The tile budget left 768 MiB for tiles over the screen buffers, while an
oblique view needs about 0.9 GiB at 960x640 and 1.2 GiB at 1440x900; over
budget Navara coarsens the view and stops loading, so the start view stayed
at imagery level 15 where Cesium shows 17. The room for tiles is now
Navara's device default (a quarter of device memory, at most 2 GiB).
Terrain tiles also split past the keyless mesh's level 14, so imagery can
refine beyond level 15, and OSM stops at its level 19.
Esri answers a tile it has no imagery for with a grey "Map data not yet
available" image (over open ocean from level 14). Cesium discards it and
keeps the parent level; on Navara that tile now fails the same way.
Brings in 190 commits from main. Integration on the engine contract:
- News and Shipwrecks draw, pick and keep their selection through the
contract and shared map markers, so they follow engine switches.
- Aircraft tracking uses main's segmented trails with contract frame
numbers.
- Decoded weather tiles move into the Cesium adapter and are injected
through the renderer registry; Navara hides those sources on globe
surfaces (docs/KNOWN-ISSUES.md).
- Street-vehicle GPU depth occlusion stays Cesium-only; Navara keeps
the geometric visibility checks (docs/KNOWN-ISSUES.md).
- Transit includes the heading-neutral atlas icons.
A live source may now declare a timeline: it can answer a snapshot at a
given time (source.timeline = { coverage(), stepMs, subscribe? }, read by
readSnapshotTimeline). While such a layer is on it follows the world clock
and registers the coverage. At the live edge it polls exactly as before;
away from it the layer asks for getSnapshot({ ...query, atMs }) at the step
at or before the clock time and shows that snapshot instead of polling.
src/time/snapshotTimeline.js owns the following: a seek, time going
backwards or leaving live replaces every contact with the snapshot at the
new time (play waits on it, for at most 8 s); playing on keeps one
snapshot ahead and gates play on it, applying those snapshots as polls so
contacts move smoothly; outside coverage the layer is empty; back at live
it is cleared and polls at once. One load at a time, and a load the
source drops for a newer one is not a failure.
While following, aircraft are drawn at the clock time with no render
delay, records and absences are judged at the clock, track requests carry
atMs, and vessels move between fixes along the great circle. The world
clock counts jumps (seeks and returns to live) in getState().
A source with no timeline never follows the clock, and every layer stays
live, unchanged.
The detection overlay only hid a contact behind the Earth's curve, so a
street vehicle behind or under a building, a 3D tile or a hill still had
its bracket and "VEH-nnnn" callout painted over that geometry, while its
own dot was depth-tested and correctly hidden.
A contact a layer marks `depthTested: true` (only Street Traffic does) is
now tested against the depth the scene drew and left out of the overlay
while a drawn surface stands in front of it by more than max(3 m, 2% of
its range). Other layers are untouched: their markers are drawn over the
scene on purpose.
The test runs on the GPU and never waits for it. Each tested contact is
one point drawn inside an occlusion query; the vertex shader looks up the
frame's packed pick depth at the contact's pixel, decodes it and drops the
point when the surface is nearer than the contact's threshold. Query
results are collected on later frames. Reading depth back was measured
and rejected: `scene.pickPosition` per contact costs 0.2 to 1.4 ms each,
and Chromium answers `getBufferSubData` with a round trip to its GPU
process even behind a signalled sync object (1.5 ms on a quiet machine,
15 to 25 ms on a busy one).
- Contacts are retested every 100 to 200 ms, spread across frames, at
most 256 a frame; with more in view the pace stretches so none starves.
- Hiding is immediate. A hidden contact returns 600 ms after its last
"hidden" answer, so one on a silhouette edge does not blink, and a
contact new to the view is held back until its first answer (400 ms at
most), so orbiting does not flash brackets behind buildings.
- A callout whose contact goes behind something is dropped at once, not
carried to the next solve and not faded out on the facade.
- These times run on painted frames, so a hidden tab does not expire them.
- A control point in every batch must draw; if it does not, the batch is
ignored and after three the probe stands down for ten seconds. With no
pick depth (2D, Columbus view, depth picking off) nothing is hidden.
A WebGL 1 context reads depth back synchronously about 8 times a second.
The renderer internals this depends on are read in one function,
`readSceneDepthSource`, with the engine version they were checked against.
Main added the News layer at the same places: both layers are kept in the
layer names, panel order (News stays last of Events), standalone sources,
voice manifest, credits, DATA_SOURCES and the CURRENT-STATE table. News took
share-link token 00 first, so Shipwrecks moves to 01 (ledger, registry,
CURRENT-STATE and the token tests). The catalog now has 40 layers and the
share-link registry 38. Shipwreck strings are added to the new Hebrew
catalog; the other catalogs keep both sides' strings.
They were listed as owned by the ghostmaps and ghostmaps-source packages,
which they are not part of; ghostmaps-source is meant to stay free of
rendering code, and shipwrecks/index.js imports Cesium. The app groups that
own every layer keep them.
The trail no longer joins a day of history into one line. It lifts the pen
across a gap of more than 48 minutes, an implied speed over 1,500 kt, or a
report without altitude (a parked aircraft within 0.1 nm for under 3 h stays
joined), and every history point is kept as its own segment. History never
gets an invented altitude. The live end is the path the aircraft was drawn
along, thinned to its shape, so it no longer cuts corners through turns; a
late report trims what was drawn after it. Old segments are frozen in chunks,
so a 2,000-point trail costs about 0.1 ms a frame.
A tracked aircraft that misses an update shows STALE and is dimmed. It keeps
flying on its last speed and heading for five minutes (at most 1,500 kt),
then stops, dimmed, labelled SIGNAL LOST with the time since its last report,
and is kept for ten minutes after that report so the camera keeps its target.
Aircraft on or near known ground hold at their last report instead of
coasting. Military flights follow the same rules.
The orbital view, sensor zoom, ground focus and satellite switching run on
the contract: camera.setFovy sets the vertical lens on both engines and the
fake, including sub-degree angles on Navara through the guarded core
bridge, with a contract check and mutants. Cesium's imagery look-ahead
stays inside its adapter. KNOWN-ISSUES keeps the two parts Navara still
lacks: imagery prefetch for the orbital view and finest-detail meshes for
the ship bridge waterline.
src/globe/surfacePlacement.js takes over the shared surface placer and
adds readSurface (tiles through sampleHeight, otherwise terrain, with a
plausibility range), surfaceTriggers (one subscription to camera moves,
tile loads, the photoreal switch and terrain revisions) and a forced full
pass. Points now re-place when the terrain provider changes, which they
did not before. GhostMaps, cables, OSM Places and Street Level move to the
new module unchanged.
src/i18n/, src/geo/ and the engine-neutral part of src/globe/ are owned by
directory entries, and aggregate groups use a directory wherever their
list covered it, so a new shared module no longer needs a manifest edit.
Small groups keep explicit lists, now checked as a snapshot: the check
fails on a reached module that is not declared and on a declared one no
longer reached, and prints both per group. --write keeps directory
entries as they are.
The manifest goes from 4,737 entries to 3,494; src/i18n accounts for 87
of them instead of 666.
A generated tileset (a plate at ellipsoid level, 50 m and 200 m blocks and
a pitched roof, 27 KB) replaces Google's tiles when a QA page passes
?qaTileset=/path; only a root-relative path on the page origin is
accepted, and the Google credit is not shown for it. The strict harness
serves the set from disk, so nothing reaches Google.
qa-tiles-fixture checks on both engines that sampleHeight reads the block
at 50 m, clampToHeight lands on the roof, tiles and terrain drapes draw
where they should, tilesLoaded turns true, surface points land on the
roof within their budget, and pick renders per pass stay within it.
Every camera find_alpr_cameras returns comes from the tile source, which builds
it through normalizeAlprNode, so the tags are already one line and capped.
Drop the second pass in the tool, its duplicate test and the boundary entry,
and the function check normalizeDisplayText never sees.
The contract suite is split into one module per area and grows from 87
to 114 checks, each named after the member it exercises: a gesture during
a flight does not cut it, a render request inside a frame draws the next
one, a pure rotation raises changed, ground corridors are as wide in CSS
pixels at device scale 2, imagery replans on resize, drapes draw only on
their surface, sampleTerrainMostDetailed keeps order and passes on a
rejection, terrainRevision changes on a new terrain and on fallback,
follows raise followChanged on retarget, and a mesh, a ground fill and an
imagery pixel draw the colour asked for. In the browser Cesium passes all
113 it runs at device scale 1 and 2; Navara passes 110 of 111 and reports
its single-colour mesh, drawn before tone mapping, as a known failure.
scripts/contract-mutants.test.mjs breaks one promise of the fake engine
at a time, 30 mutants in all, and names the check that catches each; the
list of contract members the suite never calls is down to five.
maps.errors reports tile and terrain errors where an engine has them
(the tileErrors capability: Cesium yes, Navara no), and the map
controller falls back to another map only where it does; movers.list()
rows tell whether a set is shown, as the other families do; FollowHandle
.name and PickResult.primitive are marked deprecated. qa-contract-suite
takes --dpr and the ANGLE backend from QA_ANGLE.
src/data/geoImportPlan.js decides what an imported file draws and with
what: raised or clamped geometry, rings and holes, rotation near the
poles and across 360 degrees, absolute and relative altitudes, draw
order, canvas budgets and the vertex and ring limits, with no engine and
no DOM. drawImportedFile draws that plan through the globe contract and
keeps pin placement, yielding and cancellation. Every limit of the import
lives in src/data/geoImportLimits.js with what it protects.
scripts/fixtures/geo-import holds 13 files covering those cases, and
reference.json what Cesium's own KML and GeoJSON data sources make of
them; scripts/qa-geo-import-reference.mjs records it and compares the
plan with it. Property tests on seeded random polygons check area,
bounds, ring closure and winding. Writing the plan fixed three cases:
overlays below the ellipsoid kept a negative height, raised plain-colour
overlays now draw, and the vertex limit no longer depends on whether a
ring repeats its first point.
Conflicts:
- docs/CURRENT-STATE.md: keep main's hidden-tab polling paragraph and the
display-text normalization paragraph.
- scripts/package-boundaries.json: keep main's voice panel and text chat
modules and add src/data/displayText.js.
- src/voice/gevActions.test.mjs: keep main's set_panel_open and partial
analyst card tests and the display-text tests.
Main's work since the last merge runs on the globe contract where this
branch draws through it:
- Satellites take main's classes, sensors, ground-track bearings and
world time through contract points and rings; launches take the world
timeline. The satellite's cockpit subject dims, copies positions and
refocuses through the contract.
- Flights, military and vessels keep main's photos, retry cool-downs,
world-clock motion and the ship bridge. Trail seeding in the shared
aircraft tracking reads each layer's clock.
- Traffic's flow lines and single build draw as ground sets; transit
keeps dated credits and the new cities; CCTV opens the nearest camera,
and the Quebec overlay projects and tests the horizon through the
contract. Bikeshare's manager cleanup moved into its current lifecycle.
- Draw keeps main's mark actions across engine switches.
- The world clock in src/time is plain numbers again: Julian dates come
from src/geo, the frame loop is injected, and the binding to Cesium's
viewer clock lives in src/globe/cesium/worldTime.js. Navara's sky and
sun follow the world time; live feeds keep the real time.
- The cockpit's orbital view, with its sensor zoom, ground focus and
satellite switching, stays Cesium-only for now: Navara hides its entry.
Its Cesium calls live in src/globe/cesium/orbital.js. The ship bridge's
finest-detail waterline sampling is Cesium-only too; Navara samples
loaded geometry and detailed terrain. Both are in KNOWN-ISSUES.
- Main's tests that built a Cesium scene by hand run on the fake engine.
- The boundary manifest is regenerated, the coupling baseline records
the orbital and bridge code, and the main-chunk baseline is raised to
the merged size: main itself grew its main chunk by 93,959 bytes
(gzip -9) since the previous merge base.
Tracking a civil flight asks Planespotters for that aircraft's photo, once
per aircraft, straight from the browser. The CONTEXT panel's CONTACTS view
shows it under the flight's name with the photographer's credit, as a plain
link to the photo's page. Only images on plnspttrs.net and pages on
www.planespotters.net are ever shown or linked. Military aircraft and a
converted TR-3B are never looked up.
Flight lookups (type, route, photo) whose source failed temporarily are
asked again after a 60-second cool-down; an answer, found or not, and a
permanent refusal stay final for the session. A route belongs to the
callsign it was looked up for, so a callsign change drops the old route.
Browser QA scripts that track a flight answer the photo lookup themselves,
so they never call the third party.
- KML/KMZ import: a per-file canvas budget and cache for ground overlays,
limits on rings and vertices of raised polygons, overlays across 360° and
near the poles, degenerate holes, polar and single-latitude fills,
absolute overlays at their altitude, overlay order by drawOrder as on
Cesium, and pins on tiles keep their measured height.
- Points on the surface: OSM Places and Street Level cones share the
surface sampler and sit on the terrain before they are measured, within
a sampling budget per pass; Cockpit's contact-dot bank keeps the dot
atlas, and a failed fleet model loads again after a pause.
- Startup: the loading screen lifts after the share-link view is restored,
the retry screen names the failed engine and takes focus, and the
fallback toast stays for 6 s.
- Engines: an engine switch lands a running camera flight at its
destination (camera.completeFlight on both engines), the sighting card
leaves focus on the engine picker after a switch, and a Cesium model
error stays with the instance that failed.
- The MCP panel builds again: its wind-mesh alias resolves to the stub.
- QA: the wildlife switch waits for the camera to rest, GBFS --w2 says it
needs the dev server, and test fixtures are shared instead of copied.
- Add the full catalog (2940 entries keyed by the English text; 75 plural
entries carry one/two/other per Intl.PluralRules('he')), reviewed for
meaning, terms, separators and label length.
- Register he: he-IL and he map to it; it joins RTL_LOCALES so <html
dir=rtl> mirrors the chrome through logical properties and rtl.css.
- rtl.css: Noto Sans Hebrew (loaded with the other Google fonts) leads the
mono and sans stacks, tracking released like Arabic, HUDs keep JetBrains
Mono for Latin telemetry.
- same-as-english: instrument codes and HUD readouts stay as printed
(HDG/BRG/CRS/FOV/CRT/NVG/FLIR/CAL/SRC, MW units, Sentinel-2, km/h...).
- docs/translations/he.md: style, plurals and glossary.
- docs: TRANSLATORS.md and CURRENT-STATE.md updated; normalizeLocale test
covers he-IL.
Navara now matches Cesium where the review found it diverging: a flight
holds the camera inputs, camera.changed counts turns as Cesium does,
render requests made during a frame survive it, follow pinches zoom,
ground widths follow the pixel ratio and Cyber sonar reaches glyph caches
and models. On Cesium, a staged imagery swap keeps loading tiles, setView
replaces a running flight, and a cleared weather detail goes away.
The contract gains MoverSet.failed, maps.terrainRevision,
render.maxTextureSize and surface.sampleTerrainMostDetailed on every
engine, so layers retry failed models, key road heights by terrain,
size weather textures to the GPU and place Bhote Koshi cards on detailed
terrain.
Engine switches keep the live view, follows, focus, layer options and
cards: a pick hands over the exact pose, applies no address options and
pans no carried-over card; a history visit still applies its entry.
Layers put GhostMaps pins, cable landings, OSM Places and directions on
the measured surface without per-frame sampling, draw the Cockpit contact
dot, hide disabled CCTV on attach, and keep ALPR running outside secure
contexts. KML imports draw raised polygon fills, rotated and coloured
ground overlays and north-referenced icon headings, and report files
with nothing to draw.
QA scripts read the contract instead of Cesium internals (test:track is
109/109 again), wait for the camera to rest before clicking, and give
cold switches and parked engines realistic time. Docs follow the code.
The orbital Context card is gone and the Contact card stays hidden in orbit
again. Live Signals page 1 had nothing in orbit but TRACK ACQUIRED, because
orbit feeds no Contacts snapshot; it now lists what is around the
satellite, built the way the aircraft rows are, five rows with stable keys
rebuilt once a second. Rows already shown keep their places while they
still qualify, so a site passing under a low orbit is not pulled from under
the pointer.
Site rows come from the Mapped Installations layer (getNearby, asked only
as far as the reach extends) and focus their site, or release it on the
focused site's row, which reads as the subject does. A site the layer drops
during its frequent reloads stays listed for 10 s, since sites do not move.
With the layer off nothing is listed and a SITES STANDBY row says why.
Satellite rows come from the satellites layer's bounded nearest query,
measured from the satellite itself, and ride the satellite; RETURN goes
back to the one before, skipping one that has left the catalogue. While
the camera hops to a new satellite only TRACK ACQUIRED and RETURN show, so
nothing is measured from the old position. An arrow pressed while a focus
eases back takes the arrows' own clock. Site and satellite rows name their
own actions.
In the satellite orbital view a single click on a mapped ground point (a
ground point of a layer that is on, such as a Mapped Installations site, a
fire or a local data pack feature) keeps the camera on it. Each frame the
look is aimed from the satellite's position to the point's Earth-fixed
position in the track-up nadir frame the arrows tilt, so the point stays
at the centre as the satellite moves. The swing onto the point, or between
points, uses the arrows' easing on a slower clock and then holds the point
exactly; reduced motion cuts.
A click elsewhere, FOCUS on the focused point, NADIR, Home or 0 ends the
focus and the view eases back to nadir; an arrow ends it and tilts on from
where the view is. The focus also ends, easing back, when the point leaves
the ±45° reach or the satellite's horizon, when its layer is turned off or
its record is dropped, and with the view. While focused the readout gives
the single off-nadir angle to the point (OFF-NADIR 23.4°), names it in a
polite live region, and the swath reads approximate as for a tilt. Clicks
still reach Cesium's own handler and never untrack the satellite; a click
a tool owns is ignored.
Under the readout, IN REACH lists the mapped points facing the satellite
that are within the reach or in the frame, nearest to nadir first and at
most 12, rebuilt once a second. A row opens the point's position, source
and OpenStreetMap tags (or its layer's properties); FOCUS focuses it,
switches to it, or ends the focus held on it, and is unavailable for a
point in frame but beyond the reach.
An imaging satellite with a known sensor can now zoom past 8x, up to 128x,
as far as the view stays at least as wide as the sensor's narrowest
standard swath at its live altitude; only swaths confirmed on a cited page
open those steps. The top step follows the altitude and window shape and
steps back when it falls below the zoom shown.
The view draws the sensor's swath to scale along the ground track with a
SENSOR SWATH readout (approximate off nadir), and a SENSOR button zooms to
exactly that width when it lies in range at nadir. From 16x the imagery
tiles the view will need 5 to 20 s ahead along the ground track are
pre-loaded at the matching level. The mouse wheel steps the zoom one notch
at a time, and the identity line shows how old the TLE is. The sensor
figures come from a new bundled pack of 46 Earth-observation platforms,
each value cited to its operator or agency, with its own data credit.
With a satellite tracked and no aircraft tracked, Cockpit can ride the
satellite. Entry needs only the Satellites layer; debris can be tracked but
not entered, and voice stays aircraft-only. Double-clicking the tracked
satellite or the entry buttons (reading ORBIT) enter it.
The camera sits on the satellite's per-frame SGP4 position and looks
straight down, turned to the ground track (north-up for a satellite with
none). The optical zoom steps 1x to 8x with a focus pull, the arrows tilt
the view up to 45 degrees off nadir, and Home or 0 returns to nadir. Off
nadir the tilt reads its total angle first and the swath readout becomes an
approximation marked with ≈. While in orbit the map leases flat Esri
imagery, the cloud pass stays off and the view opens in the optical style.
The orbital HUD names the satellite with its owner's flag (from the
flag-icons package, loaded per country) or an organisation badge, its
operator, class, NORAD and COSPAR ids and TLE epoch with the element lines
on request, altitude, orbital speed and the sub-satellite point. CURRENTLY
OVER names the country and its first-level division, or the sea, and from
8x the most prominent populated place nearby, looked up offline from the
bundled Natural Earth packs (populated places are a new pack). The
satellites layer hides the ridden satellite's own marks, dims the other
orbit rings and ignores scene clicks and Escape while held. Exit restores
the camera's field of view and the satellite's follow camera.
C no longer enters or leaves Cockpit, bridge or orbit: it is left to the
app-wide CCTV shortcut, and Escape or the exit button leave first person.
Only catalog and the tracked satellite were saved for the layer, so a class
chip lasted until the next reload and no share link carried it. The choice
is now the satellites option `classes`: one letter per class shown
(`sitpd` by default, `0` for none), with Starlink left to `catalog`, and
the layer takes the letters back through setParams. A partial request
(`{debris: false}`) hides only what it names.
The satellites layer now sorts every object into Stations, Imaging,
Telecom, Starlink, Probes or Debris, each with a chip and a live count on
the layer row. Stations and Imaging show by default; the Starlink chip is
the former DENSE catalog switch. Hidden classes keep their points, so a
toggle needs no reload, and they leave the detection overlay and the voice
position lists.
A class comes, strongest evidence first, from CelesTrak SATCAT (object type
and status, served by a new /api/celestrak-satcat/<group> proxy route and
cached for a day), the bundled UCS Satellite Database pack (operator, owner
country and purpose), name patterns for later launches, and the CelesTrak
group. The core catalog also reads the resource, weather, planet, sar and
science groups. Payloads with no evidence of life launched 20 or more years
ago are filed as debris. The UCS pack, its build script and its data credit
are added.
The ship bridge rules were appended to cockpit.css, taking it to 2705
lines; they move unchanged to cockpitBridge.css, imported right after it so
the cascade is the same. cockpitBridge.css and cockpitRadar.css join the
format scope and the styles package.
Layer display tests no longer re-check Cesium primitives in every layer:
they run on the fake engine, and one Cesium translation check per ground
kind lives in the Cesium engine's ground tests. Their harnesses are shared,
and the fake-engine, Cesium-engine and flights tests that repeated the
contract suite are gone, 340 tests in all.
The press-travel measurement is shared by the selection dispatcher and
CCTV, which loses its ScreenSpaceEventHandler shim and world-click module.
ALPR, installations and bikeshare share one selection keeper and selected
marker write. Layers with their own click rules keep their handlers.
Navara's pulse is a class like the ground pass, the ground and imagery
passes share the surface uniforms, and the colour, log-depth and occlusion
GLSL lives in one chunk module, each caller keeping its depth bias. The
ground suite runs inside the contract suite (--only=ground:), and fixture
and script names say what they test.
Behaviour is unchanged on both engines.
Every layer draws on both engines, so the per-engine availability
machinery always answered yes. The renderer manifest is now the id to
display map plus the wind flow backend; the catalog filter, the voice and
QA-probe refusals and the engine-kind checks go with it. The Navara
notice lists the same layers in the same words.
Paths that read the Cesium viewer behind an engine that is always there
are gone: the Recent Imagery viewer fallbacks, the shell's imagery-host
service (one shared host module remains), the engineless map controller,
viewer gates in voice actions, world packets and camera verbs, the
orientation controls' tracked-entity fallback, and the Bhote Koshi and
FIRMS Cesium-shaped paths. setCameraInputs, which nothing called, leaves
the contract; suspensions remain.
Smaller duplicates go too: one tracker-release helper for navigation and
location, the no-op frame latch, missing-engine guards, one name for the
injected display, flat bikeshare and installations displays, one overlay
host and binder, one name per world-overlay method, and stub-viewer
guards in the Cesium engine. Layers that still read the viewer keep the
shared init(viewer, context) entry.
Behaviour is unchanged on both engines. Viewer lines fall from 849 to
750 and engine-name literals from 32 to 27.
Remove code that nothing in the app, the runners or CI uses:
- the one-off migration compare, look and bench scripts and the vector
ground prototype they measured, with the bench tables in the docs;
- the unused Navara-named shared render modules (polygons, ground picks,
mover bank, ground ring);
- the Cesium-shaped aircraft and vessel test renderers, and the trail,
engineFromViewer and horizon-occluder helpers only they kept alive.
Their tests now run on the production displays over the fake engine;
- Navara's unread diagnostics, fallback ports, dead exports and
parameters, stale effect-chain anchors and the unread capability flags.
The QA scripts share one keyless guard and readiness wait, the session
harness reuses the strict harness's browser flags and block list, and the
harness honours QA_ANGLE and QA_RESERVED_PORTS. The bikeshare fixture
keeps only what its scenario reads. The docs name the engine picker where
it is (VISUAL PRESETS) and drop dated run results.
No behaviour changes on either engine. Navara layer renderer lines fall
from 189 to 0 and viewer lines from 891 to 849.
German Khokhlov