changing displa to do "line at a time"Tex

Hello,

> > refinement : allow  larger value for width  $linel, to take into account the fact 
> > that TeX display might be more compact than standard Maxima display) 
> 
> Or the TeX display might be narrower, as in 2-columns. 

Yes, you are right, 2-columns might be a good idea, it could more readable.

> insert a break.  By the time this can be computed, all the computation
> for typesetting the expression exactly as TeX would typeset it,
> namely the size and position of each
> character, is known.  You ask then that all this information be discarded
> and that it should be recomputed by TeX.

So (I am not sure I understand the sentence?), do you think it might be feasible, 
and it might be with not too much extra work ?
 
> > This way, we could benefit at the same time from the ability of Maxima to split 
> > long formulas (as compared to TeX's inability to split them) and the ability of 
> > TeX to produce beautiful output (as compared to limited ability of Maxima) ?
> 
> Another point that caused problems in the past for me, but perhaps
> is now fixed? I recall trying to typeset
> 
> (a+ b+ .... c+d)
> As 2 Latex commands,
> 
> (a + b +
> 
>    + c +d)
> 
> But encountered problems:  Tex REFUSED to do this because   $\bigl(a + b + $
> is not balanced.  This does not seem to be a problem in latex2e, but maybe
> there are other, similar problems?

Yes you are right, this is a good point. One ought probably have to write a 
generalized procedure that produces correct TeX output from incomplete formulas 
with unbalanced parenthesis. I could propose add "\left." to the left and 
"\right." to the right, either systematically (brute force) to each subformula or 
more intelligently, by examining if there are unbalanced parenthesis and adding 
either "\left." to the left, or "\right." to the right of each subformula when 
necessary. The brute force way does display correctly:

\scrollmode
$$ \left. \left(a + \right. $$
$$ \left. +b \right) \right.$$

>    A better solution to large typeset expressions may be to use arrays of
> aligned formulas.

Yes of course, you are right, a second approximation in refinements would be to 
use arrays of aligned formulas, although I imagine this would bring some extra 
difficulties if the notion of page would have to be used (for printed results?)  
rather  than strip of fixed width and indefinite height (during interactive 
computations) ?
 
> However there are various buffers in latex that sometimes
> run out of space.

No, if independent lines are produced ; Yes, you are right, it might happen with 
veeeery long arrays of aligned formulas, I suppose.

>    There are always tradeoffs between using someone else's program and
> writing one that solves exactly the problem you have in mind.  Sometimes
> NIH  (not invented here) syndrome is a waste, but sometimes it is right.

Yes, I quite agree. I think (dream?) that, if producing Maxima output as several 
TeX subformulas is easily feasible, it would provide, with Emacs and imaxima.el, 
an interesting interface for Maxima : formulas computed by Maxima -> transformed 
on the fly by TeX into bitmaps -> included as bitmaps in Emacs "à la imaxima.el". 
This would be especially interesting if the outline mode of Emacs could be used 
to display only a few images / formulas (is it ? by the way, Emacs 21.1 is 
available for Windows, now).

Best wishes,

Michel.Lavaud@univ-orleans.fr
http://www.univ-orleans.fr/EXT/ASTEX
ftp://ftp.univ-orleans.fr/pub/tex/PC/AsTeX
liste de discussion: astex@univ-orleans.fr
Abonnement à la liste: envoyer un message de contenu
"sub astex Nom Prenom Etablissement" à sympa@univ-orleans.fr