environments.  Move the conversion to esis2sgml.py.

Reported by Corran Webster <cwebster@math.tamu.edu>.

Updated comments on the documentation.

to make it easier for people to find what they're looking for.

Add a few more comments about mkhowto.

Based on comments from Albert <hat@se-46.wpa.wtb.tue.nl>.

documents probably aren't written by Guido.  ;-)

Michael Hudson's documentation for bytecodehacks exhibits incorrect
usage.

Added some comments about how to set up the mkhowto script for use
with documents not contained in the documentation source tree.

		  This reflects the changes made in
		  ../texinputs/python.sty.

		 entry based on the module type if the type is known.
		 Otherwise, spit out a warning the the module type
		 parameter was bad and generate a "plain" index entry.

		 Without this, the bad module type was emitted as
		 text.

Problem noticed by processing Michael Hudson's docs for
bytecodehacks.

since they shouldn't be here.  Only ignore the output tarballs in this
directory.

	      <piers@cs.su.oz.au>.

	      <piers@cs.su.oz.au>.

			  called on destruction is non-NULL.

When refering to chapters, use \ref, don't hardcode the chapter
number.

		     l2hinit.perl version of the function (which
		     overrode a version in latex2html), but
		     do_cmd_index() was overridden.  Since
		     do_cmd_index() was the only caller, we can trash
		     this.

define_indexing_macro():  New function used during setup to support
			  more dynamic dispatch for indexing.

process_index_macros():  New function; implement dynamic dispatching
			 of indexing support functions, allowing many
			 targets used in indexing to be shared by
			 multiple index entries.  I.e., something like
			 \index{foo}\bifuncindex{bar} can now share a
			 single target anchor.

These indexing changes reduce the amount of clutter in the generated
HTML files.  In the reference manual, this cuts the number of targets
by 44%.  Targets are less affected when generated by environments
rather than macros.

		     was never called in this incarnation.  The
		     other version is simpler as well.

page of the section.

All suggested by Eric Raymond.

in a fragment rather than the main document object.

Add controls for some of the esoteric indexing macros.

In the __coerce__ description, None should be returned if the
conversion is *impossible*, not *possible*.  Reported by Robert Kern
<kernr@ncifcrf.gov>.

	want to push nodes into the document, but into a document
	fragment, since we're not necesarily working with something
	that's well-formed.

least there's a token object in here now!  ;-)

(M.-A. Lemburg and GvR).

This still doesn't understand anything about multiple source files or
checking time dependencies.

updates by Tim Peters, markup by FLD.

1. Fixes float divmod so that the quotient it returns is always an integral
value.

2. Fixes float % and float divmod so that the remainder always gets the
right sign (the current code uses a "are the signs different?" test that
doesn't work half the time <wink> when the product of the divisor and the
remainder underflows to 0).

(The difference between grid_remove() and grid_forget() is that the
former remembers the options for the slave.)

(I tweaked it slightly so examples are allowed to have no date too.)