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
Both CelesTrak routes shared one cooldown, and any failure set it: one bad
SATCAT answer made /api/celestrak return 502 "cooling down" for a minute,
failing the Starlink load and stalling every TLE group past its cache.
Throttling (403/429) still holds both routes off, since CelesTrak throttles
the client; any other failure holds off only the route that failed.
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.
Recent community wildlife observations, as a new Wildlife group in the
layer panel.
- Server: keyless iNaturalist proxy (v2 observations with only the
projected fields; undocumented v1 UTFGrid density tiles, decoded on the
server) sharing one paced queue (burst 4, ~1/s, 60/min, 8,000/day budget,
up to four in flight), in-memory caches and upstream-only cooldowns;
Acartia Salish Sea marine-mammal feed rounded to 0.01 degrees with
comments dropped.
- Layer: grid counts drawn as a soft view-ranked density glow; real sampled
sightings as taxon badges at their true positions with ambient photo
thumbnails and labels; obscured records stay privacy cells; a clicked
sighting stays pinned through camera moves and refreshes.
- Anchored photo card per selection with the single iNaturalist link;
attribution in the layer row and the Data attribution popover.
- Filters (CONFIRMED, taxa presets) round-trip in share links with token 9.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Review follow-up for #1552:
- environmentProviders: assert exactly 3 upstream calls per refresh again,
so a returning fourth source would fail the test.
- firmsProxy: build the test's SOURCES from server/providers/firms.js, so
there is one list instead of a mirrored copy.
- firmsProxy: drop the unused fourth count entry.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
NASA stopped delivering Suomi NPP data on 2026-11-02 (FIRMS news page).
With a live MAP_KEY on 2026-11-04, VIIRS_SNPP_NRT came back ok with zero
rows while NOAA-20, NOAA-21 and MODIS returned 57k, 49k and 11k, so the
source only spent one transaction per refresh. The proxy now requests
NOAA-20, NOAA-21 and MODIS NRT. The "N" satellite label stays, so cached
or archived Suomi NPP rows still read SNPP.
Tests count sources from the list instead of assuming four, and a
regression test keeps the retired source out of the production list.
Refs #849
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The disk-cache test asserts the cache directory is 0o700 and its files
0o600. Windows has no POSIX mode bits (the directory stats as 0o666), so
the test always failed there. Both checks now run only off Windows, as
localServices.test.mjs and mcpPanelKey.test.mjs already do for their
0o600 checks. The cache code is unchanged.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
/api/osm/features keyed its rate limit and its one-search-per-client slot
by the full peer address, so an IPv6 host could rotate addresses inside
its /64 to take both global slots and slip the 12/min limit. It now uses
clientNetworkKey, as /api/overpass does.
Picks up acd87e6c: the keyless cross-site gate on /api/osm/features
(582bfce6) and main at 93804d5d. SECURITY.md, the format-scope list and
the schema-test comments take the branch's side; the place-search budget
wording from the timeout fix is kept.
Without a limit parameter the route read null, Number(null) is 0, and the
clamp turned that into a one-feature list. An absent or empty limit now
gets the documented default of 500.
The voice answer waited 25 s for /api/osm/features, while the server let
Overpass run for 25 s and itself waited up to 40 s per endpoint and 45 s
in all. When the browser gave up first it closed the request, the route
cancelled the upstream query, and nothing was cached, so the OSM_TIMEOUT
retry started the same query from scratch and timed out again.
The server now tells Overpass 20 s, waits 22 s per endpoint and 22 s in
all, which ends inside the client's 25 s. A test pins that ordering.
#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.
From review:
- The browser client for /api/osm/features gets its own tests (request
shape, abort, network failure, coded refusals), and a 200 whose body is
not JSON is now a failure instead of zero places.
- An end-to-end test runs resolve_area, osm_query and an analyst follow-up
through the real runner and checks the follow-up keeps the area; it fails
with the wiring reverted.
- The Realtime token test pins Overpass off and compares the full tool
schemas again instead of names taken from the environment.
From review:
- /api/osm/features now refuses cross-site browser requests like the other
quota-bearing routes, so a page elsewhere cannot spend the operator's
Overpass. SECURITY.md lists the route, its limits and its disk cache.
- One client runs one search at a time and two run at once in all, instead
of one global slot any single client could hold.
- A search gets 45 s across every configured endpoint in turn, answered 504
when it runs out.
- A cache write no longer waits on the disk prune; prune() waits until the
cache is idle for callers that need it.
The /api/osm/features cache key was the query alone, kept for a day, so after
an operator repointed OVERPASS_UPSTREAMS the old instance's answers were still
served. The key now includes the configured-upstream tag, as main's other
Overpass caches check it, so a new instance is always asked.
Every name ending in "Zone" was kept without a level and refused as a retired
administrative identity. That is right for Nepal, which replaced its zones in
2015, but zones are current units elsewhere: Ethiopia's "Gamo Zone" was
refused. The selector now keeps the verified level unless display_name places
the zone in Nepal or names no country, and the client no longer reads "Zone"
as a retired level when the source verified one.
Both stack parents (#837, #842) are on main, so the voice geometry branch
now builds on main directly. Conflict resolutions keep main's behaviour and
add the geometry work on top: osm_query gating on every provider, restored
admin identity for Nominatim outlines, merged analyst follow-up memory, and
the cancelled Globe View reset instruction ahead of the generic cancellation.
Resolve the CURRENT-STATE.md layer table by keeping main's realigned
table and adding this layer's row. Main added an Arabic locale, so
translate the 12 layer strings and list ar.json in the drone-airspace
package.
A building word ending the ask ('… Chapel', '… Library', '… Hall') now
passes the last-word rule only when the result is one a building ask
would accept, so a campus, zoo or district whose name merely starts the
ask (Duke University, Old Town) no longer answers it.
- A building word ("Burj Khalifa Tower", "Sagrada Familia Church") passes
the last-word rule on a building-like result; enclosures stay out.
- The anchor chooses only among the best-named outlines, so a worse name
never wins by containment alone.
- An ask without a quadrant takes only the quadrant of the matching street
nearest the point; a leading quadrant ("SE Main St") may go unsaid too.
- On an unbuffered tile edge a fragment shows as an edge along the
boundary, so a whole building touching it at a corner is not cut.
- Drop two orphaned comments and correct the place-name key wording.
- One place-name key for server and tiles, so "St. Paul" and "Saint Paul"
agree in Nominatim name checks.
- Tile street names: a lone letter is a direction only beside another name
word ("E Street" is not "East Street"), and a trailing quadrant ("E St NW")
may go unsaid in the ask.
- A building cut at the tile edge that is nearer the point than any whole
one leaves the point instead of a farther neighbour standing in; an edge
without a tile buffer counts as a cut, so a cut area never answers grounds.
- A batch building install re-judges only results that installed; a late
timeout during caption conversion keeps the pending point drawn; a failed
plain recaption keeps the old mark addressable.
- An interrupted Nominatim dispatch is reclaimed by the next lock owner ten
minutes after the start its marker names; node:sqlite loads on first use
without its experimental warning.
- PHOTON_URL accepts only https, which the page's connect-src allows.
- Shared OpenFreeMap tiles decode each layer when a consumer first selects
it; outline selection reuses the shared ring area and simplification, and
ground probes the shared ring centroid; the grounds word list lives once.
A street ribbon that holds neither its ring mean nor the anchor is centred
on the street's nearest point when the street runs within 100 m.
Nominatim outline selection ranks an exact name, then one holding every
asked word, then an overlap, and refuses an overlap whose name lacks the
ask's last word unless that word is in the result's address or type, so
"Central Park Zoo" is no longer Central Park. The server caches every
accepted outline and answers each ask with the one holding its resolved
anchor (sent as alat/alon), read across the antimeridian.
A duplicate mark that fails to clean up no longer tears down the new
outline it duplicates.
Since the UI moved to single-file components, every unit-test process
compiled the views it imports from scratch, and the loader ran off-thread
(module.register), adding a round trip to every module a test imports,
including the ones that are not components. Core's unit tests took about
twice as long, and a host running Core's suite on smaller runners passed
its CI time limit.
register-vue now installs synchronous hooks on the loading thread
(module.registerHooks). Compiled components are kept in the system temp
directory, keyed by the file's path and text, the Vue version and the
loader's own code, so each one compiles once across test processes; the
compiler itself loads only on a cache miss. Cached modules run as code, so
the directory is used only when it is a real directory owned by the current
user and writable by no one else; on Windows it lives in the user's local
app data instead of a temp directory that may be shared. The cache never
sits inside the package tree, and GEV_VUE_CACHE=0 compiles every time.
On four cores the full suite went from 441 s to 240 s from a cold cache.
The application views install one vue-i18n instance that holds the active
locale and the catalogs. src/i18n/index.js stays free of Vue: bindLocale()
keeps its locale in the instance's ref, so t(), tn() and the Intl formatters
read it from there and a view that translates while rendering renders again
when the locale changes. A tag written straight to the ref is normalized.
Message formatting, catalogs and call sites are unchanged, and switching
language still reloads the page.
The instance outlives the views: unmounting one no longer disposes it. The
browser build defines vue-i18n's feature flags and drops its message
compiler, which Core does not use.
PanelChrome, its collapse-preference restore and the phone shell read a
panel's disclosure state back from the gev-panel-chrome store
(_shownPanel), falling back to the markup's classes before the panel is
first presented, instead of reading the classes v-panel-state draws. The
share link and the phone shell share one rule for a panel the person keeps
closed (prefersCollapsed). The rails, the accordion and the hover
disclosure still read the classes, which the directive draws in the same
task as their layout.
The Display panel, the dock's style buttons and the clean-view exit call
the display store's actions, and DisplayBindings connects their handlers
to it, replacing the bindDisplayControls listeners. The detection
allocation and 3D model coverage choices render through a new kit
UiSegmented, which takes v-model, options, label,
stacked and disabled, and adds a keyboard policy, item classes and
per-option attributes.
Shell element fields and facade accessors that only the old listeners
read are removed. The package boundaries now own UiSegmented and the
UiPanel kit files, and the format scope drops the deleted test.