Comments on "Revision of the Buffer Length Derivation for a Modified Ek/D/1 System by Maritsas and Hartley