Python version: 2.5
OS name and version: Windows XP SP2
Email: dfugate@microsoft.com

The datetime.datetime subtraction operator seems to
have a bug in it:

Python 2.5 (r25:51908, Sep 19 2006, 09:52:17) [MSC
v.1310 32 bit (Intel)] on win
32
Type "help", "copyright", "credits" or "license" for
more information.
>>> import datetime
>>> x = datetime.datetime(2006, 9, 29, 15, 37, 28, 686001)
>>> y = datetime.datetime(2006, 9, 29, 15, 37, 28, 686000)
>>> x - y
datetime.timedelta(0, 0, 1)
>>> y - x
datetime.timedelta(-1, 86399, 999999)

Here, the result of y - x should be timedelta(0, 0, -1).

Logged In: YES 
user_id=31435

Sorry, nope.  As the timedelta docs say:

"""
...

days, seconds and microseconds are then normalized so that
the representation is unique, with 

0 <= microseconds < 1000000 
0 <= seconds < 3600*24 (the number of seconds in one day) 
-999999999 <= days <= 999999999 

...

If the normalized value of days lies outside the indicated
range, OverflowError is raised. 

Note that normalization of negative values may be surprising
at first. For example, 

>>> d = timedelta(microseconds=-1)
>>> (d.days, d.seconds, d.microseconds)
(-1, 86399, 999999)
"""

Which is exactly what you're seeing.