I can't say much about the transition, since I never really looked
into Box2d-js in detail. So, thanks for sharing your experience.
I'm currently replacing the for/each loops in trunk.
> When rendering, box2d-js gives you the world positions of the
> vertices of your shapes where as Box2dWeb gives you the original local
> vertices and then you can apply the angle from the body, with some
> trigonometry magic. Maybe I'm missing something, but I found it odd
> that now this calculation needs to be done on the rendering side.
Then I'm not the only one who had problems with drawing :) . But it
wasn't my decision - Box2dWeb is ported from Box2dFlash.
> As far as performance, goes it's probably
> still a bit early to tell since I'm still dealing with some issues,
> but right now it appears to be a bit slower than box2d-js
It's very hard to tell whether Box2dWeb is faster or slower than
Box2dJs. The first dev builds seemed to be a bit faster until I fixed
a bug which influenced behavior. Also, increasing the gravitation
makes everything "looking faster", as well as different drawing
scales.
Secondly, you cannot measure FPS in browsers, since rendering-rates
are different from timer execution rates.
I think that the physics engine performs very fast itself (thanks to
Erin Catto's great work). The actual bottle neck is always rendering.
Thanks for your compliments ;)
Cheers,
Uli