incomplete; I don't remember what else I thought I was going to put
in, but it looks o.k. to me know.

for the manual class go one deeper than it previously did (\subsection
entries will now be numbered); no change for howto documents.

This makes it easier to refer precisely to smaller sections using \ref.

to make it clear that these are instance attributes, not module data.

Make reference to the copy_reg module a hyperlink.

		    cannot be byteswapped.

exceptions.

When raising NotANumber, pass the string that failed as the exception
value.

versions.  XXX comments only need to be grep-able.

implemented in Python.

similar constructs elsewhere (al/AL, gl/GL/DEVICE).

Add an index entry.

to 4.

additions from Moshe's version.  Used my table for describing the
chunk format, and added some markup and index entries.

Break some other cycles too (and destroy the root when done).

user what they want first if there's unsaved stuff, and may cancel).
It closes more than before.

Add unload_extensions() method to unload all extensions; called from
_close().  It calls an extension's close() method if it has one.

Unregister everything at closing.
Don't call close() in __del__, rely on explicit call to close().

Make argument names equal to what is used in the documentation of the
file object, since chunks are supposedly file-like.

I just noticed that the changes below also apply to cmpcache.py, which
is virtually identical to cmp.py.

1. Fix incorrect file open mode on Win32 platforms (use "rb" instead
of "r").

2. Add shallow parameter to cmp.cmp().  If false, deep file
comparisons are made.

The module should be 100 percent backwards compatible.

avoid being fooled into thinking that fcntl and FCNTL are the same
thing -- they aren't!  (fcntl is the extension, FCNTL.py is h2py
output that defines all the constants).

(XXX The example is still weird -- I think there's a more portable way
to do locking now.  That's for someone else to fix...)

Urllib makes the URL of the opened file available through the geturl
method of the returned object.  For local files, this consists of
file: plus the name of the file.  This results in an invalid URL if
the file name was relative.  This patch fixes this so that the
returned URL is just a relative URL in that case.  When the file name
is absolute, the URL returned is of the form file:///absolute/path.

[I guess that a URL of the form "file:foo.html" is illegal...  GvR]

be using self.maxdict.