ae(filt(None, {'deadline': '-2d;'}), ['5', '6'])
  File "/usr/lib/python2.5/unittest.py", line 334, in failUnlessEqual
    (msg or '%r != %r' % (first, second))
AssertionError: ['6'] != ['5', '6']


Debian Lenny ppc, Package: python Version: 2.5.2-3

Similar failed test today (not close to midnight) with Roundup 1.5.0+:

Failure in test testFilteringRangeGeekInterval
(test.test_postgresql.postgresqlDBTest)
Traceback (most recent call last):                 
  File "/usr/lib/python2.7/unittest/case.py", line 332, in run
    testMethod()
  File "/home/intevation/roundup/roundup.hg/./test/db_test_base.py",
line 1350, in testFilteringRangeGeekInterval
    ae(filt(None, {'deadline': '-1d;'}), ['6'])
  File "/usr/lib/python2.7/unittest/case.py", line 516, in assertEqual
    assertion_func(first, second, msg=msg)
  File "/usr/lib/python2.7/unittest/case.py", line 745, in assertListEqual
    self.assertSequenceEqual(list1, list2, msg, seq_type=list)
  File "/usr/lib/python2.7/unittest/case.py", line 727, in
assertSequenceEqual
    self.fail(msg)
  File "/usr/lib/python2.7/unittest/case.py", line 413, in fail
    raise self.failureException(msg)
AssertionError: Lists differ: [] != ['6']

Second list contains 1 additional elements.
First extra element 0:
6

- []
+ ['6']


The same change was already reported for 1.3.0:
http://osdir.com/ml/bug-tracking.roundup.devel/2006-11/msg00038.html

But it is not easily reproducible, it did not show up on 10 further runs of:
python run_tests.py . testFilteringRangeGeekInterval

Bernhard recently reported this in issue2550835 again and Ralf fixed it
in changeset 74b24b14071a

Sorry for the duplicate. I do remember vividly having searched for 
this  issue, but I did not find it.