Commit Graph
1594 Commits
Author SHA1 Message Date
German KhokhlovandClaude Opus 5.5 b00c7c12c8 refactor(sky): cut what the sky does not need
- qa-anime-sky keeps only the Milky Way case; the satellites and keyless
  cases checked an anime shader this branch no longer changes.
- Drop qa-sky-sun-moon.mjs: the viewer-options unit test covers it.
- VisualEffects sets the sky's visibility directly: the viewer always
  carries a StarrySky and never outlives the effects.
- Drop the unused star intensity knob, the malformed-row tolerance for
  our own pack, the test-only Transforms parameter and a private-name test.
- DATA_SOURCES points to sky/README.md instead of repeating it.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-20 13:45:36 +04:00
German KhokhlovandClaude Opus 5.5 d94cc35c81 fix(sky): keep the Milky Way out of model lighting, test the sky wiring
Scene reads skyBox._cubeMap as the environment map that lights models, in
the TEME, flipY layout of Cesium's own sky box; the ICRF Milky Way would
light them rotated. Rename the field so models keep Cesium's default.

Test the anime fade on a real StarrySky and that the viewer gets the sky
as an option (skyBox: false would drop the Sun and Moon). Correct three
comments on the star limit and the Milky Way levels.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-20 13:37:43 +04:00
German KhokhlovandClaude Opus 5.5 86ae5e9cf4 fix(sky): blur the faint Milky Way glow wider so unresolved stars leave no clumps
The fine 0.4 degree blur kept the residual glow of unresolved stars as
mottled patches at the band's edges. Blend towards a 2 degree blur where
the coarse copy is faint (smoothstep over 0.5-8 % of the range), so the
core keeps its dust lanes and the faint sky goes smooth. Faces 85 -> 62 kB.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-20 13:02:48 +04:00
German KhokhlovandClaude Opus 5.5 e8fd1f2d01 fix(sky): ease the Milky Way black level in at the median sky
Away from the galactic plane the ESO panorama keeps a mottled glow of
unresolved stars and dust. With the black level at the darkest quarter of
the sky and a hard cut, it showed as grey clouds with black holes around
the globe. The black level now sits at the median sky and eases in with a
soft knee on luminance, keeping each pixel's hue: high galactic latitudes
are black, the band keeps its glow. Faces shrink from 247 to 85 KB.

The changelog fragment also notes the anime fade and the new asset weight.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-20 00:42:23 +04:00
German KhokhlovandClaude Opus 5.5 8deb47560e fix(sky): restore the Milky Way on teardown, tidy the anime QA script
- VisualEffects.destroy() puts skyBox.milkyWayVisibility back to 1, as it
  already restores the borrowed bloom stage, so a surviving viewer does not
  keep a faded sky.
- Rename scripts/qa-anime-space-mask.mjs to qa-anime-sky.mjs: rewrite the
  header for the current approach, drop --space-expr and --repo, import
  puppeteer as a package, reject unknown cases and count the milkyway case
  as reached only while the anime stage is on.
- Put both sky QA scripts under format:check.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-20 00:42:22 +04:00
German Khokhlov 65b98b4971 Merge remote-tracking branch 'origin/main' into claude/hal-627-starfield-cb6fa2 2026-10-19 23:06:49 +04:00
German KhokhlovandClaude Opus 5.5 e652e50d5c fix(anime): fade the Milky Way out instead of masking the sky by depth
The far-plane mask in the anime shader read the post-process depth, which
cannot tell the sky from far geometry: with a layer that clamps the near
plane to 0.1 m (satellites) the whole Earth lost its cel bands, keyless
terrain above the ellipsoid horizon did too, and the atmosphere, Sun and
Moon stopped banding.

Restore the anime shader from main and fade the Milky Way glow with the
style instead: VisualEffects sets skyBox.milkyWayVisibility to
1 - anime intensity. scripts/qa-anime-space-mask.mjs checks all three
cases headless and that the anime sky shows no bands.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 22:27:28 +04:00
Said Turksever 5c513684f5 Merge pull request #1201 from Danielslay86/cctv-stills-aspect
cctv: draw stills at source aspect
2026-10-19 17:46:05 +01:00
Said TurkseverandClaude Opus 5.5 ba5a1e25ee Merge branch 'main' into cctv-stills-aspect
Resolves the scripts/format-scope.json conflict with #1200 by keeping
both new CCTV test entries.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 17:32:04 +01:00
Said Turksever 7fe95b76e9 Merge pull request #1200 from Danielslay86/cctv-live-icon-color
cctv: color live video cameras apart from stills
2026-10-19 17:29:05 +01:00
German KhokhlovandClaude Opus 5.5 adeee12209 docs(changelog): add the starfield sky entry
Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 20:11:32 +04:00
Said Turksever 2120130997 Merge branch 'main' into cctv-stills-aspect 2026-10-19 17:10:53 +01:00
German Khokhlov 7d4da26905 Merge remote-tracking branch 'origin/main' into claude/hal-627-starfield-cb6fa2
# Conflicts:
#	scripts/format-scope.json
2026-10-19 20:09:55 +04:00
Said TurkseverandClaude Opus 5.5 29fb756a5c test(cctv): cover the focus de-emphasis icon colors
refreshCctvFocusStyles passes cameraColorFor as baseColorFor, but only
refreshCoverageStyles was tested. The new rendering.test.mjs runs the
focus pass through createRendering with stubbed focus services and
checks that an idle video camera fades from violet, a still from cyan
and the active camera from amber. It fails if the focus callback goes
back to the old active/idle check. Also adds the missing semicolon on
the policy.js import in cctv.test.mjs.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 16:50:21 +01:00
Said TurkseverandClaude Opus 5.5 35c612c5f6 test(cctv): lock the still draw path and fit-aware dedupe
Drive drawProjectionFrame with a fake runtime and a recording canvas:
a 640x480 still draws at its fitted rectangle over the matte, the same
pixels at the same size skip the redraw, and the same pixels at a new
aspect redraw at the new fit. The last case fails on the pixel-only
signature; the first fails if drawImage stretches to the canvas again.
Expected rectangles now read the canvas size from policy.js.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 16:50:21 +01:00
German KhokhlovandClaude Opus 5.5 fb81e155b7 fix(sky): keep the Sun and Moon, settle outside the ICRF tables
- Pass the custom sky box to the Viewer instead of `skyBox: false`, which
  also stopped CesiumWidget from creating the Sun and the Moon. Guarded by
  scripts/qa-sky-sun-moon.mjs.
- Outside the IAU 2006 tables (before 1975, after ~2050) the preload
  resolves without data; remember that span and stay on TEME instead of
  preloading and rendering every frame.
- The anime style no longer cel-shades the far plane, where four bands
  turned the faint Milky Way into blotches.
- Paint the planets out after the star-removing opening and with the
  ring median, so no brighter plateau is left; faces rebuilt.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 19:43:25 +04:00
Said Turksever 86eca5abf3 Merge branch 'main' into cctv-live-icon-color 2026-10-19 16:26:45 +01:00
Said Turksever 0f51eb21ba Merge pull request #816 from kvnloo/fix/context-radio-owner-invariant
fix(ui): dismiss compact Radio on Context rail restoration
2026-10-19 16:23:26 +01:00
Said Turksever 9f6e0ef60f Merge branch 'main' into fix/context-radio-owner-invariant 2026-10-19 16:15:15 +01:00
Said TurkseverandClaude Opus 5.5 d80683c44b docs(changelog): end the #1200 fragment with a newline
Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 16:05:38 +01:00
Said Turksever 56c5d03b65 Merge upstream/main into cctv-stills-aspect 2026-10-19 16:05:15 +01:00
Said Turksever b8bfeb7aa2 Merge upstream/main into cctv-live-icon-color 2026-10-19 16:04:58 +01:00
Said Turksever 7214f71878 Merge pull request #880 from kvnloo/feat/world-packet-boundary-clean-20260929
feat(agent): extract model-free World Packet boundary
2026-10-19 16:00:23 +01:00
Daniel Slay 8e635c9a3a test(cctv): pin the camera icon colors; say video feed, not live 2026-10-19 10:50:22 -04:00
Daniel Slay 1959b33c48 fix(cctv): test the still fit, redraw on size change, share the matte color 2026-10-19 10:39:41 -04:00
Said TurkseverandClaude Opus 5.5 daaaa348b7 test(agent): pin World Packet to the pre-extraction payload
The parity test compared getWorldPacket with voice, which now calls
getWorldPacket. Compare both against a fixed payload captured from main's
get_current_view_state before the extraction, key order included, and
cover a non-finite now.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 15:19:07 +01:00
Said TurkseverandClaude Opus 5.5 0f237a32d4 fix(agent): age World Packet feeds against the clock for non-finite now
A default parameter only covers undefined, so now: null aged every feed
against the epoch and reported stale data as just now.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 15:19:07 +01:00
Said Turksever 4ecbe54997 Merge bilawalsidhu/gods-eye-view main into feat/world-packet-boundary-clean-20260929 2026-10-19 15:19:04 +01:00
Said Turksever 3d8e1ef00f Merge pull request #605 from Danielslay86/cctv-florida-stills
feat(cctv): FL511 Florida statewide camera pack (stills)
2026-10-19 15:11:43 +01:00
Said Turksever 326c41df9c Merge branch 'main' into cctv-florida-stills 2026-10-19 14:56:33 +01:00
Said Turksever beca41c676 Merge pull request #1056 from quanticsoul4772/feat/tools-area-breakdowns
feat(tools): count and summarize every match in the area tools
2026-10-19 14:53:34 +01:00
Said TurkseverandClaude Opus 5.5 a29d5a6874 docs(changelog): credit kvnloo in the #816 fragment
Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 14:47:13 +01:00
Said Turksever 839442b20b Merge branch 'main' into feat/tools-area-breakdowns 2026-10-19 14:46:41 +01:00
Said TurkseverandClaude Opus 5.5 c174742e2c fix(cctv): FL511 written-consent wording and robust catalog paging
- State FL511's terms the same way everywhere: individual, personal use
  only, and any other use (a public deployment included) needs FDOT's
  express written consent. .env.example, README, dataCredits, the source
  licence string, the catalog comment and the changelog fragment said only
  "individual, non-commercial use".
- Page the ArcGIS catalog on exceededTransferLimit alone and advance the
  offset by the rows each page returned, so a lower server-side max
  record count no longer cuts the catalog short at the first short page.
- Order pages by OBJECTID_1 (this layer's object id; a plain OBJECTID is
  rejected with HTTP 400) so offset paging is stable.
- Raise the page ceiling to 40 and warn when it is hit.
- Tests: short pages with exceededTransferLimit keep paging, orderByFields
  is sent, and a service that ignores resultOffset stops at the ceiling.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 14:32:58 +01:00
Said TurkseverandClaude Opus 5.5 4c383214f7 fix(tools): average aircraft speed over airborne aircraft only
aircraft_in_area stats summarized altitude over airborne aircraft but
speed over every match, so parked aircraft pulled the mean speed toward
zero. Both figures now use the same airborne subset, and the on-ground
fixture is stationary so the test can tell the difference.

Reserved AIS ship-type codes 100 to 255 now read as unknown, as type 0
does, instead of showing up in the breakdown as their own type.

Adds tests for an area where every aircraft is on the ground and for
find_vessel type and destination normalization, and notes in the
changelog fragment that every vessel answer is now normalized.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 14:29:11 +01:00
Said Turksever 5d480a5fc2 Merge upstream/main into fix/context-radio-owner-invariant 2026-10-19 14:23:54 +01:00
Said Turksever 7712ecf6ed Merge pull request #975 from Sunil56224972/fix/terrain-same-site-gate-on-686
fix(server): refuse cross-site requests on the terrain-heights route (stacked on #686)
2026-10-19 14:14:54 +01:00
Said TurkseverandClaude Opus 5.5 256d7c273e Merge #686 head 558b0275 so its entry lives in changelog.d
#686 moved its changelog entry from CHANGELOG.md into
changelog.d/686-terrain-heights-bounds.md. Take #686's side so this
branch no longer edits CHANGELOG.md directly and no longer conflicts
with #686.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 13:56:15 +01:00
Said Turksever d921a1b801 Merge branch 'main' into fix/terrain-same-site-gate-on-686 2026-10-19 13:51:37 +01:00
Rudy Celekli ada7d8852f docs(changelog): record launch cache provenance PR 1199 2026-10-19 13:30:27 +01:00
Rudy Celekli 7844e765b4 fix(launches): retain stale-cache provenance in layer status 2026-10-19 13:30:27 +01:00
Rudy Celekli f747c4828e docs: add Launch Library changelog for PR 1197 2026-10-19 13:11:58 +01:00
Rudy Celekli a310cbc2f0 fix(space): align Launch Library requests with provider guidance 2026-10-19 13:11:58 +01:00
German KhokhlovandClaude Opus 5.5 6549344946 feat(sky): replace the blocky default sky box with catalogue stars
Cesium's default Tycho-2 sky box is six 1024² JPEGs stretched 3–6× around
the whole-globe view, so space read as an over-compressed JPEG (HAL-627).

The scene now draws its own sky in the Scene#skyBox slot: the 9,096 stars
of the Yale Bright Star Catalogue as point sprites sized in CSS pixels
(magnitude → width and opacity, B−V → colour), over a soft, star-free
Milky Way cube map built from ESO/S. Brunier's panorama (CC BY 4.0). Both
layers share one ICRF→Earth-fixed rotation at the scene clock, with the
TEME rotation filling in until Cesium's IAU 2006 tables load.

Assets total 482 KB, built by scripts/build-bright-stars.mjs and
scripts/build-milky-way.mjs. Frame time and the idle render governor are
unchanged.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-19 11:08:46 +04:00
daikaginza 558b0275a4 docs(changelog): move the entry into a changelog.d fragment
Main now takes pending entries as changelog.d/<pr>-<slug>.md
files (#787), so the entry leaves CHANGELOG.md for changelog.d/686-terrain-heights-bounds.md.
2026-10-18 13:36:06 +07:00
daikaginza 50559723a5 Merge upstream/main 93ea088f4 2026-10-18 13:36:06 +07:00
quanticsoul4772andClaude Opus 5.5 d39256e700 fix(tools): ship types read as their family; letterless destinations are unknown
AIS sends ship types as ITU-R M.1371 codes, so vessels_in_area grouped
"70", "74" and "79" as three types, and type: "tanker" matched nothing on
feeds that send codes. Codes now read as the family the vessel cards already
show (CARGO, TANKER, TUG), for rows, group_by and the type filter. The code
table moves to src/data/vesselTypes.js, which has no i18n import, so the
server-side tools can use it; vesselLabels re-exports it.

A destination with no letters ("0", "---") now counts as unknown.

Live, Singapore: "most by type: CARGO (3), TANKER (3), PASSENGER (1)"; "how
many tankers near Singapore" found 3.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-17 15:49:02 -07:00
Sunil56224972 97816a35de Merge remote-tracking branch 'upstream/main' into fix/terrain-same-site-gate-on-686 2026-10-18 03:36:52 +05:30
quanticsoul4772 d06e485c12 Merge branch 'main' into feat/tools-area-breakdowns 2026-10-17 14:49:25 -07:00
quanticsoul4772andClaude Opus 5.5 7bf268ec42 fix(tools): read an airline only from a callsign of seven characters or fewer
ICAO flight-plan Item 7 caps aircraft identification at seven characters, so
the flight number after the designator is one to four characters, not five.
Philippine RP-C1234 and Cuban CU-T1240 registrations read like a designator
plus a flight number once the hyphen is dropped; they now count as unknown.
Cubana's CUB123 still counts as CUB.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-17 14:37:07 -07:00