jpmschweitzer/settled-reach
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore(meta): T-1112/T-1113 to review — PR #179 up

Browse Source
This commit is contained in:
jpmschweitzer
2026-07-16 19:08:47 +02:00
parent c2f236e26e
commit a0ca09c22c
2 changed files with 27 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.pql/changelog/ticket_history/2026-07.sql
+2
View File
@@ -1403,3 +1403,5 @@ INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, chang
INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, changed_by, changed_at, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXF0JB8E6PDQRASQ18E9ER', 'status', 'backlog', 'in_progress', NULL, '2026-07-16 16:29:07', '2026-07-16 16:29:07.859', '2026-07-16 16:29:07.859', NULL, '0982ecc37d460add179eae8491ce496d', 2) ON CONFLICT(hash) DO NOTHING;
INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, changed_by, changed_at, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXF0JB8E6PDQRASQ18E9ER', 'assigned_to', NULL, 'dudley', NULL, '2026-07-16 16:29:14', '2026-07-16 16:29:14.525', '2026-07-16 16:29:14.525', NULL, 'eda3d1391d0ea48a59277fd6486cdbe5', 2) ON CONFLICT(hash) DO NOTHING;
INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, changed_by, changed_at, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXEM9NM5M5SX9WQAAA40SC', 'assigned_to', NULL, 'tyre', NULL, '2026-07-16 16:29:22', '2026-07-16 16:29:22.162', '2026-07-16 16:29:22.162', NULL, 'c9cfbbab6a28aa1e01b21c60f30624f5', 2) ON CONFLICT(hash) DO NOTHING;
INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, changed_by, changed_at, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXEM9NM5M5SX9WQAAA40SC', 'status', 'in_progress', 'review', NULL, '2026-07-16 17:08:34', '2026-07-16 17:08:34.002', '2026-07-16 17:08:34.002', NULL, '164946c94d6d95e2e7d6b1403d70d68a', 2) ON CONFLICT(hash) DO NOTHING;
INSERT INTO ticket_history (ticket_record_id, field, old_value, new_value, changed_by, changed_at, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXF0JB8E6PDQRASQ18E9ER', 'status', 'in_progress', 'review', NULL, '2026-07-16 17:08:34', '2026-07-16 17:08:34.005', '2026-07-16 17:08:34.005', NULL, 'bc87116dbb3c7aa92437d29d457a2b51', 2) ON CONFLICT(hash) DO NOTHING;
.pql/changelog/tickets/2026-07.sql
+25
View File
@@ -1889,3 +1889,28 @@ Refinement (2026-07-16, /whats-next Si review): READY, one scoping note made exp
- PRECEDENT TO IMITATE (five touch points, each with a working example): server AtlasLayerResponse field + build_quarter_footprint_layer() mirroring build_district_grid (keyed off non-empty state.quarters); client OVERLAY_DEFS gen_* id (atlas_viewer.gd), _draw_gen_* fn (atlas_marker_overlay.gd, imitate _draw_gen_district ~185-188), GENERATION_LEGEND entry (atlas_legend_panel.gd). These are named in the amendment for the wiring follow-up — not implemented in this ticket.
- PARTICIPANT SPLIT: Tyre owns the aggregate-stats data shape (fields, structure, D-010 integer discipline); Araminta owns the overlay encoding (outline vs filled block, color axis per the shape-vs-color convention in atlas_legend_panel.gd) — the same division D-226 originally used. Two coordinated agents, not a workshop round.
- pql metadata quirk noted: the ticket''s decision_ref D-226 resolves to a wrong title in show-summary; `pql decisions read D-226` returns the correct record.', 'in_progress', 'medium', 'tyre', 'client', 'D-226', '2026-07-13 21:50:30.989', '2026-07-16 16:29:22.162', NULL, 'c875a58d0f49225f768dd2e7efbee3e8', 2) ON CONFLICT(record_id) DO UPDATE SET type=excluded.type, parent_record_id=excluded.parent_record_id, title=excluded.title, description=excluded.description, status=excluded.status, priority=excluded.priority, assigned_to=excluded.assigned_to, team=excluded.team, decision_ref=excluded.decision_ref, updated_at=excluded.updated_at, deleted_at=excluded.deleted_at, hash=excluded.hash, canonical_version=excluded.canonical_version WHERE excluded.updated_at > tickets.updated_at OR (excluded.updated_at = tickets.updated_at AND excluded.hash > tickets.hash);
INSERT INTO tickets (record_id, type, parent_record_id, title, description, status, priority, assigned_to, team, decision_ref, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXEM9NM5M5SX9WQAAA40SC', 'task', '06FB0TNSRZXCHGS16BFHSSGSV4', 'Design: coarse quarter-footprint Atlas layer (L4 skeleton on the planetary map)', 'Si audit of T-960 (2026-07-13): L4 quarter/block skeletons (T-957/T-976) auto-generate server-side into BodyWorldState.quarters, but carry full per-block zoning/street/tag detail — too fine for a planetary map, and unlike the district grid there is no existing coarse shape to reuse. DESIGN ticket (Araminta encoding + Tyre data shape) before any wiring: define a QuarterFootprintLayer (city outline + aggregate stats?) for AtlasLayerResponse and its overlay encoding per the D-226 amended (2026-07-13) overlay pattern. Wiring is a follow-up once the design is recorded.
---
Refinement (2026-07-16, /whats-next Si review): READY, one scoping note made explicit.
- DELIVERABLE FORMAT (was unpinned): record the design as a **D-226 amendment** — the established mechanism (the 2026-07-13 overlay pattern itself was recorded that way); no docs/design Atlas-overlay doc precedent exists (T-1046 shipped design-in-ticket+code-comments). The design pass ends at the recorded amendment; the wiring follow-up ticket is filed by the lead afterwards, not by the design pass.
- BINDING CONSTRAINT: D-226 amendment (d) is a hard ceiling — planetary Atlas tops out at settlement/quarter-skeleton granularity with AT MOST aggregate stats. No per-block zoning/street/tag detail at planetary scale. City outline + aggregate stats only.
- DATA SOURCE: BodyWorldState.quarters stores QuarterWorldState{skeleton: QuarterSkeleton, block_tags: BTreeMap<(u8,u8), Vec<BuildingPropertyTag>>}; QuarterSkeleton carries [[BlockSkeleton;4];4] with zoning/district_type/density_pct/landmark + reservations/corridors. All aggregatable (density avg, dominant zoning/district-type mix, landmark count, corridor count) without inventing new simulation truth — same derivation class as build_district_grid/build_settlement_layer.
- PRECEDENT TO IMITATE (five touch points, each with a working example): server AtlasLayerResponse field + build_quarter_footprint_layer() mirroring build_district_grid (keyed off non-empty state.quarters); client OVERLAY_DEFS gen_* id (atlas_viewer.gd), _draw_gen_* fn (atlas_marker_overlay.gd, imitate _draw_gen_district ~185-188), GENERATION_LEGEND entry (atlas_legend_panel.gd). These are named in the amendment for the wiring follow-up — not implemented in this ticket.
- PARTICIPANT SPLIT: Tyre owns the aggregate-stats data shape (fields, structure, D-010 integer discipline); Araminta owns the overlay encoding (outline vs filled block, color axis per the shape-vs-color convention in atlas_legend_panel.gd) — the same division D-226 originally used. Two coordinated agents, not a workshop round.
- pql metadata quirk noted: the ticket''s decision_ref D-226 resolves to a wrong title in show-summary; `pql decisions read D-226` returns the correct record.', 'review', 'medium', 'tyre', 'client', 'D-226', '2026-07-13 21:50:30.989', '2026-07-16 17:08:34.002', NULL, '7090cccaa4a8519967f005e59e106c30', 2) ON CONFLICT(record_id) DO UPDATE SET type=excluded.type, parent_record_id=excluded.parent_record_id, title=excluded.title, description=excluded.description, status=excluded.status, priority=excluded.priority, assigned_to=excluded.assigned_to, team=excluded.team, decision_ref=excluded.decision_ref, updated_at=excluded.updated_at, deleted_at=excluded.deleted_at, hash=excluded.hash, canonical_version=excluded.canonical_version WHERE excluded.updated_at > tickets.updated_at OR (excluded.updated_at = tickets.updated_at AND excluded.hash > tickets.hash);
INSERT INTO tickets (record_id, type, parent_record_id, title, description, status, priority, assigned_to, team, decision_ref, created_at, updated_at, deleted_at, hash, canonical_version) VALUES ('06FNTXF0JB8E6PDQRASQ18E9ER', 'task', '06FB0TNSRZXCHGS16BFHSSGSV4', 'Wire region climate stack into the cascade (CascadeLayer::Region + proxy exposure)', 'Si audit of T-960 (2026-07-13): region_profile.rs (T-1078 — RegionProfile, derive_regions_for_body, region_baseline_at_district) is implemented and unit-tested but never called from cascade.rs: no CascadeLayer::Region variant, nothing on BodyWorldState, nothing in the layer proxy. Prerequisite for any climate Atlas overlay (explicitly out of T-960 scope per 2026-07-13 ruling): add the cascade layer, populate BodyWorldState, expose via AtlasLayerResponse following the T-1046 district-grid precedent. The climate OVERLAY itself is a small follow-up once this serves data.
---
Refinement (2026-07-16, /whats-next Si review): READY, no gaps.
Key implementation facts (Si-verified against post-PR#178 main):
- region_profile.rs (T-1078) exposes exactly what''s needed: RegionProfile{pos, clock: RegionClock{season, weather, mean_temp_c}, latitude_deg, moisture_q}, derive_regions_for_body(...) -> BTreeMap<RegionPos, RegionProfile>, region_baseline_at_district(...). Deterministic, unit-tested.
- CRITICAL: district_profile.rs::derive_all_districts (~1434-1509) ALREADY builds the region cache internally (incl. neighbours for edge-fuzz) and discards it — this ticket is retention, not new derivation. Chosen shape: option (a) — independent CascadeLayer::Region step in run_cascade_from_heightmap (one-layer-one-concern, mirrors existing layers; accept the cheap double-derive).
- CascadeLayer Ord today: Heightmap < Topography < Settlement < DistrictProfile < RoadGraph. Region slots as its own step.
- BodyWorldState: add pub regions: BTreeMap<RegionPos, RegionProfile> per the districts precedent (same determinism rationale, same empty-until-run convention). Footprint trivial (~195×98 max).
- layer_proxy.rs doc comment (76-79) already earmarks T-1113 as one of the two final layer fields. Mirror build_district_grid: RegionGridLayer + build_region_grid(state) -> Option<...>, called from handle_atlas_request cache-hit branch. Encoding: dense row-major grid per the T-1046 precedent (implementer''s low-stakes call on exact fields/quantization).
- Client wire-decode IS in scope (one line): protocol.gd::atlas_response_from_raw raw.get("region_grid") passthrough, matching district_grid. The VISUAL overlay (atlas_viewer/marker_overlay/legend) is explicitly OUT of scope — small follow-up per D-226 amendment (a) incremental-overlay pattern.
- Edge-fuzz (D-243 §4) is consumer-side sampling behavior already implemented in region_baseline_at_district — not this ticket''s concern.
- Q-105 resolved into D-243; RegionClock tick-phase callbacks stay deferred — only the mean-state fields are served.', 'review', 'medium', 'dudley', 'server', 'D-243', '2026-07-13 21:50:34.130', '2026-07-16 17:08:34.005', NULL, '3e46d1a1deedd537dea1d2d214b26c37', 2) ON CONFLICT(record_id) DO UPDATE SET type=excluded.type, parent_record_id=excluded.parent_record_id, title=excluded.title, description=excluded.description, status=excluded.status, priority=excluded.priority, assigned_to=excluded.assigned_to, team=excluded.team, decision_ref=excluded.decision_ref, updated_at=excluded.updated_at, deleted_at=excluded.deleted_at, hash=excluded.hash, canonical_version=excluded.canonical_version WHERE excluded.updated_at > tickets.updated_at OR (excluded.updated_at = tickets.updated_at AND excluded.hash > tickets.hash);