Python 2.3.x candidate.

was a mistake of mine in updating patch #975056 (I think).

this test failed under the combination of passing -O to a debug-build
Python.  Now all 4 of those pass ({debug, release} x {-O, no -O}).

rangeobject.c has an #ifdef that is reached only when LONG_MAX != INT_MAX

Easy doc corrections from Dan Gass.
Also repaired the LaTeX for optional keyword arguments throughout, which
was (incorrectly) nested instead of flat.

73: Restore support for Python 2.1.
74: Factor the previous change differently
79: Conditionalize usage of PyTuple_Pack.

want to use pywin32 instead of _subprocess, you have to edit the file.

Python's own setup.py that did the same thing (and tested on Solaris,
where LDSHARED is needed...)

detection to configure &c.

a few remaining calls to signal() were converted to PyOS_setsig().

non-ascii characters.

This support was only a first cut, and doesn't deserve to be in
a released version (where we have to support it in an ongoing
manner)

it's rather expensive to create new processes.

"Khalid A. B." on python-dev)

(should probably remove all traces of _xmlrpclib, but I'll leave that
for another day.)

debug build on Windows now.  More applications of the helper may be needed
on non-Windows platforms.

Turns out the mysterious "expected output" file contained exactly N dots,
because test_poll() has a loop that *usually* went around N times,
printing one dot on each loop trip.  But there's no guarantee of that,
because the exact value of N depended on the vagaries of scheduling
time.sleep()s across two different processes.  So stopped printing dots,
and got rid of the expected output file.  Add a loop counter instead,
and verify that the loop goes around at least a couple of times.  Also
cut the minimum time needed for this test from 4 seconds to 1.

of the test.  It's testing stdout in a different process, so it has to
print something, but I didn't find "banana" to be self-explanatory.

tester that a DOS box is expected to flash.  Slash the sleep from 2
seconds to a quarter second (why would we want to wait 2 seconds just
to stare at a DOS box?).

what this is trying to do.  If it's necessary for it to create > 1000
processes, it should be controlled by a new resource and not run by
default on Windows.

display a test's docstring as "the name" of the test.  So changed most
test docstrings to comments, and removed the clearly useless ones.  Now
unittest reports the actual names of the test methods.