On Sat, 2006-08-05 at 10:19 -0400, sen1 at math.msu.edu wrote:
> I noticed that at least some of the plotting routines are written in
> tcl.
>
> These days, is that the best choice?
> Would it be better to take something like "python," and try to write
> scripts (say, incorporating matplotlib or some such thing), for new
> plotting programs?
Most of the plotting in Maxima is done, by default, by gnuplot which is
written in C.
The only plotting in Maxima that I'm aware of, which can only be done
by a TCL script is plotdf. The reason to use TCL is because that's what the
author of plotdf used at the time, and we do not have anything similar
in gnuplot. The code was already there, inside a TCL script in the
Maxima repository; I just wrote a lisp function to launch that script.
Having said that, I must add that I've worked for several years in Python
and only recently started to learn TCL to be able to maintain openmath.
For the purpose of writing a plotting program, I do not see anything
wrong with TCL and I do not think that Python would imply any advantage
over TCL.
If we do not have a better plotting program it is not because we use the
wrong programming language, but rather because we do not have enough
people working on it. There are other groups who have spent a lot of time
to develop a nice plotting program, such as matplotlib, but incorporating
their library into Maxima would require some work and people
with time and interest in doing it.
>From my part, I've decided to try to improve what we already have (openmath)
rather than starting something new from scratch. I've just started a few
months ago so I do not have many results to show yet.
> Perhaps a good solution would be to open up a plot window which one
> could interact with to read data from maxima and use whatever external
> plotting program one wished.
Some external projects are working on that as someone has told you. You
can always save Maxima's data into a file and open another program to read
it and process it.
Regards,
Jaime Villate