The Gardeners' Almanac

The Gardeners' Almanac is a progressive web app that helps gardeners in a specific neighborhood or city coordinate planting times, share seedlings, and warn each other about pests. It uses a combination of crowdsourced data and public APIs (like weather and frost dates) to generate personalized planting schedules. The app is designed to be self-hosted on a cheap serverless stack (Cloudflare Workers, KV, D1) so it has near-zero running cost. Users can mark their gardens on a map, list what they're growing, and offer surplus seeds or seedlings. The seed swap is a simple listing with a request-and-coordinate workflow, no payment involved. The whole app is open-source, so any community can deploy their own instance. The interface is deliberately minimal, focusing on quick data entry and clear information. An offline mode (via service worker) allows gardeners to access their schedules even without signal, which is essential for remote plots. The app respects privacy: no tracking, no ads, no data selling. All data belongs to the users and can be exported at any time.

Big gardening apps are either generic (not locally specific) or require a subscription. Local gardening clubs are great but often rely on email lists and spreadsheets that are messy and not real-time. Gardeners need a lightweight, free tool that combines local knowledge (like 'Mrs. Smith always plants tomatoes after Mother's Day') with easy digital coordination for seed swaps and pest alerts. Existing solutions are either too broad (e.g., Google Maps with markers) or too commercial (e.g., Seed Savers Exchange).

• Provide accurate planting dates based on local frost dates and microclimates, with room for user adjustments.
• Enable seed swapping with minimal friction: list, request, coordinate via in-app chat or email.
• Allow users to report pests and diseases in real time with location tags.
• Keep the entire app free and open-source, with a clear path for others to deploy their own version.
• Ensure the app works offline in gardens (PWA) to avoid data dependency.
• Build a reputation system (optional) to encourage honest swaps.

• No monetization, no ads, no premium tiers.
• No social media feeds, likes, or influencers.
• No e-commerce, no shipping logistics, no money handling.
• No large-scale data collection; user data stays on device or in user-controlled cloud.
• No native mobile app; just a PWA that installs on the home screen.
• No real-time push notifications (to keep costs zero; users check manually).

Frontend: SvelteKit (free, fast) with PWA capabilities. Backend/API: Cloudflare Workers (free tier up to 100k requests/day). Database: Cloudflare D1 (serverless SQLite, free tier up to 5GB). Storage: Cloudflare R2 (free 10GB) for user-uploaded garden photos. Maps: OpenStreetMap with Leaflet (free). Weather API: Open-Meteo (free, no API key needed). Deployment: GitHub Pages (frontend) and Cloudflare Pages (workers) – both free. Version control: GitHub (free private repo).

The app is a single-page application (SPA) with a service worker for offline support. The backend is a set of Cloudflare Workers that handle API requests. Data is stored in Cloudflare D1, a global SQLite database. The map component uses Leaflet with OpenStreetMap tiles. Garden profiles and seed listings are stored as JSON in D1. Weather data is fetched from Open-Meteo on the client side, cached in the service worker. The seed swap system uses a simple request-response model: a user lists a seed, others can request it via a simple form, and the owner coordinates pickup via email or in-app chat (stored in D1). No real-time push notifications (to keep costs low); users check manually. The app is designed to be deployed as a single Cloudflare Worker that serves the API, with the static frontend hosted on Cloudflare Pages. All user data is encrypted in transit and at rest via Cloudflare's infrastructure.

• Reliance on free tiers may hit limits if the app goes viral; but for a hyperlocal app, this is unlikely.
• User adoption chicken-and-egg: need a critical mass of gardeners to make the seed swap useful. Mitigation: partner with local gardening clubs.
• Data accuracy: frost dates from weather APIs are approximate; users may report inaccuracies. Mitigation: allow users to manually adjust dates and share local knowledge via comments.
• Offline mode limited: maps require tiles, so offline features are text-only (garden profile, calendar). Mitigation: cache text data and allow full offline creation of new entries (sync later).
• Spam or bad actors in seed swap. Mitigation: optional reputation system; users can block others.

• Should the seed swap include a reputation system? Not initially; trust is built via local community. Possibly add later if needed.
• How to prevent duplicate listings? Simple: each user can have one listing per seed type (e.g., 'Tomato, Brandywine'). Duplicates handled by merging via admin? No admin, users can mark 'claimed'.
• Should we include a 'Garden Journal' for daily logs like watering, harvest? Maybe in v2.
• How to handle multiple neighborhoods within a city? Use geofencing or allow users to define their own area.
• Should the app include a wiki for local gardening tips? Could be a separate static site linked.

I sketched it out over a weekend but realized my own community wasn't big enough to reach critical mass. I'd need to invest time in outreach, and I got busy with other projects. The plan is solid; maybe someone else will build it.

The app is deliberately minimal to keep costs zero. All code is open-source on GitHub under the AGPL license. The design is responsive and accessible. The seed swap is honor-based; no tracking. The PWA can be installed on iOS and Android. Weather data is cached for 6 hours to reduce API calls. The app uses a single Cloudflare Worker for all API endpoints, handling CRUD operations and search. Database queries are optimized with indexes on user_id, seed_type, and location coordinates.