but since I am using LWJGL, here goes…
Just wanted to verify some things more to do with OpenGL. I have my scene with an object (say a cube) and I have a “camera” (simple gluLookAt) that is positioned in a stationary point and I use the “opposite” of the position of my target to tranlsate the whole scene. This gives me the appearance of a chase camera, which is what I want. Is this what you guys do, or do you move the object, and move the camera to follow and leave the scene stationary?
It’s all the same thing to OpenGL, realistically all you’re doing is applying one transform (the ModelView matrix) to get everything into eye space. How you actually choose to think about that is a minor detail.
I keep my scene at the same position and move the camera - except that during rendering the camera transform gets applied first, so effectively I’m moving the whole scene.
I think it would be better to keep the scene stationary. If you add another object that is being chased or that can just come into view somewhere, it will be moving at a different angle and rate of speed than your chase object.