I've been considering adding springs to my static model format. To do this I could allow the modeler to identify vertices that would become "soft" before exporting the model. Other properties of the spring system (mass, wind, etc) could be controlled from code. Something to think about anyway. It would be nice to give my wife that kite flying demo I've been promising her for ages.
Lets see now...for every vertex that is a part of the spring system I'll need to find the adjacent vertices that are also a part of the system and then maybe their adjacent vertices too depending on how flexible and computationally expensive the system needs to be. Controlling this property should be included in the interface. And for every adjacent pair of vertices I'll need to store the original length between them.
Thinking as I go along then we'll need to keep track of the static vertices which are adjacent to "soft" verts since there will be a spring connecting them. I guess normals will need to be recalculated for each vert after every update. And so on...
New Photos...
I posted some new pics on my site... sorry no more bunnies.
![](http://www.neilkemp.us/v2/i/photos/july_2005/july_1.jpg)
Physics... [looksaround]... I'll nod and agree - sounds clever though.
[headshake] I'm disappointed.
Jack