I am comparing the methods used in FPS to judge whether a player shooting another player (using an instant weapon) hit or miss. I googled and found the following methods. (1) the player presses fire key, but cannot fire immediately. Need wait for server to send back response saying "ok, you fire!" This method is old. (2) fire immediately, no lag compensation in server. pros: The shooting player needs to aim the opponent at a leading position related to latency. cons: more accurate result from point of view of the player being shot. (3) fire immediately, lag compensation in server. pros: The shooting player just have to aim directly at the target. cons: less accurate result from point of view of the player being shot. (e.g. "shot around the corner") Any other methods? Also, anyone know why some old games use method (1)? Though it''s not used today (bad responsiveness) it must have some advantages for being used in old games. But I cannot think of anything good from points of view of either the shooting player, the player being shot, implementation, or whatever.

Method (1) is used mainly because it is simpler, faster and is easier to control cheating since the server makes all the important decisions. It''s old but it''s fair. Method (3) would be ideal if all players were bots.

Here are a few thoughts to consider:
a) single shots VS burst (a la SMG) packet compression
b) randomness level in shot direction (perfect VS realistic weapons)
c) ammo travel time (i.e. rocket vs bullet vs laser)

Considering the above, the question becomes: can you make the lag feel like a normal weapon precision issue rather than a game design issue? Just a thought...|8-}

-cb