previous latest addition

it may not be the most beautiful diagram in the world (it's prettier inside xfig; i've had to shrink it in this png to fit it inside the page), but it is interesting.

this is a plot of a randomly generated drum skin model from the kultrung code (i'm amazed i got this far without ever making a picture like this before). cell boundaries appear as thin black lines. inside each cell is an orange cross - this is the "centre" of the cell. each cell centre is connected to its neigbours by springs (blue lines). the green circles are fixed points that define the edge of the drum skin.

the cells are supposed to be a voronoi teselation, the dual of the delaunay triangulation (the springs) that joins neigbouring points (cell centres and edge points). but the tesellation appears to be incorrect: each red circle identifies a cell boundary that is not perpendicular to a spring (there are more examples that aren't circled). if i understand things correctly, that should not happen.

also, there should not be springs between edge points.

hmmm. the more i stare at this, the more it seems like it is ok. i'm going to have a look at more examples...