Open dritchie opened 11 years ago
Something I realized after this test: The focus/rotate around point of the camera is not a piece of visible state to the user. Therefore, it's often the case that the user and the computer system will have different models of where the focus is. You can see this as an instance of a more general UI principle: show the system state.
If we analyze the problem this way, we should expect that even if we add a button to focus on an object, this problem will still persist. Certainly, so long as the user ONLY navigates the camera using pre-sets, they'll be fine. But as soon as they start manipulating it in a more freeform fashion, they will probably lose track of where the focus is and the problem will repeat.
One way that I think is a bad way to solve this problem, but is in keeping with the analysis, is to explicitly show the user the focus point as a point in the scene.
Sketch-up has some weird behavior with this if I remember correctly. They use the cursor as a proxy for user's focus when zooming (and possibly when rotating?). However, this does mean that the focus point is usually on the surface of some object, which seems like the germ of a more interesting idea.
Camera controls: