On Thu, 2007-03-29 at 17:36 -0500, sen1 at math.msu.edu wrote:
> 1. I have created several macros for the existing plot2d,
> openplot_curves, etc. which are part of programs that I am using
> in
> a "production" type of mode for my research. With time, I can go
> into them and
> change them.
Well, in that case we can put openplot_curves back into plot.lisp and
keep the documentation removed so we prevent more users from becoming
dependent on it. Notice that many bugs that have been reported and
fixed in plot2d-discrete remain in openplot_curves; just to mention
one, you cannot use fractions or double precision numbers.
If you are in "production mode", you should definitely be using the
stable version 5.11 and not the CVS version. The cvs version is suppose
to allow us to propose new things.
> Also, it would take some checking to see if your new plot2d actually
> allows the same facility as the old one--with appropriate new
> macros. It may be that further development is actually desirable.
I've done a lot of checking with several old examples that I have.
As you know, my lectures notes depend heavily on Maxima.
Nothing breaks down in any of the examples I tested; in some cases you
may see the graph presented in a different way (in a better way I'd say)
but I doubt that any plot will stop working. If you've used complicated
gnuplot_preamble's, they will still be passed to Gnuplot.
The only thing that may cause bigger changes is when
gnuplot_curves_styles was used. It will still be accepted but ignored.
We'll still have time to do more testing before 5.12 is released.
> So, until there is some opportunity to check that things work
> correctly, I would like to request that, instead of making a complete
> replacement for the current plot2d, you create a new name for your
> current version.
I have not made any complete replacement. You can verify in the CVS
browser that in fact some new features were added, but the bulk of
plot2d (draw2d, draw2d-discrete and draw2d-parametric) remains the same.
> instead of making a complete replacement for the current plot2d,
> you create a new name for your current version. For instance, it
> could be nplot2d or something similar.
As you also know, that's the strategy that I've been using during the
last two years. I felt I was ready to transfer some of that experience
into the standard plotting packages, which are very often pointed at
as the weakest point in Maxima.
Cheers,
Jaime