## Summary
Fixes frontend region crosstalk on Channels page by applying region
filtering to message fetches and live WS GRP_TXT handling.
## Changes
- Append `region` query param to channel message API calls in
`selectChannel` and `refreshMessages`.
- Add WS region guard in `public/channels.js` using observer→IATA map
with selected-region snapshot at handler entry.
- On region switch, reload channels and re-fetch selected channel
messages; if empty under selected region, clear pane and show `Channel
not available in selected region`.
- Bump cache busters in `public/index.html`.
- Add frontend helper tests for extracted WS region filter helper in
`test-frontend-helpers.js`.
## Validation
- `node test-frontend-helpers.js`
- `node test-packet-filter.js`
- `node test-aging.js`
Refs #280
---------
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Summary
- fix observer upsert write path in `cmd/ingestor` to persist identity
fields
- map status payload fields into observer metadata: `model`,
`firmware`/`firmware_version`, `client_version`/`clientVersion`, `radio`
- keep NULL-safe behavior when identity fields are missing
- add regression tests for identity persistence and missing-field
handling
## Root cause
The ingestor only wrote telemetry (`battery_mv`, `uptime_secs`,
`noise_floor`) and never included observer identity columns in the
upsert statement, leaving `model`, `firmware`, `client_version`, and
`radio` NULL on fresh DBs.
## Testing
- `cd cmd/ingestor && go test ./...`
- `cd cmd/server && go test ./...`
Fixes#295
---------
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Summary
Fixes BYOP modal stacking on the Packets page by preventing duplicate
global click handlers and enforcing a single BYOP overlay instance.
## Root cause
Packets page init could register document-level click handlers
repeatedly across SPA navigations. Clicking BYOP then spawned multiple
overlays, and each close action removed only one layer.
## Changes
- `public/packets.js`
- Added `bindDocumentHandler(...)` to de-duplicate document click
handlers.
- Applied it to packets action delegation, filter menu outside-click
close, and column menu close.
- Added `removeAllByopOverlays()` and call it before opening BYOP.
- Tagged BYOP overlay with `.byop-overlay` class.
- Updated close logic to remove all BYOP overlays in one click.
- Scoped BYOP result lookup to the active overlay
(`overlay.querySelector`).
- Added destroy cleanup for document handlers and stray BYOP overlays.
- `test-frontend-helpers.js`
- Added regression tests for:
- BYOP singleton overlay behavior
- one-click close removing all overlays
- document click handler de-dup logic
- `public/index.html`
- Bumped cache busters for JS/CSS assets.
## Validation
- `node test-frontend-helpers.js`
- `node test-packet-filter.js`
- `node test-aging.js`
All passed locally.
Fixes#249
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Fix: Enforce LF line endings repo-wide
### Problem
Windows-based agents produce CRLF line endings, causing git diffs to
show every line as changed (1000+ line "rewrites" that are actually
20-line patches). This has hit us on `manage.sh`, `deploy.yml`, and
multiple PRs.
### Fix
Added `* text=auto eol=lf` to `.gitattributes`. Git will now:
- Store all text files as LF in the repo
- Convert CRLF to LF on commit (regardless of OS)
- Check out as LF on all platforms
Also marks common binary formats explicitly.
### Impact
Existing files with CRLF will be normalized on their next commit. No
functional changes.
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Summary
Implements issue #286 P1 frontend timestamp features on top of P0:
- Added global timestamp timezone toggle in Display tab (`Local time` /
`UTC`)
- Added absolute-mode timestamp format presets (`iso`, `iso-seconds`,
`locale`)
- Added optional custom format input (only when
`SITE_CONFIG.timestamps.allowCustomFormat === true`)
- Extended `formatTimestamp()` / `formatTimestampWithTooltip()` behavior
to honor timezone + format settings
- Preserved server defaults with localStorage override precedence
- Bumped `public/index.html` cache busters in same commit
## Details
### 1) Timezone toggle
- New Display tab control persisted to `meshcore-timestamp-timezone`
- Reads server default from `window.SITE_CONFIG.timestamps.timezone`
with fallback to `local`
- Formatting logic now supports both local and UTC absolute rendering
### 2) Format presets (absolute mode only)
- New Display tab preset dropdown (shown only when timestamp mode =
`absolute`)
- Presets implemented:
- `iso` → `YYYY-MM-DD HH:mm:ss`
- `iso-seconds` → `YYYY-MM-DD HH:mm:ss.SSS`
- `locale` → `toLocaleString()` (or UTC locale when timezone=utc)
- Persisted to `meshcore-timestamp-format`
- Reads server default from `window.SITE_CONFIG.timestamps.formatPreset`
(fallback `iso`)
### 3) Custom format string (guarded)
- Text input only renders when
`window.SITE_CONFIG.timestamps.allowCustomFormat` is `true`
- Persisted to `meshcore-timestamp-custom-format`
- If non-empty and enabled, custom format overrides preset
- Frontend intentionally does not hard-validate the format string;
unsupported patterns fall back to preset behavior
## Tests
Executed required test commands:
```bash
node test-frontend-helpers.js
node test-packet-filter.js
node test-aging.js
```
Added coverage in `test-frontend-helpers.js` for:
- UTC output behavior
- Local output behavior
- `iso-seconds` includes milliseconds
- `locale` format behavior
All passed locally.
Refs #286
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
## Fix: FAQ save no longer wipes other home page sections
Fixes#284
### Problem
Editing FAQ in the customizer and saving caused other home page sections
(steps, footer links, hero text) to disappear on reload. Colors could
also reset.
### Root cause
`initState()` in `customize.js` used `||` (OR) logic for the `home`
object — if localStorage had *any* `home.checklist`, it took that and
ignored the server config for other fields. Partial localStorage data
replaced the full server config instead of merging on top.
### Fix
Changed `initState()` to properly layer: `DEFAULTS → server config →
localStorage` for all sections. Each field merges independently — a
partial localStorage save (e.g., only checklist) no longer wipes steps,
footerLinks, or hero fields. Same merge pattern applied to all theme
sections for consistency.
### Files changed
- `public/customize.js` — `initState()` merge logic
- `public/index.html` — cache buster bump
- `test-frontend-helpers.js` — regression tests:
1. Partial localStorage (checklist only) preserves steps/footerLinks
2. Server config values survive partial local overrides
3. Full localStorage properly overrides server config
### Testing
- `node test-frontend-helpers.js` ✅
- `node test-packet-filter.js` ✅
- `node test-aging.js` ✅
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Summary
- Add a new **Display** tab to the customizer tab bar (between Home Page
and Export / Save).
- Move timestamp-related **UI Settings** out of Branding into the new
Display tab.
- Keep Branding focused on site identity fields (name, tagline, logo,
favicon).
- Bump `public/index.html` cache busters so updated frontend assets load
immediately.
## Testing
- `node test-frontend-helpers.js`
- `node test-packet-filter.js`
- `node test-aging.js`
Fixes#293
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Problem
Repeaters with 2-byte adverts occasionally appear as 1-byte on the map
and in stats.
**Root cause:** `computeNodeHashSizeInfo()` sets `HashSize` by
overwriting on every packet (`ni.HashSize = hs`), so the last advert
processed wins — regardless of how many previous packets correctly
showed 2-byte.
When a node sends an ADVERT directly (no relay hops), the path byte
encodes `hashCount=0`. Some firmware sets the full path byte to `0x00`
in this case, which decodes as `hashSize=1` even if the node normally
uses 2-byte hashes. If this packet happens to be the last one iterated,
the node shows as 1-byte.
## Fix
Compute the **mode** (most frequent hash size) across all observed
adverts instead of using the last-seen value. On a tie, prefer the
larger value.
```go
counts := make(map[int]int, len(ni.AllSizes))
for _, hs := range ni.Seq {
counts[hs]++
}
best, bestCount := 1, 0
for hs, cnt := range counts {
if cnt > bestCount || (cnt == bestCount && hs > best) {
best = hs
bestCount = cnt
}
}
ni.HashSize = best
```
A node with 4× hashSize=2 and 1× hashSize=1 now correctly reports
`HashSize=2`.
## Test
`TestGetNodeHashSizeInfoDominant`: seeds 5 adverts (4× 2-byte, 1×
1-byte) and asserts `HashSize=2`.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
## Summary
- Adds `GeoFilter` struct to `Config` in `cmd/server/config.go` so
`geo_filter.polygon` and `bufferKm` from `config.json` are parsed by the
Go backend
- Adds `GET /api/config/geo-filter` endpoint in `cmd/server/routes.go`
returning the polygon + bufferKm to the frontend
- Restores the blue polygon overlay (solid inner + dashed buffer zone)
on the **Map** page (`public/map.js`)
- Restores the same overlay on the **Live** page (`public/live.js`),
toggled via the "Mesh live area" checkbox
## Test plan
- [x] `GET /api/config/geo-filter` returns `{ polygon: [...], bufferKm:
N }` when configured
- [x] `GET /api/config/geo-filter` returns `{ polygon: null, bufferKm: 0
}` when not configured
- [x] Map page shows blue polygon overlay when `geo_filter.polygon` is
set in config
- [x] Live page shows same overlay, checkbox state shared via
localStorage
- [x] Checkbox is hidden when no polygon is configured
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Removes the separate config.json file bind mount from both compose
files. The data directory mount already covers it, and the Go server
searches /app/data/config.json via LoadConfig.
- Entrypoint symlinks /app/data/config.json for ingestor compatibility
- manage.sh setup creates config in data dir, prompts admin if missing
- manage.sh start checks config exists before starting, offers to create
- deploy.yml simplified — no more sudo rm or directory cleanup
- Backup/restore updated to use data dir path
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
## Summary
- added API key middleware for write routes in cmd/server/routes.go
- protected all current non-GET API routes (POST /api/packets, POST
/api/perf/reset, POST /api/decode)
- middleware enforces X-API-Key against cfg.APIKey and returns 401 JSON
error on missing/wrong key
- preserves backward compatibility: if piKey is empty, requests pass
through
- added startup warning log in cmd/server/main.go when no API key is
configured:
- [security] WARNING: no apiKey configured — write endpoints are
unprotected
- added route tests for missing/wrong/correct key and empty-apiKey
compatibility
## Validation
- cd cmd/server && go test ./... ✅
## Notes
- config.example.json already contains piKey, so no changes were
required.
---------
Co-authored-by: Kpa-clawbot <259247574+Kpa-clawbot@users.noreply.github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Removes the separate config.json file bind mount from both compose
files. The data directory mount already covers it, and the Go server
searches /app/data/config.json via LoadConfig.
- Entrypoint symlinks /app/data/config.json for ingestor compatibility
- manage.sh setup creates config in data dir, prompts admin if missing
- manage.sh start checks config exists before starting, offers to create
- deploy.yml simplified — no more sudo rm or directory cleanup
- Backup/restore updated to use data dir path
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Add 1/2/3-byte selector to Hash Issues analytics page
- 1-byte and 2-byte modes show 16×16 matrix with stat cards (nodes
tracked, using N-byte ID, prefix space used, prefix collisions)
- 3-byte mode shows summary stat cards instead of unrenderable grid
- Fix "Nodes tracked" to always show total node count across all modes
- Use CSS variable colours for matrix cells (light/dark mode compatible)
- Replace native title tooltips with custom styled popovers
- Hide collision risk card when 3-byte mode is selected
- Fix double-tooltip bug on mode switch via _matrixTipInit guard
- Fix tooltip persisting outside matrix grid on mouseleave
https://dev.ve7kod.ca/#/analytics
Hash Issues
---------
Co-authored-by: Jesse <your@email.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Fixes#276
## Root cause
TRACE packets store hop IDs in the payload (bytes 9+) rather than in the
header path field. The header path field is overloaded in TRACE packets
to carry RSSI values instead of repeater IDs (as noted in the issue
comments). This meant `Path.Hops` was always empty for TRACE packets —
the raw bytes ended up as an opaque `PathData` hex string with no
structure.
The hashSize encoded in the header path byte (bits 6–7) is still valid
for TRACE and is used to split the payload path bytes into individual
hop prefixes.
## Fix
After decoding a TRACE payload, if `PathData` is non-empty, parse it
into individual hops using `path.HashSize`:
```go
if header.PayloadType == PayloadTRACE && payload.PathData != "" {
pathBytes, err := hex.DecodeString(payload.PathData)
if err == nil && path.HashSize > 0 {
for i := 0; i+path.HashSize <= len(pathBytes); i += path.HashSize {
path.Hops = append(path.Hops, ...)
}
}
}
```
Applied to both `cmd/ingestor/decoder.go` and `cmd/server/decoder.go`.
## Verification
Packet from the issue: `260001807dca00000000007d547d`
| | Before | After |
|---|---|---|
| `Path.Hops` | `[]` | `["7D", "54", "7D"]` |
| `Path.HashCount` | `0` | `3` |
New test `TestDecodeTracePathParsing` covers this exact packet.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
## Summary
The in-memory `PacketStore` had **no eviction or aging** — it grew
unbounded until OOM killed the process. At ~3K packets/hour and ~5KB per
packet (not the 450 bytes previously estimated), an 8GB VM would OOM in
a few days.
## Changes
### Time-based eviction
- Configurable via `config.json`: `"packetStore": { "retentionHours": 24
}`
- Packets older than the retention window are evicted from the head of
the sorted slice
### Memory-based cap
- Configurable via `"packetStore": { "maxMemoryMB": 1024 }`
- Hard ceiling — evicts oldest packets when estimated memory exceeds the
cap
### Index cleanup
When a `StoreTx` is evicted, ALL associated data is removed from:
- `byHash`, `byTxID`, `byObsID`, `byObserver`, `byNode`, `byPayloadType`
- `nodeHashes`, `distHops`, `distPaths`, `spIndex`
### Periodic execution
- Background ticker runs eviction every 60 seconds
- Analytics caches and hash size cache are invalidated after eviction
### Stats fixes
- `estimatedMB` now uses ~5KB/packet + ~500B/observation (was 430B +
200B)
- `evicted` counter reflects actual evictions (was hardcoded to 0)
- Removed fake `maxPackets: 2386092` and `maxMB: 1024` from stats
### Config example
```json
{
"packetStore": {
"retentionHours": 24,
"maxMemoryMB": 1024
}
}
```
Both values default to 0 (unlimited) for backward compatibility.
## Tests
- 7 new tests in `eviction_test.go` covering time-based, memory-based,
index cleanup, thread safety, config parsing, and no-op when disabled
- All existing tests pass unchanged
Co-authored-by: Kpa-clawbot <kpabap+clawdbot@gmail.com>
## Problem
The RF analytics `packetsPerHour` chart was counting **observations**
instead of **unique transmissions** per hour. With ~34 observations per
transmission on average, the chart showed ~5,645 packets/hr instead of
the correct ~163/hr.
**Evidence from prod API:**
- `packetsPerHour` total: 1,580,620 (sum of all hourly counts)
- `totalPackets`: 45,764
- That's a ~34× inflation — exactly the observations-per-transmission
ratio
## Root Cause
In `store.go`, the `hourBuckets[hr]++` counter was inside the
observations loop (both regional and non-regional paths). Other counters
like `packetSizes` and `typeBuckets` already deduplicate by hash —
`hourBuckets` was the only one that didn't.
## Fix
Added a `seenHourHash` map (keyed by `hash|hour`) to deduplicate. Each
unique transmission is counted once per hour bucket, matching how packet
sizes and payload types already work.
Both the regional observer path and the non-regional path are fixed. The
legacy path (transmissions without observations) was already correct
since it iterates per-transmission.
Co-authored-by: Kpa-clawbot <kpabap+clawdbot@gmail.com>
Root causes from CI logs:
1. 'read /app/config.json: is a directory' — Docker creates a directory
when bind-mounting a non-existent file. The entrypoint now detects
and removes directory config.json before falling back to example.
2. 'unable to open database file: out of memory (14)' — old container
(3GB) not fully exited when new one starts. Deploy now uses
'docker compose down' with timeout and waits for memory reclaim.
3. Supervisor gave up after 3 fast retries (FATAL in ~6s). Increased
startretries to 10 and startsecs to 2 for server and ingestor.
Additional:
- Deploy step ensures staging config.json exists before starting
- Healthcheck: added start_period=60s, increased timeout and retries
- No longer uses manage.sh (CI working dir != repo checkout dir)
Root cause: on the 8GB VM, both prod (~2.5GB) and staging (~2GB) containers
run simultaneously. During deploy, manage.sh would rm the old staging container
and immediately start a new one. The old container's memory wasn't reclaimed
yet, so the new one got 'unable to open database file: out of memory (14)'
from SQLite and both corescope-server and corescope-ingestor entered FATAL.
Fix:
- manage.sh restart staging: wait up to 15s for old container to fully exit,
plus 3s for OS memory reclamation before starting new container
- manage.sh restart staging: verify config.json exists before starting
- docker-compose.staging.yml: add deploy.resources.limits.memory=3g to
prevent staging from consuming unbounded memory
## Summary
Adds `distributionByRepeaters` to the `/api/analytics/hash-sizes`
endpoint in the **Go server**.
### Problem
PR #263 implemented this feature in the deprecated Node.js server
(server.js). All backend changes should go in the Go server at
`cmd/server/`.
### Solution
- For each hash size (1, 2, 3), count how many unique repeaters (nodes)
advertise packets with that hash size
- Uses the existing `byNode` map already computed in
`computeAnalyticsHashSizes()`
- Added to both the live response and the empty/fallback response in
routes.go
- Frontend changes from PR #263 (`public/analytics.js`) already render
this field — no frontend changes needed
### Response shape
```json
{
"distributionByRepeaters": { "1": 42, "2": 7, "3": 2 },
...existing fields...
}
```
### Testing
- All Go server tests pass
- Replaces PR #263 (which modified the wrong server)
Closes#263
---------
Co-authored-by: you <you@example.com>
down tears down the entire compose project including prod.
rm -sf stops and removes just the named service.
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Kpa-clawbot