Heartbeat and quiet-range boundary
Monolythium uses anchor and round terminology, and quiet network periods must not be interpreted as a failure merely because transaction-bearing anchors are sparse. Consumers should tolerate non-uniform time between application-visible commits and use the released SDK and read-only status interfaces for the exact runtime they are observing.
Publicly safe statement
Section titled “Publicly safe statement”The public documentation recognizes heartbeat and compact quiet-range concepts. It does not certify a quiet-period trigger, marker cadence, coalescing interval, wire payload, certificate envelope, pruning rule, database row shape, SQL query, subscription event, recovery invariant, bandwidth ratio, or disk-growth estimate.
Those details are runtime- and indexer-specific. Publishing stale values as an operator runbook would be more dangerous than omitting them. No public operator or database procedure is provided on this page.
Timing correction
Section titled “Timing correction”At the documented 4-second target, a full day contains 21,600 round intervals. A previously published larger daily count was arithmetically incorrect. This correction is only cadence arithmetic; it is not a claim that the runtime creates, stores, gossips, or later compacts 21,600 objects per day.
The active profile also has an 18-second epoch target. Neither target is a promise of wall-clock regularity or a substitute for observed finalized state.
Consumer guidance
Section titled “Consumer guidance”- Do not synthesize missing anchors or assume a fixed time-to-height conversion.
- Do not infer archive retention, proof availability, or storage bounds from a conceptual quiet-range description. Keep the independently documented SDK archive limits intact.
- Do not bind software to historical field names, table names, payload layouts, or disabled RPC method names.
- Pin chain ID, full genesis, runtime provenance, and SDK version before interpreting status data.
- Treat the value-less resettable testnet as development evidence, not a production operating certification.
The currently verified public RPC subset is limited to chain statistics, current-round status, peer summary, and runtime provenance. It is a status surface, not a public gap-record ABI.