Resolved Setting rigidbody.position sometimes doesn't change the position

The point of the code:
I have this code that shatters glass using a fracture pattern

The problem:
I have noticed a problem a while back where, sometimes when I shoot the glass and thus trigger the code, the new shattered glass doesn't get it's position changed to where it should be when I change the rigidbody.position. This seems to happen no matter the isKinematic value as well.
Also, the code creates many pieces, and when this issue happens, it happens to every one of them which leads me to believe it could have to do with timing. I have also noticed that when adding both of the Debug.Log()s in the code the issue stops appearing.

The rigid body (in case it matters):
rigidbody.interpolation = RigidbodyInterpolation.Extrapolate;
rigidbody.collisionDetectionMode = CollisionDetectionMode.ContinuousSpeculative;

The code:
I am not gonna post the full code as long as not requested as I believe this includes everything important, but I could be wrong.

The following is the function where I change the position of the rigidbody.

Code (CSHARP):

1.  
2. private GameObject ConvertConvexPolygon2DToShard(Vector2[] polygon, Transform relativeTo){
3. //relativeTo is the glass piece being shattered, which is used here to move the newly created polygon to the correct place
4. if(emptyShardsIn.childCount == 0){
5. CreateEmptyShards();
6. }
7. GameObject shatterPiece = emptyShardsIn.transform.GetChild(0).gameObject; //Get a pooled object
8. shatterPiece.SetActive(true);
9. GlassInfo glassInfo = shatterPiece.transform.GetComponent<GlassInfo>();
10. Rigidbody rigidbody = glassInfo.m_rigidbody; //So that I wouldn't have to do GetComponent<T>() everywhere
11. rigidbody.rotation = relativeTo.rotation;
12. MeshFilter meshFilter = glassInfo.meshFilter;
13. MeshRenderer meshRenderer = glassInfo.meshRenderer;
14. shatterPiece.transform.parent = glassPieces;
15. meshRenderer.material = brokenGlassMaterial;
16. Mesh newMesh = new Mesh();
17. Vector3[] vertices = new Vector3[polygon.Length];
18. int[] triangles = new int[(vertices.Length-2)*3];
19. Vector2 addedPoints = new Vector2(0,0);
20. for(int i = 0; i < polygon.Length; i++){
21. Vector2 point = polygon[i];
22. addedPoints += point;
23. }
24. Vector2 pointsMedianCenter = addedPoints/polygon.Length;
25. rigidbody.position = relativeTo.position + relativeTo.right * pointsMedianCenter.x + relativeTo.up * pointsMedianCenter.y;
26. //Debug.Log(rigidbody.position); //Adding BOTH of these makes the issue stop
27. //Debug.Log(relativeTo.position); //When printing these out by themselves though, everything seems alright
28. for(int i = 0; i < polygon.Length; i++){
29. Vector2 point = polygon[i];
30. vertices[i] = Vector3.right*(point.x-pointsMedianCenter.x) + Vector3.up*(point.y-pointsMedianCenter.y); //Move the vertices to counter the rigidbody moving
31. }
32. //The following is most likely unimportant
33. for(int i = 2; i < vertices.Length; i++){ //Fill the polygon
34. int triangleStart = (i-2)*3;
35. triangles[triangleStart] = 0;
36. triangles[triangleStart+1] = i-1;
37. triangles[triangleStart+2] = i;
38. }
39. newMesh.vertices = vertices;
40. newMesh.triangles = triangles;
41. meshFilter.mesh = newMesh;
42. return shatterPiece;
43. }
44.  

Note that I have filtered the code a bit to get rid of completely unimportant lines such as changing the name of the object

In the code, when logging the two positions by themselves so that the problem still appears, everything seems to be fine which, again, leads me to believe towards this being a timing problem.

I have attempted to get help for this problem multiple times (in the discord server), but without success.
Also, I apologize, as this is the first time I use this forum and I might have gotten something wrong, or forgotten to include information that I should have included.

This could totally be an oversight by me, but I haven't been able to fix this for over a week now so I have come here for help.

Linux - Unity Version 2022.3.7f1

 

According to the documentation:
https://docs.unity3d.com/ScriptReference/Rigidbody-position.html
"If you want to continuously move a rigidbody use MovePosition instead, which takes interpolation into account."

So if these shards are moving, which I think you mention is the problem, try:

rigidbody.MovePosition(position);

And see if that helps.

 

Doesn‘t get the position changed to where it should be: what should it be and what is actually happening? What exactly is the incorrect position? Does it correlate with some other object?

When do you call this method? From Update, FixedUpdate, or some other Unity event? If you suspect a timing issue you could defer the code by starting a coroutine that yields null (or WaitForFixedUpdate) once and then calls the method.

Personally I suspect it could be an odd scenario where it matters whether this code runs in Update vs FixedUpdate since you extrapolate the position. In such a case it‘s advisable to modify rigidbody properties only in FixedUpdate because during Update the body‘s position is being inter/extrapolated.

This would fall in line with the suspicion of this issue being timing related.

 

No, that's not quite what I want, I need to teleport them to the place the glass broke at first, not move them there continuously, but rigidbody.position = ...; seems to not always actually set the position.

 

This code is being ran when the player shoots at the glass, which is checked for in Update.

What should be happening is that once the player shoots at the glass, the broken pieces are generated and then positioned to the glass with a little offset, but what is happening is that the glass isn't being moved at all, it just stays at 0,0,0 where the "emptyShardsIn" objects were created to. I am yet to find any correlation with other objects or the behaviour of the player, other than maybe the timing, but I haven't been able to confirm that yet.

I suppose I can try to store the changes that need to be done in FixedUpdated in a list and let them be done the next time FixedUpdate is called.

If the FixedUpdate thing stated above doesn't help then I'll try this, but I am a little unsure if this would change anything if this is timing related and also I definitely couldn't keep it like that as the delay would probably be noticable.

 

Oh, and I have also noticed that if I change the transforms and the rigidbodies position then it works. This is most likely very inefficient and I'd much rather only have to move one of them, but moving only the transform also doesn't work and I had moved to changing the rigidbodies position due to that a while back. This feels pretty interesting to me.

 

Yeah, I don't really know why the code looks like that.
I am slightly confused of the response, but anyways, I am not changing the transform position. And also I don't have auto-sync-transforms on, I think, at least I haven't intentionally enabled it.
I had changing the unused and used shards being parented to just storing them in separate arrays or smth on my to-do list, but I didn't know that changing the parent could affect the current situation.
I'll try and see if it will fix the issue, if not, I'll continue trying the things others have suggested.

EDIT:
Yeah, sadly, but as expected, removing the parenting thing didn't change anything.

 

Yep, changing the position and rotation in FixedUpdate seems to have fixed the problem.