Would it be better to have a player\x#, player\y#, and player\z# type for each coordinate of each important object like in a lot of the samples, or should I just use EntityX( entity ), etc. instead?
Is there any real advantage to using the former?
Is there any real advantage to using the former?