[ODE] Tips for consistent physics steps ?

Rob Ruck bobruck at btinternet.com
Sat Feb 7 19:08:14 MST 2004


At the moment I'm simply passing ode as a timestep the same delta, as used 
elsewhere in my engine. There are obvious fluctuations and ode can spaz out 
with major step changes. As I recently finished some octree & bsp stuff, I 
figured I'd fix this, and start passing ode a constant time. But I can't quite 
seem to figure out in my head how its done.

I'm aiming for a 100 fps physics time (unless this is a wee bit ambitious), i.e. 
a delta of 0.01. But i cant work out what to do when I have an engine delta 
more than this value, obviously if its less I can step through until the physics 
have received a cumulative amount adding up to 0.01.

I'm not aiming to change my engine so that it is physics oriented and I'm not 
interested in screwing with the delta value at its root or changing how 
everything else is updated.

I just need to know how to play with a given system delta which is 
changeable, in order to get a physics delta that is constant.

Am I looking at this in completely the wrong way ?

Thanks for any advice
--
Rob Ruck
mailto:bobruck at btinternet.com



More information about the ODE mailing list