why vote
Every open community in 2026 — Nostr relays, DAOs, grant committees, Bitcoin protocol debates, forum moderation bodies — eventually runs into the same problem: "how do we take a group decision without a token, without a KYC vendor, and without trusting a tallier?"
Existing systems fail at least one of the five gatekeeping questions:
| system | bot-resistant | weighted | offline-tally | secret ballot | no token / no KYC |
|---|---|---|---|---|---|
| google forms | ✗ | ✗ | ✗ | ✗ | ✓ |
| snapshot | ✗ | ✓ | partial | ✗ | ✗ (needs a token) |
| helios | ✓ | = | ✓ | ✓ | ✗ (needs authority) |
| gitcoin passport | ✓ | ✓ | partial | ✗ | partial (stamps ≈ KYC) |
| NIP-88 polls | ✗ | ✗ | ✓ | ✗ | ✓ |
| oc vote | ✓ | ✓ | ✓ | ✓ (opt-in) | ✓ |
the one claim
Bitcoin UTXOs are the only credible, chain-neutral, time-aware weight function that doesn't require minting anything. A voter's weight is a function of the sats they hold at a declared block height. An attacker manufacturing 10,000 fake voters with 100k sats aged 30 days pays real opportunity cost. An honest voter pays nothing.
This property is uniquely Bitcoin. Ed25519 gets you a signed ballot but nothing to weight it with. GitHub stars are forgeable. ERC-20 balances require minting a token, which is what Snapshot does and why every community that uses it needs its own token. Proof-of-humanity requires an iris scanner or a video call.
If you remove the UTXO-weight primitive, nothing else OC Vote offers makes sense as a standalone protocol. That's the thesis.
the full rationale
The complete hypothesis-stress-test lives in
oc-vote-protocol/WHY.md.
It walks 13 load-bearing claims — from "one address = one ballot is enforceable
without a server" to "secret ballots achievable with creator-commit-reveal" to
"the Ed25519 substitution test" — and stress-tests each against alternatives. If
you're evaluating whether this protocol fits your use case, read that document
before the spec.
what v0 does not solve
- Receipt-freeness / coercion resistance. After secret-mode reveal, ballots are publicly linkable. Full coercion resistance requires a homomorphic scheme (Helios-style). Out of scope for v0.
- Threshold reveal. Only creator-held reveal in v0. tlock / drand threshold reveal is future work.
- Delegation / liquid democracy. Not attempted. Possible as a v1 extension.
- Quadratic voting. Deferred to a registry-based
weight_mode. - Post-quantum. secp256k1 and X25519 both break under a PQ-capable adversary.
- Sovereign elections. This is a signaling primitive, not a replacement for statutory voting.
These are explicit non-goals, not things we hope to sneak in later. See security.