Inside: SereneScreen Fan Forum - View Single Post

flipper · 10-23-2009, 10:36 AM

I'm more than willing to send Jim (or Feldon or whomever) the code I've been developing since I first came across MA about a decade ago. I've always felt I owed Jim and this would be one way to pay him back.

I have a number of settings that govern the behavior of my fish with regards to avoiding each other and the objects in the tank. And to my untrained eye, and to those that I've given the app to, they look fairly realistic.

The fish swim around pretty much as they please, and with less than one percent "collisions," which I track as simple sphere-sphere or sphere-box intersections. Tracking of actual triangle intersections yields almost no true mesh collisions at all, even with a large number of objects and fish in the tank.

Maybe this would be a starting point for Jim? At the very least I would be willing to write up my thoughts on how I went about the analysis as well as provide him with some pseudo-code. It's all pretty straightforward in terms of ideas, and I could provide code that he should be able to just plug into his and use as-is, assuming he's developing in C and/or C++.

10-23-2009, 10:36 AM	#20
flipper Registered Join Date: Mar 2002 Location: Texas Posts: 60	I'm willing to help for free I'm more than willing to send Jim (or Feldon or whomever) the code I've been developing since I first came across MA about a decade ago. I've always felt I owed Jim and this would be one way to pay him back. I have a number of settings that govern the behavior of my fish with regards to avoiding each other and the objects in the tank. And to my untrained eye, and to those that I've given the app to, they look fairly realistic. The fish swim around pretty much as they please, and with less than one percent "collisions," which I track as simple sphere-sphere or sphere-box intersections. Tracking of actual triangle intersections yields almost no true mesh collisions at all, even with a large number of objects and fish in the tank. Maybe this would be a starting point for Jim? At the very least I would be willing to write up my thoughts on how I went about the analysis as well as provide him with some pseudo-code. It's all pretty straightforward in terms of ideas, and I could provide code that he should be able to just plug into his and use as-is, assuming he's developing in C and/or C++.