Step 10C.10: CR-003 plan amendments, review screenshots and regression

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

DECISIONS.md

@@ -748,3 +748,9 @@ Implementation decisions not spelled out in `memory-bank/implementation-plan.md`
 - **Canvas resolution is capped at 2×.** Measuring in headless Chrome (software WebGL) showed frame time is set by pixel count, not our JavaScript: 58.5 fps at DPR 1 with the CPU throttled 4×, versus 11 fps at DPR 3. The art only has 2× variants, so a 3× canvas cost nine times the pixels of 1× for no visible gain; the cap doubled the rate to 22 fps there. A real phone rasterizes on its GPU, so the user's device check is the real verdict.
 - **Crowds of more than three ships use a hexagonal grid**, still 44 px apart, after the first busy screenshot showed ten simultaneous departures strung across 400 px of empty space. Up to three ships keep the perpendicular spread from Step 10C.8.
 - **Per-frame work stays allocation-light:** textures and Text objects are created once, Graphics layers are cleared and redrawn, and labels restyle only when their colour changes (restyling re-renders a text texture).
+
+## Step 10C.10
+
+- **Step 11.1's amendment names the generated folder explicitly** (`public/assets/map-gen/`), since the source art is dropped from the build and must not be precached.
+- **The "map with undiscovered locations" review shot is the tutorial map**, which shows Mars, Ceres and Jupiter locked; the mid-game shot has Mars discovered and Ceres and Jupiter still locked.
+- **Screenshot size:** phone shots are full-resolution (DPR 3) PNGs, about 2 MB each; `docs/screenshots/` is about 24 MB. They are kept as-is for the review; trimming or moving them out of the repository is a later cleanup.

memory-bank/change-requests/CR-001-faster-start.md

@@ -160,6 +160,7 @@ Record these in `DECISIONS.md` and in `implementation-plan.md` §0.5.
 - The map is available from the first screen, including during the tutorial.
 - Undiscovered locations are drawn dimmed with a lock. Tapping one shows "Undiscovered — deliver here to unlock" and which ship models can currently reach it from Earth (using the sim range check over purchasable models), for example "Reachable with: Freighter, Bulk Hauler".
 - **Test:** Component test: undiscovered locations render as locked; the reachability list for Mars matches the sim range check for the mocked time; discovered locations render normally.
+- **Amended by CR-003:** the lock rendering is replaced by Step 10C.7 (desaturated, dimmed sprite with a code-drawn padlock); the tap panel is unchanged.
 
 ### Step 10A.17 — Client: something waiting on return
 - If goals became ready while the player was away, list them in the "while you were away" modal with a link to claim.

memory-bank/implementation-plan.md

@@ -470,10 +470,12 @@ These answer the pre-build review. Log each one in `DECISIONS.md` as it is imple
 ### Step 10.1 — PixiJS map component
 - Add a Map route with a PixiJS canvas showing the Sun and the five locations at their current positions. Use a compressed (non-linear) radial scale so inner and outer locations are both visible. Support pan and pinch-zoom.
 - **Test:** Unit test confirms the scale function is monotonic (farther in reality → farther on screen). Manual check: all five locations are visible on a phone-sized screen at default zoom; pinch-zoom works on a touch device.
+- **Amended by CR-003:** Bodies use sprites with a roughly constant screen size (zoom^0.35, clamped 0.75–1.75×) and the Moon at a fixed screen offset from Earth (Step 10C.6, decisions 0.5.38–0.5.39).
 
 ### Step 10.2 — Ships on the map
 - Draw in-flight ships as markers along their route lines, positioned with sim interpolation every frame, and idle ships as badges at their location.
 - **Test:** Unit test: marker screen position for a mocked dispatch at 50% progress lies on the route line between the two endpoints. Manual check: ships move smoothly.
+- **Amended by CR-003:** In-flight ships use their model's sprite rotated to face their direction of travel, nudged apart when crowded; idle ships are shown only as count badges (Steps 10C.7, 10C.8).
 
 ### Step 10.3 — Map interactions
 - Tapping a location opens a panel with its details and a link to its contract board (if a bay exists). Tapping a ship shows its route, pilot, and ETA.
@@ -481,6 +483,7 @@ These answer the pre-build review. Log each one in `DECISIONS.md` as it is imple
 
 ---
 - **Amended by CR-002:** The location panel links every discovered location to "View contracts", with or without a bay (Step 10B.9).
+- **Amended by CR-003:** Every tap target is at least 44 px, hit-tested against the same screen layout the renderer draws (Step 10C.6).
 
 ## Phase 10A — Faster Start (CR-001)
 
@@ -505,6 +508,7 @@ Steps 10C.1–10C.10 are specified in `memory-bank/change-requests/CR-003-map-ar
 ### Step 11.1 — Installable PWA
 - Add `vite-plugin-pwa` with a manifest (name "Deadhead: Space Freight Tycoon", icons, theme color, standalone display) and a service worker that caches the app shell.
 - **Test:** A Lighthouse PWA audit in CI reports the app as installable. Manual check: the app installs to a phone home screen and launches standalone.
+- **Amended by CR-003:** The service worker precaches the generated map assets in `public/assets/map-gen/` (not the source art in `public/assets/map/`, which the build omits).
 
 ### Step 11.2 — Offline shell behavior
 - When offline, the app loads from cache and shows the last known state with a clear "Offline — showing last update" banner. In-flight progress bars keep advancing from cached dispatch records. Actions are disabled until back online.

memory-bank/progress.md

@@ -158,4 +158,4 @@ Spec: `memory-bank/change-requests/CR-003-map-art.md`. New decisions recorded as
 - [x] **Step 10C.7 — Location state overlays.** `map/overlays.ts` `locationOverlays(scene)` decides per location: undiscovered → desaturated (one shared `ColorMatrixFilter`), alpha 0.55 and a code-drawn padlock (replaces the 🔒 label of Step 10A.16; the tap panel is unchanged); bay → a small green house marker on the far side of the label; idle ships → count badge only. Test: `map-overlays.test.ts`. **Manual check:** `docs/screenshots/10c7-overlays-{tutorial,midgame}-{desktop,phone}.png`.
 - [x] **Step 10C.8 — Ships in flight.** Each flight draws its model's sprite rotated by `shipRotation` (art points up: right = 90°, down = 180°), positioned by sim progress every frame. Routes are faint end to end and brighter ahead of the ship. `separateShips` spreads ships closer than 44 px apart, perpendicular to their route and 44 px apart; hit-testing uses the same positions. The selected ship or location gets an amber ring (`MapCanvas` `selected` prop). Test: `map-ships.test.ts`. **Manual check:** `docs/screenshots/10c8-ships-midgame-{desktop,desktop-zoomed,phone}.png`.
 - [x] **Step 10C.9 — Performance.** `map/ticker.ts` `runWhileVisible`: the PixiJS ticker stops when the page is hidden, restarts when visible, and stops on unmount. Canvas resolution capped at 2×. Labels restyle only on change. Big ship crowds (> 3) pack on a hexagonal grid (`hexSpiral`) instead of a long line. Tests: `map-ticker.test.ts`, crowd test in `map-ships.test.ts`. **Measured** (headless Chrome, software WebGL, 10 ships in flight, panning): 58.5 fps at DPR 1 even with CPU throttled 4×; 11 fps at DPR 3 before the 2× cap, 22 fps after — fill-rate bound in software rendering, not JavaScript. **Manual check for the user:** pan and zoom on a real mid-range phone (GPU rendering). Screenshots: `docs/screenshots/10c9-busy-busy-{desktop,desktop-zoomed,phone}.png`.
-- [ ] Step 10C.10 — Plan amendments and regression
+- [x] **Step 10C.10 — Plan amendments and regression.** CR-003 §6 amendments added to plan steps 10.1, 10.2, 10.3 and 11.1, and to CR-001 Step 10A.16 (§0.5.37–41 were added with the CR). Full suite (629 tests), coverage, lint and web build green. **Review screenshots:** `docs/screenshots/10c10-final-{tutorial,midgame}-{desktop,phone}.png` (the tutorial map also shows the undiscovered locations), plus the per-step shots from 10C.5–10C.9. **Phase 10C complete — stopped for review.**