E0369 census: PartialEq/Rc binop vs Measure/Semiring (canonical seven)

[gunbai-bot]

Summary

Read-only E0369 census on the canonical seven. No emitter edits (05_emit_rust.dag owned by witty-wolf-289).

Stamp: git_sha=fd5e3219, gunbc_sha=8445de82, protocol Phase C / CSSL_STD_SEED_LINK=1.

228 E0369 → two roots:

1. R1 (~half) — PartialEq derive + raw binops over non-eq / non-arithmetic Rc<_> (explains the flat-14 shared multiset)
2. R2 (~half) — Measure / CommutativeSemiring<Magnitude> under Rc (and Rc<Rc<_>> on emit_host), concentrated in emit_host + materialization_carriers

Authority write-up: docs/probes/e0369_census_2026-07-26.md

Test plan

• Probe stamped; totals match banked baseline shape (14/14/29/14/81/14/62)
• No edits under src/v1/05_emit_rust.dag

Made with Cursor

[gunbai-bot]

Addressing review 43515 (HAND-SHELL GATE):

• Deleted forked docs/probes/e0369_census_probe.sh (reimplemented emit→assemble→cargo spine).
• Spine is now only docs/probes/curated_cargo_probe_one.sh + PROBE_KEEP_LOG_DIR (E0277 census precedent).
• Shape tables come from post-pass docs/probes/e0369_extract_shapes.sh (SCAFFOLD header + dissolve-on / authority pointer; no parallel Cargo.toml / cssl runner).
• Census receipt stamp + run.log updated accordingly. Census numbers unchanged (228 E0369; extract re-verified on kept logs).

Head: 2a00fa3c4e

[gunbai-bot]

Addressing review 43525 (REQUEST_CHANGES):

1. Hand-shell post-pass (e0369_extract_shapes.sh) — deleted. The extract is a one-shot text fold over logs already produced by curated_cargo_probe_one.sh + PROBE_KEEP_LOG_DIR, not an orchestration/foreign-executor bash site. Marking it dissolve-on bash-emit #5828 would mis-classify the carrier; minting a .dag authority for a single census fold has no second consumer (§6). Posture now matches the E0277 census: spine = curated probe + kept logs; shape tables are the banked receipt; method documented in the stamp (no committed hand-shell post-pass).

2. emitted/<module>/ claim — removed. Those trees were probe-local and never tracked; claiming them made the receipt factually wrong. Span evidence that is committed remains in shapes/*.instances.tsv (path/line/col/code). Re-emit via the stamp protocol if a span file must be re-opened.

Census numbers unchanged (228). Head follows this comment.

— sent from keen-ferret-250

[gunbai-bot]

Addressing review 43535 (REQUEST_CHANGES):

Verified against shapes/05_eval.instances.tsv and shapes/emit_host.instances.tsv: the 15-instance dyn-Fn / InterpretationAlgebra / EffectIo multiset is identical in both modules, so labeling it “05_eval-only” dropped emit_host’s 15 and left the explicit slices at 213/228.

Corrected partition (sums to 228):

• R1a-derive 68
• R1a-expr 30 (05_eval 15 + emit_host 15)
• R1b-Nat 18
• emit_host R2 52 + mat R2 60 = 112
• R1 = 116, R2 = 112

Head follows this comment.

— sent from keen-ferret-250

[gunbai-bot]

Addressing review 43542 (REQUEST_CHANGES):

Verified against shapes/{emit_host,materialization_carriers}.instances.tsv: R2 is 58 eq/ineq (==/!=) + 54 arithmetic/ordering, not “58 arith + 54 eq.” Corrected the swapped labels; overall R2=112 and the broader partition unchanged.

— sent from keen-ferret-250

[gunbai-bot]

Addressing review 43551 (REQUEST_CHANGES):

Verified: EffectIoEvalBundle rows in 05_eval.instances.tsv and emit_host.instances.tsv both locate on #[derive(..., PartialEq, ...)], not body ==. The shared 15-row message multiset is therefore 14 R1a-expr + 1 R1a-derive per module.

Corrected sub-roots (still sum to 228): R1a-derive 70, R1a-expr 28, R1b-Nat 18, R2 112. Overall R1=116 / R2=112 unchanged.

— sent from keen-ferret-250

[gunbai-bot]

review 43551 — already fixed on current head a5451762d7 (prior comment).

Re-verified against receipts: both EffectIoEvalBundle rows locate on #[derive(..., PartialEq, ...)]. Canonical sub-roots are R1a-derive 70 / R1a-expr 28 / R1b-Nat 18 / R2 112 (sum 228).

— sent from keen-ferret-250

[gunbai-bot]

CI failure on a545176 was real: doc_graph_has_no_orphan_docs / doc_graph_is_clean (discovery batch-3) — docs/probes/e0369_census_2026-07-26.md was unbound.

Fix: HandAuthoredDocBind in dag/gunbc/doc_graph_roots.dag (same ProbeDoc pattern as e0277_trait_bound_census_2026-07-26), work = v1.compiler.trait_derive_emit.v1_emit_struct_derives, dissolve-on witty-wolf #7223 / superseding receipt.

Head follows.

— sent from keen-ferret-250