Status: Proposed, 12 September 2026. Not decided. Awaits the engineering lead and an independent reviewer (blueprint R06). Usage policy: OPERATIONS.md.
Context
One issuer can fail while the other does not, so the series must stop only what fails. The blueprint (§11) proposes a mint stop and a per-component claim stop; the second adds dependence on the operator.
Decision
- —Pause minting:
setMintPaused(true, reason), resumable; exit allocations and claims continue. - —Pause claims of A or of B separately:
setClaimPaused(component, paused, reason). Pausing A does not pause B, and by the blueprint's default B is not paused because A has a problem. - —Every stop and resume emits an event with a reason.
- —No stop moves anything: no function transfers backing, rewrites a claim, sweeps surplus, changes a component, or pays anyone but a claim's holder, paused or not.
- —Two automatic stops, not the operator's:
BackingShortreverts a mint that would leave the series short;ShortfallHaltsPaymentreverts every claim of a short component until the shortfall is gone, so no fastest claimant takes the rest.
Consequences
- —A claim on a paused component waits; the other pays. The series page's drill shows both.
- —The operator can really delay one component's withdrawal (shown in the series document). Policy limits it to exploit risk, insufficient bookkeeping or a defined access obligation, and a resume needs a second review.
- —A shortfall halts that component for every holder, honest ones included, until resolved; first-come payment was rejected as a race.
What exists today
mintPaused, claimPausedA, claimPausedB, the setters with events, BackingShort, ShortfallHaltsPayment; tests T05, T06, T09, T10 and the operator-limit cases; the drill's "issuer freezes A" and "series short of A" scenarios. The Robinhood Chain treasury signers are recorded; an operator Safe on the candidate Ethereum series chain remains unverified.
Open
- —Should a resume need an on-chain time delay? (Today: policy only.)
- —Should a permanent "closed to new mints" state differ from a pause? (See ADR-001.)