Do I need custom Matrix classes in OpenGL?

I know that when manipulating the model-view matrix stack, OpenGL will know where a local vertex co-ord should be in world space.

But how can I obtain that world point, for say, a collision detection system? Do I need to maintain my own matrix system, duplicating the same transformations as are being rendered? Or Is there an OpenGL command such as: glLocalToWorldfv(aVector)?

newbie kind of thing…

You dont NEED to, but it will make your life easier. I like to save my transformation matrices because they come in handy for things like collision detection and clipping.

cheers,

I’ll add them, and when rendering, use my own transform data to manipulate OpenGL’s matrix stack.