Help with visualisation tool structure.

Hi, I''m writing a simple tool to visualize objects. At the moment all I''m trying to do is render a cube of sphere and perform the following: Mode 1: Camera centered on the object at a distance with mouse used to spherically move around it. Mode 2: Camera fixed at position when entered mode Mouse used to select the object (which will eventually do something but at the moment will simply bleep) Mode 3: Another Camera centered on the object at a distance and moves in the same fashion as mode. (this camera is attached to a light object and thus represents the viewpoint of the light). At the moment I''m trying to come up with a nice object structure for all this. So far I just have my windowing system which doesn''t handle input (doesn''t do anything with it) and my cube. I was thinking of storing all the objects and cameras in a scene graph and diverting input to the required objects via some form of state machine but the more I think about it the more hazzy it gets! Any ideas? Thanks <Fish>{