An open-source experiment: can a currency stay stable by measuring the price of water? Status: early-stage prototype. We are looking for contributors, not customers.

O is a research prototype of a water-price-calibrated stablecoin, built as a Bitcoin Core fork by O International, a French nonprofit (association loi 1901). Nothing is for sale. There is no token sale, no presale, no investment product — just code, an economic idea, and a list of genuinely hard problems.

The idea in three sentences:

1. 1 O = the local price of one liter of potable water, per currency zone (142 O-currencies: O_USD, O_EUR, O_JPY, …).
2. Randomly invited, verified humans measure water prices and observed exchange rates; the protocol publishes reference rates from those measurements.
3. Nobody is forced to follow the reference rate — but when actors drift from it, the protocol mints coins for participants who stay aligned, diluting those who don't. The offender's sanction is the reward of the offended. An incentive, not an order.

The full economic reasoning lives in the article series — start with Why Proof of Work Was Built for Security, Not Speed (the scalability architecture) and How to Secure Accurate Measurements Without Oracles (the measurement design).

The codebase (~1,400 C++ files on top of Bitcoin Core 27.0) expresses the intended flow: measurement submission, Gaussian aggregation, reference-rate publication, stabilization minting, invitation scheduling. It runs on a single-node regtest.

It is not ready for a real network. In particular:

• Stabilization recipient and endorser selection are now seeded from consensus data (fixed 2026-08), but the fix still needs adversarial tests and a 2-node regtest proof — and at least one selection path is still unseeded.
• Monetary state (which currencies are unstable, what was minted) lives in RAM and does not survive a restart or a reorg.
• Consensus-critical math still uses floating point.
• Two validators share one name: the live stabilization validator does recompute-and-compare blocks, but a dead always-true duplicate still lingers in the code and needs consolidation.
• The multi-currency output type exists but is not wired into the live payment path.
• The Layer-2 payment executor described in the articles is a design, not code.

Every one of these is written up with file/line pointers in Open Problems — which is exactly where to look if you want to contribute something that matters.

Because the unsolved problems are real distributed-systems problems, not plumbing:

• Deterministic randomness — network-agreed sampling of reward recipients from consensus data only.
• Durable consensus state — committing monetary state to the chain so restarts and reorgs are safe.
• Measurement security — stratified sampling across jurisdictions, FX-triangle consistency checks, honest uncertainty when a region can't be measured.
• A shared multi-currency Layer 2 — account-based deterministic executor with atomic FX across 142 currencies, settling commitments to L1.
• Data availability at scale — if payments batch off-chain, who stores the bytes and how are exits enforced?
• Monetary simulation — does incentive-based stabilization actually hold the peg under adversarial market behavior? Break it in a model before reality does.

If you work on consensus, payment systems, DA, zero-knowledge proofs, FX, distributed systems or monetary simulation: challenge the design. Clone the repo. Break the assumptions. Write a failing test. Propose a better rule.

→ Contributing · Open Problems · Issues

• Open Problems — the honest to-do list, with code pointers
• Contributing — how to get a change merged
• Installation · Testing Guide — build & run regtest

• Architecture Overview — system map
• Stabilization System — the incentive mechanism
• Measurement System — water-price calibration
• Invitation System — who gets asked, and why that's security
• Transaction Types — O-specific transactions
• Currency Coverage — the 142 O-currencies

• Scalability Overview — L1/L2 split, why we don't just shrink block time
• Transaction Volume Analysis — how big global payments actually are

• Universal Basic Income
• Climate Change Solution
• Immigration Impact

No marketing TPS. No claim that a prototype already carries the world. Throughput numbers will be published with their conditions when there is something real to measure.

• Website: o.international
• Email: support@o.international
• Repo: github.com/cno127/o-blockchain

O International is a nonprofit (association loi 1901). Code is MIT-licensed. Nothing here is for sale.

Last updated: 2026-08