PodPeopleDB

The open, community-built database of podcast hosts and guests: who's on which podcast, and where else they turn up. Live at podpeopledb.com and built into Pinepods.

• Reads the feeds. Shows that use the Podcasting 2.0 <podcast:person> tag are crawled and their credits kept in sync, channel- and episode-level.
• Fills the gaps with the community. Anyone can add people to a show or an episode: no account, name is the only required field, paste a list to add several. Additions go live immediately (marked unverified) and admins review them afterwards, with one-click revert.
• Open data. JSON API, full JSON dump, scrubbed SQLite download, and a changes feed for mirrors.

Running it

services:
  podpeopledb:
    image: madeofpendletonwool/podpeople_db:latest
    ports: ["8085:8085"]
    environment:
      BASE_URL: https://people.example.com
      ADMIN_USERNAME: admin
      ADMIN_PASSWORD: change-me
      SEARCH_API_URL: https://search.pinepods.online
      IP_HASH_SALT: change-me-to-something-random
    volumes:
      - ./podpeople-data:/app/podpeople-data

All settings are documented on the site at /docs/self-host.

Upgrading from v1 (Go): mount the same data folder. On first start v2 imports the old podpeopledb.sqlite into a new podpeople-v2.sqlite, keeping person ids (Pinepods stores them), admin passwords and pending submissions. The v1 file is never written to.

API

/api/v2 is documented at /docs/api. The v1 endpoints (/api/hosts/{id}, /api/host-search, /api/podcast/{id}, /api/discover/*) still answer in their old shapes for older Pinepods.

Development

Rust (axum, sqlx/SQLite, askama templates). Static assets and templates are compiled into the binary.

DB_PATH=./dev.sqlite ADMIN_PASSWORD=devpassword cargo run
cargo test   # unit + end-to-end tests, no network needed

Layout: src/feed.rs (RSS parsing), src/crawler.rs, src/store/ (queries), src/submit.rs (contributions), src/api/ (JSON API, v2 + legacy), src/web/ (pages + admin), templates/, static/, migrations/.

License

GPL-3.0