Hi All,
Gltf physics example seems to have some issues when using babylon latest version 5.
https://playground.babylonjs.com/#FD65RR
It falls through mesh and stops a bit below like the imposter is offset-ed.
Can anyone confirm?
Got the same result here on Edge, seems to be related to how the impostors are set? If I change the mesh impostor to a simple box impostor instead of a composite impostor like the one in the example, it works fine: Physic impostor changes? | Babylon.js Playground (babylonjs.com)
@RaananW did something change in the impostors between 4 and 5?
Huh. if I enable the composite impostors from before, but comment the physicsRoot.position.y = 3
line, the collisions work as expected, even if I change the y pos on the inspector, it falls and collides with the ground as expected.
EDIT: I wonder if it’s something related to parent-child calculations?
EDIT 2: Also works if I set the parent mesh’s position in a setTimeout
: Physic impostor changes? | Babylon.js Playground (babylonjs.com)
Thanks @carolhmj for looking into it. I was hoping to try out my little project in v5 and this could be nice way around before fix. Also should help find out whats the issue.
Also if you call physicsRoot.refreshBoundingInfo() after adding its children, before setting its position, then it works without the timeout too.
https://playground.babylonjs.com/#FD65RR#101
There was a recent Issue with ribbon physics that smells related. @Cedric might know?
Let me take a look. I think something went wrong with compounds and my latest change.
PR