Having an issue i cannot comprehend.
Decorator specifies to go to next task when Goal has been reached.
However the goal seems to get stuck once it has reached it and never updates it.
The goal is X =1075 Y =-71347 Z =205
The pawns location is x =1076 y =-21347 z=205
That is a 50 000 unit difference because it was meant to prove my sanity after it was not working on low values like 150.
Acceptance radius is 1.
When the game starts and the behavior tree first executes it registers that the goal has not been reached. But once its reached… it refuses to change.
I have the same behavior duplicated except then i do not rely on Goal reached but instead check for an overlap with a specific actor.
I can use the work around and not rely on goal reached but level design then becomes an issue as you now need to compensate by adding a bunch of blocking volumes in your level to tell the AI where to stop and turn around.
So i guess my question is if this is a known bug or if i am just missing something.
I have a few quick questions to help with my investigation. Can you show me some screenshots of your BT and level with the target it cannot reach? Have you tried changing the accepted radius to something much larger? An acceptance radius of 1 means that the AI has be be standing at the exact vector of the move, which may not be possible if there is even a tiny bit of collision in the way. Try making it 1000 then work your way down to something more precise. What was the template or project this is based off of? Any additional images or information may help as well.
I eventually removed the decorator for if goal reached which has acceptance value of 300 units. I tested various ones from 1000 - 1. I am not sure why my first post said an acceptance radius of 1. Perhaps a typo.
Strange thing is if i created a custom decorator “close enough” then it worked much better which i believe is very similar to goal reached. It is effectively this [Close Enough][3]
If i used the Goal reached decorator then it would reach the point and not register that is has reached it after the initial trigger. Meaning it would reach the goal… but after that it will never reach a goal again.
Unfortunately i abandoned this AI behavior due to some other issues i had with design choices. Also judging by what i have experienced in the past few days i am semi confident that what i experienced was not a bug. I am having a development week of hell where anything that could go wrong will go wrong mostly because of my own errors.