Under Google 3D tiles Navara hid only the base imagery and kept drawing
the terrain globe, which showed through the tiles wherever the terrain
mesh lies above them (grey patches over the Nepal mountains). As on
Cesium, the globe now stops drawing under the tiles while terrain
sampling continues, and the tileset is kept across map switches instead
of being fetched again.
Sea Marks and Ski Pistes drew as native globe rasters and would vanish
with it; they now go through engine.imagery, which drapes onto the tiles
on both engines, and the native overlay path is gone.
Navara opened its attribution card to keep on-screen credits visible,
covering the scene; it now shows them in a persistent strip like
Cesium's, with the card behind its button.
Main's own main chunk grew 232 KB since the baseline (the News layer and
the Swedish catalog), measured on keyless builds of main; this branch
still adds 72 KB on top of it.
The shared launcher always asked for ANGLE's Metal backend, which exists
only on macOS; on CI's Linux runners WebGL failed and the Street Level gate
never saw the app start. The default is now Metal on macOS and SwiftShader
elsewhere; QA_ANGLE still overrides it.
The tileset's transform and probe points come from trigonometry whose last
bits differ between arm64 and x64, so a fixture written on one failed the
byte comparison on the other. The numbers now match to 1e-12 of their size;
the meshes and the served files are still compared byte for byte.
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.
Both probes called waitForEngine without the engine kind, which can never
match and timed out after 120 s; waitForEngine now rejects a missing kind.
The trail probe passed no result vectors to the geo helpers, and drew its
trail below the ground, where only the dimmed occluded pass showed it; it
now draws halfway down to the ground and both engines count the same
pixels.
Translate the 148 strings main added since (mostly the News layer) and
prune the 5 it no longer uses; keep both Hebrew and Swedish in the
language lists.
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.
Roads the view cannot sample may keep their pending surface heights, as
before the refactoring; the oblique wait now asks for current coverage,
drawn traffic, finished loading and no scheduled refinement, and prints
every loading field when it times out.
Field entries in the ledger allowlist now apply to every difference row,
in the ledger and in the probe alike, and may carry maxDelta. The traffic
fill gets one with a bound of three: its simulated vehicles enter and
leave every frame, so two pages snapshotted moments apart differ by a
vehicle or two.
The FLY check slept a fixed 3 s and then expected a route flight, which
passed or failed on the same commit. It now waits, up to 15 s, for the
layer to report the route motion and the camera to move.
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.
Removing one source's wrecks on the page split merged records: a wreck both
sources list is one Wikidata record carrying EMODnet facts, so expiring
Wikidata dropped wrecks EMODnet still had, and expiring EMODnet kept its
facts. The page now keeps every shown wreck, marked stale, until the newest
source passes the limit, then clears them all. Adds a source test for the
reported ages and limit.
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 ~8,000 markers were one Cesium entity each; they are now one
BillboardCollection (as FIRMS, vessels and wildlife sightings do), ordered
at the bottom of the shared sprite stack so moving traffic draws over it.
- eyeOffset.z was the death toll in metres, so a 1,500-death wreck's marker
sat 1.5 km toward the camera and clipped or drifted when close. It is gone;
markers are added by rising toll, so a bigger toll still draws on top.
- The page asks every 6 h but the proxy changes data at most daily: the proxy
now sends a version that changes with either source's data, and the layer
rebuilds only when it does.
- Adds a controller test: enable/disable, teardown, draw order, unchanged
snapshots, a late response after turning off, and the period closing an
open card.
The scenario counted every sampleHeight call in a task as the placer's
pass, so a single sample from elsewhere in the app that landed in the
same task pushed a full pass to 17. Calls that are neither a fixture
point nor the one camera estimate are now counted apart.
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.
CCTV kept the coverage and projection geometry of every camera the
opening view passed on its way to the restored one; after each visibility
update it now releases the coverage of cameras out of view and the planes
of inactive cameras, keeping decoded media for reselection.
The cold-start comparison accepts a field entry in the ledger allowlist
that excuses one field of one set with a reason, and uses it for the
infrastructure footprint counts: that level of detail keeps features
chosen on the way to a view so they do not flicker while panning, which a
cold start cannot reproduce. Presence checks ignore field entries. The
fixtures give recent imagery a matching scene and date and traffic its
deterministic starts, and traffic counts as drawn without every loaded
vehicle visible.
The check for requests to Google and Cesium ion caught the Google Fonts
stylesheet the page loads on every run, which the map policy allows; it
now looks only for tiles, credits and ion.
With its cache budget on, Navara 0.2.1 deletes a source without emitting
the model-removal events, so the meshes of a released photoreal tileset
stayed in the scene: after switching to another map the buildings were
still drawn, and sampleHeight, clampToHeight, the shared surface
placement and ground drapes still answered from them. The adapter now
disposes the deleted layer's meshes, registry entries and retained
geometry itself, also for tiles that finish decoding afterwards, and
WORKAROUNDS.md records it. A contract check fails when a surface query
still reports tiles after leaving photoreal, with a mutant.
The CCTV camera check lost camera B from the picker when the flight to
camera A changed the cameras in view; the picker now keeps focus while A
is held. The oblique traffic wait compared exact seven-decimal view keys
that a settled camera never reproduced; it compares the coverage with the
current view within small tolerances and prints both keys on timeout.
With --report the scenario prints, on Navara, the scene, globe and
combined depths at the sample pixel with the drape hidden, shown and
hidden again, the buffer the ground pass classifies it as, and whether a
shown tile lies on the ray, so a drape on the wrong surface explains
itself.
Both pages now get the same deterministic SATCAT, NASA image and wildlife
data and advance animation clocks alike; the scenario waits until each
display has published and the registries stay unchanged for two seconds,
and late-created content is attributed to its registered owner. Known
allocation counters are normalised; every slot, count and visibility is
still compared.
Infrastructure drew on attach before the camera was restored, seeding its
level of detail at the incoming engine's default view; it now draws when
the view is restored.
A request-mode globe updates its tile state only on drawn frames, so tile
and UI readiness that never asked for one waited until something else
drew: imagery, weather and ground scenarios took several times longer.
The idle checks measure through their own passive observer and are not
affected.
The harness conversion replaced pauses that stood in for real conditions
with camera readiness: the cables and infrastructure idle checks started
measuring while the overlay was still fading in after moveEnd, directions
sampled before its maneuver dots had ground cells, three scripts clicked
before the loading cover was gone, the credits check hit-tested during UI
transitions, and the oblique traffic flight lost its height cap. Each now
waits for its condition; idle checks observe without requesting frames
and record the reasons of the frames they count; readiness waits for the
startup cover and no longer requests frames while polling.
Navara's ground pass took a pixel for a tile whenever its depth lay in
front of the saved globe depth, without checking which buffer the visible
depth came from; the scene buffer can be cleared apart from the globe
depth, so a set draped on 'tiles' also drew on bare terrain, which Cesium
never does. The pass now matches the visible depth to the globe or the
tiles before it drapes. A contract check compares a 'tiles' set hidden,
shown and hidden again over terrain, with a mutant.
A Space Missions replay flies the camera until it ends, so waiting for a
still camera after a switch let the replay finish before the scenario
looked at it. While a replay runs, readiness no longer asks for enabled
inputs or a still pose, as for the cockpit and a followed target.
While the cockpit holds camera inputs or the camera follows a target,
whenSettled no longer waits for enabled inputs or a still pose; it still
waits for the engine, the restored view, drawn frames and no flight, and
its timeout now prints what it saw (inputs, following, cockpit, last pose
change). Scenarios that waited for the camera where they needed the
picture wait for it instead: weather and Nepal imagery for their tiles
and frames, imagery capture while a box drag holds inputs, and pixel and
pick polling for drawn frames. The vessel click retries at the vessel's
rendered position.
Navara's 3D Tiles source defaults to model.height 1, which moves a
transformed tileset by a metre along Earth-fixed X: on the synthetic set
flat roofs read 0.12 m high and a sloping roof 0.68 m off, and the drawn
meshes move with them. The adapter sets the height to 0 and records the
package behaviour in WORKAROUNDS.md, with a test that loads the real
package. The downward surface ray now takes its direction from the
place's own normal, so it no longer drifts sideways.
qa-tiles-fixture gains --report (heights and drape pixels per engine,
without asserting), starts on Cesium, and checks each drape hidden, shown
and hidden again; qa-surface-heights uses the plan's 0.5 m tolerance under
tiles and names the point that misses.
qa-switch-equals-cold-start switches engines in both directions with each
layer on deterministic fixtures, then opens the share link of the same
state in a new page on the target engine, and compares the draw ledgers
and the probe's view of the layer. It runs as 19 acceptance groups
covering all 38 layers.
State a layer keeps through a switch but the share link does not carry is
declared as sessionState (the directions route, Radio's selected station
and audio state, the wind reading, local ADS-B's and OSM Places' local
toggles and results); a test checks each declaration against what the
layer really retains, so a declaration cannot hide a regression.
Tests that made their own engine object now use createFakeEngine() with
its scripting hooks or an override of the one member under test, so they
follow the contract when it changes. 125 such literals in 74 test files
go down to two that are not globe engines on purpose; report-qa-harness
counts them, and the baseline keeps the number from growing.
German Khokhlov