I did mess about improving the collision system a bit today, but I'm kind of getting to the point where I'm feeling like this project was an excellent way to testbed my ideas about implementing skeletal animation but am not actually producing a game that will be much fun to play.
I've been playing around with Box2D's test program today and am getting tempted to start on something physics based. I would really like to come up with a simple, feasibly completable idea for a game and get cracking on it, but I guess we can't all think of Crayon Physics.
I'd really like to get myself to the point where I could implement some decent physics myself and have been looking at Real Time Collision Detection on Amazon. It's quite pricey in the current economic climate - anyone have any feedback on this book?
Bit of a creativity dead zone at the moment. Might just spend some time working on modelling for a bit.