VECTOR — Self-Hosted Flight Logbook

A private, self-hosted flight logbook that replaces ForeFlight Logbook while staying fully compatible with its CSV format. Import your existing ForeFlight export, add/edit/delete flights, and export back to ForeFlight any time — no lock-in. Click any flight to see its route on a map: a great-circle line for direct airport-to-airport legs, or the full waypoint path when a route is entered.

• Frontend: React + TypeScript + Vite, Leaflet maps
• Backend: Python + FastAPI + SQLite
• Packaging: one Docker image (SPA served by the API)
• Auth: none in-app — designed to sit behind a Cloudflare tunnel / Access

Light and dark themes (with an auto option that follows your system).

Features

• Lossless ForeFlight CSV import/export (every column preserved; re-imports dedupe)
• Full CRUD for flights and aircraft, persisted in SQLite
• Per-flight map with great-circle distance, or multi-leg route polyline
• Worldwide coordinate resolution from bundled, auto-refreshed datasets — airports + navaids (OurAirports) plus multi-region enroute fixes (FAA NASR, OpenAIP, open flightmaps), with proximity-based disambiguation of colliding identifiers (e.g. SLI → Seal Beach, not the Colombian NDB)
• Totals dashboard (time by year / aircraft, PIC, night, XC, instrument, landings)
• FAA and Transport Canada currency tracking computed from your logbook
• Responsive desktop + mobile UI

Run it

docker compose up -d --build

Open http://localhost:8000, go to Import, and drop your ForeFlight logbook.csv. Your data lives in the logbook-data Docker volume; export any time from the Import/Export page.

Configuration (environment variables)

Variable	Default	Purpose
LOGBOOK_DATA_DIR	/data	Where the SQLite DB is stored
LOGBOOK_REFRESH_INTERVAL_DAYS	7	Coordinate dataset refresh cadence
LOGBOOK_TILE_URL_LIGHT	CARTO light	Map tiles for the light theme
LOGBOOK_TILE_URL_DARK	CARTO dark	Map tiles for the dark theme

Coordinate datasets are baked into the image (so first boot works offline) and refreshed on the schedule above; you can also trigger a manual refresh from the Import page.

Put it behind your domain with cloudflared

The app has no login of its own. Gate access with a Cloudflare Tunnel, and (recommended) a Cloudflare Access policy so only you can reach it.

Quick tunnel (testing):

cloudflared tunnel --url http://localhost:8000

Named tunnel (persistent, your domain):

cloudflared tunnel login
cloudflared tunnel create vector-logbook
cloudflared tunnel route dns vector-logbook logbook.example.com

~/.cloudflared/config.yml:

tunnel: vector-logbook
credentials-file: /root/.cloudflared/<TUNNEL_ID>.json
ingress:
  - hostname: logbook.example.com
    service: http://localhost:8000
  - service: http_status:404

cloudflared tunnel run vector-logbook

Then in the Cloudflare Zero Trust dashboard, add an Access application for logbook.example.com with a policy limited to your email. The app trusts that Cloudflare Access has authenticated the request.

Public demo (Cloudflare Pages)

A read-only demo with five redacted sample flights (no real personal data) can be deployed to Cloudflare Pages — a static build with no backend:

cd frontend
npm run build:demo          # builds with VITE_DEMO=1 → frontend/dist
npx wrangler pages deploy dist --project-name=vector-logbook-demo

In demo mode the app swaps the API for a bundled fixture (src/demo.ts): import/export, dataset refresh, and all add/edit/delete are disabled. The included GitHub Actions workflow (.github/workflows/deploy-demo.yml) deploys it on push — set the CLOUDFLARE_API_TOKEN and CLOUDFLARE_ACCOUNT_ID repo secrets. (public/_redirects handles SPA client-side routing on Pages.)

Coordinate datasets & the fixes pipeline

• Airports & navaids: OurAirports (CC0), fetched at build time and refreshed at runtime — worldwide, keyed on ICAO identifiers. Collisions (e.g. SLI) are disambiguated by proximity to the flight's airports, so the same logic works for routes anywhere in the world.

• Enroute fixes (multi-region): one fixes.csv is built centrally by .github/workflows/update-fixes.yml (running backend/scripts/build_fixes.py) and published as a release asset, so no instance ever pulls the large upstream datasets. It merges three sources, each tagged in a source column:

  • FAA NASR — US IFR enroute intersections (~70k; the subscription is a ~250 MB zip behind a WAF, fetched once in CI).
  • OpenAIP — worldwide VFR reporting points (per-country GeoJSON exports, CC BY-NC-SA).
  • open flightmaps (OFMX) — designated points for Europe and other covered regions (AIRAC snapshots discovered dynamically).

  The build validates the merge before publishing (assert_bundle_ok): if it is empty or the FAA baseline is missing, the job fails and the previous release stays latest, so there is always one working bundle. Point your instance at it (and the static demo / PWA at the same URL via VITE_FIXES_URL at build time):

  LOGBOOK_FIXES_URL=https://github.com/<owner>/<repo>/releases/latest/download/fixes.csv
  

  A bundled snapshot in the image is the offline / first-boot fallback.

  Coverage: IFR enroute intersections are complete for the US (FAA) and partial for Europe (open flightmaps); VFR reporting points are worldwide (OpenAIP); airports and navaids are global. Canada/Australia IFR enroute fixes have no open machine-readable source yet, so they are not included.

Development

Backend:

cd backend
uv sync
uv run uvicorn app.main:app --reload      # http://localhost:8000
uv run pytest                             # tests

Frontend (proxies /api to the backend on :8000):

cd frontend
npm install
npm run dev                               # http://localhost:5173
npx vitest run                            # tests

ForeFlight CSV format

A single CSV with two sections — an Aircraft Table and a Flights Table, each preceded by a marker row and a header row, separated by blank lines, with a required banner row at the top. The importer matches columns by header name (so it tolerates ForeFlight version differences) and is forgiving of missing From/To, quoted remarks, semicolon-delimited crew fields, and TRUE flags. Export reproduces the same structure for clean round-tripping.

Data & licensing

The bundled fixes.csv combines three sources: FAA NASR data (US public domain), OpenAIP reporting points (CC BY-NC-SA 4.0 — attribution and ShareAlike required, non-commercial use only), and open flightmaps designated points (OFMA General Users' License — attribution required). Because the bundle includes OpenAIP data, the dataset as a whole is CC BY-NC-SA: non-commercial use, with attribution, and any redistribution must keep the same license. See backend/datasets/bundled/ATTRIBUTION.md for the full source-by-source breakdown.

Currency calculations are an aid, not legal authority — always verify against the regulations (14 CFR Part 61 / CARs Part 401).