I saw you have a number of fixes around LayerManager. Some of them were reasons for us to create a new copy of LayerManager in the core library (e.g. we needed all anonymous layers to be serialized to Maya file and not only session layer). Do you think with the upcoming changes you could actually adopt core LayerManager? The big benefit of going this path would help with migration from AL to Adsk plugin, fixes will benefit both plugins...and ultimately remove code duplication.

Hi @kxl-adsk!

Again, this PR is part of AL's alignment process, and will likely have further changes applied on top soon. I think that adopting the shared LayerManager is moving the right direction, and I'll bring it up with the team. Similarly to #1731, let me know if there's anything concerning otherwise, but this purely reflects the effort to catch up with the latest internal changes in a somewhat linear manner.

I understand that this change only impacts the AL plugin. It also shouldn't make the transition to a common plugin any harder, so no concerns here. What makes me hesitate is the work you will have to put in adopting the shared LayerManager and I wouldn't want this to force us to keep both in the source tree. One way to make this transition easier would be to validate changes against the shared LayerManager and apply them to two places if possible/necessary.

Let's take this PR as an example - the same change can be made in https://github.com/Autodesk/maya-usd/blob/dev/lib/mayaUsd/nodes/layerManager.cpp#L137.

The benefit of paying this extra cost now and applying fixes to both LayerManagers is an easier transition to a shared LayerManager in (hopefully) the near future.

Can do @kxl-adsk. Will make the appropriate changes to the shared layerManager and add it to the PR.