VSJ – October 2002 – Sounding Board

Robin Jones mourns the death of the Father of Structured Programming.

Edsger Dijkstra died recently. On hearing the news, I was transported back almost thirty years to when I first read his ‘Notes on Structured Programming’. I’d noticed it because of an intriguing chapter heading: ‘On Our Inability To Do Much’. And then I became completely absorbed by it. It filled me with a disconcerting mixture of inspiration and terror. Inspiration because of its author’s quality of thinking and clarity of expression. Terror because it addressed head-on the apparently insuperable conundrum of scalability. He said, very simply, that if a program component has a probability, p, of being correct and there are n such components, then the probability of the program behaving correctly is pn. I remember idly calculating that this meant that half a million lines of code, each guaranteed to be 99.9999% correct, would yield a program that could have about a 40% chance of falling over.

Thirty years on, with the Press heavily exercised about NATS and the Child Support Agency system, I wonder whether we will ever be able to ‘do much’. On the other hand, if the Ancient Egyptians had done the maths first, would they have started a single pyramid?

[Something you’d like to get off your chest? Email me (Robin Jones) at eo@iap.org.uk.]

Comments are closed.