Been messing with the spring constraint today, trying to get it all fixed up and ready for trunk.
So far I’ve cleaned up the code (dude, tabs, not spaces), refactored it a bit and fixed a couple bugs – still need to figure out why the bone goes all wonky when you first add the constraint and look into the cache system.
That thing’s entirely too much fun to play with…
–edit–
Eh, forgot to mention Ivo Grigull (Loolarge?) was the original author of the code.
Huh, good question. Looks to be location & rotation based but if the bones aren’t in a parent/child chain they don’t do what you’d expect – they all end up at the location of the ‘root’ bone. Something else to look into I suppose…
And it’s fast (enough), on my slow-ass netbook it updates in real time when I move them around in the view port.