---

Mapping the Doctrine of Discovery Podcast

The official website repository for the Mapping the Doctrine of Discovery Podcast.

Listen now

Build

This site is built with Build Awesome (Eleventy) and Liquid templates. Config lives in buildawesome.config.js.

npm install
npm run build
npm test

Local preview runs on http://localhost:8080/ either way:

npm run dev
npx @awesome.me/buildawesome --serve

The build writes to _site/. The build also generates citation downloads and podcast.xml after Eleventy finishes. podcast.xml is built from Buzzsprout when the network is available and falls back to scripts/cache/buzzsprout-feed.xml for offline or restricted builds.

Site data (episodes, authors, pages, categories, tags) is assembled once by scripts/site-data.mjs and cached as JSON in .cache/site-data.json by scripts/site-data-cache.mjs. The cache is keyed on a fingerprint of every source file the builder reads, so edits invalidate it automatically and the config plus each postbuild/check script reuses the parse instead of re-walking src/. npm run clean removes _site/ only; delete .cache/ by hand if you ever need a forced rebuild.

Dependency install scripts are gated by the allowScripts field in package.json (npm 11.9+). esbuild, fsevents, and workerd are approved because they unpack platform binaries. sharp is denied: it arrives only through miniflare (a wrangler dependency this project does not run), its prebuilt @img/sharp-* binary installs regardless, and its install script tries a from-source libvips build on machines that have Homebrew libvips.

npm test runs the production build plus checks for required pages, ESM-only project code, feeds, referenced assets, internal links, redirects, tag pages, episode metadata, citations, and PDF metadata.

PDF metadata checks use pdfinfo and qpdf. Updating PDF metadata uses Ghostscript through npm run update:pdf-metadata.

Source Layout

• src/ contains Eleventy pages, layouts, includes, and data.
• src/_config/ contains Eleventy config modules for filters, collections, global data, Markdown setup, URL helpers, and passthrough copy.
• src/common/ contains generated common outputs such as feeds, sitemap, robots, redirects, and humans.txt.
• src/content/episodes/ contains podcast episode posts.
• src/content/authors/ contains author profiles.
• src/_data/site.yml contains site metadata that used to live in the legacy root _config.yml.
• public/ contains static files copied to the site root, including /assets/....
• scripts/ contains Node build helpers and verification checks.

Checks

npm run check:esm
npm run check:build
npm run check:headers
npm run check:feeds
npm run check:assets
npm run check:links
npm run check:redirects
npm run check:metadata
npm run check:citations
npm run check:pdfs

The asset check reports large public files for review. It currently flags public/assets/pdfs/Episode-05-Doctrine-of-Discovery-Abya-Yala-Tupac-Enrique-Acosta.pdf as the largest reviewed file.

URLs and Tags

Canonical tag pages live at /tags/ and /tags/{slug}/, for example /tags/haudenosaunee/. Old /tag, /tag.html, and special episode paths such as /special/episode-06/ generate redirect pages to the canonical paths. Search Console URL variants are handled through redirects when they represent legacy HTML or asset filenames.

Special episodes use /special/s01/ through /special/s08/.

Metadata

Episode pages render h-entry markup and all 15 Dublin Core elements: title, creator, subject, description, publisher, contributor, date, type, format, identifier, source, language, relation, coverage, and rights. Site identity, author pages, and episode author references use h-card markup.

Default podcast metadata is defined in scripts/site-data.mjs and can be overridden in episode front matter with fields such as guests, contributors, duration, transcript_pdf, coverage, source, and relations.

Citation downloads are generated for every episode as RIS and CSL JSON under /assets/citations/. PDF metadata is updated from the episode model and should include title, subject, author, keywords, CC-BY 4.0 rights text, the license URL, and basic bookmarks.

Deployment

GitHub Actions run on Node 24.14.0, matching .node-version and the engines.node floor in package.json.

• cloudflare-deploy.yml runs npm ci and npm run test:deploy, then deploys the generated _site/ directory to Cloudflare Workers.
• pdf-checks.yml installs Poppler and qpdf and runs npm run check:pdfs when PDFs or the scripts that verify them change.

Cloudflare Workers static assets allow at most 100 _headers rules. scripts/postbuild-headers.mjs generates four wildcard rules, and npm run check:headers enforces that ceiling:

1. /*.xml — canonical Link for the feeds.
2. /season:season/* — one rule covering every season via the :season placeholder.
3. /special/* — the one non-season episode category.
4. /*metadata.json — application/ld+json, plus ! Link to strip the episode Signposting links from the JSON-LD documents themselves.

Each episode rule emits its three Signposting link-values as a single comma-separated Link field (RFC 8288), which is how Cloudflare serves them anyway, so a rule costs one header entry rather than three. Rule order matters: /*metadata.json must stay last for its ! Link to unset what the episode rules added. PDF and citation metadata lives in the generated PDF/RIS/CSL JSON files rather than per-file HTTP headers. wrangler.toml does not define a build command; CI deploys the already-built and verified _site directory.

Notes

This site began from a Stackbit Podcaster theme and later a Jekyll version. The current build is an Eleventy migration that preserves the public URL structure and visual output of the previous site.

License

See LICENSE.