Research programs are the unit (see /lab/, rendered by lab.html).
_data/series.yml— single source of truth per program: question, status, outcome, verdict scoreboard, frontier, lineage (grew_from), code path, and the ordered parts and appendices. The Lab page, series TOC, prev/next nav and home cards render from it._posts/— program parts. Part 1 is the program's only entry in the home feed; Parts 2+ carryhidden: true(dropped from pagination only; archives/tags/atom keep them)._appendices/— theory appendices (series_part: A1,B3, …; the letter is the program)._articles/— formal articles; link one to its program withprogram: <series id>.research/<program>/— code and data behind every number.
Building image
docker build -t blog .
Serve blog from root dir
docker run --rm --volume="$PWD:/srv/jekyll" -p 4000:4000 -it blog jekyll serve --incremental
Open browser at http://localhost:4000/
With --drafts it will serve *.markdown files from _drafts folder.
docker run --rm --volume="$PWD:/srv/jekyll" -p 4000:4000 -it blog jekyll serve --drafts --incremental
docker run --rm --volume="$PWD:/srv/jekyll" -it blog \
bundle exec htmlproofer ./_site \
--only-4xx \
--ignore_urls "/example.com/,/ws-na.amazon-adsystem.com/,/molpharm.aspetjournals.org/" \
--ignore-status-codes "403"
Open sourced under the MIT license.