Interesting that in these all comments, nobody referred to the (un)reality of the creation of proofs. The fact that the proof has to be longer than a program -

Interesting that in these all comments, nobody referred to the (un)reality of the creation of proofs. The fact that the proof has to be longer than a program - ignored completely 40 years ago - undermines the generality of the "purely scientific" approach marketed by Dijkstra. While his critics of software business are fundamentally sound - today even more than in the '80s - his idealistic view of programming as scientific activity and negation of the importance of (to not say contempt to) software engineering weakens his message even if we keep in mind the creation date of his philippic.