radio#

small rust + solidjs radio server. the rust backend owns the xrpcs, media, audio storage, embedded read-only pds, and production static serving; the solid frontend owns browser oauth and the ui.

setup & hosting#

To run your own radio station and have it syndicated (so it automatically shows up on directories like https://radio.wisp.place):

1. Install dependencies:

   cp .env.example .env
   npm --prefix frontend install
   cargo check
   

2. Expose it publicly: Your station must be publicly reachable over standard HTTPS on a public domain name so ATProto relays and directory workers can crawl it. Setting up a reverse proxy (like Caddy or Nginx) or routing traffic through a Cloudflare Tunnel is the easiest way to expose it.

3. Configure your .env: Configure your public domain endpoints (you can use the exact same domain name for all of them!) and add your ATProto DID to the admin allowlist (comma-separated for multiple admins):

   APP_URL=https://radio.yourdomain.com
   SERVICE_DID=did:web:radio.yourdomain.com
   STATION_URL=https://radio.yourdomain.com
   STATION_ANNOUNCE_WORKERS=https://syndication.sharkgirl.pet
   ADMIN_DIDS=did:plc:abc123xyz...
   

When the backend starts up, it will automatically register your station's metadata and notify the syndication workers to crawl it.

You can also trigger a manual crawl announcement directly to the public worker:

curl -X POST https://syndication.sharkgirl.pet/xrpc/com.atproto.sync.requestCrawl \
  -H "Content-Type: application/json" \
  -d '{"hostname": "radio.yourdomain.com"}'

syndication worker#

If you want to run your own syndication worker directory instead of using https://syndication.sharkgirl.pet:

1. Configure environment:

   cp syndication-worker/.env.example syndication-worker/.env
   # Edit syndication-worker/.env as needed
   

2. Run it:

   cargo run --manifest-path syndication-worker/Cargo.toml
   

It listens on http://127.0.0.1:3300 by default. You can query its status and endpoints:

• GET /health - Overall health check.
• GET /stations - List of all crawled healthy stations.
• GET /stations/{did} - Detailed information on a specific station.
• POST /xrpc/com.atproto.sync.requestCrawl - Request the worker to crawl your public station.

development#

run the backend:

cargo run

run the frontend dev server:

npm --prefix frontend run dev

vite proxies /api, /xrpc, /.well-known, and /client-metadata.json to http://127.0.0.1:3000 in dev.

production build#

npm --prefix frontend run build
cargo run --release

npm --prefix frontend run build writes the normal vite output to frontend/dist, then copies it into the backend static/ directory. the rust backend serves static/ and falls back to static/index.html for frontend routes.

standalone frontend build#

To build a standalone frontend client (e.g., to host on a static hosting provider like Wisp or Netlify without running a local backend):

1. Run the build command with the standalone environment variables set:

   VITE_STANDALONE=true \
   VITE_BASE_URL=https://radio.yourdomain.com \
   VITE_SYNDICATION_WORKER_BASE=https://syndication.sharkgirl.pet \
   npm --prefix frontend run build
   

2. Upload the output directory frontend/dist/ to your static host.

Standalone Build Variables#

• VITE_STANDALONE: Set to true to enable standalone mode. This disables the local station entry and enables querying syndicated stations from the worker. It also prevents the build script from copying files into the backend /static directory.
• VITE_BASE_URL: The public URL where this client is hosted (e.g., https://radio.wisp.place). Used to generate the static OAuth client-metadata.json.
• VITE_SYNDICATION_WORKER_BASE: The HTTP base URL of the active syndication directory worker (e.g., https://syndication.sharkgirl.pet) to fetch live stations.

useful paths#

• src/ rust backend
• frontend/ solidjs frontend
• static/ frontend bundle served by rust
• data/audio/ default uploaded audio storage
• migrations/ sqlite migrations

environment reference#