Hi!
Thanks for your repy and suggestions about the fix.
As I have already written some time ago,
any change of the behavior of the Maxima commands
should be announced somewhere where it can be easily found
(for instance in a specific page for changes in our web pages),
so that the Maxima developers can fix their packages.
I would appreciate very much if "plot2d" could
be made backwards-compatible.
Fabrizio
On Sun, 6 May 2007, Robert Dodier wrote:
> On 5/6/07, Robert Dodier <robert.dodier at gmail.com> wrote:
>
> > On 5/6/07, Fabrizio Caruso <caruso at dm.unipi.it> wrote:
> >
> > > load(sarag);
> > > drawTopology(topology(x^2+y^2-1,x,y)[2]);
> > >
> > > should draw the topology of
> > > a circle and the expected drawing (topology)
> > > should look like a rombus
> > > (which is topologically equivalent to a circle).
> >
> > When the plot format is gnuplot, I get a figure that looks like:
> > /|\
> > -+-
> > \|/
>
> OK, on further investigation it appears the problem is that the
> treatment of plot options has changed. In particular gnuplot_curve_styles
> has been replaced by a new option "style" and gnuplot_curve_styles
> is no longer recognized (although it is still documented).
> The horizontal segments are not supposed to appear (only their
> endpoints instead).
>
> A resolution of the problem for SARAG is to replace
> [gnuplot_curve_styles, ["with lines 1", "with points 2", "with points 3"]]
> or whatever, by
> [style, [lines, [1]], [points, [2, 3]]]
> Note that the new option is "style", not "styles".
>
> I don't know to what extent this situation is intentional.
> Is plot2d supposed to still recognize gnuplot_curve_styles?
> (What about other gnuplot-specific options -- are those still recognized?)
> At this point I'm inclined to think the gnuplot options should be
> recognized for the sake of backwards compatibility.
> This topic probably came up before but I forget what I or anyone
> else said about it then.
>
> > When the plot format is openmath (in XMaxima), the plot is empty.
>
> I don't think plot2d([discrete, ...]) has ever worked with openmath
> so I don't see this as a new bug.
>
> FWIW
> Robert
>