On the problem of the mathematical foundation of the Gibbs postulate ie classical statistical mechanics