pm3d and parametric function surface plots

[Johannes Z. <joh...@ze...>]

Hello,

I just thought that coloring parametric surface plots according to the
computed z-Value is not reasonable or obvious at all. It would be
reasonable to color it according to any function which depends on u and
v (the parametric variables).

Thinking about the syntax, currently parametric surface plots expect
3 functions like

    gnuplot> splot u, v, u*v

Would it make sense to allow a fourth /OPTIONAL/ function like (where
color(u, v) is any user-defined function which computes gray values):

    gnuplot> set parametric
    gnuplot> splot u, v, u*v, color(u, v) w pm3d

if pm3d is selected explicitely like above? I belive this wouldn't break
any existing scripts, would it?

On the other hand, to be consistent with non-parametric plots, one would
expect there the syntax

    gnuplot> unset parametric
    gnuplot> splot f(x, y), color(x, y) w pm3d

But this time, the specification is ambigous. It could mean:

    1) plot f(x, y) and color the surface according to color(x, y)
    2) plot f(x, y) and then plot a second surface color(x, y), which
       is colored according to it's z-Value

Introducing an option which forces function plots to have a color
function would solve the ambiguity.

Or something like

    gnuplot> set parametric
    gnuplot> splot u, v, u*v w pm3d color(u, v)

And in the non-parametric case:

    gnuplot> unset parametric
    gnuplot> splot f(x, y) w pm3d color(x, y)

What do you think about this syntax?

Feedback is welcome!

-- 
Johannes