Commit Graph
3427 Commits
Author SHA1 Message Date
Sameh Khamis 09520bc34c Changelog fragment for #1642 2026-11-09 14:54:51 -08:00
Sameh Khamis 9bce41d3c6 Read the mesh floor where transit vehicles and grounded flights actually are
With readings kept per point, the floor a vehicle gets depends on which point
is probed and where it is read. Transit offered the sampler the first point of
each vehicle's path in a cell, usually its oldest fix there, so a vehicle that
stayed in one cell never had its current spot probed. It now offers the latest
point of the path in each cell, and the newest fix in the vehicle's own cell
(the selected vehicle's history is walked newest first).

The grounded-flight display clamp read the floor at its sticky cell's centre,
which now answers from the reading nearest the centre, not the contact. It
reads at the contact's own position, held inside the sticky cell so boundary
hysteresis still applies.
2026-11-09 14:51:35 -08:00
Sameh Khamis 04b8c856ad Keep vehicles off rooftops and bridge decks: probe the mesh where the vehicle is
On the photoreal stack the rendered-mesh floor took one probe at each coarse
cell's centre (0.001 degrees) and kept it for the session. A centre on a roof
passed the +80 m DEM window and lifted every vehicle in the cell to roof height
(a Haymarket bus drew 23.5 m above the street, on the garage roof); a centre in
the river put Red Line trains on the Longfellow Bridge 6-9 m under the deck.

The sampler now probes at the caller's point. A cell with a reading is probed
again only at spots about 17 m apart from earlier ones, eight at most, and a
spot refused only because the cell's DEM prior was cold is tried again later.
The floor keeps each reading that passes the gate at the point it was taken
(two within about 5 m keep the lower; sixteen at most per cell) and answers a
read from the reading nearest the asked point, so a viaduct and the street
beside it in one cell keep their own heights.
2026-11-09 14:51:35 -08:00
Mazeyar Moeini Feizabadi e5da073d33 Merge pull request #1638 from avaldini/fix/realtime-voice-unicode-patterns
fix(voice): send OpenAI Realtime tool schemas the API accepts
2026-11-09 20:32:38 +01:00
Mazeyar Moeini Feizabadi ddd07f13f0 Merge branch 'main' into fix/realtime-voice-unicode-patterns 2026-11-09 19:44:29 +01:00
German Khokhlov 10784fa9c9 Merge pull request #1552 from IMBrandao/fix/firms-drop-snpp
fix(firms): stop requesting the retired Suomi NPP source
2026-11-09 21:52:09 +04:00
German Khokhlov fce5ee6381 docs(firms): tighten the Suomi NPP notes 2026-11-09 21:44:09 +04:00
German Khokhlov d90da158ea Merge branch 'main' into fix/firms-drop-snpp
# Conflicts:
#	docs/CURRENT-STATE.md
2026-11-09 21:41:01 +04:00
German Khokhlov b04fcf191c docs(firms): count two VIIRS sources in the layer table and test note 2026-11-09 21:38:07 +04:00
German Khokhlov 581ef7eef4 Merge pull request #766 from BreezeDelegate/fix/node26-allocation-gates
fix(ci): run allocation gates on Node 26
2026-11-09 20:35:28 +04:00
German Khokhlov 015e3ff7fc Merge branch 'main' into fix/node26-allocation-gates 2026-11-09 20:19:29 +04:00
German Khokhlov 72a240ba6f Merge pull request #762 from monwiss/refactor/tile-budget
refactor(providers): move the tile budget helpers out of the TomTom module
2026-11-09 20:00:55 +04:00
German Khokhlov 77df600329 Merge branch 'main' into refactor/tile-budget 2026-11-09 19:52:40 +04:00
Said Turksever 0dbc657bf9 Merge pull request #1576 from bilawalsidhu/german/transit-bikeshare-empty-state
Say why Transit and Bike Share rows are empty outside coverage
2026-11-09 15:48:34 +00:00
German Khokhlov e22348a8e1 Merge branch 'main' into refactor/tile-budget 2026-11-09 19:43:58 +04:00
German Khokhlov 75e8e7ddbd refactor(data): describe tile budgets per period, drop a duplicate test, add the changelog entry 2026-11-09 19:41:51 +04:00
Said Turksever 9787d9c4cc Merge branch 'main' into german/transit-bikeshare-empty-state 2026-11-09 15:32:48 +00:00
Said Turksever 2725f307a6 Merge pull request #1628 from IMBrandao/feat/transit-more-cities
feat(transit): add 34 keyless cities across Europe, Asia, Oceania and North America
2026-11-09 15:30:57 +00:00
Said Turksever cf05a29b7f Merge branch 'main' into feat/transit-more-cities 2026-11-09 15:20:12 +00:00
German Khokhlov 45108a0b0b Merge pull request #794 from daikaginza/fix/api-method-guard
fix(server): declare the method contract on the three read-only query proxies
2026-11-09 19:17:49 +04:00
Said Turksever c1ca0f489c Merge branch 'main' into feat/transit-more-cities 2026-11-09 15:13:18 +00:00
German Khokhlov cdf52ab290 test(server): drop unused fields from the /track hand-off stub 2026-11-09 19:07:14 +04:00
German Khokhlov c3738c0fdb docs(server): tighten the method-guard comments and changelog 2026-11-09 19:07:14 +04:00
German Khokhlov ace88172d3 Merge branch 'main' into fix/api-method-guard 2026-11-09 18:51:02 +04:00
German Khokhlov 6a7a09372a Merge pull request #795 from daikaginza/feat/opensky-compress-cached-body
feat(server): compress the cached OpenSky snapshot for clients that accept gzip
2026-11-09 18:40:19 +04:00
Said Turksever 3e630cf34f fix(transit): mark route line names as kept for i18n
Métro, Calypso, LD and the metro letters are line names as signed, not UI
copy; the i18n catalog check flagged them.
2026-11-09 14:34:33 +00:00
German Khokhlov 9d7fe8fcaa Merge branch 'main' into feat/opensky-compress-cached-body 2026-11-09 18:30:32 +04:00
German Khokhlov a45d6af5f7 Merge pull request #1611 from abiyeenzo/fr-fixes
i18n(fr): typography, terminology and wording fixes
2026-11-09 18:22:18 +04:00
German Khokhlov caad795fa2 refactor(server): drop the null and String paths from gzipBody 2026-11-09 18:15:13 +04:00
German Khokhlov 09bf01630b docs(server): tighten compression comments and docs 2026-11-09 18:14:30 +04:00
German Khokhlov 2ea4a78e75 Fix the fr review follow-ups
- "destination refresh" is the refresh at the place the camera flew to,
  so restore "son actualisation à destination".
- Radio note: "que vous la demandez par la voix", "Votre adresse IP".
- "{count} / 30D": "30 J", like "{days} j".
- Starlink: "la couche complète", as in the active-state tooltip.
- "Show {name} sighting": colon form, so names starting with a vowel
  do not break the elision after "de".
- Wildlife note: Acartia is the source, not a kind of marine mammal.
- Straight apostrophes in the five strings added on main since.
2026-11-09 18:12:48 +04:00
German Khokhlov 61e9cf87f9 Merge branch 'main' into fr-fixes
# Conflicts:
#	src/i18n/locales/fr.json
2026-11-09 18:12:09 +04:00
Said Turksever 2c1b2c278d fix(transit): read Rouen and Le Havre from the operators' own feeds
Rouen now reads Métropole Rouen Normandie's TCAR vehicle feed
(api.mrn.cityway.fr) and Le Havre LiA's own GTFS-RT
(opendata.transports-lia.fr), both listed in their transport.data.gouv.fr
records, instead of the unofficial bus-tracker.fr relay. LiA ids carry an
8- prefix, so a new liaRouteLabel strips it and liaRouteMode reads the line.
2026-11-09 13:58:20 +00:00
Said Turksever 7a053c640d fix(transit): real route-id modes for Prague, Bordeaux and Rouen; dated Warsaw/Poznań credits; data.gov.my refresh
- Prague PID: map route-id ranges from PID's GTFS (metro L991-993, S/R trains
  L1000s/L1100s, P ferries L1800s, trolleybuses 50-59, X lines, funicular LD)
  instead of 'below 100 is a tram, rest bus'.
- Bordeaux TBM: trams ride numbered ids 59-62/163/164 (A-F), boats 951-953;
  the old letter match never fired.
- Rouen Astuce: TCAR:90 Metro as tram, TCAR:99 Calypso as ferry; Le Havre FUNI.
- Warsaw and Poznan: credit carries the data date (and Warsaw links its API),
  as their terms require.
- data.gov.my feeds: CC BY 4.0, refreshMs 30 s to stay under 4 requests/min;
  proxy honours per-feed refreshMs.
- Denver: quote the RTD agreement; tests use the feeds' real route ids.
2026-11-09 13:50:14 +00:00
German Khokhlov a5b40dfc3d Merge pull request #1569 from rudycelekli/fix/perimeters-sub-acre-size
fix(perimeters): show a 0.25-acre fire as 0.25 ac, not 0 ac
2026-11-09 17:39:29 +04:00
German Khokhlov be9c37bee2 Merge remote-tracking branch 'origin/main' into fix/perimeters-sub-acre-size 2026-11-09 17:07:05 +04:00
German Khokhlov 394f3216f8 Drop the sub-acre rounding comment 2026-11-09 17:06:56 +04:00
German Khokhlov c62f7ef00e Merge pull request #1524 from quanticsoul4772/upstream/tools-heading-wrap
fix(tools): wrap whole-degree headings at 360 to 0
2026-11-09 17:00:58 +04:00
quanticsoul4772 dbb082355a Merge branch 'main' into upstream/tools-heading-wrap 2026-11-09 04:47:58 -08:00
German Khokhlov da6b63b6fd docs(tools): tighten the heading-wrap changelog, comments and overview 2026-11-09 16:45:11 +04:00
German Khokhlov 3150a460aa Merge pull request #1547 from quanticsoul4772/upstream/nominatim-cache-mode-windows
test(nominatim): check cache file modes only where POSIX modes exist
2026-11-09 16:44:00 +04:00
German Khokhlov c2bff536a5 Merge branch 'main' into upstream/nominatim-cache-mode-windows 2026-11-09 16:07:00 +04:00
Andrés Valdez d52c719ef6 docs(changelog): OpenAI Realtime voice tool patterns (#1638) 2026-11-09 07:52:24 +00:00
Andrés Valdez 9f2dfdf820 fix(voice): send OpenAI Realtime tool schemas the API accepts
OpenAI Realtime now answers 400 invalid_function_parameters for a whole
session whose tool schema uses Unicode property classes, as resolve_area's
area-id pattern does, so voice never connects ("Failed to create Realtime
token"). The session sends its tools through chatSchema, the filter typed
chat adopted in #1616, which leaves out only \p{…} patterns; area ids are
still checked against the action's own contract before the tool runs.

Fixes #1618
2026-11-09 07:34:44 +00:00
Sameh Khamis 027c9616a5 Merge pull request #1637 from bilawalsidhu/sameh/orbital-followups
Orbital cockpit follow-ups: sensor strip through zoom, one owner-flag sheet, place names in the loading status
2026-11-08 20:21:29 -08:00
Sameh Khamis c1c71d1987 docs(changelog): orbital cockpit follow-ups (#1637) 2026-11-08 20:12:00 -08:00
Sameh Khamis 26718edc0b test(place-names): stand-in packs, the whole status run, and a failure within a load that stays failed 2026-11-08 20:07:04 -08:00
Sameh Khamis 2f31828d28 fix(place-names): a load starts with no failures, so packs that failed earlier do not mark a retry that lands as failed 2026-11-08 20:04:40 -08:00
Sameh Khamis c5e1951f51 fix(cockpit): one place lookup that loads several packs reads as one load, not a LOAD COMPLETE between packs 2026-11-08 19:57:28 -08:00
Sameh Khamis f6dc89b4fa fix(cockpit): a place pack retried after a failure counts as failed only if the retry fails 2026-11-08 19:49:05 -08:00