Manifesto

The Wordcel Manifesto

Water flows to you, moves around obstacles, and belongs to no one. Software should be the same: yours to hold, free to run, and shaped by the people who use it โ€” not a gate you must pass through, owned by whoever holds the keys. Three commitments follow from that.

01

Own the software, completely.

Totally owned. Totally scrutable. Costs nothing. Runs almost anywhere, asking for almost nothing.

Software you depend on should be entirely yours โ€” free to run, free to fork, and open to the last line. If you cannot read what it does, you are trusting a stranger with your work and your data. If it costs money to run or demands administrator rights, a gatekeeper already stands between you and your own tools.

The ideal is a single file you can open, read, and double-click: no install, no account, no server, no elevated permission it did not earn. It should run on the modest laptop, the locked-down work machine, the borrowed computer โ€” with nothing but a browser and the file itself. Ownership you cannot inspect is not ownership; it is tenancy.

02

Keep the data local. Share it like belts.

Local-first, on your disk, working offline โ€” and merged with friends, hand to hand.

Your data should live with you first. It should work with no network, survive being copied and edited on a plane, and leave your machine only when you decide to send it. Collaboration is not surrendering the archive to a central host; it is people handing each other deltas.

Think of waumpaum belts โ€” the wampum that memorialized agreements between nations: a signed record passed from hand to hand, woven into a shared truth without anyone opening their whole books to the other. That is the shape of it. CRDTs make the merge converge no matter the order; cryptography makes it safe to trade; and you stay the custodian of your own copy the entire time. Consensus is something a group reaches, not something a server holds for them.

03

Distrust centralization that isn't a community merge.

Concentrated compute and storage, absent living collaboration, is bad most of the time.

Some good things only emerge when many people converge โ€” a shared ledger, a common index, a settled agreement. But those are earned by merging: the active, consenting collaboration of a community, each party keeping their own copy and contributing their own deltas. That kind of centralization is a river fed by many streams.

Centralization that is not that โ€” compute and storage pooled under one operator because it is convenient, or profitable, or habitual โ€” will be bad most of the time. It concentrates cost, surveillance, and control; it manufactures single points of failure and single points of coercion; it makes one keyholder worth breaking. Prefer the federation of peers who choose to merge over the platform that quietly comes to own you. When in doubt, push the compute and the keys back to the edges, to the people.

Own it. Read it. Run it anywhere. Keep your data close and trade it freely with the people you trust โ€” and be slow to hand any of it to a center that did not earn it by merging.

These commitments are the through-line of everything here โ€” see the projects and the dev log for what they look like in code.