Hi, I wrote an A* program, but the problem is that it would only work with one object at a time. Is there a way to do dynamic pathfinding (target is constantly moving, but moving slowly) that would not require resetting all the scores each time I tried to find a path. Or would the time be so insignificant that it would not matter? ???
Thanks
Find a set of suitable points, do all-pairs shortest path between them, and then consider the N points nearest the moving source and the N points nearest the moving target.
[quote]Find a set of suitable points
[/quote]
You probably mean to place nodes in the level
to serve as start and end-points to precalculate a set of paths.
Then at runtime you only need to find a path to a start-node and
from the endnode to the target.
(else the actor can follow the precalculated path)
You still need an collision-avoidance methods though, as the precalculated path could
be blocked by another moving object. (some methods to circle around it)
In my code, I have multiple moving entities each with their own different destinations and start points. I need to do pathfinding for each of those entities. I think I figured out a solution which is to have each node store a reference to the start and end nodes. If these change, then I recalculate the score for that node (g score only, h score is always recalculated).