As per the discussion thread on c.l.p and summarized by
Matthew Shomphe

Logged In: YES 
user_id=21627

Can you give the rationale for this change, for those of us
that don't know which thread you are referring to?

Logged In: YES 
user_id=531881

Sure-

According to the discussion here:
1055808030.15325.python-list%40python.org&rnum=1&prev=/groups%3Fq%3Dos.chmod%2Bgroup:comp.lang.python.*%26hl%3Den%26lr%3D%26ie%3DUTF-8%26group%3Dcomp.lang.python.*%26selm%3Dmailman.1055808030.15325.python-list%2540python.org%26rnum%3D1">http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=mailman.1055808030.15325.python-list%40python.org&rnum=1&prev=/groups%3Fq%3Dos.chmod%2Bgroup:comp.lang.python.*%26hl%3Den%26lr%3D%26ie%3DUTF-8%26group%3Dcomp.lang.python.*%26selm%3Dmailman.1055808030.15325.python-list%2540python.org%26rnum%3D1

there was confusion as the behavior of os.chmod() on
non-Unix platforms.  Discussion led to a recommendation for
a documentation change, which was submitted as an attachment
to the bug report located here:
http://www.python.org/sf/755617

The submitter of the bug report didn't know how to submit a
patch.  I decided to capture the documentation, add it to
the appropriate tex file, produce a patch, and submit it to sf.

Logged In: YES 
user_id=21627

I don't think that patch fixes the bug - IMO, the
description of mode gets worse, not better, with this patch:
1. The mode is *not* a "four-digit number". It is a plain
integer instead. Some people find it convenient to write
this number in base8, using Python's octal number notation.

2. Other people find it convenient to use a symbolic bit
mask to compute the number. Your patch removes the listing
of the possible symbolic constants, which is bad.

3. You fail to mention some of the bits (e.g. the s-bits)
which are available in the current documentation.

4. The patch fails to address the platform differences which
atleast logistix considers important, in his comment to #755617

Logged In: YES 
user_id=21627

blunck2, are you willing to revise the patch in the
indicated direction, i.e. document the real semantics of
os.chmod on Windows, while keeping the material which
primarily applies to Unix?

Logged In: YES 
user_id=21627

I'm rejecting the patch on ground that it does not actually
fix the bug.