I've solved it! See code at far bottom of this post :)
The logic was almost sound, but not close enough ;)
Making a straight level, I noticed how every bullet my tower would shoot would land just beyond the waypoint the enemy had lastly passed.
Example A.
Blue sphere = enemy.
Red sphere = targetPos for bullet.
Yellow sphere = bullet itself.
![Somehow the picture did not show up correctly.][1]
I ran through my code with some debugs, and then I realised that the current position of the enemy would not be taken into account if and only if the prediction hadn't reached the target waypoint or beyond, this line right here:
return Vector3.Lerp(path.vectorPath[expectedTargetWaypoint -1], path.vectorPath[expectedTargetWaypoint], distanceExpected / distanceToCover);
Does not take the current position into account, which is no problem, as long as we have to look from the target waypoint and onward, but if we look at the last waypoint, say, waypoint 0, and the enemy has already passed that, then we should be calculating from where the enemy currently is, like from enemy to target with expected distance, instead of from A to B with expected distance, disregarding the enemy.
This fancy talk all rounds up in the code below, the debugs aren't necessary.
Note new bool "isPredictingFurtherThanTargetWaypoint" and its usages.
///
/// Calculates the position of this enemy in x amount of seconds and returns it.
///
/// Amount of seconds we're predicting ahead.
///
↧