/api/cctv/media relayed still-image cameras with an ordinary fetch: it followed
redirects to any host, passed through any content type and had no body cap.
Stills are served by /api/cctv/frame, which keeps redirects on the registered
origin, accepts only image/* and caps the body, and no client requests a still
through /media. /media now answers 404 for anything but HLS, MP4/WebM and
MJPEG, before any upstream request.
Ids were built from the feed's fid, which embeds an upstream server-start stamp,
so every upstream restart renamed all cameras (it already had since the PR
fixture). Ids now come from the validated frame URL (cat-sct-<stem>, or host
and path for partner cameras), which also collapses the five duplicate rows.
Frame links must sit directly in each host's frame directory with no explicit
port, and sctidcam must be a bare image filename. Numeric entity decoding
handles hex references and leaves invalid code points as written, so one bad
name no longer empties the pack. Counts in the docs are now ~130.
The catalog fetch now refuses redirects, cancels rejected bodies and reads
through the shared capped reader before the windows-1252 decode. The feed sends
cardinal codes plus B (both), O and null, not BW or diagonals; tests, comments
and docs now say so, with unchanged id-hash fallbacks. Documents the 8-1,200
per-pack range, drops the unverified low-volume compliance claim for facts and
a link to WSDOT's terms, and stops claiming the per-pack defaults fit under
the 4,000 cap: it is filled round-robin.
The camera endpoints are the eismoinfo.lt website backend, which the Via
Lietuva open-data statement does not list, so the pack now follows the NYC
opt-in pattern (CCTV_LITHUANIA_ENABLED=1) and the docs say why. A failed
catalog request now aborts its sibling instead of leaving it open until the
timeout, and a redirect is logged as not followed.
The catalog fetch now refuses redirects, cancels rejected bodies and reads
through the shared capped JSON reader. Duplicate ids are dropped before
prioritization so they cannot use up the per-pack cap. Vendor-generic model
strings (PTZ, PTZ IP, BOSCH/PTZ) are treated as no model. Comments and credits
say why WSDOT rows are skipped (their frames are WSDOT-hosted, outside the
pinned origin) and name Newcastle as the partner city that appears.
Adds the Toronto settings to .env.example and a current-state section for the
pack, restores an approximate camera count in the README, and corrects the
claim that adding a pack never trims another region: with the shared
4,000-camera cap, the largest packs give way round-robin. Fixes two stale
comments about the list's refresh and the ground-height sidecar.
Keeps main's 4,000 catalog-wide camera cap, which main pins in tests; WSDOT
shares it round-robin with the other packs. The branch's earlier cap raise and
its removal of main's cap tests are reverted in this merge.
A 429 now starts the credit cooldown from its headers and drops the body
unread, so a stalled or oversized 429 cannot skip the governor; an oversized
snapshot is not retried. A 429 without retry-after now waits the intended
2 minutes (headers.get() returns null, which Number() read as 0 s). Adds
oversized-stream tests for OpenSky, adsb.lol, CelesTrak and TomTom that fail
on main, names the six reads in the changelog, and corrects the docs on
what was already bounded.
Uses a full SHA-256 of the sorted instance list, so reordering the same
instances for failover keeps the cache. Adds the installation-cache repoint
test (tagged, reordered, repointed, untagged, unconfigured) and the changelog
and current-state notes.
Resolve the CHANGELOG conflict by keeping every entry from main and
adding the radio entry on top; the previous main merge (cebd870)
had dropped 19 of main's entries.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Node always sets statusCode on client responses, but keeping the
fallback stops a missing status from reading as 200, as main did.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Resolve SECURITY.md against the Street View provider (#966): the browser
key now loads the Maps JavaScript API as well as Map Tiles, so the
'Map Tiles only' guidance becomes 'Map Tiles + Maps JavaScript API'
here and in CHANGELOG, CURRENT-STATE and google-key.js. The CHANGELOG
entry moves to [Unreleased]; its old section shipped in 0.2.0.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Keep both: Catalonia CCTV alongside the NYC DOT TMC pack in the source
imports, README layer table and CURRENT-STATE data table, and both
CHANGELOG entries.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Clears the timeout guards once each race settles, measures the 15,000-vehicle transit ingestion with main-thread CPU time instead of wall time, and times the region-timeout case from the boundary lookup with a 2 s ceiling, below the 3 s default budget, so an ignored budget still fails.
Co-authored-by: Sameh Khamis <[email protected]>
NYC DOT asks developers who use its camera feed to sign a data-sharing
agreement, and this project does not hold one that covers downstream
installs. The pack now loads only when CCTV_NYC_ENABLED=1; unset, empty
or any other value makes no request to webcams.nyctmc.org (tested).
The NYC credit carries NYC DOT's requested application disclaimer
verbatim, and DATA_SOURCES, README, CHANGELOG, CURRENT-STATE and
.env.example describe the prerequisite.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The pack shipped a flat 90 m. Sampling a DEM at all 336 camera sites puts
the network between 76 m (Queens Quay) and 209 m (Keele & Steeles), mean
116 m: a median error of 14 m against the old value, 88 m at the 90th
percentile and 119 m at worst, with the Allen Road cameras — 170-197 m —
the furthest underground.
No constant fits a city with that much relief; the precomputed ground
heights are what place a camera properly, and until this pack has them the
prior is the fallback for a sidecar miss or an edited pose. The mean is at
least wrong in both directions rather than only one.
This value is not part of the served pose, so it changes no pose hash and
invalidates no precomputed entry.
The province's 511 pack covers Ontario's highways; this adds the municipal
network underneath it — roughly 336 cameras across downtown arterials plus
the Gardiner, DVP, Allen and Lake Shore, from the City's own open data.
The list is keyless and carries no frame URL, so frame URLs are built from
the convention the dataset documents (`CameraImages/loc####.jpg`) and a
camera number validated as plain digits. The frame proxy is therefore
pinned to the portal origin by construction rather than by trusting a
field. The list fetch refuses redirects, bounds the body it buffers, and
releases the transport on every rejection path.
The dataset publishes no camera facing. `D1`-`D4` look like a compass and
map trivially onto one, but they describe the *comparison* images — static
reference stills captured in 2016 and served from a different path — not
the live pan-tilt-zoom frame. A heading derived from them would be
confidently wrong for every camera in the city, so every camera takes the
shared id-hash fallback at low confidence, as headingless Austin, TfL,
Fintraffic and Calgary cameras do. A test pins that down so it is not
"fixed" later.
Names stay verbatim: the feed's labels are upper-case, and title-casing
would mangle MCCOWAN RD and ST HELEN'S AVE more often than it helped.
Carries the Open Government Licence – Toronto attribution the licence
requires. Disable with `CCTV_TORONTO_ENABLED=0`.
The layer compared the server's fetchedAt with the browser's clock, so a
skewed client clock could shift or void the 30-minute window. The proxy now
returns ageMs (its own now minus fetchedAt, at serve time, on fresh, cached
and stale answers); the source refuses a snapshot without a finite,
non-negative ageMs, and the layer dates rows as its now() - ageMs. fetchedAt
stays in the payload for display.
Tests: proxy cached/stale answers report ageMs on the server clock; the
source rejects missing/invalid/negative ageMs; the layer replay test drives
ageMs. qa-gnss stubs send ageMs.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Maintainer review on #855 (mazy1998, 2026-10-06):
- records: presence (seenAt) and degraded evidence (degradedAt) age
separately, so continuing healthy reports no longer keep an older
degraded flag alive past 30 minutes; an aircraft returning after the
window starts fresh. pruneGnssObservations ages the store with no new
rows.
- Rows are dated by the snapshot's fetchedAt, which the proxy keeps on
cached and stale answers: a replay never restarts the window, the
proxy stops serving a stale snapshot older than the window, the source
refuses an undated one, and cells expire during fetch failures and
while there is no view anchor.
- Normalization accepts ADS-B versions 1 and 2 only; a missing, invalid
or negative seen_pos is not a confirmed position from the last minute.
- qa-gnss: keyboard, 390x844 viewport (fresh page: isMobile reloads) and
replay-age sections; the live cell assertion bins the returned rows
with the layer's own rules instead of a raw row count.
Review on #855: the per-aircraft rule (NIC < 7 || NACp < 8, borrowed from
the US 14 CFR 91.227 ADS-B Out minima) is GEV-defined, not gpsjam's, and
gpsjam's map shows low navigation accuracy, not necessarily jamming.
- Primary cell state is now `navigation-integrity-anomaly` (medium/high) or
`nominal`; `suspected-gnss-interference` is a secondary `interpretation`
marked validated: false.
- GNSS_PROVENANCE / gnssProvenance({ windowMs }) records the classifier
(gev-nic-nacp-v1, GEV, unvalidated), the gpsjam aggregation (formula,
bands [0.02, 0.1], FAQ source) and the GEV 30-minute visited-view window,
built from the constants the code applies. Exposed as layer.provenance
and getStats().provenance; the proxy tags responses with `classifier`.
- Visible name "GNSS Integrity", legend "... low accuracy", blurb leads with
low navigation accuracy and states the threshold is a GEV choice.
- Layer id `gnss-interference` and share token `0` unchanged.
- Voice aliases add "gnss integrity", "navigation integrity",
"low navigation accuracy"; the interference aliases stay.
- Docs describe the method as GEV threshold + gpsjam aggregation.
- Tests: provenance object, renamed state/label, proxy classifier tag;
qa-gnss asserts provenance and wording in the real app.
Hex project - Crew Leo Agile dev team.
Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>
- Route the adsb.lol GNSS source through createReferenceSources and the
constructCatalog contract; app layer module is a thin pass-through.
- Describe the method honestly: gpsjam.org's published cell formula and
bands; per-aircraft thresholds are the 14 CFR 91.227(c) ADS-B Out minima
(NIC < 7, NACp < 8) plus readsb gpsOkBefore.
- Proxy: reuse the adsb.lol anchor parser, honour upstream Retry-After
(clamped 5-120 s), serve stale during cooldown, cap the body size.
- Layer: redraw only when a cell id or band changes, treat a settling
camera as a retry rather than a rejected enable, keep low cells visible,
skip rows without finite coordinates, state the 3-aircraft rule in the
legend blurb.
- Voice: gnss-interference in set_layer_visibility / show_data_layers_menu
plus common aliases; voice pins re-derived.
- qa-gnss: stub the route in-page (request interception stalls Cesium's
geometry workers), assert every cell is built, cover all bands over land
and sea, screenshot every state with the Data Layers row open.
Hex project - Crew Leo Agile dev team.
Co-Authored-By: Claude <[email protected]>
Bins airborne ADS-B v1/v2 aircraft that report degraded navigation
integrity (NIC < 7, NACp < 8, or readsb's GPS-loss flag) into 0.5 degree
cells over a rolling 30-minute window, coloured with the gpsjam.org
bands. Keyless: reads the capped adsb.lol point snapshot around the view
through /api/gnss-integrity, which returns integrity fields only, caches
60 s per whole-degree anchor, backs off on upstream 429 and serves the
last snapshot on failure. Adds scripts/qa-gnss.mjs.
Built by the Hex project (Crew Leo Agile dev team).
Co-Authored-By: Claude <[email protected]>
CHANGELOG-only conflict: the bullet stays first under [Unreleased], ahead of
the entries upstream added there. No code changes.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
fetchPinnedRadioResponse built `new Response(stream, { status })` inside
the raw https response callback, which runs outside the surrounding
promise. A Radio Browser mirror answering 204, 205 or 304 (null-body
statuses), or any status outside 200-599, made the Response constructor
throw. Nothing caught it, so the whole dev/preview server process exited
and every layer went down, not only radio. The catalog's mirror failover
never saw the error.
Mirror the guard local-receivers.js already uses: null-body statuses get
a null body, statuses outside 200-599 are rejected, and anything thrown
in the callback rejects the promise so the existing error handling and
mirror failover run.
Register the keyless webcams.nyctmc.org catalog as a live CCTV source,
synthesize frame URLs from validated camera UUIDs, and keep the nearest
cameras to the five borough cores under the existing catalog ceiling.
The Catalonia feed has no camera status field, so an offline camera keeps
its catalog entry and SCT or Barcelona answer its frame with their own
"image unavailable" still (HTTP 200). Say so in DATA_SOURCES.md and stop
the loader comment from implying every camera returns a live frame.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The login child inherited the whole server environment, including
OPENAI_API_KEY and every other provider credential. Pass only what Codex
needs to find its home, open a browser and reach OpenAI through a proxy.
Sameh Khamis