I have this experimental Coin Magnet system that uses OverlapSphere to pick up nearby coin objects. CoinSpawner and CoinMagnet are essentially in the same position.
https://paste.mod.gg/epzrdlrwhbdm/0
Ideally, coins that are not within the magnet's range should not be registered to the _magnetizedCoins list. But in the first version, in CoinSpawner where I set new coin objects to be active before setting its new position, some coins seem to be Magnetized, even though they are out-of-range. They are also disabled and with Magnetized == false when the pool releases them. When I change the order, and set the coins active only after setting their new positions, it works normally.
There is no stop in the spawn instruction, so my only guess is that OverlapSphere somehow takes over in the middle of the action. Is it probable that the Physics queries can pick up the colliders right after they are set to be active?
#Physics queries - Active object order
olive wing
dense robin
You need to call Physics.SyncTransforms() if you want the physics world to immediately update after moving an object. Otherwise, the changes will only show up after the next physics update
It's been a while since I dug into this. I remember helping someone with this for a ship-building system – it would find colliders in bogus locations
dense robin
You need to call `Physics.SyncTransforms()` if you want the physics world to imm...
Activating a game object may also cause its colliders to get updated immediately
which would explain why activating the coin after positioning causes the bug to go away
olive wing
You need to call `Physics.SyncTransforms()` if you want the physics world to imm...
Ah yes, if I SyncTransform, the bug also goes away. I guess that means the colliders still linger at the old position when re-enabled instead of following the transform's new position