Thanks @Greydon_Gilmore for posting and @lassoan for your comments!
Up until v4.10.2, RAS has been the default output coordinate system which is what our original markup placement protocol was developed on. We started noticing issues when placements in v4.11 were displayed in versions < 4.11. That’s what led us to find this coordinate system change in markup files.
As a short term solution, we have made it pretty clear to our user base to only use up to v4.10.2 for placements.
The better solution would be to update our validator to perform more detailed checks of the header for the version and coordinate system. We created an issue for our project here: https://github.com/afids/afids-validator/issues/72
Is there a rough timeline for a stable version of Slicer v5.0 so our team can plan accordingly?
Note: Our Markup files encodes details about the L and R sides of the coordinate space (based on the location of anatomical features in the brain), which could be useful for migration test cases.