Re: desired features for gp backend?

[Daniel J Sebald]

John W. Eaton wrote:
> On 17-Jun-2009, Daniel J Sebald wrote:
>
> | What is important in my mind is an inroad to using
> | more powerful features of the plotting engine.  Why limit oneself to
> | the capabilities of Matlab?  A few hidden options having meaning
> | only to the plotting engine seems plausible.
>
> By "the plotting engine" do you mean "a particular plotting engine"?
> Since there can be more than one, then it seems to me these "hidden
> options" will cause portability problems for code.  You say the
> options are "hidden", but if they are available, people will use them,
> then complain later if we make some other plotting engine the default,
> or if they try to share scripts with people who use an alternative
> plotting engine that doesn't handle the special properties they are
> using.

True, but that's the user's problem to deal with.  The situation is the same if 
I write code dependent on some function 'this_is_a_long_function_name(...)' and 
don't make that function available in the Octave distribution script tree.

Maybe it is easiest just to write something like

send_text_to_gnuplot_graphics_pipe(fig, 'set foo')

where internally this function locates the appropriate pipe and simply sends 
some more text.

Dan