C Y wrote:
> --- Richard Fateman <fateman@cs.berkeley.edu> wrote:
>
>>I haven't been installing any of this stuff, but
>>now that I see that imaxima is using breqn, my
>>interest in doing so takes a major drop. I have
>>tried breqn and found it to be slow and not
>>able to break up anything but very simple
>>expressions.
>>
>>Are people actually using this collection for
>>non-trivial calculations? If you have
>>LOOONG expressions, the splitting must be computed
>>in maxima, not after mactex does its job.
>>
>>RJF
>
>
> I am using Emaxima to write maximabook, and Jay was just
> able to add breqn support after figuring out how imaxima
> did it. It does only do trivial breaking, but that is
> enough to make it very useful for the maximabook.
>
> If we could get maxima itself to generate TeX with the
> proper breaking included it would be wonderful, but I don't
> have the faintest idea how that would work. The proper
> place for line breaks would change with things like paper
> size, fonts used, and all sorts of stuff only determined
> when the tex file is compiled. Possibly one could define
> a file where you define various common parameters and then
> tell Maxima which ones you wanted to use, but without
> something like that Maxima would have no ability to generate
> intelligent line breaks for tex. It would also involve
> extending Maxima's internal line breaking routines substantially,
> if I understand correctly how Maxima works. (I freely admit I
> may have it wrong.)
Yes. The internal display looks at
the line length. Mactex would have to look at the font/width etc
data and use similar breaking techniques. This kind of thing
is used for lisp program pretty-printing, and it is possible
that "syntax directed pretty-printing" templates could be
used. I think that mathematica has such things, though in
such a complex form that few people think of using them.
A plausible other way of doing this is
to have the maxima internal form be passed to the display
program so that the display can be recomputed from real
data. (or for MathML fans, pass a mathml version. The
nice thing about MathML is "it compresses nicely".)
RJF
>
> breqn, for all it's faults, works at the time of tex compile,
> which is why it turns out to be useful. Hopefully someday we
> can do better, for now this is a good, useful step.
>
> CY
>
> __________________________________________________
> Do you Yahoo!?
> Yahoo! Platinum - Watch CBS' NCAA March Madness, live on your desktop!
> http://platinum.yahoo.com
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima