People are already using it, so these docs are certainly better than
no docs at all. Markup is mostly missing and the layout is probably
ugly, but this can be fixed later.

Question: there are references to MS docs for the CSIDL_... constants
and the IShellLink interface. Are these pointers sufficient, or should
the MS docs reworded and repeated here?

test_resource calls resource.setrlimit() to change the file size limits.
This fails on Cygwin, which supports setrlimit() and getrlimit(), just not
changing that particular setting. (The same would apply to any other
platform that has those functions but not that particular feature.)

Since getrlimit() works and setrlimit() can be used for other reasons, a
check for ValueError was added to that part of the test.

Python 2.3 will support source code encodings which rely on the
builtin codecs being available to the parser.

Remove struct dependency from codecs.py

chars.  See the comment for rationale.

supported as the second argument. This has the same meaning as
for isinstance(), i.e. issubclass(X, (A, B)) is equivalent
to issubclass(X, A) or issubclass(X, B). Compared to isinstance(),
this patch does not search the tuple recursively for classes, i.e.
any entry in the tuple that is not a class, will result in a
TypeError.

This closes SF patch #649608.

This patch allows ZipFile.writestr() to be called with
an archive file name instead of a ZipInfo instance:

z = ZipFile("myarchive.zip", "w")
z.writestr("foo/baz/file.ext", data)
z.close()

I found the old writestr() method very inconvenient
for simple (but common) things.

If called with a file name instead of a ZipInfo
instance, the date_time is set to the current date/time,
which makes sense to me for anonymous data.

TARGET_API_MAC_OS8 (or !TARGET_API_MAC_CARBON) is gone. Also some
TARGET_API_MAC_OSX conditional code is gone, because it is no longer
used on OSX-only Python (only in MacPython-OS9).

Replace use of homebrew boolean with True/False.

Reflow lots more long lines.

The bug is a reference to co_first_lineno that should be
co_firstlineno.  The only other substantial change is to speed up
localtrace_count() by avoiding *costly* calls to inspect module.
It's trivial to get the filename and lineno directly from the frame.

Otherwise, delete commented out debug code and reflow very long lines.

build_ssl.py requires os.popen().

  * add oss_mixer_t and OSSMixerType
  * add newossmixerobject(), oss_mixer_dealloc(), ossopenmixer()
  * add _do_ioctl_1_internal() to support mixer ioctls
  * add mixer methods: oss_mixer_{close,fileno,channels,stereo_channels,
    rec_channels,getvol,setvol,getrecsrc,setrecsrc}()
  * add oss_mixer_methods list
  * add oss_mixer_getattr() (why?!)
  * export SOUND_MIXER_* constants from soundcard.h

_do_ioctl_1() so they take a file descriptor rather than an oss_t
pointer.

  * rename oss_t to lad_t, Ladtype to OSSType,
  * rename lad_*() methods to oss_*()
  * rename lad_methods list to oss_methods
Patch and impetus supplied by Nicholas FitzRoy-Dale <wzdd@lardcave.net>.

efficient.  Suggested by MAL.

configuration selection. Rework the usage message to match.

Also some minor code polishing.

Barry agreed, and I see no reason not to.

the obvious s/linuxaudiodev/ossaudiodev/ change made.

Most of these patches are from Thomas Heller, with long lines folded
by Tim.  The change to test_descr.py is from Guido.  See the bug report.

Not a bugfix candidate -- METH_CLASS is new in 2.3.

input, and 2) unicode input means unicode output.  This closes
SF bug #622831.

Change LongWordTestCase.setUp() -- remove leading whitespace from
  text string.
Comment fix.