This paper argues that our recent progress in the development of a sound programming methodology should not lead us to ignore the more difficult aspects of engineering; and that in future we should pay more attention to the quality of our designs, and not just the accuracy of their implementation.