Comment on "Non-equilibrium entropy of glasses formed by continuous cooling" [J. Non-Cryst. Solids 355 (2009) 600]

We use the general statement of the second law applied to an isolated system, the glass in an extremely large medium, to prove that the entropy of the glass must decrease with time during its relaxation towards the supercooled liquid state. This result contradicts the claim of Mauro et al and their computational result of the entropy, according to which their entropy of the glass increases. Our approach using the isolated system completely bypasses the issue of ergodicity loss in glasses, as discussed in the comment.