And then McIlroy said to Knuth... 2017-10-02
This old blog post about the contrasting approaches to programming to solve a particular problem shot past my RSS reader recently. It's a lovely read about a dialogue-by-article that occured between Donald Knuth and Doug McIlroy, as guest columnists in Communications of the ACM back in the day.
The post is short, wonderfully written, and serves as something of a meta-commentary about the nature of writing about code, and how to communicate the intent and the implementation of a computer program by way of documenting it. It has a wonderful flavour of a parable from the ages, because it's re-telling a story of how the giants from the old days solved a problem in a witty and entertaining way.
You could easily read this as a clash of ancient demigods with the victor being the last man standing, but I think that's probably a mistake. 'What problem are you trying to solve?' is one of my favourite pat-rules about program design, and I don't think the two authors here are trying to solve exactly the same problem.
What problem is literate programming trying to solve? Did it solve it? Are there any better ways to solve that? Wasn't UNIX designed for use in interactive text-processing?
It makes me think about man vs horse races . What problem are they trying to solve?