Optimization as constraint: a comment on Mazzoleni and Nelson