I think it's because physics is not updating transform anyway in blender.
Technically should use "object matrix input" instead. You can decompose it for location.
So the problem is on blender side?
So the problem is on blender side?
I don't know if I should call it a "problem"...
it may make sense why physics is not updating transform data.
Imagine if I were to animate its transform property before triggering the physics, and its transform data were to be changed due to physics afterwards, how will the graph editor looks like? It will also be weird if such information shows on transform but not on graph editor?
These are just personal thoughts, I don't know what maybe a better solution/proposal to this.
But as mentioned earlier, matrix input works fine in this case.
Maybe we should just adhere to this.
Thanks
Constraints, rigid bodies, and parenting act as modifiers on the object transforms. The object location is the actual location of the object, the object transform is the computed location after applying all constraints, simulations, and parenting. So this is the expected behavior as far as I know.
Most helpful comment
I think it's because physics is not updating transform anyway in blender.
Technically should use "object matrix input" instead. You can decompose it for location.