Gizmo misc + disable sync physics #14907
Conversation
CedricGuillemet
commented
Mar 25, 2024
CedricGuillemet
commented
- disable sync update in sync with motion type for physics V2 (a continuation from this PR : https://forum.babylonjs.com/t/physics-performance-with-many-static-bodies/48402 )
- fix regression introduced Fix parent transformations being applied to position/rotation deltas … #14708 for sixDof behavior. Introduced bug is visible with this PG : https://playground.babylonjs.com/#AZML8U#229 each new start drag flips the mesh.
|
Please make sure to label your PR with "bug", "new feature" or "breaking change" label(s). |
| @@ -199,7 +199,7 @@ export class SixDofDragBehavior extends BaseSixDofDragBehavior { | |||
|
|
|||
| if (pointerCount === 1) { | |||
| this._targetPosition.copyFrom(this._ownerNode.absolutePosition); | |||
| this._targetOrientation.copyFrom(this._ownerNode.absoluteRotationQuaternion); | |||
| this._targetOrientation.copyFrom(this._ownerNode.rotationQuaternion); | |||
There was a problem hiding this comment.
Why is position and scaling stay absolute, and only rotation uses local rotation?
There was a problem hiding this comment.
I think it's the reverse. the PR did change position to absoluteposition to fix a bug and did the replacement for rotation and scale as well to be consistent. But actually, it introduced a regression with rotation. Do you confirm @carolhmj ?
There was a problem hiding this comment.
Multipointer was also changed to absolute, is it still an issue?
There was a problem hiding this comment.
Let's wait for @carolhmj 's input on this fix and if it's ok then I'll do and test the multipoint.
There was a problem hiding this comment.
I changed to absolutePosition/Rotation/Scale because I removed this line https://github.com/BabylonJS/Babylon.js/pull/14708/files#diff-867020a1c839af098f39ec5a7d63e25c7fa37dfd2e507e2143f86bafe643e85aL202 which was using setPosition(null) to make the node's position be the absolute position. If you copy only the local rotation, wouldn't that ignore the parent's rotation?
|
Snapshot stored with reference name: Test environment: To test a playground add it to the URL, for example: https://babylonsnapshots.z22.web.core.windows.net/refs/pull/14907/merge/index.html#WGZLGJ#4600 Links to test babylon tools with this snapshot: https://playground.babylonjs.com/?snapshot=refs/pull/14907/merge To test the snapshot in the playground with a playground ID add it after the snapshot query string: https://playground.babylonjs.com/?snapshot=refs/pull/14907/merge#BCU1XR#0 |
|
Visualization tests for WebGPU (Experimental) |
|
WebGL2 visualization test reporter: |
|
Weird enough, one of my open issues with WebXR dragging is the multipointer dragging which doesn't work as expected. I am debugging this now, but I don't want to cause any merge conflicts, so I might make the changes here or ask you to do them if possible. |
sure, np. let me know how I can help. if the mesh flips upside down with starting to drag, do the same for multipoint. |
|
just in case: flipping upside down is one of the artifacts this PR fixes. the other one I've seen is flying away (current tip of the branch 8ca757b does fix that too, but I just wanted to make sure you are aware of that): https://playground.babylonjs.com/#AZML8U#234 The model (2teaboxes.glb) that reproduces the flying away artifact has such a transform and geometry that the mesh is not near the origin. |
No, it is actually moving to a different position and sometimes scales incorrectly. still trying to find the best solution |
|
This pull request has been marked as stale because it has been inactive for more than 14 days. Please update to "unstale". |
