The test for datetime module keeps failing. 

test test_datetime failed -- Traceback (most recent call last):
  File "/home/shantanu/cpython/Lib/test/datetimetester.py", line 215, in test_issue23600
    self.assertEqual(t - t.utcoffset(), u)
AssertionError: datet[22 chars]26, 11, 1, tzinfo=<test.datetimetester.TestTZI[63 chars]538>) != datet[22 chars]26, 12, 1, tzinfo=<test.datetimetester.TestTZI[63 chars]538>)

In datetimetester.py , in function test_issue23600 , line 196:
the variable DSTDIFF has been assigned timedelta of 1 hour, I guess it should be , 

DSTDIFF = timedelta(hours=0)

It passes the test then. 

The comment in the function code says -  
"""Simple time zone which pretends to always be in summer time, since that's what shows the failure.""" Was the test written to fail on purpose ?  

Thanks

No, that delta being one hour is the whole point of that test.

I've nosied Alexander.  He'll probably need more information about the environment in which you are seeing this.

@shanmbic - did you recompile C modules before running the tests?

@belopolsky - Yeah, I recompiled all the C modules successfully, still it fails the test. I am running Ubuntu 14.10

This is just hard to believe.  The symptom you describe is exactly what's expected if you got the new test suite but did not compile the new C code, both added by the fix for:

    http://bugs.python.org/issue23600

Since we have numerous buildbots on which the test passes, what's special about your setup?  The changes here were quite straightforward.  Please double-check.

Maybe @shanmbic has a clock skew and _datetime.so does not get recompiled?

I would try to do make clean or even make distclean and rebuild everything.

Yes, I find that if there's something that might be a compilation problem, doing a make distclean and a build from scratch is a good idea to confirm it.

make distclean worked. I guess too many reverts in hg caused some issue. :)