Re: [OctDev] Java/OpenGL-based graphics package for octave

[Daniel J Sebald]

I'm getting the impression that matryoshka pointers get so excessive that one is 
forced to implement all plotting capability inside Octave.  If every line of the 
lengend is to have its own handle and every stem and symbol is to have its own 
handle (why? so people can move an individual stem left or right a little bit?) 
and so on, the only thing of any use to Octave is primitives, in which case 
Octave has to have its own range routine, legend maker, etc., etc.  One way to 
handle the stem issue in gnuplot for K stems is to have K different sets of 
data, one sample point per data set.  But that is getting a bit extreme.

(more below)

John W. Eaton wrote:
> On 23-Apr-2007, Michael Goffioul wrote:
>
> | For instance, a legend in Matlab is a complete independent axes object
> | which is a regular child of the figure (you can even set its view to 3D mode),
> | while in octave it's a set of properties spread between the axes object and
> | the line objects (Matlab does not jave "key", "keypos" or "keylabel"
> | properties).
>
> Yes, I cheated on the legend for now because I didn't want to try to
> draw the legend with gnuplot.  Eventually, this will need to change so
> that the legend is represented in the plot properties as a separate
> small plot.  In any case, I'd be happy to consider patches for this
> change, but for now it must also work with gnuplot.

It can still be programmed as a child of the figure, and it could have some of 
the properties.  However, full blown property set probably is not doable.  I 
doubt people will complain if they can't have a handle to the, say, right line 
of the legend so its color can be changed to something different than the rest.

Dan