replace current physics code with calls to Blaze 2D physics engine

ChipMunk?: Uses Erin Catto's work as well, and added some new features:

http://wiki.slembcke.net/main/published/Chipmunk

also:

http://www.gamedev.net/community/forums/topic.asp?topic_id=459889&whichpage=1&#3034879

Box2D has it's own website now, and new features! Plus a comment on gphysics.com from Erin Catto that he will continue to support it.

http://www.box2d.org/

Box2D is under the zlib/png license as well!

"Box2D has grown beyond it's humble box simulating roots, and can now handle convex polygons and other shapes coming soon."

This is good news for us :)

I have an idea.

How about I start a project on dsource (or just put it in it's own svn in arclib) called "box2d4d" which will just concentrate on porting Box2D to native D code as closely as possible to the original. I'll keep tabs on Box2D and update the 'box2d4d' project whenever box2D gets an update.

From there, instead of worrying about physics features you can simply worry about the physics API integration with ArcLib.

Our Dev System will then be like...

Erin Catto Updates w/ new features --> I update Box2d4d --> you update ArcLib scenegraph integration, if needed

I think this seems like a foolproof system, and it takes the pressure off of us from adding new features to the physics engine because you can be sure the original author is going to add them himself.

Does this sound good?

~ Clay

Yes, the Box2D page looks most promising and I agree that we should try to stay up to date on it. Since there's no way of hooking up with the library directly, an as-literal-as-possible port is probably the least work intensive way of getting all the updates into Arc.

So, yes: I agree about having the box2d port in a separate project and would also help out. (I'm busy till Sept 24th though)

I don't think I'll have the time to do this. :/

The Blaze ArcLib extension will work well here.