Current: setting camera bounds before size is known will discard the call, because a size is required. Goal: store the desired bounds, set when size is known.