What if showing up on a map of the world's food ecosystem required nothing more than a browser tab?

No signup form. No backend to provision. No central database deciding who counts as a "real" store or farmer. Just a click on a map, a name, and you're on it, broadcasting to anyone else looking at the same map, anywhere in the world.

That's the Food Freedom Map, and as of today, it's open source.

👉 github.com/Djowda/food-freedom-map

What it actually does

Open the map, click anywhere, and you're prompted to drop a pin as one of ten roles — Store, Farmer, Restaurant, Factory, Wholesale, Transport, Delivery, Seed, User, or Admin. Pick an avatar, mark yourself online or offline, and optionally publish a small catalog:

• Listing — items you're selling, with prices
• Ask — items you're looking for
• Donation — items you're giving away for free

That's the whole interaction. No account creation, because there's nothing to create an account with — your identity is a keypair generated in your browser the moment you load the page.

Why there's no backend

The map is a reference client for DIFP, an open protocol for food coordination, built on top of Nostr — the same decentralized protocol underpinning a chunk of the social web right now.

Concretely, that means:

• Identity is a Nostr keypair (Schnorr/secp256k1), generated client-side and kept in localStorage. No server ever sees a password, because there isn't one.
• Location isn't raw GPS. Coordinates get mapped onto a 500-meter spatial grid (we call it the MinMax99 grid internally) before they're published, so pins resolve to a consistent cell ID rather than an exact point.
• Data is just signed Nostr events — a "pin" is event kind 30420, a catalog is kind 30421 — published to a public relay and read back by any client that knows how to ask. We ship a minimal ~150-line Nostr client to do exactly that; no SDK, no dependencies beyond a vendored Schnorr-signing library.
• The whole frontend is static files. HTML, CSS, vanilla ES modules, and Leaflet for the map. git clone, point a static file server at it, done. No npm install, no build step.

If you've ever wanted to see what "social network with zero infrastructure cost" looks like in practice, this is a pretty small, readable example of one.

Try it locally

git clone https://github.com/Djowda/food-freedom-map.git
cd food-freedom-map
python3 -m http.server 8080
# or: npx serve .

Open http://localhost:8080, click the map, drop a pin. That's it.

Make it yours

The repo ships with a CUSTOMIZING.md walking through how to fork this into your own instance:

• Swap in your own product catalog (data/products.csv + matching images in src/products/)
• Add or relabel the component types and their avatars
• Re-theme it — every color is a CSS variable
• Point it at your own Nostr relay instead of the public default

None of it requires touching the core logic. The spatial grid math, the Nostr client, and the DIFP catalog encoder are all separated out into their own small modules specifically so you can leave them alone and just edit config.

Where it's rough — and where you can help

We're shipping this as a genuinely early prototype, not a polished v1. A few known gaps, if you're looking for somewhere to dig in:

• The protocol supports a phone field and a workTime field on every pin, but there's no UI to actually fill them in yet — they're sent as empty/default values.
• A handful of leftover imports from an earlier version of the scanning logic are still sitting in app.js, unused.
• There's exactly one public relay hardcoded as the default — multi-relay support (read from several, publish to several) would make the network meaningfully more resilient.

None of these are hard, they're just undone. PRs, issues, and forks are all welcome — github.com/Djowda/food-freedom-map.

The bigger picture

This map is one small, self-contained piece of what DIFP is trying to be: a way for food infrastructure — farmers, stores, co-ops, food banks, delivery networks — to coordinate across borders without depending on any single company's servers staying online. A pin-drop map is about as small as that idea gets. It's also, we think, a pretty good place to start poking at the protocol if you want to understand it by reading code instead of a spec.

Fork it, break it, tell us what's missing.