Comments on 'Reliability and performance analysis for fault-tolerant programs consisting of versions with different characteristics' by Gregory Levitin [Reliability Engineering and System Safety 86 (2004) 75-81].