It’s often useful to make progress on a problem with fewer constraints before going back to the full-blown problem.