FAQ
Cover
This is the archive Discourse for the Processing (ALPHA) software.
Please visit the new Processing forum for current information.

   Processing 1.0 _ALPHA_
   Topics & Contributions
   Simulation, Artificial Life
(Moderator: REAS)
   Physics engine for p5
« Previous topic | Next topic »

Pages: 1 
   Author  Topic: Physics engine for p5  (Read 1265 times)
narain


Physics engine for p5
« on: Jul 28th, 2004, 8:04am »

I'm thinking of implementing a basic physics engine for Processing, and I have a few questions hopefully someone here can help me with...
 
Has something like this already been implemented for Java or P5?
If not, would building one be useful?
How hard would it be, and how much time would it take, considering I know a little about numerical integration methods (Euler, Verlet, etc.)?
 
And if making this engine is finally a good idea:
What's the minimum it would need to support, to be at all useful?
Are there any guidelines about what the API should be like?
 
Thanks, I'm looking forward to the replies. Physics is fun.
 
fry

WWW
Re: Physics engine for p5
« Reply #1 on: Jul 28th, 2004, 5:36pm »

at some point we're gonna try to include something like the physics stuff that simon greenwold did for his yale class in his hacked "multiprocessing", and it'll just be a little library, though who knows when that'll be.
 
i'd be really curious to hear what folks are interested in as well..
 
narain


Re: Physics engine for p5
« Reply #2 on: Jul 29th, 2004, 6:26am »

I actually started out trying to make a self-contained spring-and-mass sim. But then my code was getting really messy, and I realised that keeping track of all the positions, velocities and forces, and doing the numerical integration was better done by abstracting it all that out into a separate class.
 
Unfortunately I've never worked with a physics engine, so I don't know much about how it is supposed to be used: what its architecture should be, what methods it should expose, and all that stuff that separates at-least-it-works code from that-is-so-neat code.
 
I looked at ODE for C++, though, which has a pretty P5-like API (dBodyCreate(), dJointAttach(), dWorldStep()...), so I'm not completely clueless...
 
I'd like to help in developing that physics library you're talking about, fry, or if that's not on the radar right now I can start by myself. Is there anywhere I can get the source of Multiprocessing as a starting point?
« Last Edit: Jul 29th, 2004, 6:26am by narain »  
Pages: 1 

« Previous topic | Next topic »