mediapi/.env.example
Jean-Michel Tremblay ed9da9a065 Extract config into .env; add deploy.sh with overlay guard + rollback
- Move user-specific config (login, media roots, port, AP ssid/psk, service
  user) into a gitignored .env; add .env.example template.
- config.py reads .env at startup via a tiny zero-dependency parser (works for
  both `uv run` and systemd), and fails loudly if required vars are unset.
- load_secret_key falls back to an ephemeral key on a read-only filesystem
  instead of crashing.
- systemd units become .template files; deploy.sh renders them with the
  .env-derived user/paths.
- deploy.sh: refuses to run on a read-only overlay, git pull --ff-only,
  uv sync, ensure secret key, re-apply the AP from .env, install units,
  restart, health-check with automatic rollback to the previous commit.
- Scrub literal credentials/ssid out of the README; document .env + deploy.sh.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-05 22:06:47 -04:00

22 lines
674 B
Text

# Copy this file to .env and fill in real values.
# .env is gitignored -- never commit real secrets.
# Values must be simple (no spaces / shell-special chars) -- this file is both
# parsed by the app and sourced by deploy.sh.
# --- App login (custom login form) ---
MEDIAPI_USERNAME=changeme
MEDIAPI_PASSWORD=changeme
# --- Media ---
# colon-separated list of directories to browse
MEDIAPI_MEDIA_ROOTS=/localmedia
MEDIAPI_PORT=8080
# --- System: systemd services run as this Linux user ---
MEDIAPI_USER=pi
# --- WiFi access point (re-applied by deploy.sh) ---
MEDIAPI_WIFI_COUNTRY=US
MEDIAPI_AP_SSID=changeme
MEDIAPI_AP_PASSWORD=changeme
MEDIAPI_AP_CONN_NAME=mediapi-ap