November 3, 2025
Docs vs drama: Rust sparks a riot
Writing an Asciidoc Parser in Rust: Asciidocr
Rust-built doc tool drops, Typst fans cheer, Pandoc diehards demand answers
TLDR: An indie dev released a Rust-based AsciiDoc parser to avoid Ruby and gain speed. Comments split between Typst fans cheering and Pandoc loyalists asking “why not a filter?”, turning a small tool into a big debate over building fresh versus using the one converter everyone already knows.
Someone just built a document tool in Rust, and the comments turned into a popcorn event. “asciidocr” tackles AsciiDoc (plain‑text markup for docs) without relying on Ruby. The author’s “petty beef” with Ruby and joking “allergic” reaction to Go’s error style became meme fuel, while a hand‑made parser flexed speed, control, and “I’ll learn it myself” energy.
Reactions split fast. One crowd shouted “love it!” but admitted they’ve moved on to Typst, the trendy typesetting tool, with pbronez dropping a friendly nod and a TIL (today I learned) about HTMLBook. Another camp played gatekeeper with a challenge: why not just use Pandoc? Pandoc is the all‑in‑one document converter; if you can write a filter, why reinvent it? Rust fans cheered a single, fast app you can share without installing extras, while pragmatists asked for a clear “why not Pandoc” answer.
Jokes flew: “if err != nil” became “if drama == inevitable,” and folks riffed on the eternal developer romance—speed vs convenience. Verdict? A tiny Rust tool sparked a big debate: build fresh for control, or plug into the Swiss‑army converter and call it a day.
Key Points
- •The author created a Rust-based AsciiDoc parser named “asciidocr.”
- •AsciiDoc began as a Python project and gained features via the Ruby-based Asciidoctor converter.
- •The author wanted a compiled tool to avoid Ruby/Python runtime dependencies and improve performance.
- •They considered subprocessing to Ruby, using asciidoc.py, or building/finding a converter in another language.
- •They chose Rust and implemented a hand-written parser instead of using a parsing library like pest.