A talk in which Hoare reflects on his 1965 invention of the null reference, calling it a mistake that caused decades of costly software errors.