a more reasonable value.

Backport candidate.

more instances of the bizarre "del f; del m" ways to spell .close() (del
won't do any good here under Jython, etc).

with a size larger than the underlying file worked on Windows.  It
does <wink>.  However, merely creating an mmap that way has the side
effect of growing the file on disk to match the specified size.  A
*later* test assumed that the file on disk was still exactly as it was
before the new "size too big" test was added, but that's no longer true.
So added a hack at the end of the "size too big" test to truncate the
disk file back to its original size on Windows.

boundary could be found -- in a lax parser -- the entire body is
assigned to the message payload.

_handle_text(): Use _isstring() for stringiness test.

_handle_multipart(): Add a test before the ListType test, checking for
stringiness of the payload.  String payloads for multitypes means a
message with broken MIME chrome was parsed by a lax parser.  Instead
of raising a BoundaryError in those cases, the entire body is assigned
to the message payload (but since the content type is still
multipart/*, the Generator needs to be updated too).

UnicodeType, which is different between Python 2.1 and 2.2.

broken wrapping of long ASCII headers.

losing implicit double->long cast.

Bugfix candidate.

    to the command-line arguments.  Fix this by adding the source filename.

(Bugfix candidate for 2.2, and likely 2.1 as well)

just version 1.5.  Closes SF bug 577793.

the same.   Added check to verify the two names are not the same.  Does not
check the actual files to see if there is a symbolic link.

Closes SF bug 490165 and Tzot's patch 604600.

Will look at doing something similar in import.c so that the effort won't
have to be repeated elsewhere.

Closes SF patch 600152.

Add read(), write(), and readwrite() helper functions to shorten poll
functions.  Use get() instead of try/except KeyError for lookup.
XXX How could the lookup ever fail?

Remove module-level DEBUG flag.

Use iteritems() instead of items() when walking the socket map.

Reformat the functions I touched so that are consistently Pythonic.

after the last checkin.

entries on OS9 (where they are never available, and simply disabling
them might lead people to think otherwise).

Upped the stack size to 256KB. test_class ran afoul of the 64K limit, and this is probably a better fix than lowering the recursion limit.

GUSI/Threading interaction, I'm not sure, but I don't have the time to fix this right now.