Previously FbxImportRequestHandler used to be activated as part of DllMain init and never had CreateDescriptor called, which meant reflect was not called. BehaviorComponents get created and Activated as part of special SceneCore logic. Since this component now needs to be activated as part of the normal flow, reflecting it caused it to be picked up by the SceneCore activate logic, causing it to be created/activated twice
* Helios - LYN-3250 - Fixed morph targets for meshes that had multiple materials (#374)
Fixed morph targets for meshes that had multiple materials and were split by AssImp: Recombined them into one mesh in the O3DE scene graph, so the behavior would match FBX SDK.
* This simple fix pops any node that can not be registered using it name and unique ID
* Any scene file (i.e FBX) with the three or more nodes with the same name and parent will sort of work
* After the 2nd child node with the same name and parent will be dropped on export
Tests: added the new AutomatedTesting/Assets/BadAssets/three_same_named_nodes.fbx to regress test this
ATOM-15465: Fix issue with incorrect rotation on models.
* Revert "{LYN-3229} Update AssImp package with latest AssImp 3rd party source changes (#545)"
This reverts commit 06d2050ac4fe6b9d0d0c52e759cf73a9a3f7d3eb.
* bumping version to force assets to reprocess
* AssImp set to be the default FBX processor (#78)
If you encounter issues, reach out to the Helios team with details, and then switch back to FBX SDK locally by changing FBXImporter.h
* Merge pull request #219 from aws-lumberyard-dev/sceneapi_scripting
LYN-3030: Fix to export_chunks_builder.py to match the AssImp node paths
* Hide some automated test folders from the asset processor for automatedtesting. This is necessary because these assets should only be processed when running the test, and not any time AP is launched for this project.
* Putting these test assets back to visible to Asset Processor. These tests need to be updated at some point to handle that, but they won't work with this change for now. Note that until this is addressed, these tests may randomly time out if they're the first tests run on a clean asset processor, and these tests launch AP without using a unique port, so the test can also fail if an Asset Processor executable is hanging open.
Grabbed the change I missed from the 1.0 branch merge, no idea how this got lost.
* Moved from main to periodic. Allen and Fuzzy were already on board, and I think with the potential flakiness in this test, we don't want this in main.
Co-authored-by: jackalbe <23512001+jackalbe@users.noreply.github.com>
* Added error reporting for data integrity issues for non-uniform motion data.
* The actual issue was a mismatch between the end times of the morph and the skeletal animations. They need to match in EMotionFX.
* The morph target animation exported a keyframe too much.