Merge commit 34906c9a resolved the end-of-list conflict in scripts/format-scope.json with the branch's side, which dropped the 97 entries main had appended since Oct 26. Those files all still exist; 64 of them are tests that runtime discovery does not cover, so format:check had stopped checking them.
Each entry goes back at its position on main before the merge. All 1680 files pass npm run format:check.
A zoom in during a pivoting tilt stops at the range that keeps the camera 150 m above the pivot at its current pitch. A zoom that moves the range by less than half a metre (at the clearance, the 80 m floor or the 5,000 km cap) now answers not zoomed instead of ok, and a tilt reports tiltRangeAdjusted rather than orbitRadiusAdjusted.
#1308 moved /api/overpass to the keyless cross-site gate: foreign origins
and cross-site fetches are still refused, but forwarding headers are not,
so the map works behind a reverse proxy. Voice place search spends the same
operator Overpass, so it now uses the same gate; with the stricter one,
place search failed behind a proxy while the map kept working. The prompt
lint note records the merged size, 55,370 with an operator Overpass.
Picks up #1308 (Overpass admission), #114 (Sea Marks and Ski Pistes),
#1206 (radio country coverage) and #1435. Resolutions:
- SECURITY.md: main's rewritten cross-site paragraph, with voice place
search listed beside /api/overpass among the keyless map routes.
- scripts/format-scope.json: both sides. A merge on main dropped 97 entries
(being restored there); keeping them, this PR's 12 and main's 2 new ones
keeps every file in format:check. All 1,222 paths exist.
- Translation catalogs: main's two new keys and this PR's 47, no overlap.
- Layer count 36 (33 + OSM Places + Sea Marks + Ski Pistes); both schema
pins re-derived, the first-run length is exactly 30,919 + 292 + 107.
The far-zoom heatmap now uses the ½× grid (4096 × 2048, ≈10 km per
pixel, one-pixel blur), so a lone fire's spot matches its marker at the
3,000 km hand-off. The spot size slider and its share option are gone.
Markers and heatmap are both kept built at every zoom and cross-fade by
camera height between 2,400 and 3,600 km on every frame, so a fast zoom
never shows a beat with neither. The heatmap fades by imagery alpha:
hiding a Cesium imagery layer unmaps it from the globe tiles, and showing
it again re-mapped them late, which was the gap at the band edge. Under
the markers the heatmap is built 150 ms after the data or filter changes.
It takes the ICAO a position row carries, in any case, or the marker id, and selects through the same path as a click; selectAircraft now delegates to it.
A flyover, a launch-site view, a cancelled flyover or another mission now clears the pending orbit framing, so a late TLE lookup followed by a data refresh never takes the camera back.
Typed chat sessions no longer end on the voice inactivity timeout, and a
failed or stopped reply keeps the conversation: Stop cancels only the
reply in progress, and map actions that already ran stay in the model's
context. Each upstream request and map action gets its own deadline, old
tool results shrink instead of the history overflowing, and step limits,
a missing or rejected key, the OAuth lock and throttling each get their
own message.
- Server: OpenAI chat is off under GEV_PREFER_CODEX_OAUTH; chat throttles
default to the provider's throttle; replies carry API_KEY_MISSING,
API_KEY_REJECTED, CODEX_OAUTH_REQUIRED and RATE_LIMITED codes; upstream
Retry-After passes through, and a rejected key (Gemini's invalid-key 400
included) cools down for a minute
- Pinokio forwards the chat throttle settings
- UI: the reply reveal publishes only the message list; the border orbit
runs only while busy; layers attach with a button or Enter; the log is
aria-busy while a reply types out; auto-scroll respects a reader who
scrolled up; Escape closes the attach menu; the chat gear says Settings;
a note says chat bills the OpenAI key when voice uses ChatGPT sign-in
- Layout does less work while panel height animates; unused blurs removed
- Gemini voice start ends chat; the voice preview clears on every new
session
- Tests for the route gates, client limits, attachments and the fixes
Every scenario runs through scripts/qa-strict-harness.mjs: each page, frame and worker gets an ordered block list before it runs (paid hosts and routes first, then only the preview, declared fixtures, keyless probes and cached free tile hosts), switches with real input on the picker, and records its assertions per layer and engine. The runner reads the catalog from the live page, fails any layer without an assertion on either engine, can build and serve a keyless checkout itself, and adds rapid, hidden-tab, context-loss, failure-injection, cold-chunk and soak scenarios. The main chunk gets a 1% size budget.
Navara's transparent pass runs after the post effects without the scene depth, so a camera frame clamped in front of a wall still painted through it. scripts/qa-edge-cctv-occlusion.mjs serves a one-box tileset as a local fixture with no network and checks on both engines that the frame ends at the wall and is not visible behind it.
Sentinel-2 made the desktop source row wrap to two rows; the row now lays out every tile in one row (three columns on a phone, as before). scripts/qa-map-source-tray.mjs read the controller's removed imagery internals and counted five tiles between OSM and the styles; it now reads the live base provider and counts the section's tiles and engine picker. The five failures left also fail on origin/main e5d15f24.
The handoff test failed under a loaded test runner when the refined pass needed more yields than a fixed 5 ms timer allowed; it now yields until the refined paint lands (bounded by a task count, not wall time).
The Navara look stopped requesting frames once its globe is parked. scripts/qa-app-once.mjs now fails if anything moves a parked engine's camera or asks it to draw (its own resume aside).
A visual choice made while no globe is attached (a style key, a Display control, a voice command) is queued and applied to the next engine's settings after it restored the address's view, so the person's choice wins; reads answer from the released settings. The globe reset and a location lookup started without a globe do nothing, and the map-source chips and camera buttons are unbound until the next globe binds them.
The detection overlay asked the echo field for every visible flight,
military aircraft and vessel on every frame while a subject was
selected, projecting each to the ground before learning the pulses were
resting. It now checks awarenessEchoActive once per frame and builds no
echo paths at rest, and awarenessEchoAt checks the level before the
projection.
A detectionHost test with six echo-eligible flights checks that a
resting frame asks for the level fewer times than there are contacts
and paints no echo; it fails on the per-contact path.
The pulse and echo policy doc comments are attached to their constants
again, and one long comment line is re-wrapped.
The shell's scene part detaches it too, so a candidate failing in a later part (a layer's drawing, Cockpit clouds) leaves the shell free for the working engine instead of refusing every later attach. StyleManager.attachScene releases what it built when it throws, and one detach runs at a time. The controls phase awaits its parts, and a failed attachment no longer forces handoff off for the parts it releases. scripts/qa-app-once.mjs now also rolls back from a Navara candidate whose layer attachment fails.