algorithmic modeling for Rhino
Yes, I didn't realize doing things the proper way was so much slower. I replaced the boxes from this definition (http://www.grasshopper3d.com/forum/topics/voronoi-2) with planes and boolean differerence. It runs 4 times faster.
Just to a look and I am surprised at how simple this is, excellent. Maybe you just obsoleted qhull. Yours is much faster. Now the hard part, finding a logical architectural meaning to apply Voronoi aside from aesthetics.
Already on it.
I'm working on using 3D voronoi to define programme and structure for my current studio design.
Will upload pics when its working better.
Im working on simmilar goal as you for my thesis project.. I would like ( if possible) to see some update as you suggested?
For me qhull is slower dunno why, I'm on a MBP intel i7 quadcore sandybridge w/ 8gb ram. I think the benefit of the cluster though is it's ease, especially for beginners. I find qhull a tad tedious to control (most newer users cant even install it properly), and I like in the cluster how u can set a brep boundary, unless I missed that in qhull.
I agree, the cluster is ready and easy to use.
Setting a brep boundary isnt very hard with Qhull (with latest Dimitrie's component), for some reason I couldnt use directly Dimitrie's defininion properly, but I used it as seen in the screenshot, removed the first 8 open polysurfaces with the shift component and the solid intersection works well with the chosen boundary. You might have to play a little with the Bbox scale too.