[fluxus] pdata and particles

[nik gaffney]

>>the script below shows an example of using the pdata from one object to
>>'rebuild' something similar from a different primitive
>>     ( .. .which could be extended to do some simple mesh morphing,
>>    but getting the interpolation right could be a bit tricky ))
> 
> nice idea! I hadn't thought about copying pdata between primitives in that
> way.

ideally, it should interpolate the data between primitives in a number of
steps (a 'mere matter of programming') which is ok for primitives with the
same number of points, but a bit more difficult if it varies.. ,

> it would be possible to create an active object for every vertex on a
> sphere in the right position, simulate and each step copy the world space
> transforms of the cubes back into the sphere points. is this what you're
> getting at?

something like this,. but simpler :)
is there a single pdata operation that will give the position/centre, rather
than points?+

> I guess it would be a bit slow, and the sphere would just collapse on
> itself if there was any gravity.

unless there were some joints, or forces (or something else) holding it
together perhaps?

> we don't have the topological information
> to do spring dynamics really (yet). well we do with something like the
> nurbs plane I suppose.

how would this work?

> I think there is probably a good case for supporting particles directly in
> the ode code. I just haven't got that far yet :) Then this would be faster
> and a bit easier to set up.

,. . not to mention the new and exiting ways of slowing doan and/or crashing
fluxus!