That's pretty cool.

ddf and movax - thanks for your nice comments!

>> I have no idea whats happening …
movax, perhaps this will help explain what is happening:

this piece is based on a physical model; that of a grid simulating an elastic membrane stretched over a square frame.  In addition to the elastic forces between contiguous nodes on the grid, there are also forces trying to maintain the nodes of the grid aligned in the plane (an attempt to simulate the rigidity of a painter's canvas). i suspect that the implementation of the latter is faulty; but the results are interesting anyway.

the key physical parameters (e.g. elasticity, resistance to bending) are set by a sequencer that can vary these parameters smoothly.  Different mouse gestures cause different sequences of parameters.  for instance, drawing a line creates the following sequence:
-after the "cut", the bending parameters is increased, causing the grid to expand
-when the grid reaches a certain size, the bending forces are reduced, causing the elasticity of the grid to bring it back to a rest position
-somewhere on the way the cut is mended
-finally the grid returns to its resting state

under certain circumstances, such as following two mouse clicks separated by a brief pause, parameters are chosen such that the grid is unstable

to me, the interesting thing about all this is that the model is a simple 2D model.  as stephen wolfram has shown, very simple automatons can create wildly complex behavior.  the other surprising thing to me is that the model knows nothing about 3D.

yes, the JDK compliance had been set at (1.)5

I recompiled it at 1.4.  Hopefully you will no longer get the error

thanks for trying it

jmbara