What I Learned About Rust by Creating a Language
summary
The talk describes building several toy languages in Rust: kLisp, a Standard ML→Wasm compiler called WebML, and chema. It explains concrete lessons: kLisp revealed ownership issues fixed by RC vs Box, lifetime-driven choices between by-value and by-reference designs, and consolidating error handling into a single Error enum; WebML used Nom v2 with dozens of macros which made parsing and debugging hard until impl Trait and module improvements; chema is shipped as a CLI using StructOpt and Trust to cross‑compile and publish working binaries.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:08Rust background
- 3:23Three Rust languages
- 4:09Kappa Lisp ownership
- 6:40Lifetime challenges
- 8:10Error handling
- 10:24WebML compiler
- 12:15Module system
- 13:41Parser combinators
- 14:03Type system in SML
- 16:25Rust efficiency
- 16:48Keema schema goals
- 17:37CLI design
- 18:42Travis automation
- 20:17Lessons learned
- 20:40Rust’s future
related talks
36mRusty Runtimes: Building Languages In Rust
29mLearning Rust by accident
01h 14mRust: A Friendly Introduction
01h 42mIntroduction to Rust (for people who have never used a compiler)
35mLet’s Talk About Rust
27mClosing Keynote
01h 04mRust-ifying Your C# Codebase: A Tale of Adventure and Transformation
29mMacromancy: an introduction to Rust macros