[BUG-232869] Viewer code still not aligned to server code when displaying physics shapes

[sl-service-account]

What just happened?

Merging changes in Maint-N broke the previously working FS physics shape display due to a missing condition in the check.

What were you doing when it happened?

Fixing https://jira.firestormviewer.org/browse/FIRE-32359

Resizing a mesh that has been uploaded with a use hull physics shape will never trigger "convexification." Since the Maint-N release and specifically changes made for SL-2243, the viewer correctly convexifies meshes that have a triangular mesh physics when any dimension is <0.5m, but erroneously triggers this behaviour for meshes with user hull based physics.

The comments in the change seem to suggest that this was a deliberate choice. If so, I would argue strongly that it was the wrong choice because the viewer is not displaying the effective physics shape for those meshes. https://gyazo.com/eb56d368d1bcbd2104b96a0cf98de8c0

What were you expecting to happen instead?

The following videos show two versions of the same mesh model. The one on the left was uploaded with triangular mesh physics (i.e. not analysed), and the one on the right has a user-defined hull-based physics (analysed). Video 1 shows the incorrect behaviour as shown by LL viewer https://gyazo.com/0dd04d09838bec7fdbd9273070484d8b Video 2 shows the same but in Firestorm preview, post our Maint-N merge https://gyazo.com/a6db114cd50cc01ee36ece3a404454c3 Video 3 shows the expected behaviour. https://gyazo.com/5df7ebd54024e721722c82ebeccffbad

Other information

Pull Request For Fix: https://bitbucket.org/lindenlab/viewer/pull-requests/1187/change-determinephysicsshape-to-properly

Original Jira Fields

| Field | Value | | ------------- | ------------- | | Issue | BUG-232869 | | Summary | Viewer code still not aligned to server code when displaying physics shapes | | Type | Bug | | Priority | Unset | | Status | Accepted | | Resolution | Triaged | | Reporter | Beq Janus (beq.janus) | | Created at | 2022-11-04T16:55:52Z | | Updated at | 2022-11-08T19:12:16Z | ``` { 'Build Id': 'unset', 'Business Unit': ['Platform'], 'Date of First Response': '2022-11-08T13:12:16.249-0600', "Is there anything you'd like to add?": 'Pull Request For Fix: https://bitbucket.org/lindenlab/viewer/pull-requests/1187/change-determinephysicsshape-to-properly', 'ReOpened Count': 0.0, 'Severity': 'Unset', 'System': 'SL Viewer', 'Target Viewer Version': 'viewer-development', 'What just happened?': 'Merging changes in Maint-N broke the previously working FS physics shape display due to a missing condition in the check.', 'What were you doing when it happened?': 'Fixing https://jira.firestormviewer.org/browse/FIRE-32359\r\n\r\nResizing a mesh that has been uploaded with a use hull physics shape will never trigger "convexification." Since the Maint-N release and specifically changes made for SL-2243, the viewer correctly convexifies meshes that have a triangular mesh physics when any dimension is <0.5m, but erroneously triggers this behaviour for meshes with user hull based physics.\r\n\r\nThe comments in the change seem to suggest that this was a deliberate choice. If so, I would argue strongly that it was the wrong choice because the viewer is not displaying the effective physics shape for those meshes.\r\nhttps://gyazo.com/eb56d368d1bcbd2104b96a0cf98de8c0', 'What were you expecting to happen instead?': 'The following videos show two versions of the same mesh model. The one on the left was uploaded with triangular mesh physics (i.e. not analysed), and the one on the right has a user-defined hull-based physics (analysed).\r\nVideo 1 shows the incorrect behaviour as shown by LL viewer\r\nhttps://gyazo.com/0dd04d09838bec7fdbd9273070484d8b\r\nVideo 2 shows the same but in Firestorm preview, post our Maint-N merge\r\nhttps://gyazo.com/a6db114cd50cc01ee36ece3a404454c3\r\nVideo 3 shows the expected behaviour.\r\nhttps://gyazo.com/5df7ebd54024e721722c82ebeccffbad\r\n', 'Where': 'n/a - Second Life Server 2022-10-07.575585', } ```

[sl-service-account]

Dan Linden commented at 2022-11-08T19:12:16Z

Thank you for the report, Beq!