Cameras

Ok, I looked a little at the source and read some articles about cameras.

Now I got some questions:
1) Do I really have to transform EVERY vertex in my world from world space to camera space? That does slow the engine down a lot, doesn''t it?

2) If the answer to 2 is no, how can I determine which vertices have to be transformed to camera space?

3) What are the benefits from a camera? Only to see which polys are behind the camera to discard them?

4) Do I render the vertices in world space or in camera space?

5) Why do you say it would be so unbelievable hard? I looked for about half an hour over the source and read 3 articles and I now have a fundamental idea of what to do (I hope).

If that is too much to answer, please answer at least no 1...

Thanks

Pretty much what the above said. Good luck.

Martin