What are the differences between rigging with bones using bones constraints or just using object constraints?

Question

I'm specifically talking about mechanical rigging here.
I understand bones and armatures are necessary in organic forms where there is deformation in the object, but for mechanical rigging where no parts deform, but only move, does using bones help any?
To me it seems simpler to just use object constraints when rigging mechanical things.

What are the differences between using bones or object constraints?
Are there any benefits bones have over object constraints?
Is there anything that can only be done with one method?

Answer 1

If it's just separate, undeformed meshes you're using, you could use object-level constraints. But there are even some benefits to using Pose-space for mechanical stuff as well:

1. It makes things easier to revise. If you decide to change the mesh after you've constrained it, you may run into some issues if say, the pivot point has changed on that object. Same thing if you've parented it to another object instead of constraining. If you had a bone deforming one object (and then another bone constrained to that bone) you now have a revisable system.

2. It allows for some pose-based tools, such as saving poses. Granted, this benefit is only useful if you find you want to re-use key positions a lot. If it's just a piston or something fairly cyclical, pose-saving isn't all that useful.