Remarks on density functional theory as a thermodynamics