ID:2140803
Aug 29 2016, 10:22 pm
|
|
Which code is more efficient if you want to know if something is in your view and know the range
var/cview=6 or var/cview=6 My logic is telling me get_dist is more efficient since its a pure calculations while the 'in view' does some sort of looping thru view(). But I would like some verification on this. | |
thx bro
| |
If you're looking for a same-Z distance, the best of both worlds is probably checking get_dist() and then the z coordinate.
get_dist() will do some math internally to get the x, y, and z coordinates. That very same math is done when you access a coordinate individually, except it's done each time. Therefore for fastest results, you'd want to use get_dist() first, and then if that check passes, try the Z coordinate. If for some reason this check might be done more against things that are just about anywhere, then checking z first might end up being the better option; but chances are it won't matter much for performance. Also if you're going to be looping through a list of objects in a certain range, then range() is probably the better bet by far. | |
It would be nice though if there was a byond function to do a line of sight check between two objects alone. Replicating view() with your own code to check line of sight is HARD.
| |
The view calculations are done as a whole, though, so there's really no way to separate the view() call from the individual LOS check. If BYOND did true shadowcasting that wouldn't be the case, but it uses its own algorithm.
| |
But if you're okay with that, it might actually be most efficient to compare coordinates one at a time:
Even more efficient would be to not perform the check too frequently. In the case of AI checking for targets, it shouldn't happen every tick.