setbuf() if a buffer size of 0 or 1 byte is requested.

and the destructor is called early, it doesn't DECREF garbage.

button widget, not involving a __getattr__() method but a callback on
the widget.

Andrew, plase check to make sure these are the ones you want used and get
back to me; this stuff will be "published" in the near future.

is not appropriate for the howto document class.

clean:  Remove all the generatable tarballs.

at Guido's suggestion.

now we're reindented.

Remove files before renaming something to them -- for Windows.

Change check for Python home directory so it works correct on Windows.

Also added simple "clean" target (removes *.o and executable).

Add the script directory to the path.

Fix the sanity checks on the arguments so they don't mess up the -m
option; remove the requirement that the script must have a .py
extension.

Frozen packages are indicated by a negative size (the code string
is the __import__.py file).  A frozen package module has its __path__
set to a string, the package name.

We have a whole new module finder that uses the actual Python
parser and scans the bytecode for IMPORT_NAME and IMPORT_FROM.
This requires some support in import.c (that hasn't been checked in).
New command line options for this: -d, -q, -m.

data file.

will allow us to have a class of smaller documents as well.

work on Windows.

	target.

	along with the others.

html-$(VERSION).tar.gz:  Change the wildcard for picking up GIFs a little
	to avoid changes in what LaTeX2HTML generates.

Add comments about what to do about bugs.

Chunk.__repr__(), main():  Allow Chunk objects to get the buffer, to make the
representation contain text instead of a pair of indexes.  This makes debugging
a little easier.

get some funky quoting of ']' in \item[...] to work right without having to
do really ugly things to the documents themselves.

There are a lot of things relating to indexing that are commented out;
parts need to be examined and dealt with with respect to changes in
LaTeX2HTML internals.  I'll work with Ross to see what it takes to make
this sort of stuff reasonable.

order.  LaTeX2HTML just doesn't do things the same way as LaTeX, and this
wasn't the fix.

Simplify the generated HTML for \file{}.

For \samp{}, use "..." instead of `...'; many fonts make that look pretty
bad.   ;-(

a fix to LaTeX2HTML from Ross Moore.  Yeah!

it no longer appears to be needed.

position in new_buffersize(); the correct function to use is ftell().
Thanks to Ben Jackson.

instead of 'locals' I was passing 'lock_import' to import_module_ex().