Sorry about the more OpenGL rather than JOGL question but I am very intersted in knowing people’s opinion to apply changes to my JOGL project in order to maximize performances.
PS: for the jogl.utils people, I also would really be very interested if model from blender could be imported, especially the IK stuff 8)
I think both of your options are exactly the same. In OpenGL, there is not such notion as a ‘camera’, although the routine of gluLookAt() really does give that sort of impression. Instead, the ‘camera’ which you’re looking at the world through is always positioned at point (0,0,0), and to create the illusion of movement, you move the world co-ordinates in the inverse direction you wish the camera to travel.
[EDIT]: Someone correct me if I’m wrong; I only just learned about this whole camera thing a few weeks ago.
I agree with Ultraq. For me it’s the same in the perf point of view, but I also may be wrong. I can’t see why the matrix computations could be more expensive in one way than in the other. Maybe you could benchmark a little to be certain of it. But I would suggest you use the way that is the more convenient for your application.