The Enduring Legacy of Dennis Ritchie (2012) [pdf]

"Dennis also had a deliciously wry sense of humor. He was once asked: In your experience, how long does it take for a novice programmer to become a reasonably proficient C developer capable of writing nontrivial production code?

Dennis replied: I don’t know. I never had to learn C."

For decades I have loved Ritchie's design aesthetic. Some years ago I met one of his close family members without realizing it and we bonded over our love of both classical studies and indigenous issues. At some point through random chats I talked about what an influence Ritchie was on me. My friend was shocked that I had heard of her relative whom her family considered as a pleasant eccentric, and she was gobsmacked when I told her stories about him that she had heard for countless years from the man himself over the holiday dinner table which everyone had rolled their eyes at. The concept outsiders had heard of him was astonishing. I was similarly amazed that my super smart friend was so close to him. As time proceeded I became something of friends with Dennis, not close, but someone he at least knew and we exchanged some amusing letters and gifts.

When I called my friend to tell her that Dennis had passed away it was the first she heard of it.

Not everyone likes Dennis' design aesthetic but it's been valuable to me in my career to find someone else, someone well known, who had similar predilections.

Dennis was genuinely a wonderful person who had excellent design taste. I miss him though I can still meet with him any time I like by thumbing through the K&R.

I 100% agree with the paper.

However, one should not mistake historical contributions with state of the art needed today. Unix and C are still largely relevant and will continue to be so for a long time, even more so their legacy, but one should not interpret that as if nothing better is needed now. For Unix derivatives, well they evolved and are already largely not the same thing as the historical versions. For the C programming language, it is a way more complicated situation. Of course it evolved too, but I would say not as much, especially in areas (safety) where we now have pressent needs.

C has contributed greatly to the world and will continue so. But don't forget that so does Cobol.

The comments on the MIT approach versus the New Jersey style is interesting considering that by '89 MIT's X Window System was a major contributor to the success of Unix Workstations and the X Windows "Principles" seem a lot like Unix'. https://en.wikipedia.org/wiki/X_Window_System#Principles and https://en.wikipedia.org/wiki/Worse_is_better

"Small is beautiful" -- combined with the principle that each program should do one thing and do it well you the key to improving so many irrationally ambitious software projects.

> At the insistence of Doug McIlroy, the early Unix systems had on-line manual pages making the system much easier to learn and use.

I’m reminded of one of my favorite parts of the early UNIX story, the fact that adding the “BUGS” section to man pages was motivation to fix the bugs instead of documenting them. Shame is a powerful motivator.

The requested URL was not found on this web server: /~aho/Talks/12-09-07_DMR.pdf

What a sublime and witty joke.