Game Publishing/Business

Re: New XNA physics engine

bp — Fri, 16 May 2008 12:11:07 GMT

Nice work! I'm looking forward to the beta

Re: New XNA physics engine

toody — Fri, 16 May 2008 02:48:34 GMT

The license is ZLIB, which is about as unrestrictive as licenses come. I forgot to mention a lot of stuff in that post :/

Anyway, I've added a lot of features and made some improvements, so a beta release should be coming fairly soon.

Re: New XNA physics engine

Coatl Studios — Mon, 12 May 2008 17:51:43 GMT

and besides easy to use and speed .. a indie friendly license for it would be nice..

no matter how awesome a piece of software is if the license is worryingly restrictive or stupidly expensive.. it would not be used as much as a software with a more relaxed license with allows more freedom to its users even if its not as feature rich

Re: New XNA physics engine

toody — Mon, 05 May 2008 07:52:31 GMT

Thanks! I dont really have a target feature list per se, but I'd obviously like to add broadphase (O(n^2) at the moment), compund bodies, more joints, ragdolls and the usual stuff, maybe convex objects, and then maybe try soft body integration. I was thinking of trying to implement FastLSM, but its patented, so I'll have to try get permission from the patent-holder first.

Re: New XNA physics engine

Akai kaze — Mon, 05 May 2008 07:24:10 GMT

nice engine ! I don't know what you are planning to include but easy of use and speed are the main things someone's look for a physic engine.

but it's nice beginning

New XNA physics engine

toody — Mon, 05 May 2008 03:42:29 GMT

Hi, I wasn't sure where to post this, maybe someone could move it if its in the wrond place?

Anyway, I have just released an alpha of my XNA physics engine, you can grab it here: http://www.ralphmorton.net/projects/Icarus/dl.php

So far its pretty basic. It supports only convex shapes, and I havent added composite shape support yet. It uses MPR (Minkowski Portal Refinement) for collision detection, although there is also a continuous GJK implementation, so continuous collision detection can be enabled if desired. The solver uses sequential impulses a la Erin Catto (in fact a lot of the code was built using Box2D_Lite as reference). The only joint actually implemented at this point is a distance joint. And restitution isnt implemented yet!

Let me know if you like it, or if not, what you dont like about it. Hopefully it will at least be useful as a learning tool ;)

EDIT: I should probably point out that its a 3d physics engine, to avoid confusion

EDIT again: screens