Pilot & roadmap
v0.1-draft is a design document. No producer, council, market or exchange is running this. Phases are a plan, not a status report.
Phases
Paper phase — where we are
Specification, this wiki, and conversations with SGP councils and SENASA. Deliverable: a spec that a council coordinator can read and recognise their own process in. If they cannot, the spec is wrong, not them.
Attestations without payment
Issue and verify attestations with no Taler payment at all — QR on a printed lot label, verification in a web page. Tests the institutional layer (can a council sign? does key rotation work? is registration fast enough?) before adding the hardest dependency. Most of the project's risk is here, and none of it is cryptographic.
Token ceilings, still without payment
Add bounded issuance and burn-on-sale, recorded by the producer's instance. Tests whether yield estimation and supplementary issuance work at the speed a harvest demands. Answers OQ-4.
Taler payment in one market
One market, one SGP council, a handful of producers and stalls, a real exchange with soles settlement. This is where the adoption problem becomes concrete and where the project most plausibly stalls.
Measure and publish
M1, M2, M3 from Where the money goes, published whatever they say. A null result is a finding, and burying it would make everything else on this wiki worthless.
Open questions
Referenced throughout the wiki. None resolved.
| ID | Question | Why it is hard |
|---|---|---|
| OQ-1 | Can a certification quota be expressed within Taler's existing token-family classes, or does it need a new one? | details.class is subscription or
discount[1]; ours is
neither. Needs an upstream conversation. We do not want to maintain a
fork. |
| OQ-2 | Offline payment, as distinct from offline verification. | Verification is local and free. Settlement needs the exchange, and market connectivity is unreliable. Deferred settlement reintroduces double-spend risk. |
| OQ-3 | A defensible Peruvian cost curve for certification. | The USD 2,500 figure is from advocacy literature, not a local tariff. Group certification changes it. Without real numbers, M3 has no baseline. |
| OQ-4 | Supplementary token issuance fast enough for a good harvest. | If a top-up takes weeks, producers sell outside the scheme. Speed here competes directly with audit rigour. |
| OQ-5 | Key custody on shared, lost and sold phones. | Custody by an aggregator recreates the problem the project exists to solve. See spec § 8. |
| OQ-6 | Will distributors accept margin disclosure? | Arguably the question that decides the project. Cannot be answered from a desk. See Who does what. |
| OQ-7 | How does the project sustain itself? | Per-transaction fees put us in the payment path; data products violate the privacy rule; certification-as-a-service makes us a competitor to the councils we serve. No answer yet. |
| OQ-8 | Can a Fairtrade-style price floor be enforced as a contract invariant? | Attractive — it would convert a standard into something machine-checked — but it means encoding commercial terms we have no authority over. |
| OQ-9 | Language and interface for Quechua- and Aymara-speaking producers. | summary_i18n makes it technically expressible. Doing it
properly is a translation and design problem, not a field. |
Risks that could end the project
- No Taler installed base in Peru. No exchange, no wallets, no bank integration. Every other risk is downstream of this one. Phases 1 and 2 exist precisely so that the institutional work has value even if phase 3 never happens.
- Distributor refusal. Covered above. If stalls will not countersign, the producer-signs model has no point of sale.
- Council capacity. SGP councils run on volunteer time. Adding key management could exceed what they can absorb, no matter how good the tooling.
- Regulatory ambiguity. Whether a signed attestation carries any weight under the RTPO, or is merely a private representation alongside the official certificate, is not something we can decide unilaterally. Early contact with SENASA is required, not optional.
- Solving the wrong problem. If the binding constraint on smallholder income is transport, storage losses, or credit rather than verification of claims, this project produces an elegant system that changes very little. The Agroferias Campesinas findings — that short chains and local value addition are what move margins[2] — are a reminder that verification is one lever among several, and not obviously the largest.
Participating
What would be most useful right now, roughly in order:
- An SGP council willing to read the spec and say where it does not match how they actually work.
- Producers willing to be wrong about phase 1 in public.
- A market administration open to a small phase-3 trial.
- GNU Taler contributors to sanity-check OQ-1 before we design ourselves into a fork.
- Anyone with paired farmgate and retail price data for Peruvian produce by channel. This is the gap in the evidence ledger.
Contact and repository links will be published here once phase 0 produces something worth reviewing.
References
- GNU Taler, DD 46: Contract Format v1. docs.taler.net
- Springer, Sustainable circular strategies and short supply chains among smallholder producers: the case of Peru's Agroferias Campesinas (2025). link.springer.com