path node culling

[jdoran]

Can anyone tell me the reasoning behind the rules Pathing:CullNodesFromStart and Pathing:CullNodesFromEnd?

I see FindRoute come up with a perfectly good route, and then the culling throws away the first and last nodes in the route. Fortunately there is still LOS from the remaining nodes to the destination -- and the culling logic ensures that there will always be. Is this just path optimization?

I'm tempted to just change the rule values, but without knowing the intent for this culling I'm worried that I will mess something up.

The only reason I ask is that my NPC is having pathing problems once it reaches the last node (after culling), and I suspect it would do better if it were standing at the closer node.

[Derision]

The culling is there to optimize the path.

Without culling you can get situations were the nearest node to the destination (e.g. to the player who a mob has aggroed on) will cause the aggroed mob to run past the player to that last path node, then run back to the player.

Similar issues can occur at the start of a path without culling, i.e. a mob runs away from it's target to the nearest path node before beginning to traverse the node path.

The culling was an attempt to stop that happening.

[jdoran]

Thanks for the clarification.

The culling is working fine, and is not related to what I was seeing.
But anything unexpected becomes suspect.

That particular problem was due to pathing going underground after
the last node was reached. Nothing to do with path nodes. FindBestZ
"corrected" the destination Z to deep underground. I changed the code
to use FindClosestZ, and the pathing was correct.

I'm not ready to recommend that change in general. I need to study a variety of scenarios before I'm confident that I understand what is happening.