3D?

[Dennis Brown]

Based on the original request,

 > I am interested in 3D support within Revolution/DreamCard.
 > Nothing fancy, just the ability to plot points (variable
 > size and color spheres) on an X-Y-Z cartesian grid and
 > then rotate the grid around in different orientations.

I would build buttons of images of spheres for each point, sized to  
the requirement, then positioned/layered according to the view point.

It would be a bit harder than the clock challenge, but not too much  
based on what I have seen recently from this list.
In fact it would make a good challenge problem.
Might even find some application as a real cool ball clock in 3D.

Dennis


On Jul 22, 2005, at 3:40 PM, Lynch, Jonathan wrote:

> I believe that in part it depends on whether you are using the
> points/splines/surfaces approach or the 3D pixel approach.
>
> The points/splines/surfaces approach uses much less data, although the
> math is a bit more complex. I know that the 3D programs I have played
> with use points/splines/surfaces.
>
> I had the idea to store the data as 3D pixels, but where you only  
> store
> info for pixels that exist. For a small object, calculating the first
> intersecting point between a line extending from the notional viewing
> window and the object should not be too hard, and should allow the use
> of fast commands like intersect.
>
> But even for that, think about an image that is 100x100 - that's  
> 10,000
> pixels, each of which requires some calculation to figure out.  
> Pixel by
> pixel image processing in transcript is not that fast.