Original Post
My first question is it best to rotate the entire scene in order to rotate or rotate the camera using the eye and look. I have gotten the matrix rotation to work and thought it would be better to rotate the camera vectors for eye and look. This worked except for sometimes when I move the camera it goes a little wacko sometimes. The rotation or movement alone worked fine though. I did it by rotating vectors and a little math. Is this the wrong approach? Second is structure for game programming. Should you render by a timer, do physics calculation by a timer, input from the user as they press buttons. Some ways would obviously make a jumping game, what is the correct structure?