Expert Electronics' own document gives the command whole:
SPOT:RN6LHF,CW,7100000,16711680,ANY_TEXT;
16711680 is 0x00FF0000 — positive. This backend sent the same field as a SIGNED
32-bit value, from a third-party example: with the alpha byte set to FF for
opacity, 0xFFFFA500 becomes -22336, and a spot whose colour ExpertSDR cannot
read is dropped without a reply or an error.
Reported on ExpertSDR3 1.3, which speaks TCI 2.x — the protocol version was
never the problem, and my first reading of that report was wrong.
The first three spots are now logged verbatim: an ignored spot leaves no trace
whatsoever, so that line is the only evidence of what went out.
TCI meters: the S-meter came only from RX_SMETER and the transmit meters from
TX_POWER / TX_SWR — commands ExpertSDR3 does not send. The protocol's answer is
a subscription (RX_SENSORS_ENABLE / TX_SENSORS_ENABLE, §4.4 of the TCI PDF),
after which the radio pushes RX_CHANNEL_SENSORS and TX_SENSORS. Nobody had
asked, so the console's meters sat empty in RX and in TX while everything else
worked.
Cluster: the held-spot counter looked for the row it froze on. A station spotted
again REPLACES its row, so that row vanishes in the ordinary course of things
and the count fell through to 'everything is new' — 4, 5, then 500. Counted by
timestamp now, which survives both the replacement and the ring buffer.
SMTP: '535 5.7.139 basic authentication is disabled' is a policy, not a typo.
The message now says so, and says what actually helps, without claiming a policy
when the server simply rejected the password.
Every flush commits state the whole window re-renders on, and the window was a
flat 50 ms — twenty full renders a second under an RBN firehose. That is paid
everywhere else in the interface, and it is what a dropdown highlighting its
entries a beat behind the mouse actually is.
The window now follows the rate: 50 ms on a quiet cluster, where a spot should
appear the moment it arrives; 200 ms past five a second; 500 ms past twenty,
where half a second's delay on a line already scrolling past is invisible.
The pane drew its own title bar above the grid's toolbar, so the cluster spent
two lines on headers where Recent QSOs beside it spent one — in a pane often
only a few spots tall. The grid's toolbar now takes what the caller wants on its
left, and the pane passes its title, live count and Filters button there.
Reported by VK4DX.
UploadLoTW returns both a terse error ('no QSOs processed') and a Message
carrying TQSL's own lines — '17 QSO records were already uploaded', 'N QSO
records are out of date range'. The caller took the error whenever there was
one, which is precisely the half that explains nothing: the operator saw a red
line naming a condition with three unrelated causes and no way to tell which.
The sixth CW engine, and the one the TCI backend was left without. A SunSDR
keys from the WebSocket that already carries the CAT: no WinKeyer, no second
serial port, nothing to wire.
CW_MACROS:0,<text>; send
CW_MACROS_SPEED:<wpm>; speed
CW_MACROS_STOP; abort
Confirmed against the TCI command table in ars-ka0s/eesdr-tci — the source that
settled SPOT. CW_MACROS rather than TCI 2.0's CW_MSG: it exists from 1.6, and
OpsLog resolves its own variables, so the before/after callsign fields have
nothing to carry.
Commas and semicolons are stripped before sending. They are the protocol's own
separators: a comma would become another argument, a semicolon would end the
command with the message half keyed and the rest parsed as a command.
Only the MACRO speed is set, not the paddle keyer's — an operator who set their
paddle to 28 wpm did not ask for it to change because a macro went out at 25.
And there is no backspace: TCI can stop a message but cannot un-type one, so the
type-ahead correction the Flex engine offers is absent rather than faked.
Wails saves and restores in two different coordinate systems. WindowGetPosition
returns GetWindowRect — absolute. WindowSetPosition is winc's ControlBase.SetPos,
which adds the CURRENT monitor's work-area origin:
w32.SetWindowPos(hwnd, HWND_TOP, int(info.RcWork.Left)+x, ...)
On the primary monitor that origin is 0 and nothing shows. With a monitor to the
LEFT it compounds: a station with its left screen at x=-3840 saved -3844,
reopened there, had -3840 added, and stored -7684 — then -11524, one screen
further out at every launch, until the window was off the desktop entirely.
So the placement is done here, with SetWindowPos and no offset, falling back to
the toolkit when the handle cannot be found. Compact mode had the same pairing
and is fixed with it.
Clicking a spot sets the frequency, then the mode 150 ms later, and SetMode
chose USB or LSB from t.freqA — the frequency the radio had ECHOED. From 7 MHz
onto a 14 MHz spot that echo has often not landed, so the sideband was computed
for the band being left: LSB on 20 m. Clicking the same spot again, the echo
having arrived, gave USB. Reported from a SunSDR as 'the frequency is right, the
mode is wrong until I click twice'.
SetFrequency now records what was asked for and SetMode prefers it until the
radio's own echo confirms the move.
The server's 'protocol:' announcement is also kept instead of being filed under
unhandled: it names the ExpertSDR and its TCI version, and panorama spots need
TCI 1.5 — older servers accept the spot commands and silently drop them, which
is indistinguishable from a broken logger. Now it says so.
The status comes from spotEntityStatus: it means the country has been worked on
this band and mode. The label said 'Worked (this band + mode already in log)',
which reads as 'I have had this station' — so a spot the cluster called NEW PFX
looked like a contradiction, when the two were answering different questions
about different subjects.
The callsign-level fact has its own marker and its own words ('Callsign already
worked'), so the two are now told apart by reading them.
The strip deliberately shows three markers of five — the pill is 22 px and a
fourth segment makes it a colour code nobody reads. But the tooltip has room and
was filtering them out too, so the same spot read NEW PFX in the cluster and
'Worked (this band + mode already in log)' on the map. Both were true: the
entity is worked on that slot and the WPX prefix never has been. Neither view
said so.
A tester compared LogHX's two download options on the same account: two minutes
for the confirmations, twenty for the confirmations with QSO details. OpsLog was
always sending qso_qsldetail=yes — LogHX's slow option — which is the whole of
the wait people were reporting.
Marking a confirmation needs call, date, band and mode; the detail adds the QSL
date and the station's grid, state and county. Now a choice, off by default, and
forced on when the download is also ADDING the QSOs it cannot find — the one
case where those fields have no other source.
Miles and the LoTW download progress landed after v0.26.20 was tagged, so they
belonged to the next block: an entry appended to a shipped one never reaches the
operators who installed it, and misstates what that version contained. The .20
block is byte-identical to the tag again.
Deferred, it stopped at the end of the function — so 'still waiting for LoTW to
build the report' kept printing between the megabyte lines, saying the opposite
of what was happening.
The progress lines only start once bytes arrive, and on a large account nothing
arrives for minutes — LoTW builds the entire report before sending any of it.
That silence WAS the complaint: the window still just said 'working'.
A heartbeat every fifteen seconds until the headers land, then the HTTP status,
then the megabytes — the first at 256 KB rather than 1 MB, so the very first
sign that it is moving comes early.
Three reports of the same shape: a 503 for anything wider than a few days
before, and since the timeout was raised, a window that sits at 'working'
forever. Both are the same missing thing — nothing said what the transfer was
doing.
The body is now read in chunks and every megabyte is reported. A 503/502/504 is
retried twice, 20 s then 40 s, each attempt announced: LoTW answers 'busy' to a
wide report often enough that other loggers simply ask again. And the deadline
is no longer on the whole exchange, which either cut off a healthy slow download
or hid a dead one for twenty minutes — it is ten minutes to START answering
(LoTW builds the whole report first) and two minutes of silence once it has,
which is the difference between slow and dead.
Everything is computed in kilometres and converted once at display time, in
lib/units — storing miles anywhere would give the same number two sources of
truth and a rounding error that grows with every hop.
The grids capture the unit inside their column definitions, header and formatter
both, so a change is published to them the way the date format already is;
without that a toggle would only appear after a language change or a restart.
The preference is portable, like the other display ones.
Without qso_qslsince LoTW does not answer 'every confirmation' — it answers
with a few recent ones, as a 200 and a valid ADIF. 'All' therefore downloaded a
nearly empty report and said 'matched 1 of 1' on an account holding twelve
thousand. It now asks from 1945-11-15, older than any QSO LoTW accepts.
And the request had two minutes: LoTW spends several of them building a full
account before the first byte, so asking for everything ended in 'context
deadline exceeded while reading body' — a network-shaped message for a
too-short deadline. Twenty minutes, and the read cap raised past the 18 MB such
a report actually weighs.
Unscoped, the report carries every station on the account — including the calls
belonging to another profile's logbook. Those match nothing here, and with 'add
the ones not found' ticked they would pour a second log into this one. The
station callsigns this logbook actually holds are now the filter: a portable
worked here is kept, an expedition call never used here is skipped and counted.
A near-empty report is also the failure that reads as success — 'matched 1 of 1'
where the account holds twelve thousand. Under 4 KB, what LoTW answered is shown
verbatim instead.
The download passes qso_owncall so one profile does not pull another's
confirmations. That scope also hides them: a QSO made as F4BPO/P is confirmed at
LoTW and can never come back to an F4BPO profile, so it stays unconfirmed here
for good while the ARRL counts it — found while reconciling a DXCC account
against the panel, where it was the whole of the CW gap (Jersey, worked as
F4BPO/P).
Off by default: the scope is right for anyone whose profiles are separate
stations.
The filter keeps a REFERENCE for its unconfirmed slots, so its confirmed ones
come along — and green on grey means the answer is the quietest thing in the
row. Faded to a quarter under that filter only, not hidden: which band is
already done is part of deciding which one to chase.
The Challenge line says 1832 worked against 1554 confirmed and stops there —
278 band-slots waiting for a card with nothing pointing at them. Every existing
filter works per REFERENCE, so an entity confirmed on 20 m answers 'confirmed'
and its unconfirmed 15 m contact stays invisible.
The new filter keeps the references holding at least one such slot, and the
count beside the reference total is the gap itself, recomputed over whatever the
other filters left on screen and over the bands actually displayed — so it
always adds up to the columns in front of the operator.
The SOTA feeds put the summit in the spot's own text, so there is nothing to
poll: a regex on the comment gives the reference, and clicking the spot writes
SOTA@<ref> into the award references the way a park already wrote POTA@<ref>.
The regex is narrow on purpose — association/region-NNN, anchored — because
POTA (US-4475), WWFF (DLFF-0001) and portable callsigns (DL/SP9DPM/P) share
that field and none of them is a summit. Pinned by a table test.
Off by default: the reference is only there on the summit feeds, and an empty
column for every operator who does not chase them is worse than a checkbox.
The chip read a.spe ?? a.acom ?? a.pgxl, so a KPA fell through to the
not-connected fallback: red on a working amplifier, and a click that opened the
settings instead of switching to STANDBY. Reported from a KPA500 whose panel
showed band, SWR and temperature at the same time.
The card's offline text also borrowed the Acom's string, so a KPA that dropped
its link announced itself as an Acom.
The CI-V trace settles it: the IC-7851 answers 27 10 01 with FB and 27 14 00
with its mode, but refuses 27 11 in BOTH shapes — it controls its scope over
CI-V and never streams it. Its last firmware is from 2016, older than the
waveform stream itself; Hamlib's caps claim otherwise, the radio is the
authority.
Rejected both ways is a permanent answer, so it is remembered and reported:
the panadapter says the radio does not send its scope, which is the one thing a
black rectangle could not say.
Hamlib documents the spectrum scope on the IC-785x, so a rejected 0x27 0x11 is
not proof the command is missing — and taking 0x27 0x10 for the data switch
would have blanked that operator's own scope screen when OpsLog disabled it.
Both shapes were refused within 96 ms, so neither was a timeout, and guessing
the next subcommand is how the last two rounds went. The retry now logs the
second attempt's own error and the failure points at the CI-V trace.
The IC-7851 rejects 0x27 0x11 in BOTH shapes: on that radio 0x27 0x10 is the
waveform-output switch itself, not a display switch. Rejected both ways, take
0x10 as the output command and switch it off on disable too — there it stops a
stream rather than blanking the operator's own scope screen.
A mode/span/edge answer also settles the selector question directly: three bytes
or more means the rig wants the main/sub form. Worth reading, because a
wrong-shaped SET is answered with silence on this firmware, and a timeout is
something the rejection retry cannot act on — which is why the fixed/centre
sets were timing out.
The IC-7851 rejects "27 11 01" and wants "27 11 00 01" — the main/sub selector
its two scopes need. The shape was picked from a list of CI-V addresses, so an
address not on it got the wrong one and the scope stayed blank with a rejection
in the log and nothing on screen.
A rejection is an unambiguous answer, so a rejected 0x27 set is now re-sent in
the other shape and the winner remembered. A TIMEOUT deliberately is not
retried: several firmwares never ack a 0x27 set, and treating silence as a
wrong guess would flip a rig that was already working.
The header counts QSOs with the callsign through callMatch — RI1FJL/1 and
ZA/RI1FJL included — while the grid's own per-callsign columns matched the bare
string. A cell could hold one of the QSOs the header had counted and show
nothing.
The dot is drawn over all five cell backgrounds, so it cannot borrow one of
their colours; it defaults to the theme foreground with a background ring and
now takes two overrides of its own, worked and confirmed, alongside the rest of
the matrix palette.
The worked-before cell answered one question with two facts fused into it.
bandStatusCode takes the highest of them, so a confirmed entity outranks a
worked callsign — right for awards, wrong while chasing: a slot worked with the
DXpedition this morning reads "entity confirmed" and says nothing about it.
BandStatus now carries the callsign's own answer alongside the entity's, and the
cell draws a dot for it. The dot is deliberately identical everywhere — same
shape, same colour, over all five backgrounds — so it is read as one thing
rather than as five variants of the colour underneath it.
Operating conditions describe a plan; a logged QSO asks which radio was
keying. With one rig those were the same answer, with several they are
not — so each saved radio carries its own MY_RIG, ahead of the per-band
station and doing nothing at all when left empty.
With one rig, MY_RIG belonged in Operating conditions and nowhere else.
With several, that becomes the wrong place: operating conditions describe
a PLAN — 'on 20 m I use the beam and the 7300' — while the question a
logged QSO asks is which radio was keying.
So each saved radio carries its own MY_RIG, consulted BEFORE the per-band
station. When the two disagree — a second rig borrowed for one evening on
20 m — the one that is transmitting is the true answer.
Left empty it changes nothing at all: the old chain (band default, then
the profile's rig) answers exactly as it did, which is what every
single-radio station will keep doing without touching a setting.
The answer to 'I have two rigs' was 'make two profiles', which moves the
whole station — logbook included — to change which radio is connected.
Radios are a list now, like amplifiers, and the CAT chip switches between
them while everything else stays put.
The active radio still lives in the settings keys the rest of OpsLog
reads, so the consoles, the CAT sharing and the band-follow needed no
change at all.
It showed what the radio calls itself over CAT — 'Kenwood (911)' for a K3,
a bare 'CAT' for a Flex that reports nothing useful. That is the answer to
a question nobody asks once the rig has a name in the list, and it made
the two radios look alike in the one place they have to be told apart.
The name the operator typed comes first now, on the chip and in the menu,
and falls back to the radio's own identity when they typed none. The model
identity is still in the tooltip, where it answers 'what is actually
connected' without taking the place of 'which of my rigs is this'.
It only became a menu from two radios, on the reasoning that a single
radio has nothing to switch to. That hides the feature from precisely the
operator who has not made a second radio yet — the click lands on a
settings dialog they did not ask for, which is what happened on the first
try. With one radio the menu now shows that radio and 'Add a radio…'.
Right-click still goes straight to the CAT settings.
The answer to 'I have two rigs' was 'make two profiles', which is a heavy
instrument for a light question: a profile carries the whole station —
logbook, callsign, awards, cluster, macros — so switching one to change
which radio is connected takes all of that with it, including a change of
logbook when the profiles point at different databases.
Radios are a list now, the way amplifiers already are. The CAT chip in the
status bar opens it, one click connects the chosen rig, and nothing else
about the station moves.
The storage keeps the ACTIVE radio in the same settings keys everything
already reads (cat.backend, cat.icom_port, …), and switching writes the
chosen entry into them and reloads the link. So the consoles, the CAT
sharing, the band-follow and every panel see exactly what they saw before
and needed no change at all — the list is a second store beside the
configuration, not a replacement for it.
Details that matter more than they look:
- An operator who has run one rig for a year opens the list and finds it
there, because a missing list reads as the CURRENT settings rather
than as nothing. It is written on the first save, not on the first
read.
- Editing the CAT panel updates the active entry, or an edit made before
switching away would be lost on the way back.
- A new radio starts as a COPY of the current one: a second rig is
usually the same shape with one port changed, and an empty form is a
form to fill in twice.
- Switching saves the panel first, so edits to the radio being left
behind are kept.
- The chip only becomes a menu with more than one radio; with one it
opens the CAT settings exactly as it always did.
THE README WAS GONE. A commit of mine last night replaced OpsLog's readme
with the text of the TCI protocol specification — 322 lines of ours for 77
of Expert Electronics' — and nobody noticed because nothing builds from
it. Restored from the commit before, then updated.
What it was missing, all of it shipped since: the Kenwood/Elecraft and
Yaesu CAT backends and their consoles, the SunSDR console, TCI audio in
both directions, CAT sharing over Hamlib rigctl or TCI, the Elecraft
KPA500/KPA1500 and the Tuner Genius XL, Cloudlog and HAMLOG uploads, the
RDA district work, the DXCC Challenge, the editable satellite list, four
main-view panes instead of two, and Esri basemaps in place of Carto.
Also a section on importing a log, which had none — including the rule
that matters most to somebody moving years of contacts across: a deleted
DXCC entity keeps the number the file gives it, because cty.dat only knows
where a callsign is today.
It has not shipped yet, so an operator reading about the SunSDR console
should meet what it does — not a list of repairs to a thing they have
never run.
Everything here came from one operator with the radio in front of him, and
most of it was mine to fix. The drive commands need the transceiver index
('drive:0,15;', not 'drive:15;') or the radio ignores them without a word.
The console now holds a click the radio does not echo, so a working button
stops looking dead — and lets go the moment the radio contradicts it, so a
REFUSED setting still tells the truth. Filters per mode, APF only in CW,
levels on wide rows with typed values, the RIT control shared with the
Icom console, the transmit meters, and a TUNE that can be switched off
again.
Two findings worth keeping: this radio re-asserts sql_enable eighty
milliseconds after being told to turn the squelch off, and it never
answers TX_POWER or TX_SWR — both are its own doing, and both are visible
in the log rather than argued about.
The transmit meters are asked for four times a second while keyed, so
logging every arrival would fill an evening's log — and a diagnostic that
fills a log is one that gets switched off instead of read. Twenty of each
is enough to tell an answer from a silence, which is all it is for.
Reported with a Logger32 export: the record said <DXCC:3>151 and OpsLog
stored 54.
cty.dat answers one question — where is this callsign TODAY — and that is
the wrong question for a log. R1MVI was Malyj Vysotskij Island, entity
151, until the ARRL deleted it in 2012; the same call resolves to European
Russia today. Forcing the lookup onto a 2004 contact therefore destroys a
credit that can NEVER be worked again, because the place no longer counts.
So the ADIF's number wins whenever it names a deleted entity, and nothing
is corrected on those records at all — the country name and the zones of a
deleted entity are equally beyond what a present-day prefix table knows.
The list is the ADIF specification's own enumeration, sixty-odd entities
that only change when the ARRL deletes another one.
The guard sits in the forced path, so it also covers the grid's 'Update
from cty.dat' on a selection — the other way an operator could have lost
the same credits, one right-click at a time.
Two things from one screenshot, and the second is arithmetic rather than a
bug.
Editing an award refreshed the grid and not the statistics matrix, so
changing which confirmations count left the table showing the old rule's
numbers with nothing to mark them stale. That is how a setting comes to
look as though it does nothing.
And OpsLog's 3229 against LoTW's 3004: the Challenge counts confirmed
band-slots on ten bands, and 60 m is not one of them. 3229 less the 224 in
the 60 m column is 3005. So the number was right and the QUESTION was
different — the panel now answers the Challenge one too, under the table
where the comparison is made.
The two totals also say what they are on hover. 'Total' and 'Général' side
by side tell nobody which counts entities and which counts band-slots, and
3309 read as entities is an impossible number — which is exactly how it
was read.
The log shows 'tx_power;' going out four times a second during a tune and
nothing coming back — but that proves less than it looks: a reply that
arrived and failed to parse leaves exactly the same trace as one that
never came. Both are now visible.
Two faults with one cause: this radio does not echo 'tune:0,true'.
So the panel never knew a tune was running. The button stayed on TUNE and
every further press sent another START — there was no way to stop it from
here at all. The state is recorded when the command is sent now; whatever
the radio says afterwards still wins, it simply never says anything.
And the transmit meters were asked for only while t.tx, which a tune
carrier does not set: the radio reports tuning as its own state, not as a
transmission. So power and SWR sat at zero for the whole tune — the exact
carrier an operator holds a tune for in order to watch an SWR on. They now
follow PTT or TUNE, and the S-meter reads '—' under our own carrier
either way.
TX_POWER and TX_SWR are read-only commands the radio answers when asked,
so they are asked for on each poll of a KEYED radio and not at all
otherwise: a receiving station pays nothing for meters nobody is watching.
Both appear next to the S-meter only while transmitting, because showing
them the rest of the time would show the last thing that happened as if it
were now.
An SWR of 0 draws as '—' rather than as 1.0. A perfect match on an antenna
nobody has measured is the one reading an operator should never be handed.
There is no temperature. The protocol's command list has TX_POWER and
TX_SWR and nothing thermal at all — so rather than leave the question
hanging, it is written down where the next person will look for it. A
temperature invented from something else, on a transmitter, is exactly the
kind of number somebody would trust.
A spot at 10131.5 showed as DATA in the cluster list and raised an alert
announcing SSB. Two tables answered the same question: the frontend knew
10.130-10.150 is the 30 m data segment, while the alert engine knew a CW
range that stopped at 10.130 and ended with a bare 'return SSB' for
everything else.
That fall-through was the fault. A band plan that answers 'SSB' for
anything it does not recognise is not a band plan, it is a default wearing
one — and it fed the alert text, the FlexRadio spot colours and the mode
the radio is told to switch to.
The Go table is now transcribed from the frontend's, segment for segment,
and returns nothing where it knows nothing. The order matters and is now
load-bearing — the FT8 and FT4 watering holes are listed BEFORE the wide
data blocks they sit inside — so there is a test that fails if anybody
sorts the table by frequency and quietly turns every FT8 hole into DATA.
WPM counts as CW while we are here: nothing else is reported in words per
minute, and RBN puts it on every line.
The log settles it, and it was not mute. Mute works: '→ mute:0,true;' and
the radio answers 'mute:0,true'. What happens is that the squelch REFUSES
to go off — we send 'sql_enable:0,false;', the radio echoes false, and
eighty milliseconds later announces 'sql_enable:0,true' again. Six times
in the log, always the same shape. The SQL lamp coming back on right after
a MUTE click is that revert landing, not the mute doing it.
So the optimistic hold is fixed rather than the phantom. It was a fixed
1.2 s, which is wrong in both directions: too short for a setting the
radio never echoes at all (AGC), and too long for one it refuses, where it
showed the operator a lie for a second. The requested value now stands
while the radio says nothing about that setting, and the instant the radio
reports any change for it, its word replaces ours. A refusal is therefore
visible immediately, and an unechoed setting still sticks.
Also from the same session: filters per MODE — the narrow end in CW, the
voice widths in SSB, a middle set for the digital modes — because 250 Hz
in SSB and 2.8 kHz in CW are buttons nobody presses. And APF is shown only
in CW: it rings a single tone out of the noise, which is a CW tool and
nothing else.
250 now means 0-250. It meant 575-825: the width was right and it was
centred on the CW note, on the reasoning that a CW filter should contain
the note. That reasoning may be right for a radio and it is still wrong
here, because it is not what the button says — and a button that does not
do what it says is worse than one that does something simple. The two
edges are editable underneath for anything else, which is what TCI takes
anyway.
And MUTE still lights the squelch on a real radio. Nothing in this code
can do that — the button sends mute and only mute, and the two are
separate state — so the radio's own announcements are logged as they
arrive. What it says after the command will settle whether this is our
reading or its doing; no more reasoning from here will.
Three things reported from the radio, all of them mine.
THE LEVELS. Two half-width sliders side by side left each of them a couple
of centimetres long, which is not enough to set 15% with. One level per
row now, full width, with the value typed in or stepped with ± — and in
the radio's own units, so the volume is dB and the squelch a dBm
threshold, matching what ExpertSDR3's own window shows.
THE FILTERS. The button said 250 and the radio was set to 300-550: 250 Hz
wide, but sitting where no CW note is. The edges are computed from the
width now, and a narrow filter is CENTRED ON THE CW NOTE — a 250 Hz filter
from 100 to 350 would put the note outside its own passband. A button also
lights on the WIDTH the radio reports rather than on an exact pair of
edges, so moving one edge on the radio no longer darkens every button.
MUTE. Read from one shape only, while this radio reports the other
('mute:0,false'), so the button showed the opposite of the truth. Both are
accepted now.
None of this should have reached main before somebody had a radio in front
of it.
Two faults, reported from a real SunSDR.
DRIVE AND TUNE DRIVE DID NOTHING, and neither did MUTE. Those commands
carry the transceiver index — 'drive:0,15;', not 'drive:15;' — and sent
without it the radio ignores them silently: no error, no answer, the power
unchanged. The rule was in the radio's own reports all along, which is
where it should have been read from: it announces 'drive:0,85' and
'mute:0,false' at connect, while 'mic_level:100' and 'volume:-12' come
with no index at all. Sending the shape the radio speaks in is the whole
rule, and it is now written down next to the two exceptions.
AGC LOOKED STUCK ON SLOW. The panel showed only what the radio reported
back, on the principle that the radio is the truth — but ExpertSDR3 does
not echo every setting it accepts, so a working button sat unlit. Changes
are shown at once and held for a moment now; whatever the radio announces
afterwards still wins, so a clamped or refused setting stays honest
without every working one looking broken.
Also from the same report, and fair: the consoles did not resemble each
other. The Icom panel's RIT control is now a shared component both use —
chip, signed offset, ± keys, wheel, and TYPING a value straight in, which
is the thing a row of ±10/±100 buttons cannot do. Ctrl+←/→ shifts the RIT
here as it does there. And LONG is gone from the AGC row: the protocol
takes it, but it is a hang time nobody reaches for between overs.
TCI pushes everything about the radio unasked, so the console is mostly a
place to put what was already arriving and being logged once as unhandled.
Drive, tune drive, mic gain, TUNE, volume, mute, squelch, NB/NR/ANF/APF,
AGC, passband, RIT and XIT, VFO lock, and an S-meter in real dBm.
Merged with two corrections to the audio page the console's own work
turned up: the radio was being offered as a recording microphone, and the
RX monitor was enabled for a source it cannot play.
Adding it to the device lists put it in all four dropdowns, and an
operator picked it as their RECORDING MIC — which would have recorded the
station they were listening to instead of their own voice, on every voice
message and every QSO recording. It is not a microphone, and it is not a
pair of speakers either: the audio going the other way is transmit audio.
So it is listed for 'From radio' and 'To radio', and taken out of the mic
and the listening dropdowns.
And the RX monitor is disabled when the radio IS the source, with a line
saying why: the TCI stream is pushed into the QSO recorder, and nothing
plays it to the speakers. The button was there, enabled, doing nothing
audible — which is worse than a button that says it cannot.
TCI already carries the frequency, the mode, the meters and now the audio.
It also carries everything else about the radio — and OpsLog was logging
most of it once as '(unhandled once)' and throwing it away. The console is
mostly a place to put what was already arriving.
That makes it the cheapest panel here, and it is worth saying why. A K3
console costs a command and a reply for every value it shows, which is why
it reads its settings in a rotation and its meters only while on screen.
TCI PUSHES: the radio announces its drive, its filters, its noise blanker
and the rest on connect, and again whenever any of them changes —
including when the operator changes them in ExpertSDR3's own window, which
this panel therefore follows without asking anything.
What it drives: drive and tune drive, mic gain, TUNE, volume, mute,
squelch and its threshold, NB, NR, ANF, APF, AGC speed, the passband, RIT
and XIT with their offsets, and the VFO lock. The S-meter is a real dBm
reading, so its S units are arithmetic rather than the calibration guess a
K3's meter needs.
Setters never update the cached state. The radio answers with the new
value, and taking its word is what keeps the panel honest when a setting
is refused, clamped, or changed at the radio a second later — the one
exception being a slider mid-drag, held for 900 ms so it is not dragged
back by its own echo.
Capped width and centred, like the other consoles. Also offered as a
docked pane — and the Elecraft console is offered there too now: App has
always had that pane, Settings simply never listed it.
Reported on an Ultrabeam over USB: the port could not be chosen from the
dropdown, and the panel would not scroll.
That field was a Combobox — a list you can also type into — chosen so the
port of an adapter currently UNPLUGGED could still be configured. It is
now the same Select every other serial device in OpsLog uses, the one that
demonstrably works. The case it was built for is kept differently: a
configured port that is no longer detected is added to the list, so it
stays selected and visible instead of vanishing when the adapter is out.
Also hardened the Test button's guard, which called .trim() straight on
the port and the host. A settings object without them — an older build, a
save that landed short — throws there DURING RENDER, and a panel that
throws while rendering is exactly a panel with a dead dropdown that will
not scroll. Whether or not that is what this operator hit, it is a way to
produce both symptoms at once and it costs nothing to remove.
Reported with a screenshot: the map came up with API KEY REQUIRED written
twice across the middle of it, on the Light and Voyager views only. Those
are exactly the two Carto basemaps.
Nothing here could have caught it. The tiles still answer 200 with a real
image — the watermark is IN the image — so the map looks like it is
working and simply says otherwise in large grey letters.
Both views move to Esri, which is already behind Street and Satellite. That
is the point: a second key-free provider is a second chance to be cut off,
and this is now the third time a tile source has changed its terms under
us. Light is Esri's light canvas with its own label layer, and Voyager
becomes the topographic map, labelled Topo — the stored key stays
'voyager' so nobody who chose it is quietly moved somewhere else.
The canvas runs out of tiles at zoom 16 where the others go to 19, so
maxNativeZoom is set per basemap: Leaflet upscales the last real tile
instead of asking for a level that does not exist, which is the difference
between a slightly soft map and a blank grey square.
Written into the entry itself rather than added as a line of its own:
somebody reading about the feature should learn it there, not two entries
further down.
A fourth amplifier brand alongside the SPE, the Acom and the PowerGenius,
sharing the same multi-amplifier support: the linking, the fan-out and the
polling all work on it unchanged.
The protocol is the Elecraft one — a caret, letters, a semicolon — decoded
from the KPA1500 Programming Reference, with the document's own examples
kept as the test. ^WS gives forward power and SWR in one exchange, ^VI
gives volts in tenths and amps whole, and ^FL is HEX: read as decimal, the
'antenna not connected' fault matches nothing at all.
Three things this amplifier does that the others do not, each handled
where it shows: a KPA500 has no network side, the amplifier answers while
its main supplies are off, and going to OPERATE clears the current fault.
Nothing here has met real hardware yet.
Telling somebody's amplifier where to be is a write to their station, and
it now has a switch — the SAME switch an Acom uses. 'Keep the amplifier on
the radio's band' is one idea to an operator; that it needs a second
serial port and a rig emulator on one amplifier and a single ^BN on
another is our business, not a reason for two checkboxes.
On by default for a KPA, because an amplifier left on the wrong band is
the worse outcome of the two, and the operator who has wired it straight
to the radio turns it off once. The hint says exactly that case, since it
is the only one where a second voice telling the amplifier where to be is
unwelcome.
Two things the wiring got wrong, both visible on screen.
A KPA fell through to the PowerGenius branch of the amplifier card and was
drawn as one — titled 'POWERGENIUSXL · ELECRAFT KPA1500', with a
PowerGenius's meters and none of its own. It has a card now: OPERATE, ON
and OFF, band, power, SWR, temperature, supply, and the fault across the
end. The OPERATE button carries a tooltip saying it also clears the fault,
because that is the button an operator already has under the cursor when
they need it.
And the band-follow option was offered on it. That option exists because
an Acom POLLS a transceiver and has no command to be given a band, so
following it needs a second serial port and a rig emulator answering those
polls. The KPA has ^BN — OpsLog simply tells it, on the link it is already
using, and only when the band CHANGES. The option is gone from the KPA and
the telling is automatic.
Also: Acom rather than ACOM everywhere it is read, which is how the
company writes it. Identifiers, package names and log lines are left
alone — renaming those is churn with no reader.
Wired into the multi-amplifier support that already carries the SPE, the
ACOM and the PowerGenius: a fourth brand rather than a fourth mechanism.
The linking, the fan-out and the poll all work on it unchanged.
Three things the KPA does differently, and each is handled where it shows
rather than explained in a hint:
- A KPA500 has no network side at all. Choosing it puts the entry on
serial, and a configuration still asking for TCP is refused with a
line saying so instead of retrying an address that cannot answer.
- The amplifier answers ^ON while its main supplies are OFF — a sleeping
microcontroller stays awake for exactly that — so power-on stays
available over the network, unlike the SPE and ACOM which need their
serial control lines.
- Going to OPERATE clears the current fault, everything except
temperature. The widget says so under the fault, because the button
that fixes it is the one already on screen.
Defaults that leave a working configuration when the model is changed:
38400 baud, TCP port 1500, and a full-scale power mark of 500 or 1500 W
by model — a KPA500 read against a 1500 W scale looks idle at full
output.
One package for both amplifiers: they share the Elecraft command set — a
caret, letters, a semicolon, case-insensitive in and upper case out — the
same family as the K3/K4 panel. What differs is the transport and which
commands exist, not the grammar.
Everything here comes from the KPA1500 Programming Reference, and the
document's own examples ARE the test:
^WS1204 014; 1204 W and SWR 1.4:1 — power and SWR in one exchange
^VI513 061; 51.3 V and 61 A — volts in tenths, amps whole
^FL91; HEX, and 0x91 is 'antenna not connected?'
That last one is why the parsing is pinned rather than eyeballed: read as
decimal, 90 and 91 become 144 and 145 and match nothing, so an amplifier
shut down by high reflected power would report a fault OpsLog could not
name. SWR in tenths is confirmed by the reference too — 'expressed in
tenths, 123 is 12.3:1' — where it had only been inferred from Hamlib.
The client is question-and-answer under one lock, never two questions in
flight: the reference states there is no flow control and that commands
are paced by waiting for the reply. Fast cycle four times a second for
power, SWR and the fault; the rest once a second.
Faults are named in the operator's terms — 'the ATU found no match', not
'fault 92' — and an unknown code from a newer firmware still says
something rather than nothing.
Not wired to the app yet, and two commands are deliberately absent: ^TX
makes the amplifier transmit from software, and ^ON0 cuts the main
supplies with Wake-on-LAN as the way back. Neither belongs on a poll loop
or behind a button that can be pressed by accident.
The columns took the award's declared band list when it had one, so every
band the award PERMITS got a column whether or not anything had ever been
worked on it. DDFM permits 6m and 70cm; a station that has never worked a
French department on either was shown two empty columns, which reads as a
gap in the log rather than as a band nobody tried.
They now follow the bands this operator actually has contacts on for the
award. The declared list is the fallback for an award with nothing worked
yet — there, saying what the award is played on is the only honest answer
available — and the classic HF set is the fallback for an award that
declares nothing either.
Together with taking the order from the full band list, this is what puts
23cm on DDFM: the column appears because there are contacts on it.
The columns were the intersection of the award's bands with a fixed list
that ended at 70cm, so a band the award genuinely covers could not appear
whatever it said. DDFM is worked well into the microwaves and had no 23cm
column, and nothing anywhere would have produced one.
The intersection is now taken against every band in frequency order, the
same list internal/award already sorts its own band counts by. The classic
HF-plus-2m-and-70cm set stays as the fallback for an award with no band
list and no contacts yet — the full list would be twenty-nine columns of
mostly nothing.
Reported from a station running an IC-7850 over USB: with split engaged,
turning VFO B moved VFO A slightly. OpsLog writes nothing to that radio on
its own — but it shares it, and what it was sharing was wrong.
Hamlib's 'f' means the VFO IN USE. RigState follows ADIF, where FreqHz is
where we TRANSMIT, and the two are the same number until split is
engaged — so the adapter handed FreqHz over unchanged and every client
asking for the dial was given VFO B. A client that reads the dial and
writes it back, which is what WSJT-X and its like do, then wrote VFO B's
frequency into VFO A. Hence a shift the size of the split offset, and
hence 'it did not do this before': it only happens with split on.
get_split_freq ('i') already answers the transmit frequency and is
untouched. The one-line rule is lifted into shareRXFreq with a test that
includes this case, the simplex case, and a backend that reports split
without ever filling in the receive frequency — which would otherwise
answer a client with 0 Hz.
Two faults, and the first was a design mistake of mine.
The choice was written only as the award override, to keep the imported
CNTY as a record of what HAMLOG said. But CNTY is what the comparison
READS, so the contact went on disagreeing for ever: settle a hundred
rows, run the comparison again, get the same hundred back. The reasoning
was about preserving evidence; the effect was a button with no visible
consequence anywhere, which is indistinguishable from one that does
nothing. This is the operator's own log, and correcting a field in it is
the point of the exercise. The district now lands in CNTY as well as in
the award reference.
And the settled rows stayed on screen, which made the same button look
broken a second time. They are dropped as they are applied — locally,
rather than by re-running the comparison, which reads the whole log and
would be seconds of silence on a remote database to be told what is
already known. The agree/disagree counters follow.
Reported by VK4DX with the case that names itself: YB confirmed on 20m
digital, painted blue, because one unconfirmed YB station had also been
worked on that slot.
The ladder ran call_c > call_w > dxcc_c > dxcc_w, so a callsign worked and
not confirmed beat an entity CONFIRMED on the same band and mode. The grid
answers 'what do I still need here', and a confirmed entity needs nothing
whoever was worked afterwards — so the order is now call_c > dxcc_c >
call_w > dxcc_w.
Two things made it easy to get wrong, and both are fixed rather than
merely corrected. The code was a run of assignments where the later test
silently overwrote the earlier, so 'call worked' erased 'entity
confirmed'; it now takes the maximum. And the rule lived inside a scan
loop where nothing could reach it, so it is lifted into bandStatusCode
with a table test — including this exact case, and one that fails if
anybody reorders the constants.
The comparison could show two hundred contacts where the log and the
database name different districts, and leave the operator with nothing to
do about any of them but open each QSO and edit it by hand.
Now each row carries the two values as buttons: click the district you
keep, then apply. There are shortcuts for keeping one source everywhere,
and applying is a separate act — clicking through two hundred rows with
each one written as it is clicked would make a slip permanent before the
reading was finished.
The choice is written as the contact's AWARD REFERENCE, not over the
imported CNTY, and that distinction is the point. CNTY is what the
downloaded log said; overwriting it would destroy the evidence the
comparison runs on, and the next comparison would agree with itself and
prove nothing. The override is the operator's decision, it is what the
award engine counts, and it beats both sources. So the log keeps saying
what it said, the database keeps saying what it says, and the contact
counts for the district that was chosen.
The buttons show the districts themselves rather than a tick box: someone
arbitrating between RO-19 and KO-05 should be picking a value they can
read, not remembering which source a checked box stood for.
v0.26.15 had already shipped, so its entry about setting the source by
hand in ExpertSDR3 goes back exactly as released — it describes the
version operators are running. What replaces it belongs to the version
that carries the change.
The specification settles what a night of experiments could only guess at.
TRX takes an optional THIRD argument naming the signal source — tci, mic1,
mic2, micPC, ecoder2 — and TCI 2.0 says it plainly: 'The signal for
transmitting is always taken from the microphone selected in the
ExpertSDR3. If a third-party software connected via TCI wants to transmit
its audio signal, you must specify the third argument - TCI.'
Without it the radio sends no chrono at all, whatever the mode. That is
what the SSB attempts ran into, and what was misread here as 'digital
modes only' — the mode was never the rule, the missing argument was.
So OpsLog says it, following the 'To radio' device: 'tci' when the voice
keyer owns the transmission, nothing at all otherwise, which leaves the
operator's own microphone alone for every other PTT. Nobody has to find
that setting in ExpertSDR3 and set it again for every mode, which is how
it is remembered there.
Two more things from the same document. A chrono with no audio ready is
now answered with silence rather than left unanswered — the vendor calls
that preferable. And the receive stream declares float32 and two channels
instead of trusting the defaults: they are the documented defaults, but a
default is something another program sharing this radio can have changed,
and a stream in an unexpected format is heard as noise rather than as a
mistake.
The TCI section of the audio settings was an investigation: open the
stream, read what arrives, record ten seconds to listen to, key a tone.
It answered every question it was built for — the frame layout, the
sample width, that the radio asks rather than follows a clock, that the
transmit audio source decides — and confirmed 80 W on real hardware.
None of that belongs in front of an operator now. The radio is simply one
of the devices in the two dropdowns, and choosing it IS the configuration:
one control, in the place where the question is already being asked.
Keeping the tick box beside it would have been two switches for one
decision, with the second one where nobody looks.
Gone with it: the stream/record/probe bindings, the audio.tci_rx setting,
and twenty-four translation keys. The plumbing they proved out stays and
now carries the voice keyer.
The RDA and Elecraft SWR entries move with it: both were written after
v0.26.14 was tagged, so they ship in the version that carries them rather
than in the one already on operators' machines.
A SunSDR carries its audio on the same WebSocket as its commands, so
OpsLog can take it directly: no virtual cable, no second sound card, no
Windows mixer between the recording and the air. The radio appears as a
device in both audio lists and can be chosen for either direction.
Everything here was settled on real hardware over one evening, and none of
it was guessable from the documentation:
- The receive stream answers format=3 for four-byte floats, so the
sample width is derived from the frame rather than trusted from the
field.
- The radio asks for transmit audio only when the transmission is the
CLIENT'S, and only when its transmit audio source is TCI rather than
the microphone.
- The chrono is a REQUEST, not a clock: no payload, carrying the size it
wants, 47 times a second. Audio goes out in answer to it and never on
a timer of our own — a timer was the first attempt and all 234 frames
of it were ignored.
Confirmed: a clean test recording, and 80 W out of a 1 kHz tone.
The radio now appears as a device in both audio lists — 'Radio (TCI
network audio)' — so the receive audio and the voice keyer can both take
the CAT link instead of a sound card. No virtual cable, no second card, no
Windows mixer between the recording and the air.
The transmit side reuses the exchange the tone probe established, with a
WAV in place of the sine: the radio asks, we answer with the next slice,
and it sets the pace. The message is converted once, up front, rather than
per frame — a voice message is a few hundred kilobytes, and resampling
inside a callback that has 21 ms to answer would put arithmetic on the
path where a late frame is a gap on the air.
The PTT is untouched by all this: the keyer keys before and unkeys after
exactly as it does with a sound card, so a transmission is bracketed by
the same code whichever way the audio travels. And the playback runs OFF
the CAT goroutine, since everything else about the rig goes through that
one place and a ten-second message would otherwise freeze the frequency
display and the antenna following for ten seconds.
Two refusals rather than silent failure. A radio that has gone away stops
being offered as a device at all, and a radio whose transmit audio source
is still the microphone is caught within a fifth of a second — a voice
keyer that transmits silence is worse than one that says why it will not.
Elecraft's release note settles it: SW; returns the most recent reading in
transmit or TUNE as three digits in tenths of a ratio — SW023 is 2.3:1,
and SW999 is the 99.9:1 it reports instead of infinity.
This asked for four, so every answer failed to parse and the bar stayed
empty for the whole transmission. A tester saw exactly that: no SWR at
all, which reads as an unsupported radio rather than an off-by-one.
SW also comes out of the probe list. It is known now, and asking again
mid-transmission costs a round trip on the link the carrier depends on.
Transmit over TCI is confirmed on a SunSDR: 80 W out of a 1 kHz tone at
70% of full scale into 80% drive, every request answered.
Which makes the earlier rule wrong. 'Digital modes only' came from a real
observation — SSB silent four times over, DIGU answering at once — but the
mode was a coincidence. ExpertSDR3 has a transmit audio SOURCE, microphone
or TCI, kept per mode, and it was on the microphone in SSB. Refusing SSB
would have blocked the one thing a voice keyer exists for.
So nothing is refused on the strength of the mode. The radio declares what
it wants by asking for audio, 47 times a second when it wants any: key,
wait for one request, and stop within two tenths of a second if none
comes — naming the setting to change rather than theorising about it.
That is better on three counts. It works in SSB when the source is set
right, it cannot be wrong about a mode nobody thought to test (AM, FM,
RTTY), and a misconfiguration costs a quarter-second of carrier instead of
five seconds.
The transmit path works: six passes, the radio asked 231 times and was
answered 231 times, none missed, and the tone was there on the panadapter.
What was missing was power on the meter, and the cause was the level. The
tone went out at a quarter of full scale, out of caution, into a radio set
to 15% drive — enough to draw a clean signal and not enough to move a
needle. In a digital mode the radio expects a line level it can drive to
full output; the POWER is its own drive control, so sending quietly only
wastes the range. Now 0.7, short of 1.0 to leave room for the peaks.
The start line carries both numbers — ours and the radio's drive — because
a quiet transmission has two possible causes and one line should settle
which, rather than an evening of guessing. Reading 'drive' off the radio
is the only reason it is parsed at all.
Three transmissions on a real SunSDR settled how the transmit side works,
and none of it was guessable from the documentation.
The radio asks for audio only when the transmission is the CLIENT'S: with
the operator keying the microphone it sent 282 receive frames and nothing
else. And it asks only in a DIGITAL mode — keyed from here in SSB it stayed
silent four times over, and answered in DIGU immediately. In SSB the
modulator is wired to the microphone, which is also the honest answer to
'why can I hear myself but not the tone'.
The chrono turns out to be a REQUEST, not a clock. It carries no payload —
the message itself is the ask — and it names the size it wants in the
header: 2048 samples, two channels interleaved, 47 times a second, which
is 1024 pairs at 48 kHz, exactly real time.
So audio goes out in answer to a request and never on a timer of our own.
The timer was the first attempt and the radio ignored all 234 frames of
it. Answering also hands the pacing to the radio: no drift, no buffer to
tune, and the size taken from what it asked for rather than from what we
assumed. The sine keeps its phase across frames, since one restarted every
frame is a click 47 times a second.
A pass in SSB is now refused rather than attempted. It keys the
transmitter, produces nothing and teaches nobody anything — and it is
still a transmission.
The feed mechanism is the one the voice keyer will use: WAV samples in
place of the sine, everything else unchanged.
A first real transmission settled one question and raised a better one.
With the receive stream open and six seconds of transmit, the radio sent
282 frames of receive audio and NOTHING else: no chrono, no transmit
audio. So the chrono the documentation describes is not offered to a
client that merely happens to be connected while the operator keys the
microphone, and waiting for it to appear is waiting for nothing.
The reading that fits is that the radio asks for audio when the
transmission is the CLIENT'S and takes the microphone when it is the
operator's — which makes the experiment obvious. Key it from here, push a
1 kHz tone, and watch. Chrono frames appearing gives their size and
cadence by measurement instead of by guesswork; no chrono but a tone on
the meter is just as useful, because then the pacing is optional and the
voice keyer can push frames at the rate the stream already runs at.
A tone rather than silence so the answer shows on the power meter and not
only in the log.
It transmits, so: an explicit button inside a warning box, five seconds,
capped at ten, and every path out unkeys — including the panic that has
not happened yet and a socket that dies mid-tone. A transmitter left keyed
by a defect is the one fault here that would reach somebody else's band.
Writes are now serialised too. send() held the lock only long enough to
read the connection, which was enough while every command came from the
poll loop; a stream of audio frames from a second goroutine is not, and
gorilla panics on a concurrent write rather than failing quietly.
The first transmit test came back with a log that said nothing, which is
the one answer that cannot be read: either no transmit frames arrived, or
they arrived and went unlogged.
So each pass is now bounded by a line of its own, and every stream type is
counted without limit. A pass that reports 'receive audio: 240, and
nothing else' is a result — it says the radio sends no chrono unless
something more is asked of it — where a log with no transmit lines was
merely a silence. The forty-frame logging budget is also handed back to
the transmit types on each pass, since it was always spent on receive
audio long before anyone got round to keying.
The start line says whether the receive stream is even open, because a
radio with nothing streaming has no reason to send chrono, and that is the
likeliest reason the first attempt saw nothing.
Two faults, one behind the other.
RDAPanel is nested inside SettingsModal so it can read its state, and it
was written as <RDAPanel />. A nested function is a new component TYPE on
every render: React cannot know it is the same panel, so it unmounted the
tree and mounted a fresh one — and a fresh scroll container starts at the
top. It is called now, like the other panels, which produces the same
elements in place and never disturbs the scroll position.
What it was reacting to should not have reached it either. The PSK
Reporter and grid-cache poll ran every three seconds from wherever you
were in Preferences, re-rendering the whole dialog for a count shown only
on the cluster section. It now runs while that section is open, and the
one-time loads it was sharing an effect with have stayed where they were
rather than refetching on every click in the sidebar.
The visible result: a hundred contacts to correct can be read from top to
bottom.
rouggy