This is Andy Hefner's McCLIM formula display code, and an initial effort
by Raymond Toy at getting a Maxima REPL display in an McCLIM window. 

Licenses:  Public Domain for Andy Hefner's code (formula.lisp) 
           Raymond Toy's code is GPL
	   
Committer's note concerning formula.lisp:  This is mainly being 
uploaded to make sure I can find a copy sometime in the future.  It doesn't
have any ties to Maxima right now.

Below is preserved part of an IRC discussion on formula.lisp, mainly
so I (CY) remember what some of the issues were.

<starseeker> what license, by the way?
<hefner> dunno, that's a standard disclaimer for me. I guess if you
don't expect it to ever reach quite 'TeX' quality output it's pretty
sound. Right now I haven't worked out how/when it should emit
parenthesis according to operator precedence, that's the major issue.
<hefner> license is whatever, public domain I guess
<starseeker> Cool :-)
<starseeker> I don't think tex quality is called for, at least not
right away - Mathematica doesn't achieve it
<starseeker> Neither does word, for that matter ;-)
<hefner> true enough
<starseeker> Can it handle stuff like integral signs and summations?
* starseeker is busily emailing himself the links.
<hefner> not yet. I need to work out an interface to symbol fonts
across the two mcclim backends, then I'll look at that.
<hefner> but there's no reason in theory why it couldn't
<starseeker> I've wondered about that - should those be done with fonts
or drawing?
<hefner> I'm thinking fonts
<starseeker> that's the harder way, right?
<starseeker> granted it'll look better
<hefner> I think that's the easier way, actually. The fonts should
already exist. :)
<starseeker> Well, almost - http://www.stixfonts.org
<starseeker> yes - I was thinking about resizing them in response to
the expression they relate to - drawings you can tie the dimensions in
x and y to the size of the expression - how would that work for fonts?
<hefner> something creeping up my todo list is a little glue code that
lets me match text sizes with pixel sizes (which I intended to use for
trying to match the sizes of divisors/dividends more closely, but you
could use it here too)
<starseeker> Hmm.  That's a possibility, but doesn't that break down
when generating postscript/pdf?
<hefner> well s/pixel/device unit/
<starseeker> ok.
<starseeker> My worry is if you have something like x^2/(x^3+y)^2 in 2D
inside an integral, the font may or may not have an integral symbol to
fit.  Or am I underestimating fonts?
<hefner> ah.. yeah, that could be a problem. the x11 symbol font here
is bitmapped and goes up to 34 pixels high. that's too small, actually
<starseeker> I think that's why maple does it with drawing lines
<starseeker> the original macsyma gui I think did too
<starseeker> Mathematica I can't remember
<hefner> you could just pester people to install a scalable symbol font.
