The capture settled it: the 7760 keeps the transport chatting (pings answered,
no socket error, no loss) while CI-V data simply stops, the in-place reopen is
ignored, and only a fresh session brings it back. So: one reopen attempt at
10 s — kept, it is free and the log will say if a firmware ever honours it —
and at 15 s the link is failed deliberately so the manager rebuilds it,
halving the outage. Also: the recovered-acknowledgement now sits on the DATA
packet, not on any packet — pings were toggling the detector, 160 reopens in
thirty seconds all logged as recoveries that never happened.
The IC-7760 recurrently stops answering CI-V a few minutes into a network
session while the transport keeps chatting — pings answered, session alive,
commands unanswered — and the only recovery was the 30 s watchdog's full
teardown. At 10 s of CI-V silence the pump now says 'open' again on the
existing stream, which is all it should take if the rig quietly closed the
data flow; the watchdog remains as the backstop. The send counters gain a
mutex since the pump now transmits too.
The first detector counted any packet and never fired: pings and idles keep
flowing while the rig stops answering CI-V commands — which is what the 7760's
recurring silence turns out to be. Now it fires on 10 s without a CI-V payload
and reports the ages of the transport, the last scope frame and the last
socket error, so the next occurrence says exactly what the rig was still
sending.
reloadCAT builds the network audio sink gated on audioMgr, and startup created
the manager AFTER the first reloadCAT: every fresh launch connected with audio
silently off however the option was set, and the operator had to
untick/save/retick it to hear anything. The manager is idle-until-used, so
creating it earlier costs nothing.
Every audio packet was delivered as if it were the next 20 ms, duplicates and
late retransmits included. The monitor's capped ring quietly threw the surplus
away, so the speakers sounded fine — but the recorder keeps every sample it is
given, and the file came out longer than the QSO, slowed and stuttering, each
lost-then-resent packet heard twice. trackRxSeq now says whether a packet
advances the stream, and only those are handed on.
The CI-V keyer is the same 0x17 command on the same IcomController whatever
carries the link; only the warnings and enable checks insisted on the USB
backend, telling a network operator their working keyer would fail.
Two halves of the same station confusion. The entry form pre-filled MY_RIG
from the per-band default on every band change, so the log-time priority (the
radio that is keying beats the radio that was planned) never ran — the field
was no longer empty by the time the backend looked. The form now asks the
active radio first, through a new ActiveRadioMyRig binding.
And the manual record button started its capture from the From-radio sound
card unconditionally, where the automatic recorder already knew a network
Icom's audio is pushed from the 50003 stream: a manual take on the IC-7760
was a faithful recording of the Flex's DAX. Same source rule for both, and
the button is offered with no sound card at all when the network stream is
the source.
Two effects, wrong order: the ticking effect ran before the reset effect on a
new take, so its closure captured the PREVIOUS take's elapsed as its starting
point — the counter showed 0 for one second and then jumped back to thirty
minutes. Effects run in declaration order; the reset now comes first.
A real IC-7760 keeps going silent on the CI-V stream two-three minutes into a
session while the control link stays alive. At 10 s of silence the pump now
logs the last real socket error and the retransmit backlog, and notes when the
stream resumes — the 30 s watchdog that follows cannot tell a dead socket from
a rig that stopped talking, and the next occurrence should.
A known 50 W read raw ~89 where the two-segment curve said 62: the face is
not linear in watts below half scale. Watts now interpolate through the
measured points (50 W, 100 W, full-scale 250 W); refining the curve is adding
a row to the table.
The 30 s watchdog message asserted another program had taken the session; a
real IC-7760 went silent once with nothing else on the network, and the
message sent its operator hunting software that was not installed. State both
causes.
The power meter applies Icom's own meter calibration (raw 143 = half
deflection, 213 = full scale) anchored against the real radio: a measured
100 W sits at half deflection of the 7760's 250 W face — the linear guess
showed it as 140. The RF power slider says watts on a 200 W rig instead of a
percentage the operator has to convert.
And the Listening button, pressed while a network Icom is connected, restarts
the monitor render-only: it used to open a USB capture from the From-radio
device as well, and that second producer interleaved with the network pushes
chopped the audio to pieces.
The loader looked for an 'active' flag the radio list has never carried, so
reopening the panel always landed on the first entry — Radio 1's name over
whichever radio's settings were live, and every field edit (MY_RIG included)
silently rewriting the wrong entry. The backend is the only one who knows
which radio is on the air; ask it.
The console gains a DATA mode button (USB-D, what FT8 and friends ride on);
the PSK button now maps to the rigs' native PSK mode (0x12) instead of erroring
as an unknown mode — rigs without one NAK it, exactly as RS-BA1's button does
there. The power meter shows real watts on the IC-7760, whose meter face runs
to 250 W where a 100 W rig's percentage was the watts. The clickable S-meter
gets a pointer cursor so it looks like what it is.
Network audio: rxcodec 0x04 — the experiments on the real 7760 settle the
codec table (0x10 = 16-bit stereo, 0x02 = 8-bit mono, both heard as garble),
and 0x04 is the 16-bit mono the playback path actually plays.
The IC-7760 acknowledges the 1A 06 data-mode command and stays in USB-D
anyway, which left the operator unable to get back to plain USB from the
console at all. After every mode set the flag is read back; on a mismatch it
is said again with 0x26 — mode, data flag and filter in one frame, the form
the newer rigs actually honour. Only on a mismatch, so rigs that predate 0x26
never see the command.
The rig keeps sending two-channel LPCM after rxcodec asks for one — 1280-byte
payloads, 320 stereo frames per 20 ms tick, right channel all zeros in the
capture. Played as mono that interleaving is half-speed metallic garble. A
stereo-sized packet is folded to its left channel; a 640-byte mono packet from
a rig that honours the request passes through untouched.
The first real radio on the 50003 stream (an IC-7760) settles the two guesses
the experimental audio path shipped with. The payload starts at 0x18 — the
packet carries a big-endian payload length at 0x14 (0x500 on every packet) —
not at 0x16, which swallowed two header bytes into the PCM and laid a 50 Hz
click track under everything. And rxcodec 0x10 asks for TWO-channel LPCM,
which the mono playback path rendered as double-speed garble; 0x02 asks for
the one channel the monitor plays.
A rig that answers something unrecognised (a newer model, another firmware)
and a rig that answers nothing are different faults; a silent timeout hides
which. Capped at a dozen packets so a working handshake cannot flood the log.
Prompted by the first IC-7760 network attempt — which turned out to be aimed
at the console's IP, where nothing listens; the remote server lives on the RF
deck.
The S-unit mapping was provisional, waiting for a real radio. One arrived:
side by side with a K3's own display, raw 5 reads S7 and raw 9 reads S9+20,
so S9 sits near raw 6.5 — not 9 — and each raw step above it is worth ~8 dB.
The old scale showed everything two S-units low.
The SWR meter borrowed the power meter's needle inertia — hold the peak, then
close a quarter of the gap per poll. The K3 throws a brief SWR spike as an FT8
frame ends, and that decay turned one bad sample into twelve seconds of red on
the next transmission. SWR now holds the peak for the same 1.5 s and then
snaps back to the live reading: it is a warning light, not a needle.
The block's one genuinely new feature opens it — with the ClubLog enrichment,
which had not made it into the entry yet — instead of sitting seventh between
an Elecraft button row and a bulk-edit field.
Double-clicking a cluster spot tuned the slice and zoomed the panadapter in
the same breath. On a cross-band jump SmartSDR moves the panadapter itself to
follow the slice, and that follow lands after — and overrides — a zoom sent in
the same instant, so the width the operator configured never appeared. A spot
clicked ON the panadapter never hit this, because the radio had already tuned
before our zoom arrived; the two paths now behave the same. The Chase new
panel's pick is sequenced identically.
Every other spot click — cluster grid, band map, watchlist, alert list — goes
through handleSpotClick, which applies the operator's per-mode panadapter
width after the tune. The Chase new panel had its own pick handler that tuned
and stopped there; it now makes the same view change, after the tune for the
same reason.
clublog.org/watch.php, per entry: the DXpedition flag, OQRS, LiveStream, the
log's QSO total and the last-24h rate — the fields the schema has carried since
phase 1. Refreshed on DXHunter's own cadence (hourly for expeditions,
six-hourly for the rest), two at a time with a breath between requests: the
application API key is shared by every install, so a hundred-entry list must
read as a trickle at ClubLog's end. Nothing to configure — OpsLog's own key,
already used for cty and Most Wanted, serves.
A starred pattern is asked about by its base (ClubLog has no log for VK9*).
The card gains the 24h rate beside the QSO total and the Live link to the
expedition's stream.
The two-click arm goes: removing an entry is cheap to undo (type it again), so
it does not earn a confirmation its neighbours do not have. The bin reddens on
hover and the green line confirms the removal.
The two-click arm compared against the STATE variable, which each click's
handler captured at render time — a quick double-click read the pre-arm value
twice, so both clicks merely armed: the bin turned red and nothing was ever
deleted. The arm lives in a ref now, compared synchronously, and the window
grows to three seconds.
A green line under the toolbar says what was added (and as what — the contest
variant names the per-UTC-day rule); the already-exists error keeps its red
line. Both fade after four seconds: a stale 'added' from ten minutes ago reads
as a fresh one.
It was created after the (possibly remote) logbook connect, so a 10-30 s MySQL
dial left the tab saying 'Watchlist: 0' until the next refresh tick — twenty
seconds of an empty list for a JSON file that reads in a millisecond. It now
opens right after the data dir is known; only the auto-contest pattern waits
for the settings store.
The prefix matching ported from DXHunter swallowed too much: a three-letter
special-event entry N8W lit up NEEDED for N8WCR, a different station entirely.
The implicit becomes explicit — a bare entry matches exactly that call, a
trailing star makes it a family: VK9* catches every VK9…, RI0SP* the
expedition's portable forms. A star anywhere else (or alone) is refused at Add
rather than silently matching nothing. Same rule in the backend Match and the
tab's own matcher, pinned by test.
Add-controls left, filters right, justify-between: when the pane narrows, the
whole filter group drops below as a block instead of shedding its last chip.
Shorter chip labels (Active / Needed) and a slimmer search keep both groups on
one line at the tab's normal width.
The F1 panel widens 560 → 600 px, the header row drops its outer padding and
wide gaps, and the country badge tightens — so 'Central African Republic' and
the NEW SLOT chip share one line instead of the chip dropping alone underneath.
The counters are the tab's headline and now sit alone, centred, above the
toolbar; everything one interacts with lives on the second row. The inputs stop
shouting their placeholders — the uppercase belonged to the VALUE (callsigns),
not the hint text — and the auto-contest field says what it is.
A new spot rendered as Needed the instant it arrived and was withdrawn half a
second later when the worked answer landed — the list twitched on every burst.
Unsettled spots are now invisible everywhere (card lines, counters, the
Active/Needed-only filters) until their verdict is in: nobody needs a spot
150 ms early, they need it settled. The answer map is merged rather than
replaced, so a settled line can never fall back to unknown, and the query
debounce drops to 150 ms since there is nothing on screen waiting for it.
DXHunter's header row, ported: Watchlist / Active / Needed counts up front and
the All-Modes select beside the other filters (DIGI matches the digital class,
SSB folds USB/LSB). Counters, card lists and the Active/Needed-only filters all
read the same mode-filtered view, so the numbers add up to what is on screen.
And the notify alert now asks the SAME worked-slot question the tab asks —
before making a sound. It used to fire on the raw spot while the tab's verdict
arrived on a debounce, so the bell rang for a slot that showed Worked a moment
later. Judged in the backend at emit time: exact slot for named modes, digital
class for generic DATA, today-only for contest entries.
The wiring script died on an app.go anchor mid-run: the backend half was
committed, the frontend half (the pattern field, the changed-event refresh, the
i18n keys) never ran. Replayed; the field sits after Add.
A needed line already says so twice — the amber left border and the Needed!
badge — and the exclamation mark was noise. The green tick stays on worked
lines.
A field in the tab's toolbar (e.g. WWA): while non-empty, any spotted callsign
CONTAINING it joins the watchlist as a contest entry by itself — a
special-event fleet (HB9WWA, DL0WWA, F4WWA/P…) is collected as it appears
instead of typed in one by one, and each is judged per UTC day like any contest
entry. Contains, not prefix, because the event string sits anywhere in those
calls.
Global setting, cached in an atomic so the spot pipeline never touches the
settings store per spot. Emptying the field stops the collecting but keeps what
was collected — DXHunter deletes non-matching contest entries on a pattern
change, and a setting that silently empties an operator's list is the one
behaviour of the original this port refuses.
The tab also refreshes on backend auto-adds (event) and on a slow tick, so
last-seen and the counters stay honest while it sits open.
max-w-4xl left 'Central African R…' cut mid-name: the column grows to 5xl and
the country cell flexes with the row (capped at 14rem) instead of a hard 8rem.
7.0589 falls in the band plan's generic [7.04-7.1] DATA segment — an F/H
DXpedition off the standard dials — and 'DATA' is in nobody's log, so the spot
read NEW MODE / Needed while the FT8 contact sat right there. A generic mode
cannot be matched exactly; it is now judged at digital-class grain — worked if
ANY digital mode of that call is logged on that band — against a second
class-folded slot set built beside the exact one (shared when the grouping
option already folds them). DXHunter answers this with a finer frequency table,
which fails on exactly these off-dial expeditions; the class rule cannot.
The toolbar also wraps as a whole with tighter gaps instead of dropping its
last filter chip onto a lone second line.
The watchlist joins the Main-tab pane list (Settings → Main view), rendered in
a card like the other docked panels.
The three filters persist across tab switches — the component unmounts on every
switch, and filters that reset each time are filters nobody trusts.
Needed-only now hides the worked SPOT LINES as well as the all-worked cards,
which is DXHunter's own Not-Worked behaviour and the reported complaint: a
filter that says needed and still lists five green Worked rows answers a
different question than the one asked.
The add-row checkbox is relabelled 'as contest' — it marks the entry being
ADDED, and read as a filter beside the All/DX/Contest buttons it looked
redundant.
The first version asked the alerts' worked-index — keyed by RAW mode (FT8) —
with a CLASS name (DIGI). Nothing ever matched, and a DXpedition worked on five
bands showed every slot as Needed.
The verdict now reads the cluster's own WorkedCallSlotKeys set, built in the
cached status maps (now unconditionally — it used to exist only when a display
option wanted it). Same index, same mode normalisation as the grid, so the
watchlist can never contradict the cluster about the same spot. The contest
today-set is normalised through the same function.
And the cards column is capped at a reading width instead of stretching
callsign-to-badge lines across the whole window.
The concept, transplanted: a list of callsigns or prefixes being hunted,
matched against the live spot stream, one card per entry with the spots
underneath and the two questions that matter answered on every line — is this
slot still needed, and what is it worth (the cluster's own NEW badges, read
from the same status index).
The file is DXHunter's own watchlist.json, field for field, ClubLog block
included though phase 2 will fill it — a file that round-trips unchanged is the
whole of 'same format', and a test pins it with a real DXHunter entry. Global
(dataDir), not per profile.
CONTEST is per entry, not the global mode DXHunter has: a contest entry is
judged against the current UTC day — the boundary lives in the query's date
bound, so midnight needs no timer and resets nothing. Normal entries read the
same in-memory worked index the alerts use. Prefix matching is why RI0SP
catches RI0SP/MM, pinned by test.
Notify goes through the existing alert:fired event — the frontend already
toasts and sounds it — throttled to one alert per entry per two minutes,
because a DXpedition lights every skimmer on the planet.
Tab wired like NET Control: opt-in from Tools, persisted, closable. Single
click fills the callsign, double click works the spot — the cluster's own
gesture, kept.
The My-station group covered nineteen fields and not the three numeric ones.
They take their own integer path rather than the generic text setter: the
columns are nullable integers, and while SQLite would coerce '14' quietly, a
shared MySQL logbook would not — and empty must become NULL, never ''. Bounds
checked (DXCC <1000, CQ 1-40, ITU 1-90) so a slip cannot stamp zone 400 across
a thousand rows.
'None of the consoles look alike' is the complaint ShiftRow exists to answer,
and the Elecraft panel was still driving its offset with a row of nudge buttons.
It now uses the shared control — ±, wheel, typed value — plus the same
Ctrl+←/→ 10 Hz clarifier the Icom and TCI panels have. The K3 keeps ONE offset
for RIT and XIT, so both rows show it, exactly like the Icom's.
The backend gains the absolute setter the shared control speaks
(SetKenwoodRITOffset, funnelled through the same RO write as the nudge), and
the power slider moves in 1 W steps — nobody asks a K3 for 'between 10 and 15
watts' in fives.
Also per review: the chase switches are labelled 'Chase POTA' / 'Chase SOTA',
matching 'Chase new grids' beside them.
An operator who does not chase parks does not want NEW POTA shouting from every
activator spot. Two switches (Settings → DX Cluster, on by default): off, the
marker is withdrawn at the display layer — applySpotDisplay, the same chokepoint
the grid, the band map and the filter read through — so the badge, the colour,
the filter chip and the reference column all go quiet together, and a new-band +
new-POTA spot reads NEW BAND alone. The facts keep being computed; only the
telling stops, so ticking the box back on needs no rescan.
Per review: the digital-grouping sentence on the same-slot option, the
spot-lifetime explanation, the ring-buffer paragraph on the spots-kept field,
and the free-public-nodes line all go.
rouggy