ROADMAP §5: reconcile self-host state — Root-A LANDED (#5734) + measure-tower census refresh

[briansrls]

§5 uses hand-set AuthoredLine done flags (not auto-derived from PR merge like DerivableLines), so merged PRs left it stale. Edited the .dag authority (dsl/gunbc/roadmap_authority.dag) and regenerated ROADMAP.md via main_wet on generated_artifact_gate.dag — same artifact_generate() the drift gate compares, so committed == generated by construction.

Changes (§5 self-host lane only):

• 5-root-a: done: false → true — emit-seam grounding Lane A Root-A emit-seam grounding (the E0107-unmasked dominant pole): faithful-list/FreeMonoid CONSUMPTION through host Vec receivers — emitted count/first/skip/iteration render as bare Vec methods on Rc-of-Vec and fail method-resolution (16 errors in std_realization_schedule.rs: 13 E0599 + 3 E0308, #5734 merged.
• 5-cargo-green prose: the measure-tower pole is cleared — measure-Q Lane A measure-Q grounding (operator-BLESSED 2026-06-24: DROP Q/S phantom type-params in the EMITTED Rust). The Measure tower type Measure of Q,S,M uses the Quantity VALUE sum-type (Time, Length, ...) in the Q TYPE-arg position = type/value conflation, producing E0573 x2 (Time used as a type) + E009 #5735 (value-variant type-arg collapse, the E0560×12/E0614×7 std_measure.rs root) + std_serialization C0 fix(emit_rust): pull std.serialization into emit closure via explicit import #5761 landed. Re-census = 5 total errors; cargo-green emitter fixes: M: Clone bound on bare-generic returns (E0599) + Optional receiver unwrap before method call (E0624) #5777 (open, merry-gull — E0599/E0624) takes 5→2; residual 2 = E0091×2, one Compose decision away.

Deliberately unchanged: 5-cargo-green box stays [ ] — cargo-green isn't reached until Compose lands; the ▸ NOW: emitted crate cargo-builds green arrow holds.

Verified: regen produced exactly these two §5 diffs in ROADMAP.md; the other 5 gated artifacts are byte-identical.

[gunbai-bot]

Valid finding, fixed in 0bad8c0 by dropping the file from this PR entirely (restored origin/main verbatim).

You diagnosed it precisely: gunbc.host_effect/host_effect_realize did land (#5756), but host-effect-orchestration.md is a generated artifact (main_wet on generated_artifact_gate.dag writes it via artifact_generate), and its .dag Plan authority still emits the pre-#5756 [now, parallel] text — #5764 grounded the doc into a Plan but did not re-ground that Phase-A/B prose to the LANDED state. So a routine main_wet regen rewinds the marker and drops the Phase-B dissolution trigger. That stale regen leaked into this roadmap-citation PR; it has no business here.

The PR diff is now ROADMAP.md + roadmap_authority.dag only (the byte-aligned citation refresh you confirmed). The underlying drift — artifact_generate for the host-effect doc emitting stale Phase-A text — is a real parallel-representation debt in the #5764 Plan authority and is flagged for that lane to re-ground at source; fixing it here would be cross-lane and out of scope.

— sent from sharp-carp-825

[gunbai-bot]

CI fix pushed (e09cef9) — and a reconciliation note on the host-effect-orchestration.md Phase-A marker, since this resolves a tension between the two earlier reviews.

Root cause of the CI red: host-effect-orchestration.md is a gated generated artifact (in committed_generated_artifacts()), regenerated by main_wet from its .dag Plan authority. When I restored the [LANDED #5756] text (to satisfy the first RC), I made committed ≠ artifact_generate — the authority emits [now, parallel] — so generated_artifact_drift_gate_passes went Bool(false). That was the 1 failing check.

The reconciliation: current origin/main already carries [now, parallel] for this paragraph (verified: git show origin/main:docs/plans/host-effect-orchestration.md line 56). So the marker was rewound on main itself, not by this PR — and artifact_generate agrees. This PR now matches main byte-for-byte on that file (the diff only shows it because origin/main...HEAD diffs the merge-base, which predates main’s advance). Committing [LANDED #5756] is impossible without a red drift gate, because the Plan authority generates now-parallel.

The real debt (out of scope here, flagged for the host-effect lane): the .dag Plan authority is stale — #5756 landed gunbc.host_effect*, but its authoring text still reads [now, parallel], so every main_wet rewinds the done-marker. That is the #5764 parallel-representation debt; the fix is to re-ground the authoring text at source, not to hand-hold the generated .md. I merged current main and re-verified the drift gate green (first_failing_path → empty).

— sent from sharp-carp-825