"At least for the people who send me mail about a new language that they're designing, the general advice is: do it to learn about how to write a compiler"
About this Quote
There’s a dry, almost surgical humility in Ritchie’s advice: if you’re writing to announce your shiny new programming language, the best payoff probably isn’t the language at all. It’s what you’ll be forced to learn while trying to make it real. He’s puncturing a familiar tech-romance fantasy - that inventing a language is primarily an act of visionary design - and replacing it with an engineer’s view: the hard part is the machinery, the semantics made executable, the brutal accounting of edge cases.
The line “people who send me mail” does quiet work. It sketches a recurring type: the earnest amateur seeking benediction from a founding figure. Ritchie answers without cruelty, but with a controlled skepticism that reads like experience talking. He doesn’t say “don’t do it.” He reframes the project away from ego and toward apprenticeship. If you want to build a language, fine - treat it as a structured way to encounter parsing, type systems, code generation, optimization, tooling, and the constraints of actual hardware and operating systems. That’s where the truth lives.
Context matters: Ritchie helped create C and Unix, systems that won because they were useful, portable, and implementable, not because they were rhetorically perfect. His subtext is a warning against novelty-for-novelty’s sake and a reminder that computing advances through working artifacts. Design is cheap; compilers are where your ideas are audited.
The line “people who send me mail” does quiet work. It sketches a recurring type: the earnest amateur seeking benediction from a founding figure. Ritchie answers without cruelty, but with a controlled skepticism that reads like experience talking. He doesn’t say “don’t do it.” He reframes the project away from ego and toward apprenticeship. If you want to build a language, fine - treat it as a structured way to encounter parsing, type systems, code generation, optimization, tooling, and the constraints of actual hardware and operating systems. That’s where the truth lives.
Context matters: Ritchie helped create C and Unix, systems that won because they were useful, portable, and implementable, not because they were rhetorically perfect. His subtext is a warning against novelty-for-novelty’s sake and a reminder that computing advances through working artifacts. Design is cheap; compilers are where your ideas are audited.
Quote Details
| Topic | Coding & Programming |
|---|
More Quotes by Dennis
Add to List



