A note on the unified theory of least squares