Strategies for Dealing with Large, Complex Terrain

Marius
Posts: 10
Joined: Fri Nov 11, 2016 11:34 pm

Re: Strategies for Dealing with Large, Complex Terrain

Post by Marius »

I think Bullet is basically in meter units right? That one was only about 6000, but I've been running tests fine with over a 60,000 radius (hadn't tried higher). That seems to give a fairly "large" feel to it, although maybe not to an astronaut. :P For kicks I just tried 600,000 and indeed I get the jitters you mention. At reasonable speeds, collisions still worked at 600km (in my 10 second test) even with the jumping. The spheres have a 1m radius (I should try smaller ones next).

I don't think the fact that it isn't to-scale will matter too much. Planning on using various tricks to make it "feel" large enough. Scale objects down a bit (hopefully not so much to cause other physics issues, of course), slow travel speeds, creative texturing, etc. Real life planets are kinda too big anyway.

But can't we recompile Bullet with double precision to give us more flexibility if we need it?
gjaegy
Posts: 178
Joined: Fri Apr 18, 2008 2:20 pm

Re: Strategies for Dealing with Large, Complex Terrain

Post by gjaegy »

The video is looking good, have you finally based your solution on the suggested btConcaveShape-derived class option ?
Post Reply