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

[Daniel J Sebald]

John W. Eaton wrote:
> On 23-Apr-2007, Daniel J Sebald wrote:
>
> | I'll guess the latter.  Well, I assume hggroup is just a bunch of handles 
> | grouped together.  The children would be all the individual stems.  I'm guessing.
>
> No, that's not quite the way it works.
>
> | Well, there should be some OOP principle here.  Either each element will have to 
> | have its own drawing method that __go_draw_axes__ can call, or there are going 
> | to need to be special names in the elements like "__legend__" so that 
> | __go_draw_axes__ can recognize it is a legend child and then grab info from the 
> | legend child about setting up the legend.
>
> I don't understand what you mean by "special names in the elements
> like __legend__".

If after plot() there is

  get(gcf)

  ...
      children = -1
  ...

and then after legend() there is

  get(gcf)

  ...
      children = -1 -4
  ...

and both of these children say simply "type = axes" then there is no way of 
knowing that one of them is a legend.  So, either each of these children needs a 
drawing method, or they have to have special names identifying them.  I guess 
that word is...

> The child object that is the legend is a special kind of axes object
> that has the "Tag" property set to legend.  So when the object is
> drawn, I suppose that if it is necessary, it could be handled
> specially based on the "Tag" property.

"Tag".  Why it couldn't simply be "type = legend", I'm not sure.

Dan