Search Issue Tracker
Won't Fix
Won't Fix in 2023.1.X
Votes
1
Found in
2020.3.37f1
2021.3.8f1
2022.1.13f1
2022.2.0b4
2023.1.0a5
Issue ID
UUM-13027
Regression
No
Collider of Tiles containing "TilemapCollider2D" and "CompositeCollider2D" is not flat
How to reproduce:
1. Open the user’s attached project
2. Enter Play mode
3. In Hierarchy select “Dynamic Sprite” GameObject
4. In the Inspector window observe the Position of the X-axis
Expected result: The position of the X-axis doesn’t change
Actual result: The position of the X-axis is changing
Reproducible with: 2020.3.37f1, 2021.3.8f1, 2022.1.13f1, 2022.2.0b4, 2023.1.0a5
Reproduced on: macOS 12.4 (Intel)
Add comment
All about bugs
View bugs we have successfully reproduced, and vote for the bugs you want to see fixed most urgently.
Latest issues
- Query Builder items display raw code style names instead of human readable labels in search field
- “+” button does nothing and has no functionality when clicked in Human Template "Bone Template" list
- Solid gray or black areas are present when using ShaderGraph with Custom Function nodes without manual reimport
- The Multiplayer Play mode tool doesn't see the list of Play Mode scenarios when using 6000.3.x
- Input field does not update when selecting the same input field after previous input was canceled
Resolution Note:
In this set-up, the collision normals are not exactly Vector2.Up (0,-1) but are (-0.0000002,1) which means there's a *tiny* drift to the left. This kind of floating-point imprecision is expected. The change in X is also tiny (approx 0.00001) per sim-step which is why it's hard to see it actually moving.
The problem here is that these tiny inaccuracies/forces are supressed via linear/angular drag and/or the body sleeping.
In this project, all the drag is zero and sleeping is turned-off.
Turn on body sleeping to stop this happening.
Resolution Note (2023.1.X):
In this set-up, the collision normals are not exactly Vector2.Up (0,-1) but are (-0.0000002,1) which means there's a *tiny* drift to the left. This kind of floating-point imprecision is expected. The change in X is also tiny (approx 0.00001) per sim-step which is why it's hard to see it actually moving.
The problem here is that these tiny inaccuracies/forces are supressed via linear/angular drag and/or the body sleeping.
In this project, all the drag is zero and sleeping is turned-off.
Turn on body sleeping to stop this happening.