A Thousand Org Files, One Parser: A Site You Can Query from bb
Table of Contents
A draft for the proposal shelf. Not submitted anywhere. Status: Keep.
1. Pitch
What breaks when every tool reads the files its own way, and what you get when none of them does.
2. Abstract
A site of 1,017 indexed org-mode documents is published with org-publish, and every tool around it reads the same files through one parser. Search, feed, sitemap, header lint, verification report and provenance map are six consumers of one data structure. The talk opens on the failures that come from tools parsing the files separately, shows two projections end to end, and adds a seventh on stage.
3. Audience, format, venue
- Audience
- Emacs and org-mode users; babashka users; anyone keeping notes or docs at scale.
- Format
- 20 minutes at EmacsConf; 25 minutes or a 10-minute cut at Babashka Conf.
- Where
- Babashka Conf (the 2026 edition gave 25 minutes to a static site generator), EmacsConf.
4. Review
Both reviews keep it and call it the best Babashka Conf fit. Both said the first version cited the wrong notes; the three above are the ones that state the thesis. The parser itself is not in a public repository, so the live demo has to carry the talk.