Yes pretty much.
Firstly check the target object is onscreen if it isn't then that's a load of processing saved.
Then using only those object number ranges likely to interfere (ie the level objects) check if they are onscreen.
Then make a loop to check various points along the way, checking range to target object - if range is fairly close run your collision detection routine with that object.
BSP works differently, I dont use BSP so wouldn't know. Matrix ground height can also be checked at each intersection.
Pneumatic Dryll
