out by Jean Brouwers.

behaviours. I left the original test commented out (note
that that test came from #2702, which seems to have a
problem in FreeBSD and Windows, but not in Linux).

I included a new test, to watch over the now-broken
behaviour, I took it from #3179.

signs, and float('nan'), float('+nan') and float('-nan') actually all
generate the exact same NaN value.

Modified interface to _Py_[String|Unicode]InsertThousandsGrouping, in anticipation of fixing issue 3140.

in the issue 3165. Now cPickle does not fails with uncontrolled
behaviour when pickling into a very deep nested structure.

it makes the test fail in the trunk.loewis-sun buildbot.

directory creation time...

shutil.rmtree() in the trunk.loewis-sun buildbot.

without necessity.

not fixed length, it mallocs memory if needed. As a result, we
don't have a maximum for the getcwd() method.

specially a section of the subprocess documentation
that helps with the replacing of those functionss.

will contain both FieldStorage and MiniFieldStorage items.

Fixes #1817.

structures follows the common sense (and works like in 2.5 and 3.0).