gnomonlearn
Docs/RFC-0055/Design Q&A

RFC-0055 · Design Q&A

Versioned draft snapshot

RFCs, reference catalogs, and evidence reflect Spec 3b2e3e7a114fb826b1445d39a48756b8fff629cd. Counts and pending/status notes belong to that snapshot. Read the dated project update for later progress.

Source repositories are private during staging. Links marked “repository access required” need access to the linked repository.

← Read the authored RFC

Q: How should RFC-0006 diagnostic codes migrate to RFC-0055?

Status: Open

Source: RFC-0055 §7–§8; RFC-0006 §4.5 and its Lean diagnostic catalog Issue: The draft says M06520–M06524 remap to M55010–M55030, but the conditions do not line up. M06520 is an unwrapped AI field, M06521 is policy-required confidence, M06522 is an invalid assigned status, M06523 is undefined, and M06524 is a nested claim. Only the nested-claim error has an exact successor (M55030). Recommendation: Keep the defined historical IDs and meanings. Emit M55030 for new nested-claim findings after the transition while accepting historical M06524 in compatibility inputs. Use M55011 for the narrower agent-missing-confidence trigger; retain M06521 where a different policy requires confidence. Keep M06520 and M06522 until new diagnostics with matching triggers are defined, and leave M06523 unallocated. The full table and consumer acceptance criteria are in DRAFT_COVERAGE.md.

Trigger: Owner approval and external/persisted diagnostic consumer review before editing normative code identities or emitters.

Q: Should Confidence admit interval values for rule-derived claims?

Status: Open

Source: RFC-0055 §10

Original question: Should Confidence admit interval values ([0.6, 0.8]) for rule-derived claims with computable bounds? Deferred.

Trigger: Revisit when computable confidence bounds are specified for rule-derived claims.

Q: Should a superseded human claim attribute its agent superseder in asserted_by?

Status: Open

Source: RFC-0055 §10

Original question: When a human claim is revised by an agent, the new claim has superseded status pointing at the old. Should the old claim’s asserted_by field carry an attribution to the superseder? Deferred to RFC-0047 audit observations.

Trigger: Review RFC-0047 audit observations for cross-actor revisions before changing attribution semantics.

Q: Can evidence_required specify a minimum count per evidence kind?

Status: Open

Source: RFC-0055 §10

Original question: Per evidence_required kind, can the authoring surface specify a minimum count (document: ≥ 2)? Deferred.

Trigger: Specify authoring and evaluation semantics for counts such as two required documents.