ptimization of any kind. If I were to add a space voxelization algorithm to it I could probably do away with 90% of the force vectors in a large particle cloud.
http://www.youtube.com/watch?v=RmsbzTf79bc
--
David Rutten
david@mcneel.com
Poprad, Slovakia…
Added by David Rutten at 5:35am on December 16, 2009
ose space to make the louver look dense at some part.
Here' my method toward those space creating.
First, i create a numeric combination, so i got 3,9,15,30.
Second, i repeated the data until the length i want, and jitter them. Here's what i get:
30,3,9,3,15,9,30,30,15,3,3,15,30,9,3,3,15,3,3,9,9,3,15
Actually what i`m trying to achieve shown below:
3,9,3,3,30,3,15,9,15,30,30,15,30,3,9,15,9,3,3,15,3,9,3
My idea is, to make the center part look no so dense compare to others part, somehow
it's no a absolute looking in the final. What i mean is, maybe more [30] like 80% drop at the center part, 80% of [3] drop at the start and the end, meanwhile the others 20% [3] and [30] appear to be random in whole series.
Last, i will weave this set with other set which stand of the dimensions of sticks.…
closest point to the very first would be removed from the list, so the initial list reduces from 100 to 98. From the 98 i pick one and search the remaining 97 for the closest. From the remaining 96 i pick again one and search in the 95,...
(The product I want to result is:
having a number of random lines in 3D space, produced by an even number of points as discribed, this shall be the initial springs for a ("selfadjusting") tensegrity. Each one of these lines (later springs in kangaroo) get divided in three areas - that means four points. These four points again are the "attractor points" of neighbor springs, so the strut "knows" where to set the next elastic connection,...the rest I´ll have to figure out)
angelos…