How to reset pathfinding?

Development Programming & Scripting AI
, , , , ,
1

I’m trying to implement Quake-like jump pads with the help of LaunchCharacter function. The problem is I can’t get the AI to use it properly.

The most obvious solution was to use NavLinkProxy, but it does not give disirable effect: the AI approaches a jump pad from different sides and therefore are being launched earlier than it can reach the start point of a link, which leads to strange behaviour, when character cyclically tries to return to the first point of a link but ends up launched again. Analogous problem occures when the AI misses the end point of the link.

So the question is can I somehow reset the pathfinding component of a character, so it won’t try to reach the first point of a link, but try to find a new path from the landing point?

2

Have you tried calling StopMovement? It will abort current movement requests and will report back to your AI (assuming you’re observing that, or using our stock BT move tasks).

Cheers,

–mieszko

3

Thank you for the reply, it did the trick, but I guess I can use it only as a temporary solution, cuz it interrupts BT MoveTo task. Is there any way to do it “transparently” to the BT, without interrupting the current task?

4

Finally managed to solve the problem by rewriting the custom MoveTo task:

https://i.imgur.com/DO4iqX5.png

In this case task isn’t aborted by the StopMovement call. Many thanks to Mieszko!

Result: http://gph.is/2F92Jkp