Fixed a half dozen ways in which general dict comparison could crash
Python (even cause Win98SE to reboot) in the presence of kay and/or
value comparison routines that mutate the dict during dict comparison.
Bugfix candidate.

*not* define O_RDWR; get that from the os module.

meaning infinity -- but at least warn about it in the code!  I pissed
away a couple hours on this today, and don't wish the same on the next
in line.
Bugfix candidate.

means "replace everything".  But the string module, string.replace()
amd test_string.py believe a 0 count means "replace nothing".
"Nothing" wins, strop loses.
Bugfix candidate.

out of synch than I realized, and I managed to break replace's "count"
argument when it was 0.  All is well again.  Maybe.
Bugfix candidate.

mymemXXX stuff, and they were already out of synch.  Fix the remaining
bugs in both and get them back in synch.
Bugfix release candidate.

Platform blew up on "123".replace("123", "").  Michael Hudson pinned the
blame on platform malloc(0) returning NULL.
This is a candidate for all bugfix releases.

needed now that fcntl exports the constants.

warning.  This is similar to the TERMIOS backward compatbility module.

and using the constants defined there instead of FCNTL.

This closes SF bug #231328.

Added all constants needed to use the functions defined in this module
that are not defined elsewhere (the O_* symbols are available in the
os module).  No additonal modules are needed to use this now.

    PyObject_AsFileDescriptor() -- it does the same thing everywhere, so
    use it the same way everyone else does so that exceptions are
    consistent.  This means we have less code here, and we do not need to
    resort to hackish ways of getting the Python-visible function name to
    fdconv().

Thanks to Mark Favas.

Updates zipfile.ZipFile docs to mention the fact that you can create a
ZipFile instance from an arbitrary file-like object.

This closes patch #418011.

this could cause invalid paths to be returned for AF_UNIX sockets on some
platforms (including FreeBSD 4.2-RELEASE), appearantly because there is
no assurance that the address will be nul-terminated when filled in by the
kernel.

PySocketSock_recvfrom():  Use PyString_AS_STRING() to get the data pointer
    of a string we create ourselves; there is no need for the extra type
    check from PyString_AsString().

This closes SF bug #416573.

SF patch 416251 2.1c1 mmapmodule: unused vrbl cleanup

This closes SF bug #416530.

This header does not exist on all Unix flavors; FreeBSD in particular does
not include it.

This closes SF bug #422320.

negation.

This closes SF bug #421999.

bug tracker.

leaking LaTeX2HTML's internal string munging.

This fixes SF bug #420399.

cannot be determined.

Pseudo-fix for SF bug #420724

restore correct semantics.

    the right HTML file to the name about.html is needed even if the
    --numeric option was not given -- some other name may have been
    assigned due to some non-determinism in the algorithm use to perform
    name allocation.  ;-(

This closes the "About..." portion of SF bug #420216.

There is no imap module; refer to imaplib instead, since it exists.
Move the "See Also:" section in front of the sub-sections, for
consistency with other portions of the library reference.

This closes the library reference portion of SF bug #420216.

Closes SF bug #420230.

Always pass a full path name to LoadLibraryEx().  Fixes some Windows 9x problems.  As discussed on python-dev

interned string created by "string"[i].  Since they're immortal anyway,
this was hard to notice, but it was still wrong <wink>.

interned when created, so the cached versions generally aren't ever
interned.  With the patch, the
		Py_INCREF(t);
		*p = t;
		Py_DECREF(s);
		return;
indirection block in PyString_InternInPlace() is never executed during a
full run of the test suite, but was executed very many times before.  So
I'm trading more work when creating one-character strings for doing less
work later.  Note that the "more work" here can happen at most 256 times
per program run, so it's trivial.  The same reasoning accounts for the
patch's simplification of string_item (the new version can call
PyString_FromStringAndSize() no more than 256 times per run, so there's
no point to inlining that stuff -- if we were serious about saving time
here, we'd pre-initialize the characters vector so that no runtime testing
at all was needed!).