Floating rigid bodies: a note on the conservativeness of the hydrostatic effects