RI1FJL/MM resolved to Franz Josef Land while the expedition was still sailing
there, telling the operator they had worked an entity they had not.
The old behaviour was deliberate — the comment read "strict DXCC says no
entity, but the log should still show the operator's country" — and that is
wrong for exactly this reason. A home country on a maritime mobile is not extra
information; it is a false claim about where the contact happened, and it is
the kind of false claim that gets a QSO submitted for an award it cannot win.
Trailing only. A LEADING "MM" is the Scotland prefix and "AM" is Spain, so
MM0ABC and AM5X keep their entities — getting that wrong would be a far larger
error than the one being fixed. /P, /M and /B are unaffected: a portable
station is still ashore.
In the cluster the column says "Maritime mobile — no DXCC" rather than going
blank, because blank is what an UNRESOLVED spot looks like and the two mean
opposite things — one is "we don't know yet", this is "there is nothing to
know". Display only: the QSO's own Country stays empty, since that field holds
a DXCC entity name and there is none.
ADIF defines CNTY as "STATE,COUNTY" — "GA,BARROW". OpsLog wrote the bare county
name, which a receiving logger cannot resolve: county names repeat across
states, and there is a Washington County in thirty of them. The award engine
here was never affected, since it reads the columns rather than the ADIF; the
damage was to every file we hand to someone else.
One writer covers everything — writeRecord — so this fixes file exports and the
LoTW, Club Log, HRDLog and QRZ uploads together. MY_CNTY gets the same
treatment.
The county alone is still written when no state is known: a bare name is worth
more than nothing, and inventing a prefix would be worse than either. A value
that already carries a comma passes through untouched, so re-exporting an
imported record cannot double the prefix.
Import is the mirror: "GA,BARROW" fills both columns, and a file carrying the
bare county — as OpsLog's own older exports do — still imports unchanged. The
state parsed out of CNTY only fills a blank STATE, never overrides it: STATE is
the dedicated field and the more specific statement.
"An existing connection was forcibly closed by the remote host" during DATA
reads the same whether the server refused the SIZE, hit an hourly quota, or
simply dropped the socket — and only the first is something an operator can act
on. The message now carries host:port, the encryption in use and the total
attachment size, so the size can be ruled in or out without a second attempt.
A path that does not exist is not counted: the mail library skips it, so
reporting it would describe a message that was never sent.
Puts "F4BPO is on air 14,074,000 FT8 IC-7300" on hrdlog.net's front page while
the station is operating, which is what HRDLog's own "Automatic HRDLog ON AIR"
setting does.
The endpoint and its field names come from HRDLog's own library
(github.com/iw1qlh/HRDLOG-net-library, HrdProtocol.SendOnAirAsync): OnAir.aspx,
with Frequency in Hz, Mode, Radio, Callsign, Code and App. Read rather than
guessed — an invented endpoint would have produced a switch that silently did
nothing.
HTTPS where that library uses plain HTTP: the upload code is a credential and
has no business crossing the network in clear, and the sibling NewEntry
endpoint on the same host already serves TLS.
A heartbeat every two minutes, not a change notification — HRDLog drops a
station from the list when it stops hearing from it. Silent when switched off,
unconfigured, or when the rig reports no frequency: announcing 0 Hz would say
the operator is on air on nothing. Failures are logged, never toasted; a
broadcast the operator did not ask for at that instant must not interrupt them
every two minutes on a flaky link.
Its own switch beside auto-upload rather than folded into it: this is a live
status and needs the rig readable, not a QSO.
compactH was 158px, "tuned so the compact entry strip fits in a single row".
A constant tuned against a layout stops being true the moment the layout
changes, and this one outlived a strip that had since shrunk — leaving about
70px of empty window under the fields.
The frontend now measures what it rendered and asks for that height, watched by
a ResizeObserver so a strip that wraps at a narrow width is followed too. The
topbar is added as its declared h-8 rather than measured, since it is fixed.
Bounded in the backend: a measurement of zero — a layout not yet painted —
must not collapse the window, and the call is ignored unless compact is on so
nothing can shrink the normal window. Rounded to whole pixels so a sub-pixel
reflow cannot start a resize loop.
It went under the Station Control rotator; the compass that wanted it is the
docked one beside the entry strip. Both are the same RotorCompass, so the
readout moves INTO that component: every compass in the app carries it, and no
caller draws its own.
The bezel already showed the short path as a red marker, but a marker is a
direction, not a number — and the numbers in the status bar are 10px, which is
what prompted this.
Clickable only when the caller passed onGoto, since a compass rendered without
one cannot turn anything and a button that does nothing is worse than a label.
The same pair already sits in the status bar, where it is 10px and several
operators reported not being able to read it. Repeated under the compass at a
size that can be read across a shack.
Clickable, like the ones in the status bar: a heading you can see and not act
on would be a step backwards from what is already there. Disabled and dashed
when no station is entered, rather than showing a heading to nowhere.
The same answer the Details tab already gave, moved to where the county is now
typed — a badge in a tab nobody has open while working a station is a badge
that does not do its job.
Debounced at 300 ms and keyed on state + county together: the field is typed
into as well as filled by the lookup, and the check queries the log. A county
name means nothing without its state, so both have to settle before asking.
The looked-up values now sit directly under the geography line they continue —
QTH, State, locator, then county and zones — and the free-text fields close the
block. Reading order follows what the lookup fills, rather than interleaving it
with what the operator types.
Comment and Note moving onto one line left the right column two rows shorter
than the left, and this fills it: county, both zones and the ADIF entity
number. All four are filled by the lookup and were only visible after opening a
tab — the wrong place for values an operator wants to SEE are right before
pressing Log.
County is a name, not a code ("St. Tammany Parish"), so it takes the flexible
width and the three numbers get fixed narrow boxes.
Cleared numeric fields store undefined, not 0. Zero is a real DXCC number and a
real zone, so an empty box has to stay empty rather than assert something the
operator did not type.
State stays up on the geography line beside QTH and the locator, which is how
the lookup fills them and how they are read.
stateBlock was built and then not placed in the row — the field simply did not
exist on screen. Now between QTH and the locator, where it was meant to go.
Left column 210 → 230px, and Comment and Note move onto ONE line side by side:
two fields that are usually a few words each were taking two full rows of the
entry strip. That frees a row for whatever goes there next.
Name, Band, Mode and Country go from 300px to 210px, which moves QTH, Comment
and Note left by the same amount. The width is set in TWO places — the Name row
and the Band/Mode/Country stack below it — and they have to stay equal or the
two rows stop lining up, so both were changed together.
State goes between QTH and the locator: it is part of the same "where is this
station" reading, and it is what WAS and the US county work key off. Narrow,
since it holds two or three characters, and upper-cased on the way in like the
other coded fields.
The Name label in this layout was hardcoded English while its twin went through
t(); fixed while here.
A log greets an operator; it does not address an envelope. "Robert Smith"
filled the Name field with something no one would ever send on the air, and it
is the first name that gets used when the contact is answered.
So fname alone, with the surname kept only as a last resort for a record that
has no first name at all — better a surname than an empty field. The nickname
option now falls back to the first name too, which is what it should have done
from the start.
joinName went with it: nothing composed a name any more.
QRZ publishes <nickname> — the name an operator goes BY on the air — and
OpsLog was composing fname + name instead. "Bob" is what belongs in a log;
"Robert J Smith" is what belongs on a licence.
QRZ only, and deliberately so: HamQTH's <nick> already fills the Name field
that way, so the same switch there would toggle a behaviour it has no way to
turn off.
A published nickname is optional, so an empty one falls through to the
registered name. That is the whole point of it being a fallback rather than a
swap, and it is what the test pins — a blank nickname must never blank the
name.
Spots were bounded by COUNT alone, so on a quiet band a two-hour-old spot sat
on the band map looking like something to chase. Settings → DX Cluster now
takes a lifetime: presets at 5/10/15/30/60 minutes, or any value up to twelve
hours, and 0 keeps the old behaviour.
Spots are REMOVED from the shared list rather than hidden at render. The
cluster list, every band map and the counts all read that one array, so pruning
it once is what makes the setting mean the same thing in every view — the
lesson already paid for when the band map ignored the cluster's filters. It
also gives the memory back.
A spot whose timestamp will not parse is never dropped: an unreadable stamp is
a reason to distrust the clock, not to throw away the spot.
Swept every 30 seconds — close enough that a 5-minute setting is honoured, and
invisible next to the spot stream. Clamped in the backend as well as the UI.
Reported on BH2SWB and 4X9AA: both flagged NEW PFX, neither ever worked, both
vanished the moment "hide worked" went on.
The predicate hid a spot when worked_call OR status === 'worked' — and that
second one is the ENTITY's status, not the callsign's. So any spot in an entity
worked years ago disappeared, however new it was for something else.
The extra markers are orthogonal to the entity status; the band map already
says so in as many words. They now win: a new prefix, county, grid or park is
never hidden, whatever the entity says. A worked entity that is new for nothing
is still hidden, which is what the filter is for.
Extracted to spotIsWorked in lib/spotDisplay, beside the other rule the cluster
and band map share, rather than left inline in a predicate that has already
grown once.
This repo has no frontend test runner, so this is verified by build and reading
rather than by a test.
Replaces the four fixed rules with the model Logger32 uses and the operator
asked for: Worked / Confirmed / QSL sent / To be sent, and inside each, which
channels count — paper QSL, LoTW, eQSL, QRZ.com. No ticks means every channel,
because a rule the operator has not narrowed must not quietly become a rule
about nothing.
"To be sent" is FIRST in the order, and that is the substantive decision here.
A contact can be confirmed on LoTW and still owe a paper card; the colour an
operator scans for is the one meaning "something is still owed". Placed after
"confirmed" that row goes green and the card never gets printed.
"Worked" is the catch-all — nothing sent, nothing requested, nothing back — and
is off by default, since turning it on paints every remaining row.
R and Q both count as owed: ADIF says requested and queued, and both mean the
card has not gone out.
"The ON button does nothing" has three different causes — the modem-line pulse
was refused, the pulse went out and the amplifier ignored it, or the amplifier
woke and the UI missed it — and the log could not separate them: it recorded
one combined error and never said whether the amp came up.
RTS and DTR are now reported separately (not every USB-serial chip honours the
modem lines, and some refuse them without saying so), alongside the model and
transport. A watcher then logs whether the amplifier answered within 15
seconds, off the caller's goroutine so the click still returns at once.
No change to the wake sequence itself, which is the part that took a long time
to get right on the 1.3K-FA.
"Was the grid in what WSJT-X sent, or did OpsLog put it there?" is the first
question asked when a contact is logged in the wrong square, and a byte count
cannot answer it. An expedition logged at its licence holder's home QTH stayed
undiagnosable for exactly that reason.
One line per QSO, around 280 bytes, whitespace collapsed so the record sits on
a single line.
Grid: RI0FA transmitted QN35 all evening and was logged with its licence
holder's home square. WSJT-X's logged ADIF usually carries no GRIDSQUARE, so
the lookup was the only source left — and QRZ/HamQTH describe where an operator
LIVES, which for an expedition is the wrong side of the planet.
The square heard on the air is now applied before the lookup runs, so
refineGrid has something to defend. It still upgrades a 4-character square to a
6-character one from the same field, and still refuses a finer square from a
different field.
SPE: decodeCSV marks the client connected on every frame it parses, and the
poll goroutine can be mid-read while PowerOff runs. The frame from a second ago
landed after PowerOff had marked the amp offline and put it straight back,
which is why OFF had to be pressed twice.
Frames are ignored for three seconds after the off key goes out — the window is
opened before the key is sent, so nothing already travelling can beat it. Time
bounded rather than latched: an amp switched back on at its own front panel has
to reappear without being told.
Reported: ten fresh contacts upload and appear on LoTW; twenty older ones,
bulk-edited to exactly the same LOTW_SENT/RCVD state, are accepted by OpsLog
and never arrive. The operator's reading was that a failed attempt could not be
overridden. It is simpler and worse than that.
TQSL's exit codes, from its own cmdline documentation:
8 NO QSOs were processed — already uploaded OR OUT OF DATE RANGE
9 some processed, some ignored — same two reasons
14 some already uploaded, the rest signed
8 and 9 were both read as plain success. So on 8 — nothing uploaded at all —
OpsLog announced "already uploaded (duplicate)" and stamped every selected
contact as sent. They were never on LoTW and now looked as if they were, which
is exactly the reported symptom, and it is not recoverable by re-uploading
because the operator has no reason to try.
"Out of date range" is the cause that bites here: a contact older than the
callsign certificate's validity is silently left out. Older QSOs failing while
today's succeed is the signature.
Now: 8 is a failure and nothing is stamped — a duplicate left at "R" is
harmless and will be refused again, whereas a contact wrongly marked sent is
one nobody will look at twice. 9 and 14 succeed but carry Ignored, and the
caller says so in the console and a toast.
TQSL's own sentence ("20 QSO records are out of date range") is captured and
shown. It was being read and thrown away, and it is the whole answer to "why is
my contact not on LoTW".
Two BandMap instances, and cleaning up a duplicated attribute took the prop off
the docked one as well as the copy. The multi-band tab drew the badge; the map
beside the entry form — the one actually in front of the operator — did not.
Logger32 draws a green square for a LoTW user and an X for a worked one. The
square is the part worth copying; the way it is drawn is not.
Colour on this map is already spoken for — it carries the entity status, and
the left edge of each pill stacks the new-park / new-county markers. Adding a
green fill for LoTW would put two unrelated meanings on one channel, and an
operator would have to work out which green meant what. So it is the same "L"
badge the cluster list draws, in the same muted blue, for the reason written
there: whether a station uploads to LoTW says nothing about whether the spot is
worth chasing.
Worked stations already read as worked here — the pill goes grey and the
status markers speak — so no X is needed.
Switchable in Settings → Appearance, and on by default: a `configured` flag
distinguishes "saved with this off" from "saved before the option existed", so
turning it off sticks instead of being undone by the next default.
The band map received the raw spot stream filtered by band alone, so none of
the cluster's filters reached it: switching on "LoTW users only" changed the
list and left the panadapter showing everyone. The same was true of hide
worked, the spotter continent, and the status and mode chips.
The predicate is extracted and shared rather than copied — two views of one
spot stream disagreeing about the same spot is the fault lib/spotDisplay
already exists to prevent, and a second copy would have drifted the first time
a filter was added.
The BAND filter is deliberately excluded from the shared part: a band map's
band is its filter, and applying the cluster's would empty every map but one.
The first version filled the whole row at 24%, and in a real log that meant
every row was painted: nearly every contact has SOME QSL state, so colour was
present everywhere and stopped being information — a striped background with
the data behind it.
The default is now a 3px stripe down the left edge. Same signal, nothing lost
to read it. A filled row is still offered, with a strength slider, for
operators who want the block — and the rule cards in Settings preview whichever
is chosen, so the decision is made by looking rather than by imagining.
Drawn as an inset shadow rather than a border: a border would shift the cells
three pixels on coloured rows only, and the columns would stop lining up.
Style and strength are normalised rather than rejected — a value out of range
is a slider that got away, not a reason to reset the operator's colours.
The "to send" rule only looked at qsl_sent, so a contact waiting to go to LoTW
— marked R, and the commonest not-gone-out-yet state in a digital log — matched
nothing and stayed uncoloured. It now covers any route still queued, and the
label says so.
New Settings → Appearance section: whole-row colouring in the log grid driven
by QSL / LoTW state, each rule with its own colour from a palette or a free
picker. What Logger32 does, with one difference that matters — the colour is
applied as a 24% tint, not a fill. Logger32's grid is white; this one is dark,
and a saturated user-picked colour behind white text is unreadable at exactly
the moment the operator is scanning for what still needs sending.
Rules are ORDERED and the first match wins, because a contact is usually
several of these at once: one confirmed on LoTW and by card is confirmed, not
"sent, awaiting reply". The order lives in the data so the panel can show the
rules in the order they actually apply, numbered.
The colour is interpolated into a CSS color-mix(), so anything that is not
plainly #rrggbb is refused on the way in and falls back to the default.
Also fixes the QSO number column, which was empty for contacts logged from
WSJT-X: that path inserts through the repo directly and never reached AddQSO.
Rather than chase each of the remaining bulk-insert paths — the POTA hunter
import and the LoTW/QRZ "add what I was missing" passes, which insert OLD
dates and so shift every number after them — the index now checks its length
against a COUNT and rebuilds when they disagree. One indexed count beats
remembering to invalidate in a place that does not exist yet.
OrderedIDs scanned qso_date straight into a time.Time. The column holds a
formatted STRING — the repo writes it with Format(isoMillis) and reads it back
through parseTimeLoose everywhere else — so the scan failed on every call, the
index was never built, and omitempty then dropped the zero from the JSON. A
column that shipped blank.
Scans the string and parses only the newest row; the ordering already came from
SQL, so 30 000 parses were never needed.
Tested against a real SQLite file with the date stored exactly as the repo
writes it, and with the ids running opposite to the dates — the imported-ADIF
case that is the whole reason this is not the id. It fails without the fix with
the scan error itself.
Not the id: the primary key follows insertion order, so importing an old ADIF
gives the oldest contacts the highest ids. This is a rank over qso_date.
Computed over the WHOLE log, not the query result — ranking inside the result
would renumber every contact the moment a filter is applied, and QSO #1 would
change identity as the operator typed.
Held as one id-to-rank map, built from a single ordered id query and dropped
with the other derived indexes when the log changes. A contact logged from the
entry form is the newest, so it takes the next number without rereading the
log: AddQSO deliberately avoids full invalidation because a contest run would
pay for it once per QSO. A contact entered with an OLDER date belongs in the
middle of the order, so there the map is dropped and rebuilt rather than
mis-numbered.
A batch that never reaches the disk looks exactly like one that does: the
locators are in memory either way until the next restart, which is the one
moment the difference shows. The shutdown flush was the least observable of
all — it runs last, and nothing said whether it had written anything.
Audit of every field a lookup returns against what OpsLog can store, prompted
by a report that HamQTH was not fetching the email.
The email was never broken: pinned now against a captured HamQTH answer, it
parses and reaches the QSO. The reported callsign simply has no public address
on HamQTH, which the empty field could not distinguish from a fault — so the
"[email protected]" placeholder is gone. A greyed-out sample address in the one
field an operator checks to see whether the lookup found one reads as a value.
Real gaps found and closed:
- web: the qso table has had the column since migration 0003 and no provider
mapping ever read it. HamQTH sends <web>.
- picture: Result.ImageURL was documented "QRZ only" because HamQTH's element
is <picture>, not <image>. It was there all along.
- zip: sent by both providers, read by neither.
- adr_name: used only as a fallback when a record carries neither nick nor
name. Deliberately NOT preferred over <nick> — a log wants the name the
operator goes by on the air, "Igor", not "Igor Vladimirovich Getman".
The parse tests use a real captured payload, including the stray <div> advert
the server injects into its own XML.
The chips in Settings only ever shaped the PSK Reporter SUBSCRIPTION. Neither
feed path checked them: both asked bandopen.Watched, which says which bands the
detector is capable of and knows nothing about the selection. So the cluster
path announced every watched band regardless, and widening the subscription to
"+" for grid chasing let PSK Reporter do the same — a station with only 6 m
ticked got 10 m and 2 m badges.
The selection is now cached beside the on/off flag and checked on both paths,
and unticking a band puts its badge out: badges fade on a timer fed by spots
the detector no longer looks at, so it would otherwise stay lit until a
restart.
The store shipped without its main source. Locators came only from this
station's own WSJT-X decodes, which is what the whole MQTT discussion was
about.
PSK Reporter now feeds it through a new OnGrid callback, fired before any
geographic filtering: what the store wants is "which square is this callsign
in", and that is true whoever happened to hear the report.
The subscription filters on the RECEIVER's square, a level the v2 topic
exposes. Measured on the live feed: the four opening bands unfiltered are 83
messages a second, of which roughly one in a hundred survived the NearKm test
that already existed here — the rest was received, TLS-decrypted, JSON-parsed
and discarded. One ring of squares is 0.2 to 1.2 a second.
By square rather than by DXCC, which was the obvious alternative: one country
measured 1.2 messages a second (OH) against 72.5 (K) on a single band, because
a DXCC can be a continent. By square the same measurement is 0.2 to 1.2, so the
load follows distance — what the feed is actually about — and is the same for
every operator.
Grid chasing subscribes with the "+" band wildcard, so one subscription per
square covers every band instead of one per band per square.
The store gains a source column (decode | mqtt), migrated in place on an
existing file.
The callsign->grid map died with the process. Every restart began with an empty
Locator column that took an hour of listening to refill, and everything learnt
the day before was thrown away.
internal/gridcache is its own SQLite file in the data directory, not a table in
the settings database: that one sits wherever the operator chose to put it,
often a synchronised folder, and a store that rewrites itself every minute has
no business there. Deleting the file costs a few days of listening and nothing
else.
Callsign is the primary key and the newest report wins — operators move, go
portable, go on expedition, and a stale locator is worse than none for grid
chasing because it reads as a square already worked. Entries not seen in two
years are pruned at open: that is the one way this cache can be actively wrong
rather than merely empty, and it is what bounds a store that would otherwise
only grow.
Only CHANGES are queued. The feeds repeat themselves, so writing every report
would put the whole stream in the batch instead of the news in it. Batches
flush on a timer and on shutdown — a restart is exactly when the cache is worth
the most.
Rotation is switched off while the store is attached: the cap would discard
callsigns the database still holds and the lookup would then miss what we know.
Age in the store is the bound instead.
The cache dropped EVERY entry once it passed 20 000. That was survivable while
the only feed was this station's own WSJT-X decodes, which never reached the
ceiling — it is a cliff for anything larger, and every locator in the cluster
list would disappear at once, periodically, for no reason the operator could
see.
Two generations: when the current map fills it becomes the previous one and a
fresh map takes over; lookups consult both. A rotation therefore costs the
older half and nothing more. It needs no insertion order, no per-entry
timestamp and no bookkeeping on the write path — all of which "evict the
oldest thousand" would require, on a path that runs once per decode.
The cap is 100 000, measured at 82 bytes an entry: 8 MB a generation, 16 MB
for both. 20 000 was chosen when the ceiling was unreachable anyway.
Both maps now go through rememberDecodeGrid / lookupDecodeGrid. Rotation swaps
the map headers, so the read had to be guarded; routing every access through
one accessor is what makes that checkable rather than remembered.
Entries had grown to 250-450 characters, carrying the symptom, the mechanism
and the reassurance about existing setups. The panel is read top to bottom;
that belongs in the commit message, which has no length limit and is not
competing for the reader's attention.
"Watch for band openings" only ever governed the extra DATA SOURCES — the RBN
nodes and the PSK Reporter feed. The detector itself ran on every ordinary
cluster spot regardless, so an operator who had never enabled the watch still
got 10 m and 6 m opening banners from a feature he had deliberately left off.
The flag is cached on bandOpenState rather than read per spot: this is the
cluster hot path, where a settings query per spot is exactly what the rest of
this file avoids. startBandOpenFeed owns it, and it already runs at startup and
on every save, so the switch takes effect without a restart.
Switching it off also clears the live badges and the accumulated spot window.
The badges only fade on a timer fed by spots the detector no longer looks at,
so they would otherwise hang there until the next restart; and dropping the
window means switching back on starts from what is on the air rather than from
an hour-old burst. The remembered openings are kept — those really happened.
Hidden from the filter panel and forced off, without touching the machinery
behind them — it is correct and took several rounds to get right, so it stays
whole.
One flag in lib/spotDisplay does both jobs: readSpotDisplayOptions answers
false while it is down, which covers the band map, and App.tsx seeds its two
states from that same call instead of reading localStorage directly. Without
that second half an operator who already had "No colour on worked" saved would
have kept it applied with no switch left to turn it off.
The saved preferences are deliberately left in localStorage rather than
cleared, so whoever had either option on gets their choice back the day the
switches return instead of silently starting from off.
Two dates side by side put each one in a half-width Field2, which spends 120px
on its label column — the second label overlapped the first input and the row
ran off the panel. The comment ten lines above says exactly this about the
group/subgroup fields; one per row, as everything else in this editor.
Also shows a far-future end date as open-ended. RDA carries 9999-12-31 as "no
end" and printing it back reads like a real deadline.
The claim that an empty per-reference window inherits the award's is now a
test rather than a comment: a QSO before the award's own ValidFrom does not
count for a reference that has no window of its own, and a narrower reference
window still applies on top.
References are not forever: a park is delisted, a district merged, a castle
loses its number. A QSO made while the reference existed is a valid contact
and must keep counting; one made afterwards must not.
The JSON format did not need to change — ValidFrom/ValidTo were already on
awardref.Ref, already columns in award_references, already round-tripping
through export and import. They were simply never read: awardRefMetas dropped
them on the way into the engine, so nothing downstream could enforce them.
This carries them through and checks them in keepRefs.
Empty means the award's own window governs, which inScope already enforces for
every QSO in the award. That fallback is deliberately NOT duplicated per
reference — two places enforcing the same dates is two places for them to
disagree. The editor shows the award's dates as the hint under the boxes so
the operator can see what empty inherits.
Dates are compared as ISO strings rather than parsed times: the stored shape
is "2006-01-02", lexical order on it IS chronological, and this cannot fail on
a malformed value the way a parse can — a reference with a typo in its window
keeps counting instead of silently vanishing from an operator's totals.
The Explain trace names the date and the cutoff, because "did not count" on a
contact the operator remembers making is exactly when they need to be told it
is the reference that has a window, not their log that is wrong.
Only the whole-catalogue bundle was reachable from the UI, so sharing one
award meant handing over every award you have — including the ones you edited
for yourself.
Nothing new was needed underneath: ExportAward(code) already existed and
already reads the definition AND its references from the database rather than
from the embedded catalog, which is the point (a WAPC is only worth sharing
because of the province list and city regexes the operator added). The English
and French strings existed too. Only the button was missing.
Kept distinct from the catalog publish next to it: that one stamps a version
and clears user_edited to seed a release, this one is a plain share and leaves
both alone, so the recipient's copy is correctly marked as someone else's work
instead of a pristine built-in that future catalog updates would overwrite.
The band buttons in Station Control tuned to a fixed mid-band frequency baked
into the frontend. An operator who lives in the CW segment, or in the FT8
window, got the antenna resonant somewhere he never operates and had to nudge
it every time. The SteppIR's own controller has a Frequency (KHz) column per
band for exactly this; this is that column.
Stored sparsely: a band with no entry uses its default, so nothing migrates
and an operator sets only the bands he cares about. Left empty the Settings
box shows the default as its placeholder, which makes clearing it the obvious
way back.
The value is checked against the band plan before it is kept, because it goes
to the antenna as a tune command — a lost digit (1450 for 20 m) or kHz typed
as MHz would send the elements travelling to a length wrong for the band the
operator is on, and on a SteppIR that journey inhibits transmit the whole way.
A rejected entry is logged and the band falls back to its default.
Resolution happens in the backend and rides on the existing status poll, so
the widget no longer decides where a band button goes and cannot drift from
what Settings shows. Its own table stays only as a floor for the first poll.
The follow loop only ever had one behaviour — re-tune once the rig moved
further than a fixed step. The SteppIR's own controller software offers three,
and operators arrive with that mental model: every frequency change, past a
threshold, or only on a band change. They are real operating trade-offs, not
preferences: "always" keeps resonance perfect at the cost of motors running
constantly (and on a SteppIR every move inhibits transmit while the elements
travel), "band" moves them a handful of times a day.
"Always" is implemented as the threshold mode with a 1 kHz threshold rather
than as a separate branch, so all modes keep the one deadband reference that
matters: the rig frequency last commanded for, NOT the antenna's own reported
frequency — a SteppIR flips its reported frequency between the commanded value
and its home value, which would re-issue a SET on nearly every poll and leave
the operator permanently unable to transmit.
Band mode compares against the band last COMMANDED for, not the rig's previous
band, so the first move after startup and any move made by hand still get
reconciled. It applies to the immediate spot-click path too: a click inside the
band the antenna is already resonant in moves nothing.
An unset or unrecognised mode resolves to the step mode, so every config
written before this option behaves exactly as it did.
Also routes the antenna settings block through t() — it was hardcoded English.
The card is two grid columns wide and sits beside a tall neighbour in Station
Control, so a single row of four bars left the height unused and squeezed each
bar into a quarter of the width. Two rows of two fill the room already there
and double the resolution of every bar.
The PowerGenius XL card gated its whole meter row on flex.amp_available, so
on any station without a FlexRadio it rendered OPERATE and the fan selector
over an otherwise empty two-column card — reported from a TS-590 station.
Nothing was missing from the backend. internal/powergenius already parses
forward power, drain current, VSWR and temperature out of the GSCP status
frame, and the docked AmpWidget already prefers the radio's meter stream and
falls back to those. Only the full card never learnt the fallback.
Same source preference as the widget, and the same two exclusions: peakfwd
and peakid are latched maxima that are never reset and survive in the
last-known status after the amp disconnects, so the plain readings are used
and gated on the transmit state instead of freezing on the last over.
WSJT-X "Fake It" shifts the dial for the duration of each over and puts it
back afterwards, but the restore is conditional: it reads the frequency back
and only moves the dial if it disagrees with where it believes the radio
should be.
That read lands inside the window where this backend stops polling on
purpose. A Kenwood answers "?;" to IF; while it is transmitting, and treating
that as a fault used to drop the shared CAT link entirely, so the cached
state answers instead. SetFrequency wrote the command to the rig without
touching that cache — so mid-over the cache still described the pre-over
dial, WSJT-X read its own receive frequency back, concluded there was nothing
to restore, and the radio stayed on the transmit frequency. Every later over
started from there.
Reported from a session where the dial stuck at 7075500 after a full FT8
over while a bare TUNE, which never sets a frequency, worked fine.
Only simplex updates the cache. Under split, FreqHz is the transmit frequency
while the write lands on whichever VFO the operator is on, and guessing which
side moved would put a wrong number in front of the operator — a stale one
survives until the next poll.
The test reproduces the reported sequence and fails without the fix with the
same frequency the log shows.
A 2 m opening was announced at 9650 km towards Japan. The callsigns gave it
away — 7M4RRM, JA7RPC, JF1AWC — all working EME, which is routine on 2 m and
reported to PSK Reporter like anything else. Real contacts, real grids, and no
evidence whatsoever about the band: an operator turning a beam on that bearing
would hear nothing.
Removing the flat 2400 km ceiling was right; it threw away genuine multi-hop Es
on 6 m. "No limit anywhere" was the overcorrection. The limit is per band now,
and it is physics rather than a threshold: 2 m reaches a few thousand kilometres
by tropospheric duct or a chain of Es clouds and no further, so 3500 km, and
4 m 4000. Six and ten metres keep no ceiling — multi-hop Es and F2 genuinely do
go round the world, which is the case that started all this.
Pinned from both sides: the 9650 km Japanese burst produces nothing, and a
2000 km 2 m burst in one sector still counts.
The UDP path applies the station profile, the DXCC number, the Club Log
exceptions, the zone refinement and the QSL defaults — and its comment says
"same as the manual AddQSO path", which it was not. applySolar and fillDistance
were never called there.
For an operator running digital, and that is most of the traffic on most
stations, it meant SFI, A, K and distance were empty across the whole log while
a hand-logged contact carried all four. It also quietly undermines the
history-based propagation work, which needs those numbers to be there.
applySolar now refuses a QSO more than a day old, whichever path it arrives by.
The UDP feed and the ADIF monitor normally carry contacts seconds old but
neither promises it — a logger re-broadcasting its backlog, or an operator
typing last month's contact by hand, would be handed this morning's SFI as
though it had been measured at the time. A wrong reading is worse than a missing
one: afterwards nothing tells it apart from a real one.
Width should follow use, and these two are nowhere near equal: a manager's
callsign is typed constantly, a QSL message almost never. The message held 7
columns of 12 for text most operators never write, while the field beside it —
often a long "via" instruction — was the one running out of room.
Swapped, so QSL via takes 7 and the message 5. Also puts them in the order they
are reached for.
QRZ has no record under M0BFS/QRP, so the lookup falls back to the home call —
that part worked. What followed did not: the home record's location is then
discarded, on the reasoning that a portable operator is not at their registered
address. True for /P and /M, and simply wrong for /QRP, which says something
about the transmitter and nothing about where it is.
So the grid came back empty for M0BFS/QRP while the same call without the suffix
answered perfectly, which is exactly how it was reported.
The distinction is now explicit rather than lumped in with the other operational
suffixes. /P and /M keep clearing the location, because mobile and portable both
mean somewhere other than home — that behaviour was correct and is untouched.
A power suffix stacked on a portable one (F4BPO/P/QRP) still clears: one of them
moved the operator.
A NULL never equals '', so a condition written that way returned zero rows. The
question is perfectly clear — the operator wants the ones with nothing in that
field — and answering it with silence makes the filter look broken rather than
mis-stated. eq and ne with a blank value now run the empty / not-empty test.
Empty on a NUMERIC column also had a real fault behind it. IFNULL(col,'')=''
compares 0 against '', which SQLite calls false and MySQL calls true — one
expression quietly answering two different questions depending on where the
logbook lives. Numeric columns test NULL or zero explicitly now; text keeps the
string test, where '' is a real value and 0 is not.
The dialog walked a hand-written GROUPS array. "The contact" was added to the
fields and not to it, so mode, submode and RST existed, passed every check I had
written, and could not be picked — the same shape of fault as the missing column
mapping, one layer further out. Two lists that must agree, with nothing making
them agree.
GROUPS is derived from the fields now, in declaration order, so adding a group
is adding a field. A group with no translation falls back to its own name rather
than rendering an empty heading: a missing label should look untidy, not
invisible.
Also stacks the DX sunrise/sunset in the band-slot header. Side by side it cost
about 150 px of a row that also carries the callsign, the badges and the band
grid, and it was what pushed that row onto a second line. Two short times one
above the other cost a fraction of it and no extra height — the row is already
taller than one line of text. The "UTC" label goes to the tooltip: the times are
monospaced and always UTC everywhere in OpsLog, so it was spending width to
repeat a convention the operator already lives by.
A bulk-editable field passes through THREE tables: the list the dialog shows,
bulkFieldColumns in app.go, and bulkEditableCols in internal/qso. I added mode,
submode and RST to the first and the third and missed the middle one, so the
dialog offered them and every save came back "unknown field".
Nothing warns about that. Each table is perfectly valid on its own, and the
mismatch only surfaces when an operator picks the one field that falls through
the gap — which is exactly how it was found.
So both directions are pinned now: every mapped field must be writable by the
qso layer, and every writable column must have a field mapping to it. A column
nothing maps to looks supported from the inside and cannot be reached from
outside, which is the same fault wearing the other hat.
The button was the wrong shape twice over. A maintenance chore does not belong
beside the options an operator actually chooses, and nobody should have to be
told their log is missing a field before it gets filled in.
So the distance is computed where QSOs come from: on the logging path and on the
ADIF import. fillDistance is its own function rather than part of
applyStationDefaults, because the import only applies those when the operator
ticks a box — and a distance is not a station default. It is derived from the
QSO's own two grids and is true whatever was chosen about profile fields.
What is already in the log is handled by a one-time migration at startup,
recorded by a settings key. In the background: on a large log over a remote
MySQL that is thousands of row updates and startup must not wait for a tidy-up.
Marked done only on SUCCESS, so a run cut short by a closed program tries again
next time rather than leaving half the log filled for ever.
An imported value still wins. It came from the log that made the contact, which
knew the real positions rather than two four-character squares.
Every one of the 2660 references was filed under European Russia. Corrected
against the RDA reference list: 991 are Asiatic Russia (15), and the 24 KA-
districts are Kaliningrad (126), which is not Russia for DXCC purposes at all.
1645 were already right.
dxcc_filter went from [15,54] to [15,54,126] with them. That filter decides
which QSOs are even considered, so leaving it alone would have made the 24
Kaliningrad districts unclaimable — a correction that quietly removes references
is worse than the error it fixes.
BA-27 (Mezhgorye) and BO-33 (Schebekino) are not in the reference list at all
and were left exactly as they were. Guessing an entity for a district nobody
lists is how a second wrong answer gets added to the first.
Only the dxcc field changed. Names differ in transliteration between the two
sources — Maykop against Maikop — and rewriting those was not asked for and
would churn 2660 lines to no purpose.
The published page derives a distance as it renders, which fixed the empty
column there but not the cause: the field is empty in the database, so it goes
out empty in every ADIF export and leaves the same gap in whoever imports it.
BackfillDistances computes it from the two locators for QSOs that have both.
Only where it is EMPTY: a stored distance came from the log that recorded the
contact, which knew the real positions, and two four-character squares are a
worse answer that must not overwrite a better one. Whole kilometres, for the
same reason the published column is.
In the Database panel rather than beside the county backfill, which lives under
US Counties: this one touches the whole logbook, whatever country the QSOs are in.
The page was capped at 1100px — a comfortable reading width, and the right
choice for the eight default columns. Now that all 123 fields can be published,
anything past about eight sat behind a horizontal scrollbar on a screen wide
enough to show the lot.
Worse, Windows hides overlay scrollbars until something moves, so a table that
scrolled looked exactly like a table with its right-hand columns missing. That
is what was reported, and the report was reasonable: nothing on screen said
otherwise.
The wrapper grows with its content now, up to the window, and keeps a minimum so
a two-column table does not collapse on a large display. The table sizes to what
its cells need rather than being squeezed to the container first — that squeeze
could wrap a callsign while empty space sat further along the same row. And the
scrollbar is drawn permanently, thin and in the page's own colours.
Nothing computes a distance when a QSO is logged. The DISTANCE field is only
ever filled by an ADIF import that carried one, so publishing it straight gave
an empty column to anyone whose log was made in OpsLog — which is everyone who
reported it.
It falls back to the two locators, which are on the QSO already. A stored
distance still wins: it came from the log that recorded the contact, which knew
more than two four-character squares do. No grids means an empty cell, not a
zero — an empty cell is honest, a zero is a claim.
Rounded to whole kilometres. The squares are tens of kilometres across and a
decimal would assert an accuracy nobody has.
The geometry moved to internal/geo on the way. It lived in package main, which
internal packages cannot import, so the PSK Reporter watcher already had its
maths injected from main and this would have been a third copy. A bearing that
disagrees with itself between two panels is a fault nobody reports, because each
screen looks perfectly plausible on its own.
They were excluded as "per-QSO", alongside callsign and date. That rule confused
two different things: bulk edit is not for describing QSOs, it is for REPAIRING
a batch of them — an import that mapped every contact to SSB, an ADIF that
carried no MODE at all. Refusing because a hundred rows should not normally
share a value left the operator editing a hundred rows by hand.
Setting the mode CLEARS the submode. A submode belongs to the mode it was
recorded under; left behind it contradicts the new one, and "FT8 with a submode
of USB" is not a thing — worse, the submode is what most ADIF readers believe.
Band is still refused on its own, and that half of the rule stands. It travels
with the frequency through BulkSetFrequency, which writes the pair: a band
contradicting its own frequency is invalid ADIF, and every export would carry
the contradiction out into the world. Callsign and date stay out too — they
identify the contact rather than describe it.
A saved band selection outlives the code that made it. 12 m was removed from the
watched set, but every operator who had already enabled the watch kept
subscribing to it — paying for a firehose whose messages the detector then threw
away. The stored list is filtered against the offered one now.
The badge appeared on the next poll rather than on the event, up to twenty
seconds after the announcement, so the two read as unrelated things that happened
to mention the same band. It appears immediately.
And the toast is gone. It said the same thing as the badge and vanished after
five seconds — an operator who was tuning when it passed had no way back to it,
which is precisely what the badge was built to fix. Keeping both announced an
opening twice and still lost it once.
Abbreviating 1500 to "1.5k" and then appending "km" produced "1.5kkm", which
means nothing. Plain kilometres now.
Written correctly it would still have been wrong: "1.5k km" asks the reader to
do arithmetic to recover a number that was four characters long to begin with.
There was nothing to save.
It read "6M OPEN" and kept the rest in a tooltip nobody hovers — an alarm with
no content, announcing that something was happening and refusing to say what.
Band, direction and typical distance are on the badge now: the three facts that
decide whether to point an antenna, all readable without touching anything. Out
of season gets a mark there too, since that is the one an operator must not
learn last.
The compass point is computed in Go and carried on the Opening rather than
derived in the UI. Sector() already gets the wrap round north right, and it got
it wrong once — a second copy of that arithmetic is how a badge ends up naming
the opposite direction to the log line beside it.
set_split_vfo and set_split_freq both answered RPRT 0 and did nothing. WSJT-X
and JTDX in "Split Operating: Rig" send exactly that pair, believed both, and
transmitted on the RECEIVE frequency — on a pileup, straight onto the DX, while
showing the operator precisely what they had asked for. A lie that leaves no
trace in any log is the worst kind of bug this program can have.
The two commands are honoured as a PAIR. Arming alone does nothing on the radio,
because WSJT-X sends the frequency second and split armed on whatever the
transmit VFO happened to hold is worse than no split at all: it transmits
somewhere the operator never chose. The request is remembered and set_split_freq
does the work.
Kenwood gains SetSplit — FB to place the dial, then FR0/FT1 to arm, in that
order for the same reason. It writes what State() already knows how to read.
Everything else REFUSES, and that is the feature, not a shortfall. Only Flex and
Icom could even toggle split before, neither could set the transmit frequency,
and Yaesu, TCI and OmniRig have nothing at all. A refusal WSJT-X can report —
and act on, by falling back to Fake It — is worth far more than a success it has
no way to check.
Both paths are pinned: split reaching the rig as one armed call with the right
frequency, and a backend that cannot do it producing an error rather than RPRT 0.
IF reports the VFO "in use". In split that is the RECEIVE VFO while receiving
and the TRANSMIT VFO while transmitting — the backend took it as the receive VFO
in both cases. Everything therefore read correctly until the PTT closed and then
reversed, which is exactly how it was reported from a TS-590 on USB.
The IF frame has carried the transmit bit all along; parseKenwoodIF already
decoded it into f.TX. It was simply never consulted here.
This is not only a display fault. FreqHz is what a QSO is LOGGED on, so a
contact worked in split went into the log on the DX's frequency instead of the
operator's — wrong in the log, wrong in every ADIF exported from it, and
invisible until someone checked a QSO by hand.
The emulated rig in the test package can now be keyed, so the case is pinned
from both sides: verified failing without the fix (tx and rx exchanged) and
passing with it.
Every condition was joined by ONE global AND or OR. "2 m or 70 cm, in FT8, since
January" therefore had no expression at all: AND killed the two bands, OR let
every FT8 QSO through. The filter could ask simple questions and nothing else.
Rather than grow nested groups — a tree in the UI to answer something that is
nearly always "this field, any of these values" — the OR lives INSIDE one
condition and everything else keeps ANDing. Two operators, a comma-separated
value, no change to how the rest of the filter behaves.
Two edges that matter more than they look. An empty list matches NOTHING rather
than being dropped: dropping it would widen the result set, the opposite of what
someone typing a filter expects. And "is none of" wraps the column in IFNULL,
because raw SQL NOT IN discards NULL rows — a QSO with no band recorded is not
one of the listed bands, so it belongs in the answer.
SetMotorFollow went through SaveUltrabeamSettings, which tears the client down
and dials again. That is right when the transport changed and absurd for a
checkbox: the operator flipped tracking and watched the antenna disconnect and
come back, which on a remote controller is several seconds of a link that was
working perfectly.
The follow loop is a goroutine with its own stop channel and can be replaced on
its own — the connection underneath never knows. Only the two keys this call
actually changes are written, too: re-saving the whole block to change a
checkbox means re-normalising host, port, baud and bands, and every one of those
is a chance to alter something nobody asked to alter.
startUltrabeam now starts the loop through the same function rather than its own
copy. Two versions of "start the follow loop" drifting apart is how a step
change quietly stops taking effect.
It showed pattern, elements and Retract — everything except where the antenna
is pointed, which is the thing an operator changes most. Tuning it meant opening
Settings or moving the rig.
A button per band, taken from the bands configured in Settings rather than a
list invented here: a band dropped there cannot be clicked here. They point at
where people actually work, not the arithmetic centre — 20 m centres on 14175
and nobody lives there, 10 m spans 1.7 MHz of which the top half is empty.
Up and down move 25 kHz, which is also the finest tracking threshold: a smaller
nudge would be undone by the next poll while tracking is on. They work from the
ANTENNA's frequency, not the rig's, or walking the antenna across a band while
the rig stays put would be undone on the first press.
Tracking moved within reach because it is an operating decision — off to park
the antenna, on to resume — not something set up once. Its step only appears
when it is on: a threshold for something switched off is a question the operator
cannot act on.
MotorTuneKHz carries the CURRENT direction rather than resetting it. Both
controllers set frequency and pattern in one command, so a bare frequency would
silently drop a 180° or bidirectional pattern — a beam turning round is not an
acceptable side effect of clicking a band.
Club Log refuses with its ordinary web page rather than an error string, so the
new test reported a rejected login by pasting a 403 page — title, stylesheets,
navigation and all — into the status bar. The body now goes to the log, where a
real diagnosis happens, and the operator gets the one sentence there is to act
on: check the e-mail, the password and the logbook callsign.
Dropped the startyear=2099 filter with it, and that one matters more than it
looks. It was there to keep the reply small, but it was never verified against
Club Log's API — and Club Log answers an unrecognised request with the SAME 403
it uses for a refused login. An unverified parameter would therefore have made
every CORRECT password look wrong, which is precisely the failure this change
set out to end. The reply is capped at 4 KB and closed at once instead; the
status code arrives ahead of the body either way.
Band-opening detection had five entries in 0.24.5 — long paths, the PSK Reporter
feed, which bands, the densest sector, the region filter. Those are five steps I
took, not five things that happened to the operator, and the version read like a
work diary. Merged into one entry describing what the feature now does.
The 12 m mention went with them: it was added and removed inside this same
version, so it never existed as far as anyone reading this is concerned.
A PSK Reporter message says "X was heard BY Y". The watcher measured the
distance and bearing from the operator to X and stopped there, never asking who
had actually heard it — so a station 1400 km away, decoded by somebody in Japan,
counted as evidence. That proves the path from X to JAPAN and says nothing about
whether anything reaches this station. It is how a "2 m opening" came to be
announced out of a KX9X in the United States, and the operator was right to find
the callsign list absurd.
Reports are now kept only when the RECEIVING station is within 300 km. Far
enough to borrow the ears of a whole region — an opening reaches an area, not a
postcode, and waiting for a decode at one's own antenna is just working the band
— and close enough that the ionosphere doing something there is it doing the
same thing here. The transmitter still supplies the direction and the path
length, which is what was always wanted from it.
Also drops 12 m from the watched bands, at the operator's request: at this point
in the cycle it is open often enough that announcing it is a notification rather
than news, and a band that cries wolf costs the ones that do not. One fewer
subscription is also less traffic on a PC that pays for every message.
Two faults, both found from one field log: 70 000 decodes, 10 and 12 m plainly
open, nothing announced — and the two openings that DID fire named the wrong
direction.
evaluate() required EVERY distinct station in the window to fit inside one 90°
arc. That is the shape of a sporadic-E cloud and of nothing else. With 10 and 12
m open on F2 the reports arrive from all round the compass, the arc is 360°, and
the test can never pass — so the busier the band, the less likely an opening was
announced. Exactly backwards. It now finds the DENSEST sector instead, which
keeps the Es signature intact (a cloud still makes one direction dense) and lets
a real F2 opening be seen through the handful of neighbours who are always
there. Scattered-but-busy still reports nothing: the point was to stop demanding
global agreement, not to call every open band an opening.
Sector() averaged the two bearings arithmetically, so an arc crossing north was
labelled by its opposite: 353–61° averaged to 207° and went out as SW when it
was NE. Not a vague error — a reversed one, given to an operator who may turn a
beam on it. The detector has handled the 0/360 wrap since it was written; only
this label had not.
The only sign of a detection was a toast. It is gone in seconds, and an operator
who was tuning at that moment had no way back to it — for an event that lasts
hours and happens a handful of times a season, that is the wrong shape entirely.
The badge sits at the right-hand end of the status bar, before the clock: an
opening is a state of the WORLD, not of this station, so it belongs with the
time and the logbook rather than among the rig and amplifier chips.
Knowing when to go out was the real work. Nothing announces that an opening
ended, and the detector deliberately says nothing more about a band for 45
minutes after announcing it — right for a message, useless for a badge. So it is
inferred: every qualifying spot on that band pushes a deadline out, and when
they stop arriving the badge fades by itself. Fifteen minutes, comfortably more
than the detector's own twelve-minute window, so a quiet couple of minutes
mid-opening does not blink it off and on again.
Gated on the same distance floor the detector uses. Without that, a band busy
with short-range tropo would hold an Es badge lit indefinitely on spots the
detector itself had refused.
TestClublog checked that three fields were non-empty and returned "Ready — CALL
via EMAIL". Nothing was ever sent to Club Log, so a wrong password produced the
identical green message. It now signs in, through getadif.php because that is
the one authenticated endpoint that cannot change anything: a test must never
put a record into someone's log. A future start year keeps it from downloading
130 000 QSOs to prove a password, and a rejected login answers 403 before any
body arrives.
LoTW is two credentials doing two jobs and the button reported only the first.
Uploads go through TQSL signed by the certificate — the website password is
never involved — so a wrong one breaks nothing until the day confirmations are
downloaded, by which time nobody connects the two events. Both are now checked
and, more importantly, reported separately.
Also: the detector still refused 12 and 10 m while the PSK Reporter feed was
already subscribed to them, so those decodes were fetched and thrown away. The
rule was never "HF is out", it is "is an opening here an event" — 20 m being
open is the normal state of the band, 10 m opening is not.
Beside the rig and amplifier chips, before ON AIR, because it is the same kind
of fact they report: a link that is either up or it is not. Clicking it opens
the settings that control it, like the amplifier chips do.
Shown ONLY while the opening watch is on. A permanently grey chip for a feature
nobody enabled is clutter, and the bar is 28 px tall — every chip in it has to
earn its width.
Green once decodes are arriving, amber while connected but silent: those are
different states and an operator wondering why no opening has been announced
needs to tell them apart. The count itself is in the tooltip, not on the chip —
it moves several times a second on an open band, and a number flickering in the
corner of the eye is a distraction rather than information.
The detection shipped reading whatever the operator's cluster nodes happened to
carry. On VHF that is a few hundred skimmers, nearly all of them on HF: a 6 m
opening carrying 869 stations reached OpsLog as a handful of spots or none, and
Nexus flagged it on the same PC while OpsLog stayed silent.
internal/pskr subscribes to pskr/filter/v2/<band>/# on PSK Reporter's MQTT
broker. Every ordinary station running WSJT-X reports what it decodes, so the
difference is two orders of magnitude rather than a threshold. The feed's shape
suits us exactly: BOTH grids are in each message, so distance and bearing are
arithmetic — no lookup, and no DXCC-centre approximation, which is what made the
cluster path's bearings coarse. The geometry is injected from app.go so it stays
the same arithmetic the cluster path uses; two answers to one question is how a
bearing quietly becomes wrong.
Volume was the design constraint, not the protocol. Six metres open is thousands
of messages a minute and this runs on some very old PCs, so nothing is kept or
persisted in the watcher: each message is parsed, measured and handed on or
dropped, and the detector's existing window does the deciding.
And the part that was the real bug: enabling the watch now ARRANGES ITS OWN
SOURCES. It adds the two RBN nodes when missing and brings the feed up. A
feature that silently depends on sources nobody can know are needed does not
look unconfigured, it looks broken. Matched on host and port, not name, so an
operator who renamed theirs does not get a duplicate — which the detector would
read as twice as many stations, and announce an opening that is not there.
Turning it off leaves the nodes alone: they may have been wanted for their own
sake, and removing a node someone is using is worse than leaving one they are not.
The detector refused any path over 2400 km, reasoning that past a single hop the
bearing test stops meaning anything. That was wrong, and it discarded exactly the
openings worth announcing: Nexus flagged a 6 m opening at 5477 km that OpsLog
never saw, because the spots were thrown away before any test ran.
Multi-hop Es is ordinary on 6 m — 5000 km paths are common, 10000 km happens —
and it stays directional: a second hop leaves the sector the first one entered.
So the sector test, which is what does the real work here, holds perfectly well
at any distance. The ceiling was standing in for a judgement it could not make.
The floor stays at 500 km: a short 6 m contact is tropo or ground wave and says
nothing about the ionosphere. Both are now pinned by a test.
Both are ordinary Hamlib commands, and Nexus sends them around every transmit.
Refusing with RPRT -11 is permitted and a tolerant client carries on, but
nothing obliges it to — and while they sit in the log as "unimplemented" they
stay suspects every time something else goes wrong.
get_lock_mode answers 0, which is true: OpsLog never locks the dial against its
own clients. set_lock_mode is accepted and ignored, like set_vfo — there is no
lock to set, and failing would abort a client's transmit sequence over a setting
with no effect either way. stop_morse answers success because nothing is queued
here: CW is keyed by the rig's own keyer through the backend, never buffered in
this server, so "stopped" is accurate rather than polite.
Whether this is what Nexus is actually unhappy about is not established. It
removes two known irritants and two lines of noise from the log; if the trouble
persists, what remains in the log will be about the trouble.
It was sticky inside the scrolling list, and the rows went over it: a menu item
carries its own background and its own stacking context, so it wins against a
sticky sibling however high the z-index is raised.
The menu is a flex column now — a header that never scrolls, and the list
scrolling beneath it. Nothing to lose the fight with.
The first attempt laid 123 fields out on the page, sectioned by group. That
buried every other setting in the panel and was no easier to read than the flat
wrap it replaced — more surface, same problem.
One dropdown now, alphabetical, with a search box pinned at its top. Alphabetical
because any other order means hunting: the operator arrives knowing the name of
the field they want. The menu stays open while ticking, since picking eight
columns should be one visit rather than eight.
The chosen columns keep their place above it, in publication order — that list
answers "what will the page look like", which no catalogue can.
The catalogue held 23 hand-picked columns, so publishing a county, a satellite
pass or an award reference was simply not possible. It is generated from the
qso.QSO struct now — 123 fields — which also means a new ADIF field cannot be
forgotten here, as a hand-written list always eventually is.
Three keys were renamed to their ADIF names on the way (pota, sota, station).
An alias map keeps existing configurations publishing the same columns; without
it three would have vanished silently on upgrade, which is the worst way for a
setting to change.
The picker had to change with it: 123 chips in one wrap is a wall nobody reads
to the end of. It is sectioned by group, filtered as you type, and the chosen
columns sit on top in publication order — after picking eight out of a hundred,
the question stops being "what exists" and becomes "what did I pick".
The package doc said columns were curated so the operator's address could not
be published. That is no longer true and the comment now says so plainly: the
judgement moved to the operator, the default selection is unchanged, and nothing
is published that was not chosen.
Also collapses the changelog: four separate shared-CAT entries were one thing
from where the operator sits, and all of them were far longer than the one or
two sentences this project asks for.
Nexus sends set_ptt 0 about sixteen times a second, and every one of them became
an "xmit 0" to the FlexRadio — a write every 60 ms, for ever, saying nothing.
Repeating a state is not a request to change it. Only transitions reach the
radio now, and each is logged, so the next "it will not transmit" can be traced
to whoever asked rather than inferred from a wall of identical lines. The first
call always goes through: how the radio was left is not ours to assume.
This is not the whole of the reported symptom, and should not be read as such.
The log also shows "xmit 1" followed by "xmit 0" one to two milliseconds later,
off the 60 ms cadence — that is a genuine transition pair, so it still gets
through and the radio still unkeys at once. Those two commands come from the
client, not from here; the writes originate only in set_ptt, and there is no
loop in OpsLog that emits them.
Listen() succeeding is not the same as being reachable. OpsLog binds 0.0.0.0,
and Windows lets a second program bind the SAME port on the specific address
127.0.0.1. Connections to localhost then go to the more specific listener, so
every client reaches the other program while ours sits there having logged
"sharing CAT on port 4532" and never seeing a single connection.
Found with Nexus, which starts its own rigctld on 127.0.0.1:4532 and talks to
it. Neither side reports anything wrong: the operator gets a CAT timeout from a
daemon with no radio behind it, and OpsLog's log is silent because nothing ever
arrived. Three exchanges went into establishing that the connection simply never
reached us — the port table was what settled it, not the code.
So the server now dials its own port at startup and checks the connection lands
on its own accept loop. If it does not, it says which program will be receiving
the CAT connections and what to do. The counter it compares can only be raised
by our own accept loop, so a real client arriving during the probe makes the
check pass, never fail wrongly.
A client set to a RIG MODEL (Kenwood, Yaesu, …) pointed at the CAT-sharing port
speaks raw rig dialect: "ID;", "IF;". That is not rigctl, so it fell to the
unknown-command branch and got RPRT -11 like anything else.
The symptom hides the cause completely. RPRT -11 has no ';' for the client's
parser to terminate on, so it waits out its timeout and reports "reply
incomplete, got nothing" — a hard failure that reads as "OpsLog's CAT sharing
does not work", when it is one setting in the other program.
A frame ending in ';' with no space in it cannot be a rigctl command, so the log
now says what it is and what to set instead. Reported from Nexus, whose Hamlib
error named kenwood_transaction — the one word that gave it away.
cfdd24d shipped the "?;" tolerance with no changelog entry, against the
project's own rule. Placed before the CW entry: the link had to survive being
busy before anything keyed through it could matter.
CW over CAT never worked on a Kenwood. The KY implementation was written against
Elecraft's, which accepts a string of any length, so "KY OH5CX;" went out and a
TS-590SG answered "?;". OpsLog read that as "this radio refuses CW over CAT" and
told the operator to fit a serial keyer — advice that was wrong, and expensive.
The TS-590 manual is explicit: P2 has a fixed length of 24, blanks are filled
with spaces, and those spaces are not keyed. So the fix costs nothing on air; it
is simply the shape the command has. Elecraft stays variable-length, where
padding would key the trailing spaces as word gaps.
The semicolon is also gone from the allowed CW characters. It TERMINATES a CAT
frame, the manual forbids it in P2, and one in a macro would have closed the
command early and left the rest of the message to be read as commands.
Found from a log and a manual page, not from a rig: nobody here owns a Kenwood.
What is proven is the frame shape; that a TS-590SG then keys it still needs the
operator to confirm.
A Kenwood answers "?;" while it is busy, and the TS-590SG does it for a moment
after RX; — which is exactly when the poll resumes. One rejected IF; then tore
down the whole CAT link.
The code already knew this. Its own comment says a "?;" to IF; is "a transient
busy, NOT unsupported" and that latching it off "would read as lost the rig".
That reasoning was applied to the unsupported-command map and stopped there: the
call still returned an error, the poll loop still turned it into connected=false,
and the link still dropped.
What it cost was not cosmetic. WSJT-X keys through shared CAT, so it lost the
rig mid-sequence; Hamlib then sent "F 9223372036854775808.000000" - an
uninitialised 2^63 - which OpsLog refused and logged. The operator sees a
frightening frequency error whose actual cause is three lines earlier.
A "?;" is now sentinel-wrapped so the poll can tell "the rig declined" from "the
serial link is gone", and three consecutive ones are ridden out on the last
known state before the link is called dead. Under a second of tolerance: enough
for the rig to finish, far too short to hide an unplugged cable. A serial fault
is untouched and still drops immediately.
Tested against the emulated rig already in this package, with a hook that
answers "?;" on demand: the link survives inside the grace and still reports the
last good frequency, the count resets on recovery so a later busy spell gets the
full allowance, and a rig refusing forever is still reported as broken.
N0CALL is what WSJT-X transmits under when its owner never set a callsign. It
has a letter, a digit and an ordinary shape, so nothing rejected it: it was
spotted, coloured, counted as a new WPX prefix, and now that CQ grids are read
it would have put a grid into the worked index under a callsign nobody holds.
The obvious fix - adding it to looksLikeCall's reject list - was wrong, and the
test caught it before it shipped. That function answers "could this token be a
callsign at all", and the CQ grammar uses it to decide whether the word after CQ
is a modifier (DX, NA, a zone) or the call itself. Teaching it that N0CALL is
not a callsign made "CQ N0CALL JN36" skip a slot and return JN36. Shape and
policy are different questions and now live in different functions.
Chasing that turned up the real defect behind it: ANY unrecognised word after CQ
made the parser skip a slot, and a four-character grid passes every shape test a
callsign does. "CQ FOO JN36" returned JN36 as the sender - logged, spotted and
coloured as a station. A grid in the callsign slot is now refused outright.
An operator reporting "nothing happens for two or three seconds after I click"
was impossible to answer from opslog.log, because the first line applog writes
already sits inside startup(). Everything spent loading the binary and creating
the WebView2 environment happened before the log begins.
In one of the reported logs, startup() to window-visible is 251 ms end to end -
so whatever the wait is, it is not OpsLog initialising. It just could not be
proven, only asserted.
processStart is stamped on the first instruction of main. Two lines now bracket
the launch: how long was spent before startup() ran at all, and click-to-window,
which is the only number that matches what the operator experiences.
Finishes the half of this that was already computing on the backend and reaching
nobody. A Grid column in the Geo group, NEW GRID as a badge in Status, a filled
cell in the marker's own colour, and a filter chip.
new_grid joins lib/spotMarkers rather than getting colours of its own, so the
badge, the cell fill and the chip cannot drift apart - and the per-marker colour
setting will drive it with the rest from one table. Magenta: the last hue in the
categorical set not already spoken for, and one that does not read as a status,
because a new square is never urgent the way a new entity is.
The band map leaves it to the cluster, as it already leaves the prefix. The pill
is 22 px tall and its accent strip stops being readable past three segments.
A row carrying a new grid is no longer "dull", or the dimming would grey out the
one thing worth looking at.
The column is off by default, like the other Geo columns: it is only ever filled
for stations this receiver decoded over the UDP link, so for an operator who
does not run digital it would be a permanently empty column.
"No colour on worked" blanked the spot status as well as the worked-call flag,
on the theory that a spot bringing no novelty should stop painting entirely.
But the status is exactly what the cluster list reads to DIM a row. Blanking it
made isDull() fall through its "unresolved, never dim" guard, so every quiet grey
row came back at full brightness. The option meant to calm the list was the one
making it shout. It also forced a `muted` flag to exist purely so tooltips could
say that an empty status did not, this time, mean "entity not resolved".
It now removes the blue already-worked mark and nothing else, which is all it
ever needed to do: a worked entity already renders with no colour of its own and
is already dimmed. The flag, its two tooltip strings and the special-cased band
map styling all go with it.
The amplifier PUSHES status frames ("S0|state=…") on the same socket it answers
commands on, and it pushes them constantly once in OPERATE - power, SWR and
temperature all move while transmitting.
command() read exactly one line and took whatever arrived first as its answer.
One pushed frame therefore put the stream permanently one reply behind: every
later command read the PREVIOUS command's answer, and eventually one waited out
the 3 s deadline, failed, and dropped the connection. Hence a fresh TCP
connection and a fresh authentication every few seconds in the log.
The stall in transmit is the same defect seen from the other side: command()
holds the connection mutex across that whole dead wait, so SetOperate and every
other control queued behind up to three seconds of nothing.
It only appeared remotely and in OPERATE because that is when there is anything
to push - on a LAN with an idle amplifier the race hardly ever opens. Which is
why it did not do this yesterday.
readReplyLocked now reads until the R<id>| that belongs to the command, feeding
every frame it passes to parse() on the way - a pushed status is fresher than
the one we were about to ask for, so nothing is wasted.
authLocked used the same one-line read and worked around this by re-sending the
whole handshake, which is visible in the log as "auth reply=S0|state=IDLE
(try 1)" followed by a second attempt. It goes through the same reader now.
IOTA is an official ADIF field, and almost everything for it was already here:
the qso table has carried my_iota since the first migration, the insert and scan
handle it, and both ADIF import and export write it. The one missing link was
the station profile, so the reference had to be typed on every contact of an
island activation, or added afterwards in a bulk edit.
Migration 0025 adds the column; it joins my_sota_ref and my_pota_ref in the
profile, in the Station Information panel, and in the same stamp-if-empty block
that fills the other My* fields on a logged QSO.
Uppercased and trimmed on save. ADIF spells it EU-005, and an operator typing
"eu-005" would otherwise put a reference no award matcher recognises on every
QSO of an activation - the kind of mistake you only find months later.
AddQSO emitted qso:logged but never invalidated the cluster status snapshot.
The frontend did its part - it re-queried every visible spot two seconds later -
and ClusterSpotStatuses answered out of a snapshot built before the contact, so
it returned exactly the same "new band" as before. For ever, until an import, an
edit or a profile switch happened to invalidate it for another reason.
Reported twice: an E51 and then a ZD7 that stayed yellow on the band map with
the QSO plainly in the log. The first report was answered by fixing the
visibility gate, which was a real bug of its own and hid this one.
Confirmed against the operator's MySQL logbook rather than guessed: ZD7BG has
dxcc=250 on all eleven QSOs and 0 of 29579 rows lack a DXCC number, which ruled
out the missing-entity-number theory and pointed here.
Deliberately not invalidateAwardStats(): that also drops the award matrices,
which are expensive on a large log, and a contest run would pay for it once per
QSO. This index is a few DISTINCT scans and it is what the spot colours read.
Also fixes the new worked-grid query, which asked for a column named
"gridsquare". The column is "grid" - gridsquare_ext is a different one - so NEW
GRID could never have worked on either backend. Verified against the real
database: 12516 distinct grid|mode pairs.
RBN spotters report as "VU2OY-#" and cluster nodes as "DL1ABC-2". That string
went straight into the DXCC prefix matcher, which saw an unknown callsign and
gave up, so every spot came back with an empty continent.
The filter then matched nothing at all - and worse, silently: unresolved spots
are deliberately never dropped, because the status arrives a moment after the
row and filtering meanwhile makes the list flicker. So selecting AS left the
Europeans and the Americans exactly where they were, with nothing to say why.
A real callsign never contains a hyphen, so cutting at the first one is safe.
The panel had grown two shapes for the same kind of choice. Some filters were
checkboxes, some were chips, and New counties only was a checkbox duplicating a
chip - which I added, and which was the worst of it: a control that exists twice
is not more discoverable, it is one control the operator has to recognise twice.
The rule now:
SWITCH a behaviour that is on or off, and narrows nothing by a property of
the station: hide worked, group duplicates, the two display options,
LoTW users only.
CHIPS pick any number from a set; none picked means all. Status, mode,
spotter continent. Selected is solid, unselected is the same chip
faded, so the palette keeps teaching the colour code while switched
off.
Nothing appears in both shapes. The duplicate county checkbox is gone.
Spotter continent became a chip row: seven two-letter codes fit on two lines,
they read as a set the way Status and Mode do, and several can be picked at
once - which "EU or NA" needs and a dropdown cannot express.
Both Lock buttons now sit in the heading of the section they lock, instead of
floating between sections, and every multi-select section clears the same way
through the same heading slot. One section helper and one chip helper, so the
next filter cannot drift.
The panel was also entirely hardcoded English - Search call, Hide worked, Bands,
Status - against the project's own bilingual rule. All of it goes through t()
now, both locales.
The L badge is a single letter, not a word: the call column is 120 px and holds
a callsign. It keeps the muted blue of a confirmation and never a status colour,
because whether a station uploads to LoTW says nothing about whether the spot is
worth chasing - those are different questions and must not share a palette.
The spotter's continent is not the DX's. It asks whether anyone near you is
hearing the band at all, which is why it earns its own control rather than
reusing the existing Continent column. The spotter callsign now travels with the
status query so the backend can resolve it against the one DXCC prefix table,
instead of a second continent rule appearing in the frontend.
Both AND with the status chips rather than joining their OR: "a new band, and
from Europe" is the question being asked. An unresolved spot is never dropped by
them - the status arrives a moment after the row, and filtering meanwhile made
the list flicker.
New counties only is the SAME state as the NEW COUNTY chip, reached a second
way, not a second filter. County chasing is a mode you switch into, and hunting
for one chip among eight is not how you switch into it.
Backend groundwork; the columns and filters that consume it come next.
GRIDS. A CQ is the one WSJT-X message that carries a locator, and wsjtSender was
throwing that token away. It is now returned, validated as a real field+square,
and remembered per callsign. This is the ONLY grid source available: a DX-cluster
line carries the spotter's grid at best and never the DX's, and a per-callsign
QRZ lookup under an RBN firehose is not a trade worth making. So grids are known
for the stations this receiver decoded - which is exactly the FT8/FT4 watering
hole an operator is looking at while grid chasing.
RR73 is why the grid is validated rather than pattern-matched. R is inside A-R
and 73 inside 00-99, so a sign-off satisfies the Maidenhead shape exactly and
would have planted a grid that does not exist into the index, silently.
NEW GRID keys on "GRID|MODE" with the mode put through the same normMode as
everything else, so the "group digital modes" option decides whether a grid
worked on FT8 is still new on FT4 - one rule, no branch. Grids are truncated to
four characters: a log holds a mix of JN36 and JN36QU, and without that the same
square is new forever, once per subsquare.
On cost, which was the condition: one more DISTINCT scan when the status
snapshot is rebuilt, then map lookups per spot. The same shape as the county and
POTA sets it sits beside, and the snapshot exists precisely so a spot batch never
touches the logbook.
Spotter continent and the LoTW flag come from tables already in memory - the
DXCC prefix table and ARRL's user list - so they cost a lookup each. The spotter
continent answers a different question from the DX's: whether anyone near you is
hearing this at all.
parseFloat accepts a numeric PREFIX. "2026-08-10" therefore became the number
2026, every date in the same year compared equal, and since the sort is stable
nothing moved: the Date column looked as though it were simply not sortable.
The same trap caught every callsign starting with a digit - 8B81SU and 8P9AB
both read as 8 - and the times, where "09:56" became 9. The numeric test is now
anchored to the whole cell, so anything that is not entirely a number is
compared as text, which is exactly right for an ISO date.
Sorting had no way back either. Each row now carries the index it was published
at, and a third click on a header restores that order. Headers also show an
arrow: with no indicator, a column that silently refused to sort was
indistinguishable from one that had sorted into the same order.
Blank cells sink in both directions rather than leading the ascending sort. An
empty field is missing data, not the smallest value.
Tested where it can be: the page ships its own script, so a regression is silent
- the table still renders, it just sorts wrongly.
Colour moves from the text to the CELL. A filled Band cell means new band, a
filled Pfx cell means new prefix, a filled County cell means new county. This is
not the pills coming back: a pill is a box inside the cell with its own height
and padding, so it pushed the text off the row baseline. A background has no
geometry - the text does not move a pixel - and it reads from across the room,
which a tinted glyph does not. Worked-call stays text-only: it is not a novelty,
and filling it would wash most of the rows.
The colours still come from the semantic tokens and lib/spotMarkers, so a fact
keeps one colour across the grid, the band map and the filter chips - and the
per-marker colour setting will drive all of it from one table.
US County column. The backend already resolved the county from the offline ULS
store to decide NewCounty and then threw it away; it now returns it, which costs
nothing.
The Locator column is renamed Spotter locator. It always held the SPOTTER's grid
- cluster.go says so - so a column labelled Locator next to a DX callsign was
reading as the DX's grid and was mostly empty besides. The DX grid is not in the
feed at any price worth paying under an RBN firehose.
Both display options move out of Preferences into the cluster filter panel,
beside Hide worked. They are changed while working a run, not set up once, and
a preferences dialog reopened every ten minutes is a filter in the wrong place.
The slot-highlight option reused the new-slot status, which already meant
something else and something narrower: the ENTITY was worked on this band and
on this mode, but never on the two together. What the option surfaces is a fact
about the CALLSIGN - never worked on this band and mode - and the entity may be
long since confirmed there.
Overloading it cost exactly what it always costs. Forty consecutive rows read
NEW SLOT, the genuine new-slot rows drowned in them, and the Status column
stopped carrying information at all.
new-call is its own status now, with its own badge, its own tooltip spelling out
that it is the callsign and not the entity, and its own filter chip. It keeps
the caution colour for the moment: it sits at the same "worth a look, not a new
one" level as new-slot, and the colour rework will separate them.
The status filter had to move onto the displayed status too. It read the raw
backend entry, where new-call by construction never appears, so the chip would
have selected nothing at all.
The "colour the stations not worked on this band and mode" option did nothing.
The backend computed WorkedSlot correctly and the generated bindings declared
it, but the status map in App.tsx is assembled field by field from the reply,
and worked_slot was not on that list. The extra key was simply dropped, so the
frontend saw undefined, the === false test never matched, and every spot stayed
muted. Silent by construction: nothing warns that a copied-out struct is missing
a field. All three assembly sites now carry it, and the state type names it so
the compiler catches the next one.
The band map also memoised the transformed status map on spotStatusRaw alone,
so toggling an option changed nothing until a poll happened to hand over a fresh
object. The options are read at render time and are part of the dependencies.
Muting emptied the status, which dropped the spot onto statusStyle's default
branch. That branch paints bg-primary/60 - the theme's burnt orange at 60% over
a dark card, i.e. brown - and it was written for the rare unresolved entity,
where a hint of the accent colour is right because the operator should look.
With the mute option on, most of the map is muted. So the whole band map turned
brown and the spots that were meant to stop competing for attention were the
only ones tinted with the app's own "look here" colour.
Muted now lands on QUIET_STYLE: card background, grey border, grey accent bar
and leader. That is the same style the 'worked' status already used, so it is
now one constant rather than two copies - a muted spot and a worked spot are
the same statement and must not drift apart.
Three defects in the options added in the previous commit, all found on air.
Order. Mute returned early, so with both options on the mute swallowed exactly
the spots the slot option existed to surface: the mute test is entity-level, and
an unworked callsign inside a worked entity passes it. Slot promotion now runs
first, and protects itself for free since bringsNothingNew is false on new-slot.
The two were documented as composing - "mute what is done, light up what is
not" - and they did not.
Empty status. Muting emptied the status, but an empty status already meant
"entity not resolved" in the band map, so every muted spot claimed its entity
was unknown with the country printed two words earlier in the same tooltip. A
muted flag now records WHY it went empty; blanking still drives the colour, the
badges and the ranking.
Wording. The muted tooltip said "already worked" of a station never worked -
LZ8NG on a 421st Bulgarian. What is worked is the ENTITY on this band and mode,
so that is what it says now.
new-mode had no case in the band map's statusLabel and fell through to "entity
not resolved" too. Pre-existing, same one-line switch.
Two options that change what the eye is pulled towards, both applying to the
cluster list AND the band map.
Mute worked before: a spot that brings nothing new loses its colour and its
badges. It stays in the list - the operator asked for less noise, not less
information. "Brings nothing new" reuses the dimming rule the cluster list
already had rather than inventing a second notion of done, so it keeps obeying
the same-slot option and the digital-mode grouping for free. A new-band or
new-slot status is NOT muted: having worked that callsign once on another band
says nothing about the band in front of you.
Highlight unworked in this slot: colours any callsign not yet worked on this
band and this mode, whatever the entity says. For an operator filling slots a
common entity on a fresh band+mode is the whole point, and the entity-level
status flatly calls it worked. It reuses the existing new-slot status, so no new
colour, no new legend, no new badge - both panels already knew how to draw it.
WorkedSlot is computed independently of the same-slot preference: it is what
this option reads, and it must not change meaning because a different option was
toggled. The slot index is now built when either option needs it, and the status
cache is keyed on both so a toggle invalidates it.
The rules live in one module used by both panels. Marker colours already taught
us what happens when the two derive the same thing separately.
Observation, not prediction, and it needs no new data source: the cluster event
worker already enriches every spot with the great-circle distance and bearing
from the operator's grid, which is exactly what a single-hop Es detection rests
on.
The signature is four or more DISTINCT stations at 500-2400 km inside a 90
degree bearing sector within twelve minutes. Each constraint earns its place:
distinct callsigns because one station spotted by six skimmers is six spots and
one station; the lower bound because a 6 m contact under 500 km is ordinary
tropo and says nothing about the ionosphere; the upper bound because past one
hop the bearing test stops meaning anything; and the sector because a real Es
cloud illuminates a direction, which is what separates an opening from a merely
busy evening.
Fed AFTER the Historical guard in the worker. A SH/DX reply replays a hundred
past spots in a second - precisely the shape of a burst - and would announce an
opening that ended hours ago.
Season LABELS, it never gates. Both hemispheres get a summer peak and a lesser
winter one, and an opening outside those is announced with "unusual for the
season" attached: the rare one is the one an operator must not hear about last.
One announcement per band per opening (45 minute quiet period). An opening runs
for hours and produces hundreds of spots; one alert is information, forty is
noise.
Opened as soon as 0.24.2 is tagged, so the next change has somewhere to go and cannot end up described under a version that already shipped - which is exactly what happened to 0.24.1. Invisible until the build reports 0.24.3: GetChangelog drops every entry newer than appVersion.
v0.24.1 was tagged and shipped mid-session; the six commits that followed kept appending to its block, so the released version appeared to describe changes it does not contain. 0.24.1 is restored to exactly the five entries that shipped - verified against git show v0.24.1:changelog.json - and the five later ones open 0.24.2.
The band row was hardwired to 160-6 m, so an IC-9700 - the one radio in the range with no HF at all - showed ten dead buttons and none of 2 m, 70 cm or 23 cm. bandsFor(model) follows the model, exactly as attOptions already did for the attenuator steps. bandOfHz's labels had to move with it: it returned '70' where the button says '70cm', so the current-band highlight could never have matched, and it knew nothing of 23 cm.
Settings -> Web publishing. Renders the last N QSOs with the columns the
operator picks, writes the file locally, and optionally uploads it by FTP or
explicit FTPS.
Local write FIRST, upload second, always. A network failure then leaves a good
file on disk that can be published another way, instead of a truncated one on
the server. The local write itself goes to a temp file and renames over the
target, so a reader — or a syncing client — never sees a half-written page.
The HTML page is fully self-contained: inline CSS, inline sort script, no font,
no CDN, no external request at all. It has to work on hosting that blocks
third-party requests, and a page about someone's hobby should not report its
readers to anyone.
Columns are a curated set, not "every ADIF field". This is published to the
public: RST and QSL status belong on it, the operator's home address does not.
Two triggers, both debounced through one path: a QSO is logged, or the optional
timer fires. Fifteen seconds of coalescing means a run of contacts produces one
upload rather than one per QSO, and nobody reading a web page can tell the
difference.
The config is one JSON blob under a single settings key, and that key is marked
sensitive: the FTP password lives inside it, so the whole blob is encrypted at
rest with the others. A locked vault reads back empty, which correctly reads as
"not configured" — publishing must not run with a password it cannot decrypt.
The post-QSO status refresh was gated on the DX-cluster list being on screen; the band map draws the same statuses and was left out. Working a station in the ordinary layout - Recent QSOs left, docked band map right - left its pill NEW until the cluster tab happened to be opened. Reported on an E51 that stayed orange with the QSO already in the log.
OSM blocked OpsLog and every user's map filled with "Access blocked / 403"
tiles at once. Their policy is explicit: an application must identify itself
with a proper, unique User-Agent on every tile request. A program that draws
its maps inside a web view cannot do that — the browser sets that header and
Leaflet fetches tiles as plain <img> loads — so this is not something to tune,
it is a service we are not entitled to use.
Two layers still pointed there: the Street basemap and the locator map. Street
moves to Esri World Street Map, the locator map to Carto, both key-free and
both under terms that cover a redistributed application. OpenStreetMap keeps
its credit: Carto's tiles are built from OSM data.
The locator map also gains the tile restraint the world map already had —
updateWhenIdle, no fetching mid-zoom, one ring of buffer. Fetching as fast as
the pointer moves is another item on OSM's block list, and the replacement
providers are under no more obligation to tolerate it than OSM was.
The basemap preference key is untouched, so an operator who had picked "Street"
still gets a street map, from the new source.
A new county and a new park both drew --success, so the two were the same green
in the cluster list and, since the band map copied it, in the band map too.
County takes violet; POTA keeps green, the association being worth something.
Violet is a chart hue rather than a semantic token because every token was
already spoken for: red, orange and yellow are the entity statuses, blue is
"callsign already worked", green is now POTA. It is defined in both the light
and dark chart groups, so all eleven themes have it.
Changed in BOTH panels at once — separating them in one place would have left
the two views contradicting each other about the same fact, which is worse than
sharing a colour. To make that impossible to get wrong again, the marker
definitions move to lib/spotMarkers: key, colour and label in one table that the
cluster list and the band map both read. That table is what a per-marker colour
setting will drive, which is why it is a table and not three constants.
The band map deliberately shows three of the four markers. A new PREFIX stays a
cluster-list badge: the pill is 22 px tall and a fourth segment on its strip
turns it into a colour code nobody reads at a glance.
The band map coloured only the entity status and dropped the other three
markers on the floor — not for want of data: they were already in the status
entry it receives, the local type simply never declared them, so a new park on
an entity you have worked was indistinguishable from any other worked spot.
Colours are the DX-cluster list's, taken from it rather than chosen: --info for
a worked callsign, --success for a new county and a new park. The same fact must
not be blue in one panel and green in the next.
They STACK, they do not replace. These markers are orthogonal to the entity
status — worked entity plus new park is an ordinary combination — and the
cluster list already spells them out side by side rather than letting one win.
So the pill keeps the status colour and the markers take its left strip, split
into one segment each. That strip previously repeated the pill's own colour and
carried no information at all, which is what made it the right place.
Note that a new county and a new park share --success in the cluster list, so
they share it here too. Splitting them means changing both views together, which
is the moment to do it — when the per-marker colour setting arrives.
The input was bound straight to the number and committed on every keystroke, so
it could not be EMPTIED: clearing it yields "", Number("") is 0, 0 fails the
"> 0" guard, the state never moved, and value={qsoLimit} snapped the old figure
straight back. Going from 200000 down to 100 was a fight against the field — and
read as the setting refusing to persist, which it was not: writeUiPref stores it
and syncPortablePrefs makes the database authoritative at boot. It also wrote
the preference once per keystroke (1, then 10, then 100).
Raw text in local state, committed on blur or Enter, per the controlled-input
note in CLAUDE.md.
Opening the Awards panel pulls every QSO into a cached slice and kept it for the rest of the session: the cache was only ever invalidated by a logbook change, never by disuse. Each cached QSO is a 1896-byte struct plus its strings AND a decoded map of its ADIF extras - one map allocation per QSO. Harmless at 30k rows, several hundred megabytes at 132k, which is where it was reported. A janitor drops it after 15 minutes without a reader and calls FreeOSMemory, because Go hands pages back lazily and the whole point is that the operator sees the memory return. 15 minutes is deliberately generous: an awards session recomputes every few seconds and re-pulling a large remote logbook costs seconds. The heap size is now logged when the snapshot is built and when it is released - a memory report was unanswerable without a number.
Every line of cluster traffic went straight to state: a spread copy of a 2000-element array, a slice copy, and a React render, per line. An RBN feed sends hundreds a second, so that was tens of MB/s of garbage in the renderer and hundreds of re-renders - and it ran whether the console was open or not, so the cost was paid for a panel nobody was looking at. Reported on an old PC where the UI had stopped responding. Lines are staged in a ref and flushed on a 200 ms timer, the same shape the spot handler already uses. The staging buffer is bounded too, so a burst longer than the console can show is not carried in full just to be sliced away on commit.
Several operators hit "stage current exe: rename …\OpsLog.exe …\OpsLog.exe.old:
Accès refusé" and could not update again. Two separate causes, both ours to
handle.
The staging name was fixed. os.Rename replaces its target, so a single leftover
".old" that could not be deleted — a scanner holding it open is the usual
reason, and the pre-existing os.Remove was best-effort and ignored — made every
later update fail with that error, permanently, recoverable only by deleting the
file by hand. Staging now uses a unique ".old-<nanos>", which no leftover can
block, and the startup cleanup sweeps the pattern instead of one name.
Renaming a running image is legal on Windows, but some endpoint protection
(Bitdefender's ransomware remediation among them) blocks it outright, and no
retry gets past that. So the swap is deferred: the new build is parked beside
the old one and a detached helper moves it into place after this process exits,
when the file is no longer a running image. It keeps trying for ten seconds,
since a scanner tends to let go a beat after the process dies.
A short retry stays in front of both, for the ordinary case of a scanner holding
the file it has just watched being written.
If even the deferred move fails, OpsLog restarts on the CURRENT version rather
than leaving the operator with nothing — someone mid-QSO losing their logger is
worse than an update that waits — and only a successful swap passes
--post-update, so the download survives for the next attempt instead of being
swept by the cleanup.
Both band maps were pinned to a hardcoded width — 300px docked beside the
tables, 260px per card in the Band map tab. On a busy band the map could not be
given more room, and on a quiet one the log could not take it back.
The docked map becomes a resizable grid column with the grip in the gap between
the panes, so the handle costs no space; the tab cards share one width with the
grip on their right edge. Side-by-side columns of different widths read as a
mistake rather than a choice, which is why the tab has one width and not one
per card. Double-click either grip to return to the default.
The drag measures from the pointer's START position rather than the container,
so the same helper serves both edges — the docked map sits on the left or the
right depending on the operator's setting, and the grip is on its inner edge
either way. Pointer capture, like the main splitter: without it the map or the
grid under the cursor swallows the moves.
Both widths are persisted through writeUiPref and registered as portable, so
they travel with the data folder like the main splitter and the rest of the
layout.
The seven existing themes are warm beige, cool grey, sage grey, slate, warm
dark, graphite and black — every one neutral, six of the seven accented orange.
Picking a theme changed the shade of grey and little else.
These colour the SURFACES, not just the accent, and each takes a different
primary so the picker tells them apart at a glance: violet on deep indigo, cyan
on deep teal, magenta on aubergine, indigo on a crisp cool white.
The semantic colours stay recognisable as themselves. A logger is read for
hours and "red means a problem" cannot become a decorative choice, so the hue
budget went on the surfaces and the primary. Where a theme's primary would have
collided with a meaning, the MEANING kept its identity and the ornament moved:
Ocean's info is blue rather than cyan, Plum's danger is red rather than rose,
and the matrix entity ramp shifts to cyan under Indigo and teal under Nordic so
a "confirmed" cell never reads as a button.
Everything else follows for free: the grids resolve var(--…) at runtime, so
they re-skin with no re-render. The three shared registrations that do NOT
follow automatically are done — the chart palette (light vs dark ramp) and the
date-picker icon inversion, which is keyed on an explicit list of dark themes.
Prefix exists so a field holding a bare value counts for an award whose codes
carry a letter: a French operator writes "74" in STATE, DDFM's codes are "D74".
Def's own doc says exactly that. But searchOne applied the prefix AFTER looking
the token up in the reference list — after the step that had just failed — so
the bare form matched nothing and the prefix decorated an empty result. The only
way through was a regex, in a mode where the operator had chosen "code" and
explicitly not "pattern".
The token lookup now tries the prefixed form when the bare one is not a known
reference. The list stays the authority: an unknown number still matches
nothing, so the prefix completes references rather than inventing them.
Second bug in the same pass: the blanket prefix also hit codes that came
straight OUT of the list, so a field already holding "D74" produced "DD74" as
soon as a prefix was configured — for both the token lookup and the
description matcher. Branches that yield whole codes are now excluded from it;
the ones that yield a raw capture (regex, whole-field split) still get it.
A forwarder (W&P, seen in the field in front of MSHV) prepends the origin as
plain text before re-broadcasting:
"127.0.0.1:2237|" + <the original, untouched WSJT-X packet>
That puts the magic 15 bytes in, so every datagram failed on "bad magic
0x3132372e" — those four bytes being ASCII "127." — and an operator running
MSHV behind the relay saw no decodes, no callsigns and no auto-logged QSOs.
No new service type: what follows the header IS a WSJT-X packet, so the parser
and everything downstream apply unchanged, and a separate type would duplicate
decode, status and logged-ADIF handling to strip 15 bytes. ParseWSJT skips the
header instead, which also covers any other relay that wraps traffic this way.
The match is deliberately narrow — the magic must fall within the first 64
bytes AND every byte before it must be printable ASCII. A corrupt or truncated
packet that merely contains those four bytes somewhere is not resurrected into
a QSO; it fails exactly as it did before.
Test data is the real captured datagram, header included.
"WSJT parse error: bad magic 0x3132372e" named neither the sender nor the
payload, so there was nothing to act on — even though those four bytes are
ASCII "127.", i.e. some program broadcasting an address on a port expecting
WSJT-X binary.
The line now carries the remote address, the size, a printable preview and a
hex dump of the first 96 bytes. Text senders are readable at a glance; a
genuinely binary payload still shows its bytes.
And it stops after five. The reported case wrote that line about 150 times a
second: a permanently misconfigured port would fill the 10 MB rotating log with
one repeated sentence and bury every other piece of evidence — the log's whole
purpose. The fifth line names the two things worth checking, the sender and the
service type.
N1MM's parse error goes through the same path; it had no packet detail either.
The award editor offered five confirmation sources and Def's own doc comment
named five, but confirmed() had cases for three. "qrzcom" and "custom" fell
through the switch, so ticking either marked nothing as confirmed — the exact
failure the GrantCodes comment in this struct warns about: a checkbox that does
nothing is worse than no checkbox, because it is trusted.
QRZ.com reads qrzcom_qso_download_status, not the upload one: uploading a QSO
is us telling QRZ about it, which confirms nothing.
Custom names a field. Rather than a checkbox per external source, the Def gains
ConfirmField + ConfirmValue: any QSO field or ADIF extras key, and optionally
the comma-separated values that count. That one shape covers the three cases
asked for — the OpsLog card marker (APP_OPSLOG_QSL_RCVD), an arbitrary ADIF
tag, and a tag stamped by an imported club list — because all three end up as a
field on the QSO.
An empty ConfirmValue means any non-empty content confirms: the OpsLog marker
stores the date the card arrived, not a Y/N flag. A custom source naming NO
field confirms nothing, deliberately — the opposite default would silently mark
a whole logbook confirmed.
Windows parks a minimised window at -32000,-32000 with a stub size and reports
it as not-maximised, so closing OpsLog from the taskbar while minimised stored
exactly that. Seen in a log: "window: saving -32000,-32000 237x39
maximised=false".
The restore side already rejects both the impossible corner and the
below-minimum size, so nothing opened off-screen — but it fell back to the
default placement, and the operator silently lost the size, the position and the
maximised state they had set.
saveWindowState now keeps what was already stored when the window is minimised.
Two tests for that: the Wails flag, and the -32000 corner, because a window that
is mid-close can sit at that corner with the flag already cleared. A poisoned
window.json repairs itself on the next close of an on-screen window.
The stamp answers "PSE QSL", so it reads as its counterpart or not at all: a
bare "TNX" next to a QSL message says thanks for something unnamed.
Changed at the source of the {qso.pse_tnx} token, so every card picks it up
with no template edit. The live indicator in the QSO editor and the hint beside
it follow.
The previous commit added its row to the status table but left the picker
listing only the column-backed channels, so the one channel you could actually
tick was the one you could not select.
Selecting it swaps the left panel for an OpsLog editor instead of the generic
sent/received/date grid, which has no fields to bind to here: Sent is shown
read-only (OpsLog stamps it when the card goes out), Received is the tick, and
the PSE QSL / TNX indicator sits next to them since that stamp is what the flag
is for.
The OpsLog QSL marker sat under QSL Msg in "Contact's details", nowhere near
the channel it belongs to, while the QSL Info tab listed every other
confirmation — QSL, LoTW, eQSL, QRZ.com, Club Log, HRDLog — with a Sent and a
Received column.
It now has its own row in that table and the "QSL received" tick moved to the
same tab, keeping the PSE QSL / TNX indicator that is the reason the flag
exists at all: received prints TNX on the card, otherwise PSE QSL.
Written by hand rather than added to CONFIRMATIONS: that table maps QSO
columns, and this channel is backed by ADIF extras (APP_OPSLOG_QSL_RCVD, plus
the older APP_OPSLOG_QSL_CARD_SENT for the sent side). Sent stays read-only —
OpsLog stamps it when a card actually goes out, and a hand tick would record
something that never happened.
Typing RK3DWA found nothing while RK3DWA/3 found 21 QSOs, so a station's
history was only visible if you happened to type the exact form it had been
logged under — and an operator who worked it as /0, /P or /MM saw none of it.
The other RDA tools and Log4OM fold these together; this does too.
The predicate strips the suffix from what was typed and matches "call = base OR
call LIKE base/%", so it works from either end: the base call finds the portable
QSOs and a portable call finds the plain ones. Deliberately not a bare prefix
LIKE 'RK3DWA%', which would also match RK3DWAB — a different station. The '/' is
what makes it the same operator.
Settings -> General to turn it off. Default ON, hence the inverted storage: an
existing install has no key, and reading that as OFF would leave everyone with
the behaviour we were asked to change.
Contest dupe checking is untouched — it runs through ContestDupe, a separate
binding, and stays an exact match as a contest requires.
Settings -> DX Cluster: a toggle and an interval. When it is on, logging a QSO
announces the station on the master cluster — the QSO's own station callsign as
the DX, on the frequency the contact was made on — so callers find the run
without waiting for someone else to spot it. The node fills the DE field from
the login, so the spotter is us too: a self-spot.
It fires on the FIRST QSO of a frequency and then at most once per interval.
Both halves matter: announcing every QSO would flood the node and get the
station filtered out, while a pure timer would stay silent for minutes after a
band change. A drift of up to 500 Hz still counts as the same run, so nudging
the VFO mid-pileup does not re-announce.
Five minutes is the floor, clamped in SaveSelfSpotSettings as well as in the
input: the limit protects the node from us, so it must not depend on the
frontend. The interval input keeps raw text and clamps on blur — clamping per
keystroke rewrote "10" to "5" as soon as the "1" landed.
Wired into both log paths (manual entry and UDP auto-log) on the async side, so
a cluster that is slow or down never holds up logging. A send failure restores
the previous throttle state, so the next QSO retries instead of sitting out an
interval that produced no spot.
The Kenwood/Elecraft backend deliberately suspends its wire poll while PTT is
held — a K3 answers "?;" to IF; during transmit, and treating that as a fault
used to drop the whole CAT link. The consequence was that nothing watched the
transmitter: a client that crashed, was closed, or simply had its socket shut
under it left the rig on air.
And shutting the socket is routine. reloadCATShare tears the sharing server
down and rebuilds it on every settings save, so a Save while WSJT-X held PTT
was enough. A K3 operator's log shows exactly that: "TX;" at 17:53:09, no "RX;"
ever, the poll silent, and the rig still keyed 29 s later when the CAT link
happened to be rebuilt.
The server now drops PTT when a connection ends and when Stop() is called.
Stop() runs before reloadCAT restarts the backend, so the unkey still reaches
the radio. An atomic Swap keeps it once-only across the two paths.
MD6 is FSK on a Kenwood and DATA on a K3/K4. kenwoodModeToADIF decoded the
digit unconditionally as RTTY, so OpsLog contradicted the mode it had just
set: SetMode writes MD6 for a digital mode on an Elecraft, then ReadState
parsed the IF frame back as RTTY.
A K3 running FT8 therefore showed RTTY in the status bar, logged its QSOs on
RTTY, and — through the shared CAT server — told WSJT-X/JTDX the rig sat in
RTTY while they had just asked for a data mode. Found in a K3 operator's log:
every cat:state line read mode=RTTY on 21.074 FT8, two lines after OpsLog's
own "MD6;".
The digit now resolves to the configured digital mode whenever MD6 means DATA
on this rig — the Elecraft backend, and the "DATA A - MD6" data-mode option
that exists for it. A plain Kenwood still reads MD6 as RTTY.
One station reaches the log SHOUTED by QRZ, lower-cased by a hurried
operator and in whatever case an imported ADIF carried, so the same name
appears three ways across a log. Name and QTH are now title-cased word by
word; Comment and Note only get a capital first letter, because the rest
routinely holds callsigns and modes ("TNX QSO F5ABC, FT8 59") that
lower-casing would destroy.
Normalised on blur, never per keystroke — rewriting the value mid-word
fights the typist (the controlled-input trap in CLAUDE.md). Applied in
both entry layouts and in the QSO editor, since leaving the editor alone
would just reintroduce the mess on the first correction.
The Recent QSOs box only ever searches callsigns, so it upper-cases as
you type and carries an inline clear button.
Award columns were filtered out of the persisted column state entirely,
so applyColumnState({applyOrder:true}) had nowhere to put them and AG
Grid appended what the state does not mention. A column the operator
dragged to sit AFTER an award column therefore jumped back in front of
it on every reload.
Only the `hide` property is dropped now. Visibility stays owned by the
awardShown code-set — that is what stopped shown award columns from
vanishing on an awardCols rebuild — while the position survives the
round-trip.
The Settings -> Database panel had lost every action on the settings
database except "Open folder", which launches Windows Explorer — where a
.db cannot be selected at all, so double-clicking it only asks which
program should open it. The handlers were still there, just no longer
rendered; New / Open existing / Save a copy / Rename / Reset are back,
along with the restart banner (the DB pointer is only read at startup).
The backup block put the rotation field and three checkboxes on one flex
row inside max-w-xl. That fitted until "keep every backup" arrived with
its two-line hint, after which every label wrapped into an unreadable
sliver. The options are stacked now, and keep-all sits under the option
it depends on.
Stale since the 0.23.9 work: SetOpsLogQSLReceived, the backup keep_all
flag, the award PRIMARY search prefix and the default QSL message were
all missing from the generated bindings.
The Elecraft handling was buried in the Kenwood "data mode" option, so users
couldn't find it. Add "Elecraft K3/K4" as its own entry in the CAT backend
selector. It reuses the Kenwood-dialect transport and the Kenwood USB/network
settings (the K3 emulates the Kenwood command set — a separate transport would be
a near-total duplicate), with an `elecraft` flag on the client that forces
digital modes to DATA A (MD6+DT0) — no data-mode dropdown needed. RigState still
reports Backend "kenwood" so the CW-over-CAT keyer capability keeps working.
Integrates the fix/jtdx-kenwood-tx work onto main: the Kenwood/Elecraft backend's
"data" data-mode path now sends MD6 then DT0 so a K3/K4 lands in DATA A — the
audio sub-mode FT8 uses. MD6 alone could leave the rig in an FSK/PSK sub-mode
where it keyed but the rear sound-card audio never modulated ("transmits but
nothing comes out"). The data-mode option is relabelled "DATA A — MD6+DT0
(Elecraft K3/K4)". A shared Kenwood/Elecraft backend is kept (the K3 emulates the
Kenwood dialect); the Elecraft specifics live behind the data-mode seam rather
than a 7000-line duplicate backend.
Only OrRule had a Prefix ("prepended to each found reference", e.g. postal 74 →
D74); the primary search passed "" for prefix, so an award whose references are
D01/D02… couldn't prepend the D on its main rule — only on OR fallbacks. Add
Def.Prefix, pass it into the primary run(), and expose a Prefix input next to the
primary Leading/Trailing fields in the award editor.
Sub-option of "back up on every exit": when on, backup.Run/RunADIF stamp the
filename with the time (opslog-YYYY-MM-DD-HHMMSS.db) so each run is a distinct
file instead of overwriting the day's snapshot. Rotation still trims to the
newest N (raise Rotation for a longer history). Threaded as a `unique` flag
through runConfiguredBackup/backupLogADIF from BackupSettings.KeepAll
(keyBackupKeepAll); UI sub-checkbox shown only when EveryExit is on.
autoLogFromUDP (the inbound WSJT-X/JTDX/MSHV log path) inserted the QSO but,
unlike the manual LogQSO path, never called a.udp.EmitLoggedADIF — so a QSO
received from MSHV was not re-emitted to the outbound ADIF listeners (Log4OM,
N1MM, GridTracker…). Emit it in the same async goroutine. A pathological
self-loop is broken by the existing ±2-min dedup, which returns before the emit.
Received flag: new APP_OPSLOG_QSL_RCVD extra, toggled on the QSO edit window
next to QSL Message (immediate targeted write via SetOpsLogQSLReceived so it sets
AND clears reliably), plus a live PSE QSL / TNX indicator and a Recent-QSOs
column mirroring the sent one.
PSE/TNX card stamp: automatic — received → TNX, otherwise PSE QSL — exposed as
the {qso.pse_tnx} token (added to qslVars, so preview and send agree) and placed
in the default QSO-box footer; it can be moved to its own element in the designer.
Default QSL message: new qsl.default_message (QSLEmailTemplates.DefaultMessage),
edited under Settings → E-mail → QSL card e-mail. qslVars falls back to it when
the QSO's own QSLMSG is empty, so a per-QSO message always wins. Single choke
point covers both live preview and send.
New BackupSettings.EveryExit (keyBackupEveryExit) with a Settings → Backup
checkbox. When on, the three shutdown-backup gates (plannedShutdownSteps,
runBackupForShutdown, maybeShutdownBackup) bypass the HasBackupToday /
HasADIFBackupToday "already done today" check and always run, so a second
session's QSOs are captured instead of skipped until the next day. The dated
backup file for today is refreshed (overwritten) each exit; prior days are still
kept by rotation.
spotStatus updates ~20x/second under an RBN firehose; firing a full refreshCells
on each was pure churn on a slow PC. Coalesce into one refresh per 200ms. Keeps
the NEW/WORKED badge colours and the 'represents nothing' dimming — the dimming
itself is a trivial pure read now that the worked-index scan is cached, so it's
no longer a cost worth removing.
A user on a slow PC with RBN saw OpsLog at 94% CPU and 8.5 GB RAM (Logger32: 3%
/ 34 MB on the same feeds). Two runaway costs under the spot firehose:
- ClusterSpotStatuses re-scanned the ENTIRE logbook (5-6 full-table maps) on
every 50 ms spot batch — ~20×/second — its "one scan regardless of batch" doc
was untrue. On a big log that's millions of row-scans/second → the pegged CPU.
Now cached in clusterStatusCache (an immutable snapshot), rebuilt only when the
logbook changes (noteWorked on a single log, invalidateAwardStats on bulk), so
it's one scan per logged QSO instead of per batch.
- The frontend spotStatus map had no cap: one entry per call|band|mode ever seen,
and RBN produces thousands of unique calls/hour → unbounded growth to GBs, plus
a full {...prev} copy 20×/second. Now pruned back to the live (SPOTS_CAP=1000)
spots once it drifts past 2×, with a cheap same-reference bail-out otherwise.
Since the logbook was split into its own SQLite file, backup.Run was still
snapshotting a.db (the settings/config database) — so the scheduled and manual
backups silently stopped including the QSOs. The operator was backing up config
and thinking it was their log.
Track the resolved logbook file path (a.logDbPath, set in connectLogbook) and
route the backup through a new runConfiguredBackup: the CONTACTS become the
primary "opslog-*" backup (the logbook file on SQLite, an ADIF export on MySQL),
and the settings/config db is snapshotted separately as "opslogcfg-*" so nothing
is lost. backup.Run takes a name prefix; the two sets rotate independently.
The recording e-mail's subject and body round-tripped through EmailSettings and
the backend (keyEmailSubject/keyEmailBody) but had no UI editor — only the QSL
card e-mail did, so the recording mail was stuck on the default text. Add the
same subject/body editor to the E-mail panel, above the QSL block, sharing the
{CALL}/{DATE}/{BAND}/{MODE}/{MYCALL} template variables.
A K3 (and other Kenwood-dialect rigs) answers "?;" to an IF; status poll while
it is transmitting. ask() latched that as "IF unsupported" and ReadState
returned an error, so the Manager dropped the whole CAT link mid-transmit — the
log shows connected=false / "IF rejected" the instant JTDX sent TX;. That tore
down the shared rigctld CAT the digital-mode software keys through.
Two fixes: (1) track PTT (SetPTT) and skip the wire poll while it's held —
ReadState hands back the cached last state, capped at 30 s so a missed
SetPTT(false) can't freeze it; (2) never latch IF/ID as unsupported — those are
universal on Kenwood/Elecraft, so a "?;" to them is a transient busy (mid-TX),
not a missing command.
The Row helper was defined inside AwardsSelectionPanel, so each render minted a
new component type and React remounted the whole list on every click — flicker,
and clicks lost between mousedown and click (hence needing many tries). Inline
the row markup so the keyed buttons keep a stable identity.
New per-profile "tracked awards" selection: Settings → Awards (under User
configuration) is a two-column transfer list — every defined award on the left,
the ones you follow on the right, click to move either way. The Awards tab's
list is narrowed to the followed set; an empty set means "show them all" so the
tab is never blank.
Backend: app_awards_tracked.go adds keyAwardsTracked (per-profile JSON array of
award codes) with GetTrackedAwards/SaveTrackedAwards; saving emits
awards:tracked-changed so the Awards tab re-filters live. Award definitions stay
global — only the follow selection is per profile.
The follow filter correctly leaves the antenna put on an un-ticked band, but the
TX-inhibit loop ran independently: it gagged FlexRadio transmit whenever the
antenna reported "moving" (a stray polled Moving flag, or a move finishing from
the previous band), even on a band OpsLog was deliberately not managing. An
operator working 15 m on another antenna had their FT8 cut mid-transmission by
an "antenna moving" interlock.
motorTXInhibitLoop now takes the covered-band set and forces moving=false when
the current rig band isn't in it — un-ticking a band means hands off entirely:
no tune AND no inhibit. Also correct the RCU-01 references to RCU-06 (the actual
controller behind the short 11-byte status frame).
A contiguous FreqMin/FreqMax range can't drop a single band while keeping its
neighbours — so it couldn't express "40 m and 20 m yes, 30 m no" for an antenna
whose 30 m extension isn't fitted. Replace it with a Covered bands selector
(40 m–6 m) that applies to both the Ultrabeam and the SteppIR.
The follow loop and immediate re-tune now gate on band membership
(motorBandAllowed) instead of a MHz range: only bands in the operator's set are
followed, and 80 m/160 m are never followed (outside a beam's reach) regardless.
The legacy FreqMin/FreqMax is migrated to a band set on load and still
round-trips, so existing configs are unchanged. UI swaps the two number inputs
for a row of toggle chips.
The follow logic skips out-of-range frequencies only when it knows BOTH edges
of the antenna's range. The Ultrabeam path took those edges from the
controller's status frame, but an RCU-01 behind an RS232-to-Ethernet bridge
answers with the short frame that omits FreqMax → the guard saw FreqMax==0 and
disabled itself → OpsLog forwarded every rig frequency, so 80 m (un-tunable)
reached the controller, which clamped the elements down to its lowest band
(~30 m).
ubAdapter now carries the operator-configured FreqMin/FreqMax from Settings →
Antenna (like steppirAdapter already did) and uses them as a hard floor/ceiling,
falling back to the controller-reported bound only when one is left unset. This
guarantees both bounds are present so the guard actually engages. The Tunable
range editor in the Antenna panel, previously SteppIR-only, is now shown for the
Ultrabeam too, with a generic bilingual hint.
New internal/rotator/dcu1 client speaking the DCU-1 command set (AP1nnn / AM1
to go-to, AI1 to read bearing), over a serial COM port (4800 baud default) or a
raw TCP serial-over-IP bridge. Azimuth-only; Stop re-commands the current
bearing since the base set has no stop opcode.
Wired through app.go (normRotorType, default port 4001, GoTo/Heading/Stop/test
switches, park returns the same "PstRotator only" note as ARCO) and exposed in
the rotor settings as "Hy-Gain DCU-1 (RotorCard DXA, Rotor-EZ, Green Heron)"
with the ARCO's serial/TCP transport chooser and a bilingual hint. Protocol is
implemented from the standard DCU-1 spec and still needs field confirmation
against a real RotorCard DXA.
The pop-up listing the QSOs behind a Stats band/mode square gets a cleaner header
(band chip, hint), banded/hover rows and roomier spacing. Callsigns are now
clickable and open the QSO editor (App.openEdit threaded through DetailsPanel →
BandSlotGrid → SlotQSOModal as onEditQso), matching the double-click-to-edit of
the log grids.
Network relay boards (WebSwitch/KMTronic/Dingtian) were always reached at
http://<host>, so a board on the LAN behind a reverse proxy — the common remote
setup where the operator's 80/443 already go to Nginx Proxy Manager — couldn't be
reached from outside. relayBase() now keeps a scheme the operator supplies
(https://relay.example.com, optional sub-path) and only defaults to http:// for a
bare host/host:port. UI hint + test added.
setSpotStatus({}) blanked EVERY spot's status, so logging one contact wiped all
the other NEW/NEW-BAND badges and they popped back a moment later. Instead,
re-fetch the shown spots and OVERWRITE each status in place (merge) — the just-
worked spot updates while the rest stay put, no flicker. Same visibility gate
and 2 s debounce; backend cost is one status batch (one logbook scan) as before.
A just-worked call/entity/POTA/prefix/slot kept its NEW badge because per-spot
status is cached and never expires. Drop the cache on qso:logged so the grid
re-evaluates. Kept cheap for old machines / big logbooks: only fires while the
cluster is on screen (a log while hidden marks it dirty and refreshes once on
next open), and debounced 2 s so a fast run coalesces instead of re-scanning the
logbook per QSO.
The "dim spots that represent nothing" feature drove the dimming with getRowStyle
and called api.redrawRows() on EVERY spotStatus update to re-apply it — a full
grid re-render on a firehose of status updates, which pegged the CPU on a busy
cluster (reported as PCs slowing down since 0.23.6). Move the dimming to a
defaultColDef cellClassRules ('opacity-40') that the existing light
refreshCells({force:true}) re-applies; drop redrawRows and getRowStyle.
Confirmed live: "setup fanmode=CONTEST" is accepted (reply code 0) while the
bare "fanmode=CONTEST" we switched to earlier is rejected (0x50000015), so the
fan mode snapped back and never changed on the amp. Restored the "setup " verb
and now check the reply code, returning an error if the amp rejects the set.
The Ctrl+↑/↓ keyboard-nav useEffect sat AFTER `if (!range) return`, so on a band
with no range — e.g. a spurious 33 cm frequency briefly reported by an Icom over
CAT — the component returned before that hook and the hook count changed between
renders, crashing the whole window with React #300. Moved the effect above the
early return; lo/hi already fall back to [0,1] when there's no range.
An older controller behind an RS232-to-Ethernet bridge returns the status frame
without its trailing FreqMax byte (11 bytes, not 12). The packet checksum is
still valid, so it's a real short frame, not a fragment. queryStatus now accepts
>= 10 bytes and defaults the missing FreqMin/FreqMax rather than rejecting it as
"too short", which reconnect-looped the link. Whether the byte LAYOUT is
otherwise identical across controller versions still needs confirming from the
protocol reference.
A spot whose entity/band/mode is already worked, whose callsign isn't in the log,
and that carries no POTA/county/prefix novelty is dimmed (whole row, opacity) so
the eye skips it. "Nothing" follows the status, which already obeys the "same
slot" option and digital grouping. getRowStyle drives it; the spot-status effect
now redrawRows so the dimming re-applies when a status lands.
The WORKED filter also matched the entity status 'worked' (entity/band/mode
already worked, shown as a plain "—"), so selecting WORKED still listed spots of
calls never worked. Match the 'worked' key only via worked_call (the blue
WKD-CALL flag); a worked call still shows under NEW BAND / NEW SLOT.
The filter logic already matched worked_call spots against a 'worked' key; add
the chip to the status row so worked spots can be shown or isolated, not only
dropped via the "Hide worked" checkbox.
New DX Cluster setting (Settings → DX Cluster). Off (today's behaviour) a call
worked anywhere reads as already worked; on, the WORKED-call flag needs the same
band AND mode. It folds through the digital-mode grouping (Settings → General):
grouped, a 20m FT8 contact also marks a 20m FT4 spot worked; ungrouped they are
separate slots. Backend: qso.WorkedCallSlotKeys builds CALL|band and
CALL|band|mode(grouped) keys; ClusterSpotStatuses uses them when the option is on
(a spot with no inferable mode falls back to same-band).
Five operator-requested items:
- Alt+W clears the QSO entry. Handled before the `typing` guard and above
the keyer's key routing, so there is always one key that clears whatever
else is running — Esc is not that key when the CW keyer reserves it.
- The Grid box no longer pops outside the entry panel on a narrow window.
Row 2 needed 300+130+76+gaps = 538 px inside a panel whose min-width is
520, so Grid was pushed out and clipped at every narrow width, not just
extreme ones. QTH's min-width drops to 80 and the row wraps rather than
overflowing if it ever still can't fit.
- Selecting a QSO in the log now drives the Stats (F1) matrix. Uses its own
WorkedBefore call into separate state, NOT runWorkedBefore: that one owns
the entry form's wbRef and can trigger a field backfill, which browsing
the log must never do. The entry form wins whenever it holds a call.
- Clicking a coloured band/mode square lists the contacts behind it. Returns
the exact callsign AND the rest of the entity, because that pair is what
the cell's colour encodes; the call's own QSOs are bolded. The DXCC arm
matches the stored dxcc column only — reconstructing it from the callsign
here would disagree with the matrix above, which is built from that column.
- The Awards DXCC list shows each entity's primary prefix in its own sortable
column. Derived live from cty.dat into Ref.Group rather than stored on the
reference row, so an existing installation needs no re-seed.
Switching an SPE amp off left the already-open connection reading as connected
(the amp stops answering once off, but the stale open handle kept the status at
"connected"), so the panel never showed OFF and PowerOn — which only wakes a
DISCONNECTED amp, and must NOT pulse RTS/DTR on a running one — did nothing
until the operator restarted OpsLog. PowerOff now drops the connection and marks
the amp offline, exactly as a fresh start would; the poll's plain reopen never
wakes a switched-off amp, so it stays off until the ON button's RTS/DTR edge.
Two ways a bogus contact could be logged:
- Enter in the TX/RX frequency field bubbled to the entry container's
onKeyDown, which calls save() on Enter-in-an-input. Typing a frequency and
pressing Enter tuned the rig AND logged the QSO. The freq fields now
stopPropagation so Enter only tunes.
- WSJT-X/JTDX broadcasts its DX-call field over UDP; when it holds a bare
number ("1"), applyUdpCall dropped it into the callsign, and the next Enter
logged a QSO with callsign "1". Reject any DX call with no letter.
Removed the long explanatory hints under the Kenwood COM/baud/host, data-mode,
CAT-trace and PTT-hotkey fields; shortened the CAT-sharing hint to just the
client setup line; and replaced the DTR/RTS hint with a one-liner. Less clutter
in the CAT settings.
JTDX polls get_ptt DURING transmit to confirm the rig is keyed (WSJT-X does not,
which is why it worked and JTDX cut after ~2 s). We answered a blanket "0" (RX),
so JTDX concluded PTT had failed and dropped the over. get_ptt now echoes the
last PTT state commanded via set_ptt — always consistent with the client's own
command. Trace showed T 1 → three t→"0" polls → T 0.
The optimistic get_freq echo added in 0.23.5 (to stop the "Fake It" dial creep)
stopped JTDX/WSJT-X from going into transmit: the client reads frequency back
during its TX sequence, and echoing the commanded value instead of the rig's
live value interfered with it. No-TX is far worse than a 0.5 kHz drift, so revert
to reporting the rig's live frequency. get_freq/get_split_freq return s.rig.Freq()
again; the echo state, noteSetFreq, reportedFreq and the drift test are removed.
The antenna does NOT echo our sequence number — its replies carry their own
counter — so the previous seq-matching drained every reply and stalled status
updates. Revert to reading one reply per command, but flush any bytes left in
the stream before each command (drainStale): a reply left by a timed-out command
is discarded so the next read stays 1:1. readPacket also resyncs to the next STX.
This fixes the intermittent disconnects, phantom frequency jumps and wrong
element-length readings without depending on the seq.
Also: report motion for a short window right after a commanded move, so the
"moving" indicator and the Flex TX-inhibit fire the instant a band/pattern is
clicked instead of a poll (~2 s) later; the real motor state takes over once
polled.
1 MB (~6500 lines) fetched, split and re-rendered every second made the window
lag. 512 KB (~3200 lines) keeps a useful backlog — double the original 256 KB —
while staying responsive.
sendCommand discarded the reply's sequence number and accepted whatever frame
came back. On a slow remote link a command that timed out left its late reply in
the stream, and the NEXT command read it as its own — crossing STATUS with
READ_BANDS/PROGRESS. That surfaced as phantom frequency jumps (a spurious
follow-loop re-tune), intermittent "reply too short" disconnects, and wrong
element-length readings (READ_BANDS getting the status frame).
Now every reply is matched to its request seq and stale/malformed frames are
drained (bounded), with readPacket resyncing to the next STX. Tested with a
net.Pipe that injects a stale reply ahead of the real one.
The viewer fetched only the last 256 KB (~1600 lines) of opslog.log. During a
busy trace the oldest lines scrolled out of the window while the operator was
still reading them. Raise the tail to 1 MB (~6500 lines); the backend already
allows up to 4 MB.
Checking every profile made a denied call unrecoverable — the process quits
before the operator can change it, and no profile switch could get back in.
Gating on the active profile alone keeps "--profile <other>" as a way back in,
and still turns away a denied operator whose active call is the blocked one.
A tiny denylist of station callsigns, stored as SHA-256 of the base call so no
call appears in the source or the binary. Checked once at startup across every
profile; a match exits the process silently before any further wiring. Not in
the changelog.
A digital app (JTDX/WSJT-X) sharing OpsLog's rig in "Fake It" split follows
the dial by polling get_freq. Freq() is the last polled value and lags a
set_freq by a poll cycle, so right after a transmission the client read the
still-shifted frequency, mistook it for a manual QSY and adopted it — the dial
crept down every over and never came back. get_freq now echoes the last
commanded frequency until the rig confirms it (or a short deadline passes),
closing the race. Backend-agnostic, so it fixes every rig, not just Kenwood.
QRZ hands back the whole logbook; a confirmation download only acts on the
QSOs QRZ marks confirmed. The old "(of 7950 returned)" read as if QRZ had
returned only a fraction of a 26k-QSO logbook. Now it states "N of M are
confirmed — the rest aren't confirmed yet", so nothing looks lost.
The LoTW/QRZ confirmation download ran on the app-lifetime context, so it
kept going after the QSL Manager was closed and a new download didn't stop the
previous one. A slow QRZ sync therefore bled its "flag cleared" log into a
freshly started LoTW download — the operator saw QRZ activity during a LoTW
run. Now each download gets a cancellable context: starting one cancels the
previous, closing the window cancels it (CancelConfirmations), and the parse
loops bail promptly on cancel. (A LoTW http 503 is separate — ARRL's server.)
Changing Normal/180/bi over a remote connection reverted to the old pattern
after ~4s: the fixed 4s optimistic window expired while the motors were still
flipping the elements, and the lagging remote status poll then overrode the UI.
Hold the commanded direction while the motors are moving, then a short grace
window after they stop for the confirmation poll to land (SteppIR already used
a generous 45s hold, which is why only Ultrabeam-remote hit this).
The keyer panel synthesised a "connected" status for the icom/flex/yaesu
rig engines but not kenwood, so the Kenwood/Elecraft engine fell back to the
(disconnected) WinKeyer hardware status and always showed "not connected"
even though its CAT link was up.
No single Kenwood mode digit fits every rig for a data mode: a K3/K4 wants
DATA (MD6), a TS-590SG/TS-990S data mode is a USB modifier set on the rig,
and MD6 on a plain Kenwood is FSK/RTTY. So the operator now chooses in
Settings → CAT: USB (default), DATA (MD6), or leave the rig's mode unchanged.
Only soundcard/data modes (FT8, PSK, JT…) follow this; RTTY/FSK stays its own
native mode, and voice/CW are untouched (isKenwoodDataMode + test).
The frontend engine loader had no "kenwood" case, so a saved engine of
"kenwood" fell through to the WinKeyer default — the Kenwood/Elecraft
CW-over-CAT engine never ran even though the setting showed it selected.
Also clarify the Kenwood data-mode comment: there is no single mode digit
right for both a K3 (DATA=MD6) and a TS-590SG (MD6=FSK), so data stays on
USB as the safe common base; a rig-specific DATA mode is a follow-up.
- PGXL: remote AUTH password; direct-link meter fallback (VITA-first) with
250 ms poll + peak-hold; fan mode uses bare "fanmode=" (was ignored).
- TCI: spot colour sent as decimal ARGB (ExpertSDR dropped the hex string);
spot click handled via CLICKED_ON_SPOT / RX_CLICKED_ON_SPOT to fill the entry.
- FlexRadio: clicking an OpsLog spot on the panadapter now applies the mode too.
- Filter presets: the trash icon deletes again (Radix pointer-down intercept).
- PTT hotkey: keys over CAT when a backend is active (was hijacked by a stale
Audio-tab RTS/DTR serial setting); AltGr keys allowed; presses logged.
- CW macro <LOGQSO>: logs after the preceding CW is sent, not on the first letter.
Settings -> Rotator is now a list like the amplifiers: add/remove rotors,
mix PstRotator / Rotator Genius / ARCO, and a Rotator Genius can drive both
its ports (one entry = two rotors). The compass gains a rotor selector, and
a per-rotor motorized-antenna flag so the boom/pattern paths show only for
the rotor carrying the Ultrabeam/SteppIR.
Also in this batch: a keyboard PTT hotkey (hold-to-talk or toggle, reusing
the audio PTT method with a CAT fallback), and TCI spot push to the
panadapter with click-to-fill of the callsign.
The 0.22.7/0.22.8 breakage came from a Xiegu-reported fault being fixed
inside the Yaesu and Kenwood backends. The lesson is structural: nothing
a backend does may be steered by another backend's report or another
backend's setting.
The lower-lines checkbox introduced for 0.22.9 was still one shared key
applied to both Yaesu and Kenwood — the same reflex in miniature. It is
now cat.yaesu.low_dtr_rts and cat.kenwood.low_dtr_rts, each checkbox in
its backend's own settings block, each wired only to its backend. The
shared key never shipped, so nothing needs migrating.
The rig-panel sliders take a non-passive wheel listener and acted on
hover, so scrolling the panel fed the gesture to whichever control sat
under the cursor — RF power included, mid-QSO. An operator scrolling to
read the bottom of the panel could change his transmit power without
touching anything.
They now require focus, i.e. a deliberate click on that slider first.
Without it the event is left alone and the panel scrolls, which is what
the operator meant. Same fix on the Flex and Yaesu panels, which share
the pattern.
This is a fault in its own right. It is NOT an explanation for the
"transmit power moves since 0.22.8" report: that release changed nothing
in the Icom path, and this behaviour is older than it.
Both defaults break someone. 0.22.7 lowered the lines on Yaesu and
Kenwood and silenced a TS-990 whose interface needs RTS raised to
transmit; 0.22.8 stopped lowering them and an FT8 station reported its
output power wandering, cured by closing OpsLog — Windows raises both on
open and the interface reads one as PTT.
So it is now a checkbox on those two backends, off by default: that is
how they behaved before any of this. Xiegu keeps lowering them
unconditionally (its own report, its own explicit keying setting), and
the CI-V backend keeps the behaviour it has always had.
Selecting CW showed FT8 and SSB contacts. The filter ran client-side over
the reference list and nothing else, so a reference kept because it had
one CW contact still displayed the band cells it earned on SSB, still
counted its SSB confirmations, and opening it listed every contact
regardless of mode.
The class now narrows the log BEFORE the engine runs, so the bands, the
totals and the confirmations all describe the selected mode, and
AwardCellQSOs takes the same class. The panel cache is keyed by
"CODE|MODECLASS" — one key per award showed the previous mode's numbers
after switching.
The snapshot is cached by log revision, so this costs a matching pass and
no database read.
Reported by one in the field. Kenwood's documentation gives 024 for that
radio; both are mapped rather than betting on which is right, and the log
no longer reads "Kenwood (022)".
d4bfd30 fixed a Xiegu G90 behind a DE-19 that keyed on connect, and I
applied the same line-lowering to Kenwood and Yaesu with no evidence that
either needed it. It silenced them: many USB-serial interfaces will not
transmit with RTS low — hardware flow control, or an output stage the
line enables. A TS-990 on COM3 opened cleanly and answered nothing, which
reached us as "CAT stopped working after the update".
Xiegu keeps the behaviour: that is where the fault was reported, and that
backend now has an explicit setting for which line keys the rig. The CI-V
backend keeps its own, which predates all of this.
Kenwood also distinguishes a silent port from one sending data that never
answers, and quotes what arrived — "the rig is not answering" sent an
operator checking the power switch on a radio whose frames were visibly
on the wire.
A TS-480 connected, answered ID, then never reported a state:
discarding " 000000000010000000;" while waiting for IF
connected on COM3 @ 115200 baud
cat:state → connected=false freq=0
A headless frame tail can only come from bytes that were thrown away.
ask() buffered into a local slice, so everything left after the matched
frame — including the head of a frame still arriving — went with it, and
the link could sit one answer behind its questions. The buffer now lives
on the backend, and Connect drains whatever the rig said before AI0 took
effect rather than inheriting it.
Also: the cat:state diagnostic prints RigState.Error. It read
"connected=false freq=0" and said nothing about why, so every report of a
CAT failure arrived without the one fact that explains it — the reason
only ever reached a tooltip.
New work moves to 0.22.8; 0.22.7 is released.
The Main tab was a fixed 50/50 grid, but a map and a cluster list want
very different widths and which one deserves the room changes with what
the operator is doing.
The share is clamped to 15..85: a pane can be made small but never
dragged out of existence, because recovering from that would mean
grabbing a divider no longer on screen. Double-click restores even.
The drag uses pointer capture on the divider — without it Leaflet
swallows the moves as soon as the cursor crosses the map. Persisted
through writeUiPref, so it travels with data/ like the other UI prefs.
The cluster already answers this question for a spot; the entry panel did
not answer it for the station actually being worked. A county hunter
needs to know before the QSO ends — by the time the county shows up in
the awards table the station is long gone.
qso.CountyWorked asks about one county instead of loading the whole
worked set, narrowed to the state so it stays a few dozen rows on a
remote MySQL. It compares through award.USCountyKey on both sides rather
than in SQL, because logged spellings vary ("Los Angeles" against
"LOS ANGELES, CA") and that key function is what resolves it everywhere
else.
An empty or non-US county is not "new" but unknown, and shows nothing: a
badge on a guess is worse than no badge.
Radix gives the content element focus when a tab is selected, so the
focus-visible ring fired on an ordinary click and outlined the whole
lower pane — visible as an orange rule under the tab strip. A tab panel
is a container; the controls inside it carry their own focus styling.
At two grid columns the labels wrapped onto a second line, which set the
row height for everything beside them. The unit is obvious from the
neighbouring DXCC # field and the numbers themselves.
Six equal columns gave a US county name the same width as a two-digit CQ
zone: the county truncated while the zones, the prefix and the DXCC
number sat mostly empty. The row is twelve columns now, spans set per
field.
The bearing and the two distances are read-only and never exceed
"12345 km", so they move to a flex row at a fixed width and the address —
the one field here that was never wide enough — takes the remainder.
The award editor's Test tab matched DK2FJ on its OR rule "qth / exact"
and reported the two Aachen DOKs as ambiguous. The F3 panel, on the same
contact, showed nothing — it builds its own QSO for ComputeQSOAwardRefs
and that object carried address, state, county, zones and continent only.
Any award keyed on the town could not fire there, so live detection was
silently blind to a whole class of awards and the operator only saw them
after logging.
It now passes the entry-strip text as well: QTH, name, country, comment,
note and grid.
The refusal to guess was only visible in the award editor's test tab and
in the awards table — both of them places an operator goes long after the
contact is over. It has to appear while the station is still in front of
you, so ComputeQSOAwardRefs now reports the candidates it declined to
assign and the F3 panel lists them as buttons: click one and it is
written to award_refs, replacing any sibling picked for that award.
They are never auto-added — adding one would be exactly the guess the
setting exists to refuse.
Also: the dark-theme fix for the native calendar glyph named only
input[type=date], so date-time fields kept an invisible icon. All the
temporal input types now share the rule.
Two DARC DOKs are both called "Gießen" — two clubs share the town. Exact
matching finds both and both are right as far as the matcher can tell, so
it kept both and the operator got a QSO credited to a DOK that may well
have been the other one.
Picking one is not an option: it writes a reference into the log with no
evidence behind it. So a Def can now be marked one reference per QSO, and
an ambiguous match on such an award assigns none, names the candidates in
the trace and leaves the contact under missing references — where the
operator's choice now sticks (06e3437).
Off by default, and deliberately NOT a revival of the old `Multi` flag: a
field holding several references (an n-fer POTA activation, a VUCC grid
line) is several references, correctly, and stays that way.
Reported on the DARC DOK award, with DL2FDM: a German address matches two or
three DOKs, the operator deletes the wrong ones and assigns the right one, and
the next recompute puts the wrong ones straight back.
The manual reference was APPENDED to what the matcher found. Storage already
kept exactly one manual entry per award — setOverrideRef drops the previous one
— so it was an override everywhere except at the one place it mattered. The
operator was arguing with the matcher and could not win.
It now replaces. The trace records what was superseded, so a corrected QSO can
still be explained rather than being merely different. A QSO with no correction
keeps every automatic reference, which the test also pins: an override must not
quietly disable matching for everyone else.
This does not address the over-matching itself — an address containing "Berlin
Mitte, 10115 Berlin" genuinely matches two DOK descriptions. That is the next
question, and a separate one.
I took the callsign field's extra width from the RST pair, on the assumption
that five characters was the most they would ever hold. A signal report of
"59+30" is exactly five characters plus its padding, and it was clipped.
The RST fields are back to their original width. The callsign keeps its extra
room — the row has space for both.
Reported on a Xiegu G90. With the blue programming cable — three wires, no
modem lines — CAT works perfectly. Behind a DE-19, which carries audio, CAT and
PTT, the radio went into transmit the moment OpsLog connected and stayed there.
Windows raises DTR and RTS when a serial port is opened, and the DE-19 reads
them as PTT; Xiegu's own documentation asks for both low. The CI-V backend has
dropped them since it was written, for the same reason on Icom rigs with USB
SEND mapped to a line. Xiegu, Yaesu and Kenwood did not. They do now.
The second half of the report: WSJT-X through OpsLog's rigctld decoded fine and
never transmitted. Nothing was broken in that chain — set_ptt reaches the
backend, which sends the CI-V PTT command, which a G90 ignores. That is why
Xiegu keys on a hardware line instead. The backend can now do that: Settings →
CAT → Xiegu → how the rig is keyed (CI-V / RTS / DTR).
Untested here — no G90 in reach. The operator who reported it offered to try.
Nine entries had been appended to the 0.22.5 block after 0.22.5 was already out,
so what shipped as 0.22.6 had no block at all — and the release script said so:
"changelog.json has no 'fr' entries for 0.22.6 - skipping". No Discord
announcement, and no What's-new for anyone updating.
The split is not guesswork: the changelog at tag v0.22.5 carried exactly one
entry, so everything after it belongs to 0.22.6.
Reported by an operator whose node sends little: nothing for two minutes, then a
reconnect — losing the login, the filters, and any spot that arrived during the
gap.
The read had a 120 s deadline and ANY error ended the session, a timeout
included. That reads a silence as a verdict on the link. It is not: a cluster on
a dead band legitimately says nothing for minutes.
Only a real error ends a session now. A genuinely dead peer is still caught, and
by the mechanism meant for it — TCP keepalive probes an idle connection and its
failure surfaces as an error on Read, not as a timeout. The dialer sets it
explicitly rather than inheriting a default, since it is now what the design
depends on.
One trap the advice this came from did not mention: ReadString returns the bytes
it HAD read along with the timeout. Discarding them would lose the first half of
any spot line straddling a deadline, so the partial is carried over.
Tested: a listener that goes silent past the deadline and then sends a spot —
the spot arrives on the ORIGINAL connection, and the listener never accepts a
second one. The tick is a var so that test runs in four seconds instead of
sixty; a slow test is a test that gets skipped.
A panadapter shot settles it: answering an S5 station, the replayed recording
goes out several times wider and taller than the station being answered — in WAV
as well as MP3, so this is not the encoder.
One setting served two sources that have nothing in common. A voice-keyer
message is a microphone at speaking distance and legitimately wants 195%; a QSO
recording is a receiver's line output, which is far hotter. The same 195% put it
into the transmitter flat out, ALC pinned, noise and voice alike.
The QSO playback now has its own level, defaulting to 100%. The voice keyer
keeps the setting it had, so nobody's F-key messages change.
The operator split it in one test: WAV is fine, MP3 is not. The MP3 path is the
only one that resamples — 16 kHz up to 32 kHz, because the encoder emits broken
frames at MPEG-2 rates — and it did so by averaging neighbouring samples.
That is an upsampler in name only. Every component at f is mirrored to
16 kHz − f, and the test measures the mirror just 7 dB below the wanted signal:
a second copy of the whole spectrum. On speech it adds brightness; on receiver
noise, which is broadband, it lays a second noise floor over the top of the
band. Hence "a hiss that is not the QRM, louder than the voice" — and hence its
absence from the WAV, which resamples nothing.
Zero-stuffing plus a 63-tap windowed sinc at the old Nyquist puts the image
56.8 dB down instead of 7, at a cost measured in microseconds against the MP3
encode that follows. Two tests pin it: the image suppression, and that the level
is unchanged — a fix that quietly halved every recording would be a second
surprise on top of the one being repaired.
A station with two radios has two sets of audio endpoints, and the recorder will
happily capture the one nobody is listening to — a Flex DAX channel while the
operator works a Yaesu over USB. The file is then full of hiss with no relation
to what they hear, unaffected by the rig's AGC, and nothing anywhere said which
receiver it came from.
The device names are now logged, resolved to their friendly names, once when
capture starts.
Reported precisely: on the air the voice is well clear of the noise; in the
recording the noise is louder than the voice. Same audio, so the fault is in how
it is captured, not in what is captured.
AUDCLNT_STREAMFLAGS_SRC_DEFAULT_QUALITY was set alongside AUTOCONVERTPCM. The
name reads like a sensible default and is not: it selects the CHEAP sample-rate
converter, for cases where quality does not matter. Taking a rig's 48 kHz stream
down to our 16 kHz, it folds everything above 8 kHz back into the audio band —
and receiver hiss is mostly high frequencies. The voice keeps its level, the
noise arrives twice.
Dropped, so Windows uses its normal converter, which filters before decimating.
This affects every capture and render path in the app, the voice keyer included.
The monitor followed the slider, but two takes made at 10% and 300% came out
identical: starting a recording re-reads the STORED settings and calls SetGains
with them, so the live value was overwritten the moment a take began.
The slider now persists as it applies. Saving Settings still works as before —
this only means the stored value is already right when a take starts, which is
what every path downstream reads.
Reported again after the monitor was wired up: 300% still sounded like 10%. The
monitor fix was right but could not show, because the sliders only changed the
settings panel's own state — the value reached the audio engine when the
operator clicked Save.
A level is set by ear, by dragging. It now applies on every move, to the monitor
and to the recorder mix alike; saving still persists it. Both halves were needed
and neither works alone.
Padlocks. LockBtn was declared INSIDE the App component, so every render gave
React a new component type and the button was unmounted and rebuilt — about four
times a second while the CAT polls. It flickered under the pointer and swallowed
clicks, because the node being clicked no longer existed when the click landed.
The padlock now lives at module level and is passed its state.
From-radio level. It reached the QSO recorder and nothing else, so an operator
listening through OpsLog heard no difference between 10% and 150% — the control
was not weak, it was disconnected. It now scales the monitor too, applied on the
captured chunk so the network-fed path keeps its own levels, and a running
monitor picks up a change immediately: making someone restart the monitor to
hear the slider is how a working control gets reported as broken.
Callsign field widened to w-56, the room coming from the RST pair which never
needs more than five characters. It is the one field always typed into, it
carries the REC badges, and a long portable call has to stay readable.
A second press restarted the message from the beginning. What the operator wants
there is to cut it short — enough heard, or pressed by mistake with the
transmitter keyed — and then to be able to send it again at once.
It now shows a stop square while playing and releases the PTT through the same
path the natural end uses, so nothing is left keyed. Together with the device
handover fixed a moment ago, sending again immediately afterwards works.
Reported precisely: press play again and it stops the current message, but you
then have to wait the whole length of the message before it will play — otherwise
the radio is keyed and the PTT released at once.
Play only SIGNALLED the previous playback to stop and started a new one straight
away. The old goroutine still held the WASAPI render client for a moment, so the
new client could not start; it returned immediately and the PTT tail was all that
was left. The log now shows both halves of that: a play line, then a release
120 ms later.
Stopping now waits for the goroutine to have returned — the point of stopping
before starting again — bounded at 1.5 s so a wedged device cannot freeze the
operator's click, and saying so in the log if it ever comes to that.
From an operator's log, playing a recording on the air repeatedly:
18:52:40.148 play → 18:52:48.420 PTT released 8.2 s, the whole take
18:52:57.759 play → 18:52:57.879 PTT released 0.12 s, nothing came out
Those 120 ms are the PTT release tail alone: the render device refused to start
and playPCM returned an error that was thrown away. The call reported success
every time, which is what "it plays once and then never again" looks like from
the outside — and why the search went to the recorder, which was innocent.
The error is now logged with the device name.
An operator emptied "Recording mic" and got "no audio output to the radio is
configured", and read it as a restriction: that a recording of the other station
alone cannot be played back. No such rule exists — the mic is an INPUT, one more
source to record, while this message is about the OUTPUT that feeds the
transmitter.
The message now names the setting exactly as the panel labels it, and the log
line carries the other two device values so the next report says which field was
actually empty.
Reported: the recording plays once and never again, "as if it threw the sound
away after playing it". PeekQSO is what the playback reads and it is meant to
copy — TakeQSO is the one that clears — so this pins that, and that the copy is
independent of the recorder's own buffer.
It passes, which is the useful part: the take survives, so the one-shot failure
is downstream of it and the search moves to the file and the player.
The grid already badges them; only the filter was missing.
They are not entity statuses — a new park or county is another dimension of the
same spot — so the filter key is now a superset of the status, and matching is
an OR across all of them, as it already was for a previously-worked callsign.
The chips carry the colours their badges use in the grid: a filter that does not
look like what it selects has to be learned twice.
All / CW / Phone / Digital, stacked on top of the worked/confirmed filter. The
question an operator actually has — which DXCC entities have I worked on CW but
not confirmed on CW — needs both at once; answering only one of them is what
sends people to a spreadsheet.
References carried bands but not modes, so the award computation now aggregates
mode CLASSES per reference, worked and confirmed separately. Classes, not ADIF
modes: nobody chasing CW cares whether the digital side was FT8 or RTTY, and a
list of twenty mode names would not answer the question. An unrecognised mode
counts as data rather than as nothing, so a new digital mode does not vanish
from the filter the day it appears.
Two rules the filter needs to be honest:
- Not-confirmed means not confirmed ON THIS MODE. An entity worked on CW and
confirmed on SSB is still a CW entity to chase, and that is exactly what the
filter is for.
- "Not worked" plus a mode is a contradiction — an unworked reference has no
mode — so the mode filter stands aside there instead of emptying the list.
The operator's own fetch settles it. One record, both fields:
<app_qrzlog_status:1>N <qrzcom_qso_download_status:1>Y
QRZ says the QSO is NOT confirmed and sets the download field anyway — it marks
what was handed back, not what was confirmed. OpsLog read it as a confirmation,
so eighteen QSOs came back "UPDATED" and green, none of them confirmed on QRZ.
The site's own logbook page shows them without a star.
Only app_qrzlog_status = C counts now. The test that asserted otherwise is
corrected and carries the real record.
And the download now takes back what it wrongly gave: when QRZ explicitly says
not-confirmed and our log says confirmed, the flag is cleared and reported on
its own line. These QSOs count towards award slots, so leaving them green until
someone noticed was not an option — and nothing else in the app would ever have
undone them. Only on an explicit no: a record with no app_qrzlog_status is QRZ
saying nothing, and silence is not a retraction.
Closes#9.
It was clamped against a GUESSED height — 230 px with the upload submenu, 110
without. The menu is far taller than that with a selection, so right-clicking on
a row near the bottom hid half the entries, delete among them.
It is measured now: placed above the cursor when it does not fit below, pinned
to the top with its own scrollbar when it fits neither way. Its height depends
on which actions the caller passes, so no constant could have stayed right.
The backend honours the setting: it normalises the log side and the spot's own
mode through the same grouper, so an FT4 spot on a band worked in FT8 reads as
worked. That was not the fault.
Spot statuses are cached per call+band+mode and the cache never expired. Ticking
the box in Settings therefore changed nothing on screen — every spot already
listed kept the answer to the old question, and only new arrivals were judged by
the new rule. The cache is now dropped when settings are saved.
Also: writeUiPref swallowed a failed database write. The interface kept working
from localStorage while the BACKEND — which reads some of these keys, this one
among them — still saw the old value. The two disagreeing in silence is exactly
the shape of this bug, so the failure is now logged.
The rule between the band-driven columns and the mode-driven ones was an extra
empty cell added to the body rows only. Seven cells per row against six in the
header slid every heading across: Phone stood over the TX antenna, CW over
Phone.
The rule is now drawn on the first power cell itself, in the header and the body
alike, so the two cannot drift apart again.
Antennas and power were two tables listing the same bands, one under the other,
so configuring a band meant matching rows by eye between them.
They are now one row per band. That the two settings are stored separately and
applied on different triggers — antennas on band change, power on band or mode
change — is a backend detail, and no reason to split what is a single decision
per band for the person configuring it. A rule in the header keeps the two
groups readable.
Settings → FlexRadio, beside the per-band antennas and applied the same way:
when the band or the mode changes. 1.5 kW that is fine on SSB has no business
going into an FT8 signal on the same band.
Three classes rather than one row per mode, because what decides the power is
duty cycle and what the amplifier will take of it, not the mode's name — FT8 and
RTTY punish an amplifier the same way, SSB and AM do not. The classification is
pinned by a test: a digital mode sorted as phone would deliver exactly the
1.5 kW this exists to prevent.
An empty box means "leave the power alone", never zero watts, and an unknown
mode changes nothing — setting a transmit power from a name we do not recognise
is not a good guess.
Unlike the antennas, a locked band does not skip it: the lock means "do not let
the rig drag my log entry around", not "let the amplifier keep whatever the last
mode left".
Both faults reported on the air, both reproduced in synthetic signals first —
the existing tests all passed because each starts a fresh decoder and sends
textbook spacing, which is neither of the cases that hurt.
Speed change. A 32 wpm station followed by a 14 wpm reply decoded as
"TTT TTTT TTTTT TT…" for most of the over. Not the element classifier: the
CHARACTER-gap threshold is 2.2 dits, so at the stale fast estimate it stood at
81 ms while the newcomer's element gaps were 86 ms. Every single element was
flushed as its own character, and a lone element with no contrast reads as a
dah. A silence long enough to end an over now drops the estimate back to the
seed, which is exactly the state a freshly started decoder is in — and that case
was always fine. The threshold scales with the current estimate (12 dits, floor
600 ms) because a fixed one cannot serve both 10 and 40 wpm.
Callsign split. The word boundary sits at 4.6 dits, the geometric mean of a
3-dit letter gap and a 7-dit word gap. It assumes textbook spacing; a fist
leaving 5 dits between letters had every letter turned into a word, so a
callsign arrived as "O Y 1 C T". The boundary now also follows the letter gaps
this operator actually sends, whichever is larger. A wide sender's words may run
together — a far smaller price than a callsign in pieces.
The operator asked the right question: why wait 1.5 s at all, when the mode is
known from CAT?
Because the wait is not the problem — what happened during it was. The drift
guard, meant for two sound cards running on their own clocks, saw a starved
source as a one-second skew and trimmed the radio audio to match. So the opening
of every CW recording was thrown away although it had been captured.
The guard now applies only while BOTH sources are actually delivering. Waiting
then costs nothing: everything buffered is written whole the moment the silent
source is written off, which a test now pins with three seconds of audio.
Using the CAT mode to skip the wait would have been a second mechanism for a
case this one already covers — and it would not cover the others: a device
switched off, unplugged, or held by another program.
The log read:
recorder: the second audio source is silent — recording the radio alone
recorder: no audio at all from "DAX Mic" after 3 s — nothing is streaming
Both true, and together they read as a fault while the recording was going
perfectly well — it was saved seconds later. On CW the mic channel legitimately
delivers nothing; the mixer already said what it was doing about it.
The watchdog now mentions the mic only when the radio is silent too, which is
the case where the recording really is empty.
The operator said it from the start: in CW there is no sidetone, only reception.
My previous fix did not cover it — it wrote off a source that STOPPED, but a
source that never starts has no last-delivery time, so it stayed forever "not
yet dry" and held the mixer at min(A,B)=0. Three empty recordings while the
radio audio streamed perfectly.
A source with no delivery is now measured from when capture started. Tested both
ways: nothing is written off in the first moments, and a source that has never
spoken is written off once the grace period has passed.
Third empty recording, same station, and the sequence was as plain as it gets:
type the call, click record, log the QSO. The device opened and never streamed.
Endpoint ids are what gets stored, and a DAX channel that is reconfigured,
disabled or recreated comes back with a DIFFERENT id. The stale one opens
without complaint on some drivers and simply never delivers — indistinguishable
from a quiet band until the file turns out to be empty.
The recorder now checks the id against the current endpoint list when it starts,
which takes milliseconds, and says outright that the configured input is gone —
on screen, not only in the log. If it IS still there, the answer is equally
useful: the device exists and the fault is upstream, in whatever should be
feeding it.
The watchdog fired exactly as intended and said:
no audio at all from "{0.0.1.00000000}.{6a27abfd-f277-496b-b6f1-f92ec284c636}"
which tells the operator nothing they can act on. "DAX RX 1 (FlexRadio DAX)"
points straight at the DAX panel.
Endpoint ids are what gets CONFIGURED and stored, so they are what the recorder
holds; the friendly name is resolved only when something has gone wrong. When
the endpoint cannot be found at all the id is still shown — a device that has
disappeared explains an empty recording too.
"No recording was running at log time" says the take had already ended, but not
who ended it. The only path that does is the callsign being cleared — which also
happens when a clicked spot replaces it — and that path was silent.
Now it names itself, so the log shows the moment the recording was lost instead
of only its absence at the end.
Second empty recording from the same station, and this time NEITHER source had
delivered a sample — so the fix for a silent second source could not apply, and
the log still said only "recording was empty" at the end of the QSO.
Two blind spots, both now closed:
- captureStream's error was discarded. A device that cannot be opened — renamed,
unplugged, held by another program — looked exactly like one that is merely
quiet.
- A WASAPI device can open cleanly and then produce nothing at all. A DAX
channel with no stream behind it does precisely that, and it is
indistinguishable from silence until the file turns out to be empty.
A watchdog now reports both, three seconds in, and the first goes to the OPERATOR
as a toast rather than only to the log: they are mid-QSO, and by the time they
find out at save time the audio is gone for good.
From an operator's log: a Flex with From Radio on DAX RX and Recording mic on
DAX Mic, a full CW QSO recorded, and at log time "snapshot failed: recording
was empty".
The mixer took min(len(A), len(B)) samples. In CW the mic path does not run, so
DAX Mic delivered nothing, min stayed 0, and the drift guard then threw the
radio audio away a second at a time. Two devices configured meant the recording
depended on the quieter one being alive.
A source that has delivered nothing for 1.5 s is now treated as absent and the
other is recorded alone, with one line in the log saying so — half a recording
is worth having, silence is not. Both are mixed again the moment the missing one
comes back.
Neither is written off before it has EVER spoken: at startup both timestamps are
zero, and declaring one dead then would clip the opening of every recording.
That case is tested too.
In CW the transmitter builds its tone from the keyer and ignores the DAX TX
audio path, so the button keyed the rig and sent silence. Confirmed on a Flex:
our side reported the playback as done while nothing went out.
Only the button goes. The row stays, or a CW operator who stopped their
recording would be left with no way to resume it — and the recording itself is
still worth keeping.
"I click play and nothing happens" is what an operator sees when a function
returns an error into a promise the interface swallows. Every refusal in the
playback path is now logged as well as returned, and each names its own cause —
audio not initialised, the take still running, the take empty, the device
rejecting the file.
One of those was not even checked: playing back needs an OUTPUT device INTO the
radio, which is a different setting from the recorder's input and is very likely
unset on a station that has only ever recorded. It now refuses with that
sentence instead of failing somewhere inside the player.
CW has moved in and out of this list twice — excluded for the missing sidetone,
restored because the other station is captured regardless. A test says which,
so the next reading of that comment does not undo the decision.
CW was excluded because SmartSDR does not route the operator's own sidetone
through DAX, so the recording sounds one-sided. The audio path is open all the
same and the other station IS captured — which is the half worth keeping.
The frontend had ONE list serving two purposes: which modes may be recorded, and
which modes the voice keyer may transmit on. Adding CW to it would have let the
DVK key the rig with speech on a CW frequency. They are now two lists: PHONE_MODES
for the keyer, RECORDABLE_MODES (phone + CW) for the recorder.
Digital modes stay out of both: their audio is a modem tone nobody will replay.
SmartSDR moves its active flag onto the TX slice as soon as the transmit
frequency is touched, and OpsLog followed it. But the operator is listening to
the DX on the other slice: the S-meter, audio level, filter and DSP they are
adjusting all belong there, so following the transmitter handed them the
controls of a receiver they were not using.
In split the receive slice now wins. Simplex is unchanged, and two same-band
slices in different classes (SSB alongside FT8) are still not a split.
A slice clicked IN OPSLOG is remembered and overrides both the split rule and
the radio's focus — that is the operator speaking. Activating a slice on the
radio or in SmartSDR deliberately does not move OpsLog. The pin is dropped when
that slice stops being in use, so a closed receiver cannot strand the display.
The panel now highlights the slice OpsLog is working with rather than the radio's
focused one; reporting the radio's flag would light up one slice while every
control acted on another.
The 0.22.3 block had grown into explanations: why a fault happened, how it was
found, what it does not cover. That belongs in the commits, where it already is.
An operator opens this list to learn what changed for them. Sixteen entries now
say it in a line each.
Settings → General, beside the language. An operator reading 2026-07-30 as the
7th of the 30th month is reading their own log wrongly, and that is a display
problem.
Storage stays ISO/ADIF, deliberately and permanently: it is what ADIF
specifies, it sorts correctly as text, and a stored format that followed a UI
preference would make the log unreadable the day the preference changed. Exports
and uploads are untouched.
Two details that decide whether it actually works:
- The columns are rebuilt when the format changes. The formatters are captured
inside AG-Grid's column definitions, so nothing else would notice and the
table would keep the old format until something forced a rebuild.
- main.tsx re-reads the choice after the portable prefs are pulled from the
database. The module reads localStorage at import time, which is BEFORE that
sync — so a copied data/ folder would have shown ISO until the next launch.
UTC is not part of the choice and never will be: a logbook is kept in UTC, and
showing local time would make the display disagree with the QSO's own record
twice a year.
Y green, N red, R blue — colour only, no pill or badge. These columns sit among
callsigns and dates, and a row of coloured boxes would shout louder than the
callsign it belongs to.
Twelve columns: paper QSL, LoTW, eQSL, Club Log, HRDLog, QRZ.com both ways, and
OpsLog's own card and recording flags. All four views the operator named —
recent QSOs, worked before, NET Control, QSL manager — turn out to share this
one grid, so they are all covered by construction rather than by four edits that
could drift.
Semantic tokens rather than fixed colours, so the four themes follow and the
contrast stays right on the light ones. ADIF's I (ignore) is deliberately left
in the default ink: it is neither good news nor bad, and colouring it would put
it in one camp or the other.
Says what the switch does, then names the one limit that changes the decision.
Drops my "no undo" line: deleting a QSO is already understood to be permanent,
and repeating it reads as a warning about the checkbox rather than the QSO.
Four sentences of API behaviour under a checkbox. Two facts actually change the
operator's decision: it cannot be undone, and QRZ.com cannot reach QSOs uploaded
before OpsLog started keeping their record number.
The rest — how Club Log matches, that the local delete happens regardless, that
the log records each answer — is true, and belongs in the log and the code where
it already is. Documentation in a checkbox hint is documentation nobody reads.
It applies to QRZ.com and Club Log both, but it was rendered after the tab
strip — so it read as a QRZ setting when the QRZ tab was open, appeared to
repeat itself on the Club Log tab, and was invisible to anyone opening a third
one.
Above the tabs it belongs to the panel, which is what it actually is.
Opt-in, off by default: neither service can undo it, and a local delete is not
always a statement about the world — a duplicate cleared out of a contest log
was never meant to change the DX's log.
The two APIs are not alike, and the difference decides what is possible:
- Club Log deletes by MATCH (callsign, exact timestamp, band number), so it
works for every QSO in the log, past or future. The band table is its own
list and is pinned by a test: a band mapped to the wrong number does not
fail, it points the delete at a different QSO.
- QRZ.com deletes ONLY by logid. So markExtUploaded now stores the LOGID it
already received on every upload — it was being logged and thrown away. A
QSO uploaded before this cannot be withdrawn from here at all, and the log
says so by name rather than leaving the operator to assume it worked.
Both run BEFORE the local delete, because both need the QSO's own fields to
find their copy and there is nothing left to ask with once the row is gone.
Neither can block it: the operator asked for the QSO to go, and a refusing
website is not a reason to keep it.
The sort choice is now written through writeUiPref and both keys are registered
as portable, so they travel with a copied data/ folder like every other UI
preference.
Registering them is the part that matters: writeUiPref stores to the DB, but
only keys on the portable list are read BACK at startup. A key written and never
restored looks exactly like one that was never saved — which is what would have
happened here.
The grey-line toggle had the same gap. It called writeUiPref from the day it was
written but was never added to the list, so it came back off on the next launch.
Found while adding the award keys beside it.
Reference stays the default: that is how award lists are published and how a
paper list is read. Description answers the other question an operator actually
has — "have I worked anything in Alicante" — which reference order scatters
across eight thousand rows.
Sorted where the filtered list is built, so the grid and list views inherit it
rather than each growing its own copy that could disagree.
localeCompare rather than a byte comparison, because these lists are Spanish,
French and Italian place names: "Ávila" belongs beside "Avilés", not after
"Zaragoza". References compare numerically for the same reason in reverse —
08019 must not sort between 0801 and 081.
It was the same red as the recording indicator beside it, so at that size the
two read as one control. Green is also what the theme already uses for a
go-ahead action (the amplifier ON buttons), and it is a semantic token, so it
follows all four themes rather than pinning a colour.
Two silent returns in the save path: no take was running at log time, or the
mode carries no audio worth keeping. Both simply produced no file.
An operator reporting "the sound is not in the folder" is describing an absence,
which is the same absence in either case — but they are different problems with
different fixes, and neither leaves anything behind to inspect. Now each says
which one happened, and for which callsign.
The success line already existed ("qso-rec: saved <path>"), so between the three
the log now accounts for every logged QSO.
Two absolutely-positioned boxes at guessed offsets: right-14 and right-2. They
collided as soon as the counter passed a minute and the text grew — a layout
that only looked right for the first 60 seconds, and did not even manage that at
this font size.
One flex row with a gap, so spacing is a consequence of the content rather than
an arithmetic guess about it. The callsign field also gains right padding while
the badges are shown, so a long call runs under them instead of being typed over.
The situation, in the operator's words: working VP2MAA, recording, and he asks
to hear his own signal. So: stop, play, and the recording is still saved with
the QSO.
Stopping FREEZES the take rather than ending it — nothing more is captured,
nothing is discarded, and logging writes the file exactly as before. Playback
reuses the voice keyer's path: same output device, same PTT keying and release
with its generation guard, so a replay cannot be cut short by a stale unkey.
Playing is refused while still recording. The transmitted audio would feed back
into the same take on any rig monitoring its own transmission, and stopping is
one click the operator has already made — it is their statement that the take is
finished, not a hoop.
The elapsed clock freezes with the recording and resumes from where it stopped,
because it now shows the LENGTH OF THE FILE rather than the time since the QSO
began. Only a fresh take returns it to zero.
A red dot in the slot the counter will occupy. Click it and the counter starts,
the dot disappears, and logging the QSO writes the file exactly as an automatic
recording would — the save path already keyed off an active take, not off the
setting.
With automatic recording disabled the capture engine is not idle, it is not
running at all, so a manual take starts it. Two consequences handled here:
- It is released again when the take ends — logged, cancelled, or dropped for
being a digital mode. An operator who records one contact does not expect
their microphone held open for the rest of the session. The release is a
deferred call at the top of the save path so every exit goes through it,
rather than one that has to be remembered at each return.
- There is no pre-roll. An automatic recording opens with the seconds BEFORE
the callsign was typed; a manual one can only begin now. The button's
tooltip says so, because otherwise the difference between two files is a
mystery to the person holding them.
The button only appears where it can work: devices configured, automatic
recording off, and a mode whose audio is worth keeping. That state is asked of
the backend rather than inferred from a preference in the interface, so the two
cannot drift apart, and it is re-read when settings are saved.
The list read "Yaesu (native CAT)", "Kenwood (native)", "FlexRadio / SmartSDR
(native)". Native says how OpsLog is built, which is our business; the operator
opening that menu is answering a different question — how is my radio plugged
in.
So: OmniRig, FlexRadio (API), Yaesu (USB), Kenwood (USB, network), Xiegu (USB),
Icom (CI-V USB), Icom (CI-V network), TCI. Shorter, and each entry now tells
someone with a cable in their hand which line to pick.
The same ASCII stream over a socket instead of a wire: ser2net, an
Ethernet-serial adapter, a Raspberry Pi at the radio. One TCP transport
presented as a serial port, so the backend keeps a single code path.
Explicitly NOT the radio's own RJ45. A TS-890 or TS-990 speaks Kenwood's
KNS/ARCP there — session, authentication, a different protocol — and that needs
one of those radios in hand to write honestly. The setting's help text says so,
because an operator who plugs in their TS-890's Ethernet port and types its
address deserves to learn that from the interface rather than from silence.
Two details that decide whether this is usable or maddening:
- A read deadline expiring is reported as "no data yet", not as an error. It
is exactly what a serial read timeout means to the caller; as an error it
would make ask() abandon a rig that is merely thinking.
- The log line names what was connected to. "connected on @ 0 baud" after a
network connect sends someone hunting a serial fault that does not exist.
The host wins over the COM port when both are filled: it is the more deliberate
setting, and silently preferring the wire leaves someone staring at an address
they typed and a radio that never answers.
From a Kenwood trace: IF answered mode digit 1 on 7.160 MHz, the backend read
LSB correctly, and the entry form showed no mode at all while the frequency
tracked perfectly.
Nothing was wrong in the backend. Radios report the SIDEBAND, and the station's
mode list holds SSB — so the value pushed into the selector was not in its own
list of options and nothing could match. Yaesu and Icom report the same way, so
this was never Kenwood-specific; it only surfaced there.
Normalising at the point where a rig-reported mode enters the form fixes every
backend at once, rather than editing three mode tables and the next backend
someone writes.
It is also the ADIF-correct direction, which matters more than the selector:
LSB and USB are SUBMODEs in ADIF, not modes. MODE=LSB is not a valid value and
is what would have gone to LoTW and eQSL.
An operator who has deliberately put LSB or USB in their own mode list keeps
them — the list is their statement of what they log, not ours.
The grey line has never been released, so its off-by-one-copy edge was never
something an operator saw. An entry for it would describe a defect that only
existed between two commits of the same unreleased feature — noise in the middle
of the list someone reads to find out what changed for them.
Same case for the bulk-update progress bar: the feature and the fix that made it
non-blocking are both new in 0.22.3, so they are now one entry describing what
it does. The menu rename goes there too rather than standing alone.
The map repeats the world sideways, and zoomed out several copies are on screen
at once. The rings span −180…+180 once, so the shading ended in a hard vertical
edge wherever the next copy began: it read as a rectangle laid over the planet
rather than as night.
Each ring is now drawn again shifted a full turn either way. Three copies cover
every zoom this map reaches and cost nothing — three more polygons on a canvas
layer.
Zooming the map to fit one world would have hidden it rather than fixed it, and
would have taken away the free pan and zoom the operator has.
Grey line — the day/night terminator and its twilight band, off by default,
redrawn every minute so a map left open all day is not silently wrong by
evening. Its own Leaflet pane below the overlays, non-interactive, so it never
hides the path or beam nor swallows a click.
The solar maths is pure and was checked against known positions before being
wired to anything: both solstices, the March equinox, the subsolar longitude at
12 UTC, and day/night at Paris, Tokyo, New York, Sydney and Reykjavik across
seasons and hemispheres. That last set is what caught the real bug: the equation
has TWO solutions per meridian and the naive branch put New York in daylight at
02:00 in December — the terminator at +63° where it belongs at −63°. A shaded
map that looks entirely plausible and is exactly mirrored. It now anchors on the
classical terminator and takes the branch nearest it, which is also what makes
the twilight band follow the terminator instead of jumping hemisphere.
Separately, from a Xiegu user: the protocol-trace checkbox does not stay ticked.
It is React state initialised to false on every mount, so a trace still running
came back unticked — the operator ticks it to "switch it on", which switches it
OFF, and the log they send contains no trace at all. Worse than a cosmetic bug:
it silently defeats the one tool asked for to diagnose their radio. Both boxes
(CAT and WinKeyer) now read their real state from the backend.
Two faults in the progress bar shipped an hour ago.
It was a modal. A hundred QSOs is a minute; a contest log is two thousand and
ten minutes, and for all of it the operator could do nothing else — while their
radio is on. It is now a card in the corner: same bar, same count, same
callsign, none of the imprisonment.
And the menu said "Update from QRZ.com" while the code has always walked EVERY
configured provider, falling through to HamQTH when QRZ answers not-found. The
label is now "Update from the callsign databases", which is what happens.
Noted while reading that code, NOT fixed here because it trades against the
user's API quota and is their call: this path goes through the lookup cache, so
a second run on the same callsigns re-reads the cache rather than reaching the
provider. The cache comment already argues the opposite for deliberate clicks —
"a lookup the operator asked for by clicking is a deliberate act and must reach
the provider" — and a right-click on a selection is exactly that.
Selecting the 102 contacts of a freshly imported contest log and updating them
from QRZ.com is 102 network round trips. Nothing moved on screen while it ran,
so the only honest reading was that the application had frozen — and the natural
response to that is to kill it, halfway through.
The backend now emits a progress event per QSO and the frontend shows the same
overlay the ADIF import uses: a bar, the count, and the callsign being queried.
The callsign is reported BEFORE the lookup rather than after. With a slow
provider that is the difference between "waiting on F4BPO" and a name that lags
one QSO behind whatever is actually taking the time — which would be the field
someone stares at while deciding whether it is stuck.
The overlay is cleared by the closing event and again in a finally, so neither
a completed run nor a failed one can leave it on screen.
Reported from a Flex in Kenwood CAT mode: frequency read perfectly, split never
appeared. Not every rig speaking this dialect fills IF's split bit.
Rather than guess which column that firmware populates, ask the question that
DEFINES split — is the transmit VFO a different VFO from the receive one — which
is exactly what FR and FT answer, and the same rule the Yaesu backend settled on
after several wrong turns. FR is also trusted over IF for which VFO is in use:
they are asked in the same breath, and a rig vague about the split bit may be
just as vague about the VFO field.
Cost is bounded: a rig that rejects FR/FT answers "?;" once and is never asked
again, so this is two short commands per poll only where it works. Both paths
are tested against the emulator — split found through FR/FT with IF silent, and
IF-reported split still working on a rig that refuses FR/FT without re-asking.
Also extends the CAT wire trace to the Kenwood backend, ASCII quoted so an empty
reply is visible as such: "" and ";" look identical unquoted, and telling them
apart is the whole question when a rig half-supports a command. If this fix is
not the whole story on real hardware, the trace is what will say so.
Nobody here owns a Kenwood, and a backend that has never completed a single
exchange is a guess however carefully it was written.
But this repository already contains the other half of the conversation:
internal/catemu ANSWERS this dialect, pretending to be a TS-2000 so an ACOM
amplifier follows OpsLog. The test's responder replies exactly as catemu does —
same FA/FB widths, same 38-character IF layout, same ID019 — so the two halves
are checked against each other instead of against my reading of a manual.
Connect adds a dialPort seam for this; nothing else changed in the backend.
Covered: model identification, simplex on A, everything following the VFO in
use when on B, tuning writing to that VFO rather than blindly to FA, split with
TX and RX the right way round (ADIF FREQ is the TRANSMIT frequency — swapping
them writes the wrong frequency into every logged QSO), the mode digit, and a
silent port reported as not answering rather than as a connected radio. The last
three are each a fault that actually reached a user through the Yaesu backend.
What this does NOT prove: that a real TS-590 answers on the same timings, or
fills every IF field the way its documentation says. That still needs a radio.
They searched only the rows on screen. With the grid holding the most recent
page, filtering a column for RSGB-IOTA across a 28 000-QSO log returned nothing
— which reads as "the log does not contain it", not "it is not on this page".
That is how an import that had worked perfectly came to be reported as broken:
the tool answered a narrower question than the one being asked, and said so
only by being empty.
The advanced filter queries the whole logbook and is the honest instrument. The
headers also get back the width the filter menu icon was taking, which is the
smaller reason but a welcome one on a grid this dense.
No saved filter can be stranded by this: the grid persists widths, sort and
visibility, never a filter model.
CONTEST_ID was reported as not imported. It is imported: this drives an
operator's own IOTA-contest record through a real SQLite logbook and reads the
row back, confirming contest_id=RSGB-IOTA and iota=EU005 in the database.
The conversion-level test was already green while the report stood, which is
exactly why it was not enough — a promoted field can be dropped at any of five
places between the file and the grid, and only the round trip says which. The
answer here turned out to be none of them: the Contest columns are simply hidden
by default in the QSO grid.
An operator worked the RSGB IOTA contest in another logger and got back records
carrying <STATE:5>EU005: N1MM-class software stores the received exchange in the
column its contest module uses, not the one ADIF reserves. Imported verbatim,
every QSO gains a US-state field reading "EU005" and the IOTA award sees
nothing — the reference is in the log, just not where anything looks for it.
The import dialog gains optional source → destination rows, prefilled with
STATE → IOTA since that is the case that prompted it. Applied to the RAW record
before conversion, so it works for promoted columns and Extras alike and the
destination is parsed exactly as if the file had carried it there.
Two rules, both from the same principle that the file outranks our guess:
- the source is CLEARED, so a reference does not linger in STATE where it
would show as the contacted station's US state in the grid and every export;
- a destination the file already filled is kept, judged against the record as
it ARRIVED — otherwise the result would depend on Go's map iteration order,
which is deliberately random.
The swap case in the test is what forced that second rule to be stated: "never
overwrite" and "exchange two fields" cannot both hold silently, so a destination
that is itself a mapping source is treated as an explicit swap and nothing else
is overwritten.
Two white-screen reports are open — one after logging a 10 GHz QSO, one on
starting a DVK auto-call — and neither could be investigated, for the same
reason: a render error emptied the window and left NOTHING. No line in
opslog.log, no dialog, nothing the operator could send but the words "white
screen". A fault the user cannot report is a fault that cannot be fixed.
So before hunting either trigger, the reporting path:
- An error boundary catches a render throw, writes the error and component
stack to the app log through the new LogUIError binding, and shows it on
screen — selectable, with Copy and Reload.
- Global handlers catch what a boundary cannot: throws from timers, event
handlers and rejected promises. An auto-call loop lives entirely in
callbacks, which is exactly where the second report came from, and those
leave no trace at all today.
It deliberately does not try to resume: React cannot promise sane state after a
render throw, and a half-working logger would be worse than a clear stop.
This does not fix either crash. It makes the next occurrence arrive with its
cause attached, which is the step that has been missing.
Reported on an IC-9100: the MOX button does not transmit, it sets split. The
source sends the PTT command (0x1C 0x00) and nowhere sends the split one (0x0F),
so the discrepancy is between what OpsLog sends and what the rig acts on — and
that link has already been shown, in the same operator's log, to lose frame sync.
Rather than guess from the command table, this adds what settled the WinKeyer
bug in one line: the actual bytes. Settings → CAT → Log the CI-V protocol writes
every frame in both directions as hex. RX is traced BEFORE framing, since the
bytes as they arrived are what reveals a lost boundary — a decoded view would
hide exactly the fault being hunted.
Session-only, like the keyer trace: it is a diagnostic, and a log full of hex
helps nobody who forgot it was on.
I am not claiming a cause yet. The last time I inferred one from a protocol
document rather than from evidence, I inverted a digit and broke the case that
worked.
From F4JND's log: a rig sitting on 145.550 MHz FM produced status lines reading
16445550000, 8364550000, 5817408364 and 12640 Hz between good readings — the last
of those even resolved to a 6 cm band.
BCDToFreq treated the nibbles 0x0A..0x0F as the digits 10..15, so a byte stream
that had lost frame sync still decoded into a number. The same log shows why sync
was lost: PTT round trips timing out and a scope stream the rig kept rejecting
(0x27), both competing with the poll on one serial line.
Non-decimal nibbles are now refused. Each caller keeps its last good value rather
than publishing noise: the frequency reader returns an error, the sub-VFO reader
returns not-ok, and the scope keeps its previous edges. Garbage that LOOKS like a
frequency is worse than a refused frame — it reaches the display, the band slots
and eventually the log, and it is the operator who then has to disprove it.
The test builds its fixture with FreqToBCD rather than by hand, because my
hand-written one was byte-reversed and the test caught it.
This does NOT explain the white screen that was also reported, and I have not
claimed it does: it removes one source of absurd values reaching the UI, which is
worth doing on its own evidence.
The 0.22.1 tag was cut at d98bb2e, and six commits landed after it. Five of them
had appended their entries to the 0.22.1 block, so the released changelog
promised things the released binary does not contain: the 4O3A marks, the
status-bar logbook path, the Yaesu per-VFO mode, ERC rotator support and the
Kenwood backend.
Those five move to a new 0.22.2 block, in the order they were written. The
0.22.1 block is restored to exactly the twelve entries the published tag holds —
verified against `git show v0.22.1:changelog.json` rather than by eye.
Worth noting for next time: entries are appended as work lands, so any commit
after the release tag needs its own block opened first. The mistake is silent —
nothing fails, the What's New dialog simply describes a version that never
shipped those changes.
A sixth CAT backend, talking to a TS-590/890/990/2000 over its serial port with
no OmniRig in between — frequency, mode, VFO, split and PTT. Elecraft K3/K4 and
the "Kenwood" setting on other radios speak the same dialect.
The protocol was not researched from scratch: internal/catemu already ANSWERS
these commands, pretending to be a TS-2000 so an ACOM amplifier follows OpsLog.
The IF frame layout here is read from the same format string catemu emits, so
the two halves of the repository agree by construction — and the test caught my
own fixture being one character short, which is exactly the error that layout
invites.
Design notes worth keeping:
- IF; is the poll. One frame carries frequency, TX state, mode, VFO and split,
so simplex operation costs a single round trip; the other VFO is only asked
for when split is actually on.
- FA/FB take ELEVEN digits here where Yaesu uses nine. That is the likeliest
place to copy the Yaesu backend and be wrong by a factor of a hundred, so it
has its own test.
- Every lesson the Yaesu backend learned the hard way is built in from the
start: replies matched to the command that asked, "?;" remembered so a poll
stops paying a timeout for an unsupported command, the serial handle closed
before reopening, and a rig that answers nothing reported as absent rather
than "connected".
Untested on hardware — I have no Kenwood here. The frame parsing is covered by
tests against catemu's own format.
An ERC (Easy Rotor Control, incl. ERC Mini) emulates Yaesu GS-232A/B, which is
exactly what the backend written for the microHAM ARCO already speaks — azimuth
out (Maaa), azimuth in (C), stop (S), which is the whole of what was asked for.
So this is two small gaps rather than a new backend:
- Serial speed was hardcoded to 9600. An ARCO's virtual COM ignores it, but an
ERC runs at whatever its own configuration sets, and a mismatch reads as a
dead rotator. It is now selectable beside the COM port.
- The entry was called "microHAM ARCO", so an ERC owner would never have found
it. It is named after the PROTOCOL now: "GS-232A controller (microHAM ARCO,
ERC…)".
The help text states the condition plainly in both languages, because it is the
one thing that will otherwise waste an evening: an ERC left on Hy-Gain DCU-1
speaks a different command set and simply will not answer.
Untested on hardware — I have no ERC here, and the GS-232 path itself is proven
on a real ARCO.
Reported, and reasoned out from the half that already worked: a spot click now
tunes the right VFO, but it set the mode with MD0 — main — whatever VFO the
operator was on. Working from SUB, the spot changed the mode of the VFO NOT in
use and left the one in use as it was. The author inferred the sub half from
seeing main change, and was right.
The read had the same fault, so the console displayed main's mode while showing
the sub VFO's frequency.
Mode now addresses the receiver in use — MD0 for main, MD1 for sub — in the poll
read, in SetMode, and in the console's raw-mode setter. A rig without a sub
receiver never sees MD1: curVFO only becomes B where the rig reports its receive
VFO in the first place.
Spotted by the author: on SQLite the database chip showed a.dbPath — the
settings/profile database — under a tooltip reading "Local SQLite logbook".
Those became two separate files when the logbook was split out, so an operator
checking where their QSOs live, or which file to copy before a trip, was pointed
at the wrong one. MySQL was already right.
It now resolves the logbook the same way connectLogbook does: the active
profile's own file if it names one, then the default logbook.db, and the settings
db only in the case where it genuinely doubles as the logbook.
Also moves the 4O3A mark beside its menu label instead of the far edge of the
row, as asked.
The wordmark I put there was not what the author expected: the brand is "4", a
green waveform of stacked lenses, then "3A".
Only the waveform is SVG. The digits stay HTML text so they keep the sidebar's
own typeface at any theme or zoom — an SVG <text> would pick its own font and
drift from the menu around it. Seven ellipses, tallest in the middle, match the
shape of the mark.
Still drawn rather than fetched: an image off the web has no verifiable
provenance, and this stays one small component to swap if the official SVG is
preferred.
Antenna Genius and Tuner Genius now carry a small 4O3A wordmark, so the two
panels that drive that hardware are recognisable without reading the labels.
Drawn inline rather than shipped as an image file, following the language
picker's flags: it scales with the row, follows the theme through currentColor,
and adds no binary asset to the repository.
It is a plain wordmark, not the manufacturer's logo artwork — I did not pull an
image off the web, since I cannot verify the provenance or licence of what I
would find, and redistributing a company's mark inside the product is the
author's call. The component is one function and takes the official SVG in its
place if that is preferred.
Not applied to the Amplifier panel: it covers both the 4O3A PowerGenius XL and
the SPE Expert range, so a single vendor mark there would be wrong.
Reported: a 6 s auto-call gap gives clearly less than 6 s between the end of one
CQ and the start of the next.
The wait after sending watched only the keyer's busy signal. That signal travels
from the keyer to the window as an event, and the code gave it one second to
appear before giving up — if it had not arrived and cleared by then, the wait
ended immediately and the gap ran from the START of the call. A 6 s gap after a
4 s CQ becomes about 2 s of silence, which is exactly what was heard.
The message's own duration is now a FLOOR: the wait runs at least the estimated
sending time, and the busy signal only extends it (slow link, long buffer),
capped as before. The estimate can only be approximate, but it cannot be skipped
by a late event, and erring long costs a slightly wider gap rather than a call
that steps on the previous one.
Reported on an FTDX101: with RX and TX moved together to the sub VFO, OpsLog
showed split — and took the MAIN frequency as the transmit one, which would log
the wrong frequency.
ST is a bare flag. It says "split" without saying which VFO transmits, and this
rig raises it whenever the transmit VFO is the sub one, whether or not the
operator is also listening there. FT names the transmit VFO, so where the receive
VFO is known as well (FR), split is derived from the pair: they differ or they do
not. That is a fact about the rig's state rather than a flag whose meaning varies
by model.
The fix is therefore in the probe ORDER, not in the reading: FT is asked first
when FR answered, and ST stays the fallback for rigs that have neither. A test
pins the order and both halves of the trap — that FT gets this case right, and
that ST alone gets it wrong.
The seven minutes of reconnects in the operator's log turned out to be the radio
switched off — no defect. But the log said otherwise:
20:14:42 yaesu: ID query failed (timeout) — continuing…
20:14:44 yaesu: connected on COM7 @ 38400 baud, model="FTDX101D"
Opening the port was treated as connecting, and the model name was left over from
the previous session, so a powered-down rig produced a line claiming it was
connected as a named model. That is what sent me looking for a software fault.
Connect now tracks whether ANY probe answered. If none did, it clears the stale
model, says so in terms an operator can act on — is it powered on, is the CAT
rate right — and returns an error so the reconnect loop keeps trying rather than
believing it succeeded.
Connect assigned a fresh handle to y.port without closing the one it already
held. It is called again on every reconnect, so a rig that fails to answer leaks
one handle per attempt — every few seconds, indefinitely.
Windows opens a serial port EXCLUSIVELY, which makes this visible as "yaesu: open
COM7 @ 9600 baud: Serial port busy" in an operator's log: OpsLog was competing
with itself for the port.
This is a real defect found while reading the reconnect loop, but I want to be
straight about its limits: it does NOT explain the seven minutes of "ID query
failed — timeout" in the same log, where the port opened and the rig stayed
silent. That one is still open, and the operator alternates between two separate
radios (a TCI SDR and the Yaesu) rather than sharing one port, so a conflict
between backends is not the explanation either.
Reported both ways round, which is what settles it. With RX and TX on SUB
everything worked; with them on MAIN the frequency froze and a spot click tuned
the sub VFO. The log shows the rig answering FR01 in the MAIN case.
So on an FTDX101 the state is the FIRST digit — 0 = main — and the second is a
separate parameter. Yesterday I changed this to the LAST digit, which read FR01
as SUB and inverted the whole thing. The fault that change was meant to fix ("SUB
shows MAIN") came from reading VS instead of FR, and probing FR had already fixed
it on its own; the digit change was an unnecessary guess layered on a real fix.
Nothing else needed changing for the operator's two questions: the display and
SetFrequency already follow curVFO, so a spot now tunes whichever VFO holds RX
and TX, and the main frequency is read when on main. A test states that in those
terms — active VFO and which command a spot writes — rather than only as a byte,
because the byte is what I got wrong while believing the logic was right.
The log named the cause in one line: yaesu: receive VFO is read through FR
(answered "FR01;"). An FTDX10 answers "FR0;" — one digit — so reading the digit
straight after the prefix worked there and took the 0 of "01" on an FTDX101.
The rig was on SUB and OpsLog said MAIN, which is the reported "Sub shows the main
frequency". It also explains the second half: split is "TX VFO differs from RX
VFO", so a receive VFO that alternated between right and wrong made the split flag
flip between consecutive polls and swap freq and rx in the log.
The state is now taken from the LAST digit before the terminator, which is correct
on both forms, and the same reading applies to ST/FT/VS rather than leaving the
next two-digit rig to be found in the field.
On the operator's question of whether freq should simply be the active VFO: no —
freq is the TRANSMIT frequency by ADIF convention (FREQ is the QSO frequency,
FREQ_RX the receive one, set only when split). In simplex they are the same, so it
does read as the active VFO; in split it must stay the transmit VFO or the log
records the wrong frequency. What the operator saw was this bug, not that rule.
"Showing 10000 of 23683 matches · 28648 total" was read as a filter matching more
than the log holds. The numbers were right — 23 683 matches out of 28 648 QSOs,
capped at 10 000 displayed — but the middle one belonged to "matches" while
sitting where "of N" normally means the total.
Each number is now named where it appears: shown · match the filter · in the log.
Thousands separators too, since 23683 and 28648 are hard to compare at a glance.
No counting change: both counts run over the same table, so matches can never
exceed the total.
Reported: after a confirmation download they all turn to Y.
The test accepted qsl_rcvd = Y as a confirmation. That field is the operator's
OWN paper-QSL flag, which they uploaded to QRZ in the first place — so QRZ handed
it back and OpsLog read it as QRZ's answer. On a log full of paper cards that is
most of the log, which is exactly what was seen.
Only QRZ's own statement counts now: app_qrzlog_status = C, or the ADIF field
qrzcom_qso_download_status = Y that QRZ sets for it. LoTW and eQSL confirmations
are explicitly not accepted either — they are other services' answers, and the
column says QRZ.
This status feeds award slots, so a false Y is a QSO counted as confirmed when it
is not. That asymmetry is why the rule is narrow rather than generous, and the
test lists the negatives as deliberately as the positives.
The changelog warns that earlier downloads may already have set the column
wrongly — the fix stops it happening again but does not undo it.
Asked for as a missing feature. It is not missing — the "fill my station fields"
option has stamped the default QSL / LoTW / eQSL / Club Log / HRDLog / QRZ.com
statuses on empty fields since v0.14. What was missing is any mention of it: the
option's description talked only about MY_* fields, so the behaviour was
undiscoverable except by comparing a log before and after.
The description now says it, in both languages, and names why it matters — a
WSJT-X log carries almost no QSL fields.
The rule is also pinned by a test now, which meant splitting the fill from the
settings lookup: blanks are filled, existing values are never touched.
Overwriting them would erase confirmations the operator has actually received,
and that is the half a future edit is most likely to get wrong.
Describes what the operator gets: a spot click in DXHunter now moves the radio,
not just the callsign field. Notes explicitly that a packet without <FREQ>/<MODE>
behaves exactly as before — that is the reassurance anyone already using
remote_call needs, and it is the part a feature description usually omits.
Appended at the bottom of the block, per the rule adopted with 0.22.0: a version
is read top to bottom.
The operator's description separates the two cases precisely, and that is what
identifies the fault. RX alone on SUB displayed VFO B correctly; RX AND TX on
SUB — entirely on the sub receiver — displayed VFO A.
VS was being read as "which VFO is in use". On an FTDX101 it does not answer that
question: with both RX and TX on sub it still reported the main VFO. FR is the
command that selects the RECEIVE VFO, and it is now asked first, VS remaining the
fallback for models that lack it.
Split falls out of the same correction: split means the TRANSMIT VFO differs from
the RECEIVE one, so a wrong receive VFO made the comparison wrong too — which is
the second half of what was reported. A test covers all four RX/TX combinations,
including the one that is NOT split (both on sub) and reads as split if you take
the transmit VFO alone.
Translating the schema correctly does nothing for a database that already
exists. This operator's had been created by an earlier attempt, so qso.state was
already TEXT and every launch died on the same statement — CREATE INDEX … ON
qso(state), error 1170 — with the connection failing at startup and no way
forward from the UI.
Columns that MUST be VARCHAR (the indexed and keyed ones) are now converted
before the migrations run. It reads information_schema, touches only the columns
in varcharColumns, and only when they are actually TEXT — a no-op on a healthy
database, and it preserves each column's NULL/NOT NULL.
A failure to repair IS reported: it is the difference between a logbook that
opens and one that does not. A failure to READ information_schema is not — a
server that will not answer it can still be running a perfectly good schema, and
refusing to open would turn a non-problem into an outage.
The quoting fix was necessary but not sufficient: "column 'op_name' can't have a
default value" persisted.
The translation worked line by line and rewrote only the FIRST column on each
line. Our own migrations put one column per line, so nothing showed there — but
sqlite_master stores a table's CREATE statement with every ALTER-added column
appended to the SAME line, and the baseline is built from exactly that. On a
logbook that had been through upgrades, most columns therefore kept their TEXT
type and MySQL rejected the schema.
Two consequences fixed together: every match is now rewritten, and each one is
decided from ITS OWN declaration — the text between the surrounding commas —
rather than from the whole line. Deciding per line would have made every column
sharing a line with a default-bearing one into VARCHAR(255), which breaches the
InnoDB row-size limit the TEXT rule exists to respect.
Tests cover the many-columns-on-one-line shape and pin that a neighbour's DEFAULT
does not leak.
Two errors from a shared MySQL logbook, both fatal to the schema and neither
naming its cause:
CREATE INDEX … ON qso(state) → 1170 BLOB/TEXT column used in key
specification without a key length
op_name TEXT … DEFAULT '' → 1101 TEXT column can't have a default value
One root cause. TEXT columns are rewritten to VARCHAR(255) when they are indexed
or carry a default — that is exactly what those two MySQL rules require — but the
pattern only matched a BARE column name. The SQLite baseline dump quotes its
identifiers, and mysqlDDL turns those quotes into backticks before the TEXT rules
run, so every column coming from the baseline kept its TEXT type and MySQL
rejected the schema.
The operator saw none of this: just "MySQL is enabled but the connection failed
at startup" and a silent fallback to local SQLite — which is the worst shape for
a shared logbook, because each operator then logs into their own copy.
The pattern now accepts a quoted or bare name and puts the quoting back. A test
covers both forms, plus the case that must NOT change: an unindexed column stays
TEXT, since VARCHAR everywhere would break MySQL's row-size limit. Reverting the
one-line change makes it fail.
Reported on an FTDX101: the panel showed split ON with the radio OFF, and the
reverse — while the same model behaved correctly for another operator, and an
FTDX10 was fine throughout. That pattern is the clue: one operator was on MAIN,
the other on SUB.
The two commands say different things and were read alike:
ST is a split FLAG — ST1 means split, whatever VFO is in use.
FT names the TX VFO — FT0 = transmit on A, FT1 = transmit on B.
Split is on when the rig transmits on a DIFFERENT VFO from the one it listens to.
Reading FT1 as "split" is therefore correct only on VFO A, and exactly inverted
on SUB. It is now compared with the current VFO.
Writing had the identical fault, and it was worse: sending FT1 for "split on"
while the operator listened on SUB CLEARED the split it was asked to set. Both
directions are pinned by tests over each VFO state.
Reported on an FTDX101 (F4NBZ): listening on the SUB VFO, clicking a spot moved
the radio to MAIN.
SetSimplexMode is the first thing SetFrequency tries, and it means "receive and
transmit here, simplex" — OmniRig applies that to the MAIN VFO by definition. It
is there because several Icoms ignore direct FreqA/FreqB writes, so it stays for
the main VFO; on SUB it is now skipped entirely and FreqB carries the change.
The generic "Freq" property is skipped on SUB too: on several rig files it IS the
main VFO, so writing it would move the VFO the operator is not using — or pull
them back to it, which is the bug being fixed.
An operator on SUB put themselves there deliberately. A spot click asks for a
frequency, not for a change of VFO — that distinction is the whole fix, and a
test pins which property each VFO state writes.
Reported with screenshots: a dozen records lifted out of wsjtx_log.adi into a new
.adi imported as "0 imported, 0 ignored, 0 total". The parser only emitted
records after <EOH>, and a hand-assembled file has no header — so the whole file
was read as header text and thrown away. The count says the file was empty; there
is nothing on screen to suggest a missing tag, which is why this was reported as
"import does nothing".
An <eor> is proof that a record ended, header or no header. Fields are now
collected always and DISCARDED at <eoh> — that tag is exactly the statement
"everything before this was the header", so real headers stay out of the data
just as reliably, and a headerless file keeps its first record instead of losing
it to the same rule.
An existing test asserted the old behaviour. It encoded a defensible reading of
the spec, but the file it rejects is one operators really make and the rejection
is silent, so the test is rewritten with the evidence that changed it rather than
deleted.
New entries were being prepended, so a version block read backwards: an operator
met "the Yaesu meters are corrected" and "the console follows the mode" several
entries before learning that a Yaesu console existed. The fixes made sense only
to someone who had followed the development.
0.22.0 is reordered so each feature appears before what was fixed on it, and
CLAUDE.md now says to append at the bottom.
Restored the entry that INTRODUCES the Yaesu console: I had overwritten it when
editing the meters entry in place, so the release described corrections to a
panel it never announced.
The byte trace from a real WK2 settled in one line what no amount of reading
could: "TX 11 32". Command 0x11 is SET KEY COMPENSATION in milliseconds, not the
dit/dah ratio — that is 0x17. So a neutral ratio of 50 asked for 50 ms of extra
key-down on EVERY element. At 25 wpm a dit is 48 ms, so elements more than
doubled and ran into each other: the reported "it sends one element, then a long
pause".
The ratio now goes out as 0x17, and the compensation is explicitly set to 0 —
merely stopping the wrong command would leave an affected keyer misbehaving,
since it keeps the value in EEPROM until something writes over it.
The init sequence is now built by a separate function so the BYTES are testable,
and a test pins each command number. These numbers are the contract with the
hardware, and a wrong one produces a fault that cannot be diagnosed from the UI
at all — this one cost the operator weeks and needed a trace to find.
This is why the fix waited for the trace rather than being guessed: the plausible
guesses (mode register, sidetone, WK1-vs-WK2 differences) were all wrong, and any
of them shipped blind would have broken the keyers that work today.
Renamed from 0.21.9 at the author's request: this release adds native Yaesu and
Xiegu CAT, CAT sharing over Hamlib NET rigctl, a Yaesu console and a fifth CW
engine, which is a minor version rather than a patch.
The Yaesu keyer had two entries — one written when it was added, one when the
FTDX10 turned out to refuse it. Merged into the single statement an operator
needs: what it does, where to select it, and which rigs accept it.
The weak audio was not a level at all: the radio was still modulating from its
front microphone, so almost nothing of the USB feed reached the air. On an FTDX10
that is MENU → SSB MOD SOURCE = REAR.
Worth putting in the settings hint rather than leaving in a conversation — every
operator wiring a voice keyer for the first time meets it, and no amount of gain
in OpsLog can compensate for a rig listening to the wrong input.
The messages went to the radio exactly as recorded. Nothing in OpsLog could
raise them, so a microphone captured quietly drove the rig quietly and the only
remedies were the radio's own USB input menu or the Windows mixer — which is
where the operator was heading. There is now a level from 10 to 400 %, applied
with clamping (a wrap would turn loud speech into noise on the air), and Play
previews at that same level so the adjustment is made against what will actually
be transmitted.
The PTT method list also named only OmniRig, Flex, Icom and TCI, so choosing a
native Yaesu left "CAT" with no backend beside it — which reads as "there is no
CAT PTT for my radio" and sends the operator to RTS on a COM port that has
nothing to do with the rig. Same list-needing-every-member shape as three
earlier bugs in this feature. The TestPTT log line had the same rot: it said
"CAT via OmniRig" whatever backend was running.
Receiving cleared the two bar percentages but not the values the bars now
actually draw from: the watts and the SWR ratio, added when the meters were
corrected. So the power bar sat pinned across the panel with the rig plainly
receiving — the dash in the label said one thing and the bar another.
Every transmit value is cleared now, and the peak-hold state with them: a peak
left in the holder would have carried the last transmission's reading into the
start of the next one, which is worse, being wrong while it looks live.
The panel also draws both bars from zero unless the rig reports transmitting, so
a value that has not been refreshed yet cannot show as output.
One point could not reveal the curve. Scaling 207 = 100 W straight down read
30 W where the radio showed 10, and 75 where it showed 50 — wrong everywhere
except at the single point it was fitted to.
Three readings taken against the rig's own display give the shape:
raw 62 → 10 W
raw 155 → 50 W
raw 207 → 100 W
Interpolating between them reproduces the radio exactly at those points and stays
close in between. I did not fit a formula: three samples can be made to support
several curves, and the operator can check a table against their own meter.
Above the top the last segment's slope continues rather than clamping, so a rig
driving an amplifier does not sit pinned at 100 W. A test pins the measured pairs
and the monotonicity, so a later change that breaks them fails against the radio
rather than against taste.
A quarter-per-poll decay covers the milliseconds between CW elements but not the
gaps that matter on the air: between the words of a CQ, in CW as in SSB, the
meters genuinely read 0 for most of a second and the bars fell with them.
A peak now stands for 1.5 s before it starts to fall, and still rises instantly —
a needle goes up fast and comes down slow. The SWR RATIO is not updated at all
from a zero reading: showing 1.0 during a word gap is worse than showing a stale
figure, because it looks like good news.
A test caught a real defect on the way: the proportional decay stalls on
integers. With the needle at 13 and the truth at 10, a quarter of the gap rounds
to zero and the meter sat three units high for ever. It now always steps down by
at least one, so it converges.
An FM carrier could not answer it: constant by definition, so nothing to
correlate. CW at 100 W did, because the keying itself varies the output:
key down: RM4=25 RM5=207 RM6=13
key up: RM4=25 RM5=0 RM6=0
RM5 follows the RF envelope exactly — it IS the power meter. RM4 sits near 25
whether the key is down or up, so it is not measuring output at all, and reading
it as power is what showed 8 W on a 100 W transmission. The operator's hunch was
right and my first reading of the ramp was wrong: what I took for a needle
rising was RM4 drifting, not tracking.
Watts are now derived from that meter (207 = 100 W, measured) instead of the
power SETTING scaled by a percentage — the setting says what was asked for, the
meter says what left the radio.
The bars also hold their peak with a gentle decay. In CW the meters genuinely
read zero between elements, so following the raw value made them flash to nothing
several times a second; a needle has inertia, and this only ever holds a value
the radio really reported.
One transmission cannot name it. RM4 rose while RM5 stayed flat — which points at
RM4 — but RM5 differed BETWEEN transmissions, 208 then 105, which points at RM5.
Both readings are consistent with either answer, so picking one now would just be
the FT-991A table mistake again in a new place.
The survey line now carries the rig's power setting, and the sample cap is raised
from 12 to 40 so two transmissions fit. Keying at, say, 10 W and then 100 W makes
the answer a one-line comparison: the wattmeter is whichever index moves with the
setting.
The displayed mapping is left alone until that comparison exists.
A second measurement at a known mismatch settled both the index and the scale.
At SWR 1.1: RM6=0. At SWR 1.5: RM6=52, while RM4 kept tracking the power.
52/255 = 0.204, which is the reflection coefficient of a 1.5 SWR to three
decimals. So the raw value is rho scaled to 255, and the ratio is
(1+rho)/(1-rho) — physics, not a curve fitted through two points, which is why
2.0 and 3.0 fall out of it correctly without ever having been measured.
The panel now shows that ratio, the number the operator reads on the rig, rather
than a percentage of meter travel — and a dash while receiving, since a stale SWR
from the last transmission reads as a live one.
Both measurements are recorded in the code and in a test, so the mapping is
evidence rather than a table borrowed from another model — which is exactly how
it came to read 81 in the first place.
Measured on the radio, steady carrier for three seconds:
RM1=0 RM2=unsupported RM3=0 RM4=9→18→21→22 RM5=208 flat RM6=0
RM4 is the index that RAMPS with the output, so RM4 is the power meter. RM5 sat
at 208 from the first sample to the last, unmoved by the power — that is not SWR
on an operator reading 1.1, and 208/255 is exactly the 81 that appeared on the
bar. Borrowing the FT-991A's table, which puts SWR on RM5, is what put it there.
SWR now reads RM6. It stayed at 0 throughout, which is CONSISTENT with a 1.1
match but does not prove the index — only a deliberate mismatch would, and I am
not asking for that. A bar at zero on a good antenna is honest; 81 was actively
misleading.
The measurement is written into the code next to the mapping, so the next person
sees the evidence rather than a table copied from another model.
The one-shot survey fired as the transmission began and caught the meters still
at rest — RM4=13 and everything else zero, which identifies nothing. And the
operator's reading points the other way: the bar showing 81 tracks his 100 W,
while the one labelled power sat at 8.
What names a meter is which index FOLLOWS the power over a few seconds of steady
carrier, so the survey now samples on every poll while transmitting, capped at a
dozen lines. Two seconds of tune will settle it.
Still not guessing at the mapping: the numbers will say which index is power and
which is SWR on this radio, and it gets corrected then.
Speed: with DTR/RTS line keying the PC does the timing, so the Yaesu console's
slider — which sets the rig's internal keyer — changed nothing audible and looked
broken. Both entry points now go through one handler that drives the engine
actually sending, and additionally sets the rig's own keyer whenever a Yaesu is
on CAT, so the radio's front panel shows the same figure.
SWR: an operator reads 80 on the bar with a real SWR of 1.1. That is the shape of
reading the WRONG METER — ALC, say — not of a scaling error, and which RM index
carries which meter is not consistent across the family. Rather than guess again
and move the wrong number somewhere else, the first transmission now logs RM1
through RM6 raw, once. The log will say which index is which on this radio, and
the mapping can then be corrected as a fact.
Tested on the radio: DAKY does not help, KY is refused whatever PC KEYING is set
to, and the "Serial port (DTR=CW / RTS=PTT)" engine on the second COM port keys
correctly. So my earlier "try DAKY first" was wrong, and the order is now the
other way round: name what works, mention the model that refuses.
The KY engine stays. It is documented for the FTDX101 / FT-991A / FT-710 family
and costs nothing to keep — a rig that refuses it now says so in one clear
sentence instead of transmitting nothing for no stated reason.
The operator found the menu I said to look for: it offers DAKY, RTS and DTR. KY
is the CAT route, so DAKY is the setting that would enable it — which makes my
previous "this rig has no CAT keying command" too strong a claim to leave
standing, since it was drawn from a refusal at the DEFAULT setting.
The error and the settings hint now name DAKY first, and keep the serial line
keyer as the fallback. That order matters: the CAT route costs nothing, the
fallback costs a second COM port.
Marked as untested rather than verified — I have no confirmation that DAKY makes
KY work on this model, only that it is the option that should.
Confirmed on the radio: it answers "?;" to KY, so this is not a setting to find.
My previous message sent the operator looking for a MENU → CW → PC KEYING entry I
named without checking it exists on that model, and they could not find it.
The message now names the path that does work: the "Serial port (DTR=CW /
RTS=PTT)" keyer on the rig's OTHER COM port — the standard one — while CAT keeps
the enhanced one. Same for the settings hint and the changelog.
The KY engine stays: it is documented for the FTDX101 / FT-991A / FT-710 family.
It now fails loudly and usefully on the models that lack it, which is the
difference between a dead feature and a wrong one.
The log timed it exactly: the CW send at 13:16:17.591, the CAT dropping at
13:16:17.633, forty milliseconds later.
An accepted KY says nothing, but a REJECTED one answers "?;" — and nobody was
reading it. The frame sat in the buffer until the poll loop's next query picked
it up, failed, and the Manager tore the link down. That is the disconnect on
every macro click, and it also explains the silence: the backend was being
rebuilt underneath the send.
Two changes. The KY write is now followed by a short read: silence means
accepted, "?;" means refused — and the operator is told so, naming MENU → CW →
PC KEYING, instead of getting nothing with no reason. And FA; retries once when
it meets a stray rejection, because a "?;" arriving there is almost never about
FA: the rig answers frequency queries perfectly well, it is the previous
command's refusal being attributed to this one.
This does not yet prove the FTDX10 accepts KY at all. It makes the next run say
so plainly either way, which is the point.
The log settled it: the rig answers "?;" — its "unknown command" — to KY; and to
MG;. So the keyer status query cannot work on this model, and every send spent
four seconds waiting for an answer that was never coming.
"?;" is now recognised as what it is. A command refused once is never asked
again: models implement different subsets, and re-asking costs a 600 ms timeout
on every slow beat for a control that will never answer — which is also why the
panel felt sluggish.
Without a buffer status there is still a real constraint: one KY command carries
24 characters and the rig DROPS the rest silently. So the send is now paced by
how long the text takes to key, from PARIS timing at the rig's own speed. A long
macro goes out complete instead of losing its tail.
That leaves the question the log cannot answer: whether KY <text>; itself is
accepted. If the rig also replies "?;" to it, the CAT menu's PC KEYING setting is
the next suspect — but nothing in the code will be guessing at it.
Clicking a macro dropped the CAT for a few seconds, keyed nothing, then came
back.
ask() returned the first ';'-terminated frame it saw, whatever command it
belonged to. KY produces NO reply, so the next query — FA; from the poll loop —
collected a leftover frame, failed to parse it as a frequency, and ReadState
reported an error. The Manager reads that as "lost the rig": disconnect, wait,
reconnect. Hence the drop and the automatic recovery a few seconds later.
Replies are now matched to the command that asked for them: anything else is
discarded and logged, so a stray frame costs one log line instead of the link.
This also explains the silence — the send never got a clean run at the port
while the backend was being torn down under it.
"the keyer buffer stayed full for 4s" and nothing keyed, on a rig whose CAT is
working. The buffer was not full: the check demanded the reply be exactly "KY0;"
with the digit at byte 2, so anything the FTDX10 phrases differently — a space
before the digit, another command's reply arriving first — read as "still full"
on every poll until the deadline.
The test is now asymmetric on purpose: only a clear "1" holds the send back.
Anything unrecognised goes ahead. Refusing to transmit because a status line was
phrased unexpectedly is the worse failure — the operator gets silence with no
explanation, where at worst sending early truncates a long message, which the
chunk loop then recovers from.
The reply is also logged once per run, since we have no verified sample of it —
that line is what turns the next surprise into a fact instead of a theory.
Three places decide what a rig keyer is, and the Yaesu was only in two of them.
The CW panel's status came from a list naming icom and flex; anything else fell
back to the WinKeyer status, which reports disconnected because no WinKeyer is
attached. So the panel said the rig CAT was offline while the console beside it
was reading the FTDX10 perfectly.
The send loop had the same gap: a rig keyer BUFFERS the whole message, so the
wait before <LOGQSO> is a length estimate, while WinKeyer watches a busy echo
that will never arrive here. Auto-call would have raced the transmission.
This is the third time the same shape has bitten in this feature — a list of
engines that needs every member named, with a silent fallback for the rest.
Grepping for the pair "icom || flex" is what finds them.
Choosing "Yaesu (rig keyer)" saved correctly and still produced a WinKeyer panel
asking for a COM port. The engine is normalised on load through a chain that
names icom, flex and serial and maps EVERYTHING ELSE to "winkeyer" — so the
value came back as WinKeyer on every read, whatever the settings said.
That is why the panel offered COM3 then COM10: it genuinely believed the engine
was a WinKeyer. The list now names yaesu too.
The same shape of bug is worth watching for: a normaliser with a silent default
turns an unknown value into a plausible one instead of an error, and the symptom
appears far from the cause — here, in a panel three components away from the
setting.
Selecting the Yaesu keyer fell through to the WinKeyer branch of the settings,
which offers a serial port picker and a Connect button. So the operator went
hunting for the right port — trying the standard COM and the enhanced one in turn
— and neither connected, because this keyer uses NEITHER: it keys over the CAT
link already configured in Settings → CAT.
It now has its own branch, like the Flex one: a speed field, the warning when the
CAT backend is not a Yaesu, and a line saying explicitly that keying rides the
CAT link. Speed changes go to the rig's keyer instead of a WinKeyer that is not
there.
Reported on an 8x2: port A correctly on the 80 m vertical, then a few seconds
later both A and B shown on the same beam.
The display preferred the TX antenna, falling back to RX. On an 8x2 only ONE port
can hold the transmit antenna, so the switch reports the same txant on both — and
every keepalive poll redrew port A as whatever port B transmits through. The RX
antenna is the per-port selection, and the only thing Activate sets, so it is
what a port shows; TX stays the fallback for a port reporting no RX antenna.
A port change is now logged with the raw line — but only on CHANGE, since the
device pushes state every few seconds and logging each one would bury the rest.
Without any trace, "port A jumped to the wrong antenna" cannot be checked.
Two tests pin it: rx wins over a disagreeing tx, and a message about one port
never moves the other — which is the shape that was actually on screen.
The radio has a keyer and a command to feed it (KY), so an FTDX10 needs no
WinKeyer and no second cable, exactly as the Icom CI-V and Flex CWX engines
already do. The rig keys with its own timing, which is why the spacing is right
where a PC keying a line through USB latency drifts.
Text is filtered to what the keyer can actually send: an unsupported byte does
not produce an error on a Yaesu, it can abort the whole buffer, so the rest of a
macro would vanish silently. It is then fed in 24-character pieces, waiting for
room between them — the rig DROPS what does not fit, again with no error, so a
contest CQ would lose its tail.
STOP is the honest gap. Yaesu documents no buffer-clear, so it drops the
transmitter (TX0) instead: nothing queued reaches the air, which is what Escape
means to an operator. It deliberately does NOT send "KY0;" — a plausible-looking
clear that the rig would read as the CHARACTER zero and transmit.
A test caught a real one on the way: tabs and newlines were dropped as
"unsupported", gluing the words either side together, so a macro written on two
lines went out as CQCQ. Whitespace now becomes a word gap before filtering.
Settings warn when the Yaesu keyer is selected without the Yaesu CAT backend —
otherwise it simply never keys, with nothing on screen saying why.
Send path follows the CAT reference and how Hamlib drives these rigs; NOT yet
verified on the air.
Confirmed on the FTDX10: receiving on 14.018.98 and pressing SPLIT put 14.019.98
in the big display with "TX 14.019.98 (0 kHz)" under it. The radio was right —
RX 14.018.98, TX 14.019.98, up 1 kHz on CW — the panel was not.
RigState follows the ADIF convention where freq_hz is the TRANSMIT frequency, so
under split it is the OTHER VFO. Taking it as the main display showed the
operator the frequency they transmit on, and then an offset of that frequency
against itself: zero.
The header now reads the listening frequency (freq_rx_hz when split, freq_hz
otherwise) and the TX line shows the real transmit frequency and offset.
Microphone gain and VOX are meaningless in CW: the rig ignores both, so showing
them is showing dead controls. They are hidden, and a CW card takes their place
with keyer speed (KS), break-in (BI) and ZIN (ZI), the zero-in that retunes so
the station being received lands on the operator's own pitch.
ZIN is a one-shot with no state, so it is a plain button rather than a chip that
would look latched, and no settings read-back follows — the frequency change
arrives through the normal poll like any other.
The keyer values are read on the slow beat whatever the mode, so the card is
already populated the instant the operator switches to CW instead of filling in a
poll cycle later.
Which controls show is decided by the RIG's mode, not the logged one: the logged
mode can be a digital sub-mode the radio knows nothing about.
The bar was one colour for its whole travel, so nothing distinguished a signal
that is merely readable from one that belongs in the log as 59+20.
Green to S9, amber through the S9+ range, red from +20 dB. The thresholds are
computed from the SAME S9 point the label uses rather than hard-coded
percentages: the S9 position is still a hypothesis on this rig, and when it is
corrected the colours have to move with it or the meter would say 59+20 in
amber.
Reported on the FTDX10: listening on 14.244 with VFO B still holding 18.115 from
an earlier session, pressing SPLIT threw the transmitter onto another band. The
button only flipped the rig's split flag, and the other VFO is stale by nature —
the only transmit frequency that makes sense is one derived from where the
operator is listening NOW. SPLIT therefore places the TX VFO too: up 1 kHz on CW
and the data modes, up 5 kHz on phone, the offsets operators actually call. The
+1k / +5k buttons remain for anything else, and a test pins the mapping.
The panel also drew its own flat meters while the Flex and Icom consoles use the
shared LED-segment MeterBar. Two instrument styles in one application is just
inconsistency — it now uses the shared component, and the local one is gone.
And the three consoles are named alike: "Flex Console", "Icom Console", "Yaesu
Console", in the tabs and in the Main-view pane list, in both languages.
Three corrections from the operator's second pass.
The sideband gesture was wrong: I used double-click, which hides the action.
Clicking a button that is ALREADY active now flips its sideband — CW-U → CW-L →
CW-U. One button, one finger, nothing to discover.
A lit SPLIT chip does not tell an operator anything useful: it says split is on,
not where they transmit. The header now shows the TX frequency and the offset in
kHz whenever split is active.
And the offset that matters is set in one action: up 1 kHz on CW, up 5 kHz on
phone. Doing it by hand means swapping VFOs, retuning and swapping back — exactly
the fumbling a panel exists to remove. Both are offered rather than picked from
the mode, because which one is idiomatic is the operator's call, and the button
turns split on at the same time.
The offset is measured from the RECEIVE frequency and written to the VFO we are
not listening on, so it stays correct when the operator works on VFO B, where the
roles are mirrored.
Capping the width without mx-auto pinned the console to the left edge with an
empty window beside it. It now uses the exact wrapper the other two panels use —
h-full min-h-0 overflow-auto bg-background, then max-w-5xl mx-auto p-3 — rather
than a second layout of my own invention.
First look on the FTDX10 turned up six things:
The S meter printed a raw percentage — "57" tells an operator nothing, and it is
the S number that goes into a report. It now reads S1-S9/S9+dB, the same value
the click-to-fill RST already used.
CW, RTTY and the data modes exist on BOTH sidebands and the operator is the one
who knows which they want. The buttons now carry the rig's actual sideband and a
double-click flips it; PSK is added, riding the rig's DATA mode as it does on the
radio itself. This also means the mode row drives the rig directly (MD0 with the
exact mode) instead of going through the ADIF path, which could only pick a
sideband by convention.
The attenuator is a 6/12/18 dB pad on these rigs, not the single step I assumed —
two thirds of the control was unreachable.
Sliders had no visible filled side: --muted is barely lighter than the card it
sits on, so the whole track read as one bar. They also came in three kinds (two
bare range inputs among them). One component now, explicit track colour, and it
takes a min/max so power in watts and DNR 1-15 look like the rest.
The panel stretched across the whole window; it is a column of controls, so it is
now capped and every row stays readable.
And it gets its own Yaesu tab, like FlexRadio and Icom, rather than only being
available as a Main-view pane.
A console pane for the native Yaesu backend, in the same shape as the Icom and
Flex ones: S/PO/SWR meters, band and mode rows, AF/RF/squelch, AGC, the IPO/AMP1/
AMP2 front-end selector, ATT, NB, DNR + level, narrow filter, power in watts, mic
gain, VOX, split and ATU tune.
Three decisions worth keeping:
Panel reads are STAGGERED and live in their own file, away from ReadState. Meters
poll every cycle; settings only change when someone turns a knob, so they refresh
every 8th cycle and right after any set. Polling all of it every cycle would put
twenty queries a second on the serial link the frequency display shares.
Band buttons use the rig's own band memory (BS) rather than a frequency we pick,
so 20 m lands where the operator last was on 20 m — what the radio's own band
keys do.
A set is followed by a read-back, so the panel shows what the RIG ended up with,
not what we asked for; the two differ whenever a value is out of range or the
mode forbids the control. And a control the model lacks keeps its previous value
instead of dropping to zero, which reads as a setting that reset itself.
The S9 point of the S-meter scale is a hypothesis (the manual does not state it)
and is commented as such — one number to correct if reports come out an S unit
off, rather than a fudge spread through the RST helper.
Also removes the FlexRadio settings blurb, which explained the backend to
someone who had already chosen it.
Xiegu speaks CI-V with a reduced command set: frames, BCD encoding, addressing
and the opcodes for frequency, mode, PTT and split are Icom's, so this reuses
internal/cat/civ wholesale instead of re-deriving a codec.
It is a SEPARATE backend rather than the Icom one at address 0x70, because what
the two rigs do NOT share is the deciding part. The Icom backend reads the
spectrum scope, the DSP block, data mode via 1A 06 and the model id via 19 —
none of which a Xiegu implements. Pointed at a G90 it would poll every cycle for
answers that never come, and spend its silence tolerance on commands the radio
was never going to support.
Two consequences of the rig's smaller mode table are handled explicitly rather
than left to fail: there is no data mode, so a digital QSO is set to plain
sideband (what the operator does on the radio anyway) instead of being refused;
and the split TX frequency is NOT reported, because reading the unselected VFO
needs 0x25, which the Xiegu table does not list — a split flag carrying a wrong
TX frequency is worse than the flag alone, since the frequency is what gets
logged.
The published command table has rows that slipped during typesetting (0x07 and
0x0F share a block). Where it contradicts itself the Icom meaning is used, the
rest of the table matching Icom exactly, and every unexpected reply is logged
raw so a first on-air run settles it.
NOT yet verified on a radio.
MSHV worked immediately, JTDX answered "Hamlib error: Invalid parameter while
setting frequency". The two use different Hamlib dialects: MSHV sends
"F 14074000", JTDX names the target first — "F VFOA 14074000". The VFO name
landed in the slot the frequency was read from, the parse failed, and we returned
RPRT -1, which is exactly the error JTDX reported.
A leading VFO name is now stripped from every command's arguments. It costs
nothing: OpsLog follows the rig's own VFO selection, so the name carries no
information we act on — but refusing it locked out a whole family of clients.
Both dialects are covered by a test, the plain one included, since this is an
addition and must not become a swap.
A rejected frequency is also logged with the raw line now. The client only shows
"Invalid parameter", which says nothing about what it actually sent — that is
why this took a screenshot to diagnose rather than a log.
A native CAT backend owns the rig's serial port, and Windows gives a COM port to
one process — so choosing native CAT locked WSJT-X, MSHV and JTDX out of the
radio entirely. That is the cost of dropping OmniRig, which was itself a sharing
layer, and it has to be paid back.
OpsLog now becomes the server, as wfview does. It speaks the Hamlib net rigctl
protocol, which every one of those programs supports natively (rig model "Hamlib
NET rigctl", 127.0.0.1:4532) with no driver to install. It sits in front of the
MANAGER, not a backend, so an operator on OmniRig, Flex, Icom or TCI gets the
same server.
Two details that decide whether a client works at all rather than degrading:
dump_state is parsed positionally and WSJT-X refuses to proceed without a
well-formed block, so it is written out in full and its shape is pinned by a
test; and set_vfo / set_split_vfo answer RPRT 0 rather than an error, because
OpsLog follows the rig's own VFO and a refusal makes WSJT-X abandon the
connection. Unknown commands answer RPRT -11 — never silence, which hangs a
client instead.
The whole protocol is tested against a fake rig, plus one end-to-end exchange
over a real socket, since the framing is as much the contract as the text.
Also: the Yaesu backend is confirmed working on a real FTDX10 (frequency, mode,
VFO, split), so its "not yet verified" note is now wrong and is corrected.
Every Yaesu fault reported so far came from OmniRig's interpretation layer, not
from the radio: a rig file that never exposes the VFO, a Freq property meaning A
on one model and B on another, a split flag that alternates between polls. This
talks to the rig directly, so what the radio answers is what is shown.
Modern Yaesu CAT is plain ASCII with a ';' terminator — FA/FB for the VFOs, MD0
for the mode, VS for the selected VFO, TX to key. Frequency is written to the VFO
the operator is actually on, not always to A, which is the failure that made a
display disagree with the radio.
Two things are genuinely uncertain across the family and are treated as such
rather than guessed. SPLIT is read through ST, then FT if the rig ignores ST —
whichever answers wins and the choice is remembered, because the two commands say
DIFFERENT things (a split flag vs which VFO transmits). If neither answers, split
is reported OFF and the fact is logged, rather than invented. Unknown model ids
and mode bytes are logged raw for the same reason.
Split resolution, frequency parsing and the mode mapping are pure functions with
a table test — the OmniRig equivalent is where every Yaesu bug lived, and it had
no test until late.
Written from the CAT reference; NOT yet verified on a radio.
The log settled it: the backend was right all along. Every set produced
"readback +1.5s FreqA=21140000 Freq=21140000 -> shown 21140000" and published
cat:state with the new frequency and band. The display did not follow.
The fault is in the frontend. Changing band does two things at once: it opens the
1.5 s freeze that protects what the operator is typing, and it commands the rig.
The rig's answer comes back in ~170 ms — inside that freeze — and the handler
DROPPED any snapshot arriving during it. The backend only emits on change, so
nothing came afterwards, and the strip kept the old frequency until the VFO was
nudged. Changing band from the radio always worked because no freeze was open.
A dropped snapshot is now kept and replayed when the freeze closes, rather than
discarded. Further typing simply defers the replay again.
An import that drops the locator leaves it missing on a whole batch, and fixing
that one QSO at a time is exactly what bulk edit exists to avoid.
Added on all three sides in lockstep — the dialog field, the field-to-column map,
and the repository whitelist — because they are separate lists and offering a
field the repository refuses fails only at Apply, after the operator has selected
the QSOs. The existing contract test covers it.
The callsign and RST stay excluded, as before: bulk-setting those corrupts a log.
The locator does not carry that risk — it identifies a place, not a station, and
a wrong value is simply overwritten again.
An operator reports that "QRZ.com received status = N" returns rows showing both
N and Y (issue #5 follow-up). The SQL is a plain col = ?, but reading it cannot
distinguish a wrong query from a UI that kept the previous rows after an error —
so this runs it: five QSOs inserted, filtered, and both the list and the count
asserted. The backend filters correctly, list and count agree. The fault is not
in the query.
Found while looking: the filter builder tested an ADIF date with /^d{8}$/ instead
of /^\d{8}$/. The escape was missing, so the branch never matched and the
calendar input was handed "20260728", which type=date rejects — an empty box
over a value that was really stored.
Adds the IC-7300MKII at CI-V 0xB6.
Doing so exposed a drift between the two hand-kept copies of the model table: the
settings offered the IC-7700 at 0x88 and the IC-7800 at 0x80, which are the
IC-7100's and the IC-7410's factory addresses. Picking either set an address the
rig never answers on — the symptom is a radio that simply stays silent — and the
backend then named it as the other model. Corrected to 0x74 and 0x6A, and the
four models the backend already knew (IC-7100, IC-7410, IC-7600, IC-7851) are now
offered too instead of forcing a manual address.
A test reads the model list out of the .tsx and asserts civ.ModelName agrees, so
the next model added on one side alone fails the build rather than someone's
radio.
The network backend treated "control link alive but no CI-V reply" as the rig
being in standby, and tolerated it WITHOUT BOUND. When another program takes the
CI-V session — WSJT-X through OmniRig, or the Remote Utility — the rig goes on
answering pings on the control stream while sending us nothing at all. Alive()
stayed true, so ReadState returned the cached frequency with err == nil, the
Manager never saw a failure, never reconnected, and re-published a frozen number
with a fresh timestamp on every poll. Only restarting OpsLog cleared it.
Bounded now, on the last SUCCESSFUL read. Past the grace the error is reported so
the Manager tears the session down and reconnects, which re-takes the CI-V
stream. Also fatal immediately: the CI-V reader goroutine having exited — no read
can ever succeed after that, however healthy the control link looks.
The grace backs off to minutes when the silence persists, because the two cases
pull opposite ways: a stolen session recovers on the first attempt, while a rig
switched OFF is silent for hours and re-tearing its session every 30 s would
blink the panel — and its ON button — away continuously. A good read resets it.
The decision table is pinned by a test; the standby case (never answered since
connect) keeps the old tolerate-for-ever behaviour.
ULS was wired into AddQSO alone, so an operator who downloaded it and has no
QRZ.com/HamQTH account saw nothing while typing a US call: cty.dat gives the
country, the zones, and a 4-character grid that is the ENTITY centroid — a
thousand kilometres from the station. The county and the real square were stamped
after logging, too late to point an antenna with.
The enrichment runs last in the lookup wrapper and only fills blanks, so a
provider always wins. ULS carries no name and no address, so it completes a QRZ
record and can never replace one.
The grid needed a rule of its own. refineGrid keeps what is there unless the new
value extends it — correct for a QRZ square, wrong against an entity centroid,
which is simply a different square. A per-callsign FCC square therefore beats any
4-character grid, while a 6-character one already present is left untouched.
Lat/lon are recomputed only when we actually moved the grid.
Portable calls are skipped: W1AW/4 is not what the FCC licensed.
The profile's MY_* metadata is derived from the callsign through cty.dat, and the
effect that derives it ran on every load — so opening the settings counted as
"the source changed" and overwrote whatever the operator had typed. Reported by
an operator in CQ 4 / ITU 4 handed 5 and 9: he corrected them, and each restart
put 5 and 9 back.
cty.dat gives the zones of the ENTITY, and a large country spans several, so the
automatic value cannot be treated as authoritative — it is a starting point. A
load now fills only fields that are EMPTY; a recompute that overwrites happens
only when the callsign or grid itself changes, which is the case the derivation
exists for.
The load-vs-edit distinction is a ref holding the profile id the values were last
derived from — first sight of a profile is a load.
"Last download" and an operator-chosen date are different questions, and they
now ask QRZ different ones. The automatic run wants everything TOUCHED since it
last ran (MODSINCE) — that is what returns a 2015 QSO confirmed yesterday. A date
typed by the operator names a period of OPERATING, so it maps to BETWEEN date and
today, on the QSO date.
The client-side date filter is skipped only under MODSINCE, where it would throw
away the older-but-newly-confirmed records the query exists to fetch. Under
BETWEEN it agrees with the server and stays on as a backstop.
The date window was sent as OPTION=AFTER:<date>. That was a guess, and QRZ
rejects the request outright — RESULT=FAIL with no reason — so the confirmation
download failed completely for everyone, whatever the account.
The documented option is MODSINCE, and it is the better window anyway: it selects
records MODIFIED since the date, so a QSO from years ago that was confirmed
yesterday comes back. A QSO-date window structurally cannot return those, and
confirmations are the whole point of this download — which is why the client-side
date filter is now skipped when the server did the windowing, instead of throwing
those same records away on arrival.
A refusal now falls back to ALL rather than leaving the operator with nothing.
Reported as a regression — reopens on the wrong screen — but the window code is
untouched since it was verified working: the only change to main.go since is the
background colour. So the fault is in the DATA, and there was no way to see it.
Both ends now log their coordinates, plus where the window actually landed after
the un-maximise / move / re-maximise dance. That separates the cases, which need
opposite fixes: nothing was captured, something wrong was captured, or the right
corner was captured and Windows moved the window anyway.
Reverts 30d8827. Ctrl+C / Ctrl+V already work in the app, so the WebView's
default menu only added Refresh / Save as on the neutral areas — noise in an
application window, for nothing gained. Changelog entry dropped with it.
Wails disables the WebView context menu in production, so an operator could not
paste a cluster address, an API key or a callsign anywhere in the app. Ctrl+V
did work — nothing intercepts it — but the menu is where people look, and a key
copied off a web page is exactly what you paste rather than retype.
EnableDefaultContextMenu turns it back on.
Every port box was parseInt(e.target.value) || <default>. Delete the contents
and parseInt gives NaN, so the default is written straight back: the digits
reappear under the cursor and the only way to enter a different port is to
overwrite it character by character. Reported on the cluster editor; the same
line existed in eight other places.
One PortInput component now holds the raw TEXT as its state and only tells the
parent about a value that is actually a port. An empty box stays empty while you
type; on blur an invalid one falls back to the default, so nothing is ever saved
as 0 or NaN. It adopts a value arriving from the backend only while the operator
has not started typing — re-syncing mid-edit is the behaviour being fixed.
This is the controlled-input trap the project notes already warn about: keep the
raw text in local state and derive the stored value from it.
Same slip as before the 0.21.6 release: I appended to a block that had already
shipped. 0.21.7 is back to its 12 published entries; the microwave-band fix
opens 0.21.8.
Reported on 3 cm. cat.BandFromHz stopped at 23 cm, so a 10 GHz frequency came
back EMPTY — and an empty band is not cosmetic: the QSO is logged without one,
counts in no award slot, and exports with no BAND field. The low end was short
too (2190m / 630m / 560m).
Four band tables had to be extended because four exist: the CAT one, the award
plan in app.go, the cluster spot classifier, and the frontend's. Plus the places
that LIST bands — the QSO editor and award-definition pickers (you could not set
3 cm by hand either), the statistics axis, and the award band ORDER, where a
missing band sorts to the end instead of in frequency order.
Ranges and names are ADIF 3.1.7, which is what an export has to carry.
A test now pins one frequency per band across the Go tables and asserts they
agree — that is what was missing, four hand-maintained copies with nothing
checking them. It also pins that an out-of-band frequency stays empty rather
than snapping to the nearest band, which would file a QSO under a band the
operator never used.
Left out as "internal", but the log is something operators read and send —
and the Flex meter removal, which is the same nature, was listed. Consistent
now.
An FTDX10 operator reports that changing band from OpsLog moves the rig but the
displayed frequency stays on the old band until they nudge the VFO. The existing
readback is taken immediately after the write, so it always shows the old value
and proves nothing: OmniRig queues the CAT command and sends it over serial.
A second readback is now logged ~1.5 s later, from ReadState (the goroutine that
owns the COM apartment), together with what OpsLog concluded. That separates the
two candidates, which look identical to the operator: the rig ignored the write,
or the rig moved and its rig file never re-reads the frequency.
The backend already parsed the atu object and exposed ATUStart / ATUBypass /
SetATUMemories through the Flex controller and the App bindings — only the UI
was missing, so the radio's own tuner was the one thing on the panel you still
had to reach SmartSDR for.
Placed under TUNE, which is the order of operations: the tuner needs a carrier
to measure against.
The status is decoded into words rather than shown raw. That is the point of the
block: TUNE_FAIL and "never tuned" leave the radio looking identical on its own
front panel, and the operator transmits into a bad match either way. Failed is
red, tuning amber, tuned green.
Switching station is the frequent act and editing a profile the rare one, but
the callsign opened the settings — so changing station took four clicks and a
scroll. It is now a dropdown of every profile: pick one and it activates. The
active one is ticked and disabled, and "Manage profiles…" stays at the bottom
for the rare case.
The list reloads on every profile change, so a profile added or renamed in the
settings appears without a restart.
The UTC clock leaves the header for the status bar, next to the database: in the
header it sat beside the frequency in the same weight and competed with it for
the eye, while being something you consult rather than watch.
0.21.6 was published before these landed, so its notes must stay exactly as they
shipped — a released block is a record of what an operator actually got, not a
place to keep appending. Restored it to the 11 entries in the release commit and
opened 0.21.7 for the eight that followed.
The bulk-edit line was the one subtlety: it was published in 0.21.6 and then
REWRITTEN in place to cover the filter builder. 0.21.6 keeps its published
wording; the filter half is new and gets its own 0.21.7 entry.
An operator reported migrations being very slow and sent a log with three
interleaved runs — and no way to tell one database migrated three times from
three databases migrated once. Each line now carries its target: the SQLite file
name, mysql:<database>, or baseline:memory for the in-memory pass that derives
the schema for a FRESH MySQL database (that one is fast and is not a real
database — in the report it sat between two slow runs and looked like a third).
A pass also announces how many migrations it will apply, so a run that applies
nothing is visibly distinct from one that does.
Three faults compounding, all visible in one operator's report (117k QSOs):
1. We asked for OPTION=ALL every time, so QRZ serialised the entire logbook —
56 MB — and cut it short. The parser stopped at record 27832 with "unexpected
EOF" and everything past that was never read. The window is now sent to the
server (OPTION=AFTER:date); the client-side date filter stays as a backstop.
2. The last-download date was stored RIGHT AFTER the fetch, before knowing
whether the records had been processed. So the truncated run still advanced
the window past data it had never read — and every later run skipped it. It
is now written at the end, and only when the ADIF parsed cleanly. A window
that moves past unseen data is worse than one that re-reads.
3. Which is why the run in the report ended "Confirmed 0, added 0 (of 0
returned)" despite parsing 27832 records: an earlier truncated run had
already pushed the window to yesterday, so every record was filtered out
client-side. The operator could not recover without clearing the key by hand.
A truncated download now says so in plain words, and says the window was not
moved.
My mistake yesterday: the new confirmation columns went into uploadStatusCols —
the QSL Manager's FILTER whitelist — instead of bulkEditableCols. The dialog
listed QRZ.com received status and the ten dates, and the repository refused
them at Apply, after the operator had selected 54 QSOs.
Moved to the right map. And two fields have been broken this way since long
before: State and County were offered under "Contacted station" (whose IOTA /
POTA / SOTA / SIG siblings are all allowed) but were missing from the whitelist,
so choosing one always failed. Added — they are exactly what an import loses and
what a whole run shares.
The real fix is the test. Two lists in two languages had to agree by hand and
nothing checked it, so the failure only ever surfaced as an error message at the
last click. The new tests read the ACTUAL field ids out of BulkEditModal.tsx and
FilterBuilder.tsx and assert the repository accepts every one — a field added to
the UI alone now fails the build instead of the operator.
They arrive many times a second and were dumped every 5 s, drowning the rest of
the log. They only ever existed to confirm the meter NAMES during development,
which the subscription line already does — and that one is now a single summary
instead of one line per meter, a Flex announcing dozens of them at each connect.
Grid on the UDP path — WSJT-X and MSHV can only send a FOUR-character grid,
that is all the FT8 protocol carries. The enrichment rule there is "fill only
what is empty", so the coarse JN05 always won and QRZ's JN05JG was thrown away:
roughly 100 km of accuracy discarded on every digital QSO. The lookup grid is
now taken when it EXTENDS the received square. A lookup that disagrees outright
(JN18 against JN05) is not a refinement — the station may be portable and what
came over the air is then the better record. Both rules are pinned by tests.
Manual fetch in the QSO editor — it read the CACHE, valid for 30 days. An
operator who upgraded their QRZ subscription went on getting the thin
free-account record and clearing the cached row by hand was the only way out.
The button now forces a lookup that bypasses the cache, reaches the provider and
REFRESHES the stored row, with a 15 s budget instead of 2 s: a deliberate click
must reach the network, where giving up early would fall back to cty.dat and
look like nothing happened. The automatic type-ahead lookup keeps the cache,
which is where it earns its keep.
Same fetch: it used `??`, which only guards against null. Go marshals an unset
string as "", so a QRZ record with no grid BLANKED the grid already in the QSO.
The lookup still wins — that is the point of asking for it — but an empty
result no longer erases a good value.
"Upload" was wrong for half of them: a paper QSL is not uploaded. Every entry
now says whether it sets a STATUS or a DATE, and in which direction — "QRZ.com
sent status", "LoTW received date" — which is also how the ADIF fields read.
Added: the QRZ.com received status (the column existed, it was simply never
offered) and the ten confirmation DATES — QSL / LoTW / eQSL / QRZ.com in both
directions, Club Log and HRDLog outbound only, those two having no return
channel. Dates use the same picker as the QSL Info tab, with a Today (UTC)
button; clearing the field writes an empty value, which is how a wrong date is
removed from a batch.
Fields are grouped by CHANNEL now, each status followed by its date, in the same
order as the QSL Info tab — a longer list, but nothing to hunt for.
The repository whitelist is extended to match: without it the new columns would
have been refused at write time. The old ids (qrz_upload, clublog_upload,
hrdlog_upload) still resolve, in case anything still holds one.
Statistics — the activity chart ignored the period selector: picking a year up
top left it showing the last 7 days, two controls contradicting each other. The
buttons were four fixed rolling windows anchored on the newest QSO, not
granularities. They now choose the BUCKET SIZE over the selected period, with an
Auto default and a step-up when a bucket would be too fine (the backend drops a
series past 750 buckets rather than ship 6000 unreadable bars). The hour-of-day
histogram covered a single day, too little to read anything from; it moves to
its own card with a weekday view, over the whole period.
Band map — the CW/data/phone sub-bands were shaded with the STATUS tokens, the
same amber that means "new band" on the pills drawn on top of them. A sub-band
is an identity, not a state: categorical hues now, validated in both themes
(violet failed on the dark steps — 1.9 ΔE from blue for a protan reader) and
added to the legend, since identity must never be colour-alone.
Frameless window — the OS title bar was a dead 32px band above a window that
already has its own. Minimise/maximise/close move into the app header, which
carries the drag region and double-click-to-maximise. The drag opt-out is by
ROLE in CSS, so a future button in that bar keeps working without anyone
remembering the rule.
Fixes:
- Saving settings froze the window while a device was slow: restarting a link
waits for its poll goroutine, which can sit 45 s inside an uninterruptible COM
Connect when another program holds the rig. The restart is now off the UI
path; a slow one is logged.
- Multi-screen: a MAXIMISED window was never repositioned, so it came back on
the primary screen every launch. The corner is now recorded even when
maximised (it names the monitor) and re-applied while still hidden.
- The Callsign field is marked out at rest — operators were typing the call into
Name, which sat beside it and looked identical.
- QSL dates get a real picker (native, for the OS calendar and locale order); a
malformed old value stays in a text box rather than vanishing behind an empty
one.
Also: a Support OpsLog entry in Help, and a WinKeyer protocol trace. The WK2
report (one element then a ten-second pause, sidetone that will not switch off)
is NOT fixed here: the WK1/WK2/WK3 command sets differ and a guessed fix would
break the operators it works for today. The trace logs every byte with the
command name and spells out the firmware family, which is what will settle it.
Moving the folder (C:\OpsLog → D:\OpsLog, or onto a stick) broke everything:
config.json and each profile's logbook path were absolute, so on the new
machine the pointer named a drive that no longer applied.
The logbook case was the dangerous one. db.Open CREATES what is missing, so if
the stale C: path happened to be creatable, the operator silently got a NEW
EMPTY logbook instead of an error — with their QSOs sitting untouched in the
folder they had just copied.
A path inside the application folder is now stored relative to it and resolved
against the current location at read time. A path OUTSIDE it (a synced folder,
a chosen drive) stays absolute and untouched — that is a deliberate choice; it
is only re-rooted when it has gone missing AND a file of the same name exists in
this install's data folder, i.e. the copied-folder case. With no such twin the
path is left alone so the failure is reported rather than papered over with an
unrelated database. Both rules are pinned by tests.
The Main-tab pane and the full Recent QSOs tab are the same component sharing
one storage key, but the pane is about half as wide — so it wants fewer and
narrower columns. Whichever was opened last rewrote the other's widths.
The pane now stores under "mainpane", which also scopes its award-column
visibility and widths. The full tab keeps the historical key, so its layout is
untouched; the pane starts from defaults once.
1. A language change rebuilt the column defs, which sets the anti-clobber
guard, but the effect that clears it listened to only two of the rebuild's
causes. Guard stuck on → every column change was silently discarded for the
rest of the session. Now keyed on the rebuilt defs themselves, so any future
cause is covered.
2. The cluster grid had no guard at all: the same rebuild wrote its defaults
over the saved layout, in the cache AND the database. Unrecoverable.
3. Worked-before and cluster read their state before the active profile was
known, so they looked up the unscoped cache key (always a miss) and then
saved under the scoped one. They now wait for the scope and remount on a
profile switch, like the main log.
4. GetUIPref returns an ERROR, not "", while the settings store is not yet
scoped — deliberately distinct from "unset". The frontend read it as "no
preference", rendered defaults, and the first column event overwrote the
good copy. It now retries instead of giving up.
5. The QSL manager had no storageKey, so it shared the main log's layout and
each rewrote the other.
Also: the DB write is debounced (a resize drag fired dozens of writes) with a
flush on close, and a rejected write is retried rather than dropped.
On its CAT/AUX connector an ACOM is the MASTER: it polls a transceiver and
changes band from the reply, so nothing can be pushed to it. internal/catemu
answers those polls on a second serial port, in ACOM command set 5 (Kenwood /
Elecraft): FA;, FB;, IF; and ID;, and nothing else — a wrong-length reply is
worse than none, it desynchronises the amp's parser for the following poll.
Also offered for SPE. Some amps do not poll at all but read the radio↔PC CAT
line in parallel; those never hear a responder, so an optional unprompted send
(500/1000 ms) covers them.
The TX frequency is what is sent: in split the amp must be tuned where we
transmit. Frame lengths are pinned by tests — the failure they guard against is
silent and only shows up as an amp that mistunes.
Untested on hardware.
The release prompt was answered with a missing dot and the script propagated
it verbatim into both constants, so the shipped exe reported "021.4".
That is not cosmetic: versionLess splits on dots, so "021.4" parses as [21,4]
and compares GREATER than "0.21.5" — a client on that build would be told it
is up to date forever and never offered another update. It also broke the
What's new dialog on a fresh install, where the changelog entry ("0.21.4") is
matched against appVersion by string equality.
Same code as the release otherwise; only the version string changes.
OmniRig reports the VFO pair (AA/AB/BA/BB) on the whole Yaesu range and never
the single-letter form. Only the latter was honoured, so every rig in that
family stayed pinned to VFO A: pressing SUB moved the radio but not OpsLog,
and a QSO worked on SUB was logged on the main VFO's frequency. The first
letter of the pair is the VFO being listened on — Log4OM reads it and gets the
right frequency on the same rigs, which is what showed the data was there.
The enum now wins over the Yaesu Freq==FreqB inference, which is only a
fallback for a rig file that names no VFO at all (the stock FTDX10 one answers
neither VS; nor FR; usably — verified on the air).
Split: the ON flag is still latched to survive a rig file that flips it on its
own, but the latch is now ARMED only after 8 flips in 30 s. A first cut at
3-in-15s was armed by the operator toggling split while testing, imposing the
6 s clearing delay on a radio that did not need it; a misreading file flips a
dozen times in that window untouched, so the two cases separate cleanly.
Distance (km) column added to Recent QSOs and Worked before (shared catalog).
Computed from the QSO's OWN my_grid/my_lat/lon first, falling back to the
current profile's locator: a log spans years and portable outings, so the
station a QSO was made from is not necessarily today's.
Locator: a precise QRZ/HamQTH grid is no longer overwritten by the cty.dat
entity centroid. The lookup runs several times per QSO and the provider gets
2 s; a slow second answer fell back to cty.dat and downgraded JN05JG to JN16
while name and QTH survived (they are only written when non-empty).
The OmniRig diagnostic line now logs what OpsLog concluded, not just what
OmniRig reported. icomnet.go: gofmt alignment only.
- SCP/N+1: new internal/scp downloads the community MASTER.SCP master list and a
docked two-column widget shows, as you type a call, the known calls containing it
(Partial) and the calls one edit away (N+1) — click to fix a busted call. Opt-in
in Settings → General; top-bar toggle; queried debounced on callsign input.
- Flex: OpsLog's GUI-client detection was too loose and could bind to "SmartSDR CAT"
(or DAX) — both carry "smartsdr" in the program name — instead of the real GUI
client, making SmartSDR CAT drop and reconnect in a loop while OpsLog was open.
Now it binds only to a real SmartSDR/Maestro GUI client (e.g. a FLEX-8600M's
integrated screen) and excludes cat/dax; dropped the risky empty-program fallback.
- Telemetry: on a fresh install the callsign isn't set at launch, so the once-a-day
heartbeat recorded the machine UUID. Now it waits (~10 min) for the operator to
enter their callsign before sending, falling back to the UUID only if none appears.
For an entity resolved only via cty.dat (e.g. a French call not on QRZ/HamQTH),
the provider block sets a coarse country-centroid grid. The worked-before backfill
now overrides that with the precise locator from the last QSO with this call, and
takes that QSO's lat/lon (derived from the grid when the record stored none) so
the map and saved record stay consistent. A real provider grid still wins.
Moved the (unreleased) worked-before backfill entry into a new 0.21.2 changelog
block, since 0.21.1 is already out.
When a callsign resolves only to cty.dat (not on QRZ/HamQTH, or no lookup service
configured) — or the lookup errors — enrich the entry from the most recent QSO
already in the log with that call. Fills ONLY the fields the provider left empty
and the operator hasn't edited (name, QTH, grid, country, address, state, county,
lat/lon, zones, continent, email, QSL-via), so a real QRZ/HamQTH hit is never
overridden. Uses the worked-before entries (qso_date DESC, [0] = latest) via a
live ref so it works regardless of the lookup/worked-before debounce ordering.
CSS `zoom` re-lays-out and pixel-rounds each element, which opened ~1px seams
between map tiles (a faint grid over the map when zoomed). Switch the persistent
UI zoom to `transform: scale(z)` from the app root's top-left, with the root
counter-sized to (100/z)vw × (100/z)vh so it reflows to fill and scales back to
the window. transform scales the subtree as one composited layer, so tiles stay
seamless. Removed the now-unneeded tile-overlap CSS hack.
Under our CSS zoom the map tiles scale by a fractional factor, opening ~1px gaps
between adjacent tiles that read as a faint grid over the map. Tag the document
with data-uizoom and, only while zoomed, enlarge each 256px tile by 1px so
neighbours overlap and cover the seams. The 100% view is untouched.
CSS `zoom` doesn't rescale viewport units, so the 100vh app root left an empty
strip below the UI when zoomed under 100%. Counter-size the app root to
(100/z)vw × (100/z)vh so that, after the zoom scales it back down, it exactly
fills the window again. Zoom still applies to the whole document so portalled
menus/modals scale with it.
- Zoom: our own Ctrl+wheel zoom (CSS zoom on the root, 50–250%), persisted in
localStorage and restored at startup; Ctrl+0 resets. Replaces the non-persistent
native WebView2 zoom. The freq-digit wheel now ignores Ctrl so it passes through
to zoom.
- Award column widths: they were stripped from AG Grid's saved column state (that
strip fixes a visibility desync) which also dropped their width. Now persisted
separately per award code (localStorage + portable DB copy) and re-applied on
rebuild/reopen.
- CW keyer widget: macros padded to 9 (F1–F9 slot) and empty macros hidden like
the voice keyer, with the F-number kept tied to the real macro index so the
F-key shortcuts still line up. New QRZ? default for F9.
They took vertical space in the CW Keyer panel for text operators already know.
Kept the actionable CAT-backend mismatch warnings and the checkbox labels.
With ESM enabled (Settings → CW Keyer), the keyer active and mode CW, Enter in
the entry strip fires a macro by QSO stage instead of logging:
- callsign field empty → F1 (CQ)
- callsign entered (in call field) → F2 (report), then focus jumps to RST-sent
- Enter in RST-sent / RST-rcvd → F3 (TU), which logs via its own <LOGQSO>
Works for every keyer engine (WinKeyer / serial DTR-RTS / Flex CWX / Icom CI-V)
since it routes through the shared macro sender. Enter in other fields still logs
normally, and ESM off keeps the classic Enter-to-log behaviour.
Backend: new `esm` flag on WinkeyerSettings (winkeyer.esm). Frontend: esmHandleEnter
state machine keyed off data-esm markers on the call/RST blocks, a Settings
checkbox + hint, and i18n EN/FR.
Rolling the mouse wheel over the hundreds / tens / units-of-kHz digit of the
header (and compact top-bar) frequency steps the frequency by 100 / 10 / 1 kHz
(wheel up = higher). The display updates optimistically on every notch and the
rig QSYs over CAT, debounced so a fast scroll doesn't flood the link. Only the
kHz digits are wheel-sensitive; MHz and Hz stay static.
New FreqWheelDisplay component (replaces the plain fmtFreqDots span in both the
full header and the compact top bar) + nudgeFreqHz handler.
The FlexRadio panel, AmpCard and TunerCard each carried their own copy of the
MeterBar (LED-bar) component. Even small drift between the copies made the tuner
meters look slightly off in height/LED size vs the others. Extracted one shared
components/MeterBar.tsx (segments, bar height, padding) and imported it in all
three, so every meter renders at exactly the same size. No behaviour change.
FlexRadio panel refinements from feedback:
- TRANSMIT and RECEIVE now share one collapse state — folding either folds both
(Card gained an optional controlled open/onToggle mode; the parent owns the
shared txrx state, persisted).
- Meter sizes: the Tuner Genius PWR/SWR meters used a 2-column grid (wider than
the Flex/amp meters). Switched to the same grid-cols-2 sm:grid-cols-3 so every
meter across the METERS, AMPLIFIER and TUNER cards is the same width.
- MIC and COMP meters are hidden outside phone modes (shown for SSB/AM/FM only).
- S-meter dBm moved inline next to the S-value (was a second line under the bar),
and the redundant dBm line under PWR removed — keeps only the watts. Saves height.
Addresses three points of feedback on the Tuner Genius work:
- Meter sizes: the amplifier meters were rendered `compact` (smaller than the
FlexRadio meters). Dropped compact so the amp, tuner and Flex meters are all
the same size.
- Collapsible cards: the FlexRadio panel Card, AmpCard and TunerCard now fold
from a chevron in the header, state persisted per card (opslog.cardOpen.*).
The amplifier cards share the "amplifier" collapse key across their SPE/ACOM/
PGXL variants so folding sticks regardless of the shown model.
- TGXL responsiveness: the tuner's device poll dropped 1500ms→400ms and the
three UI pollers 1500ms→500ms, so the SWR/power meters track TX without the
2–3s lag behind the amplifier the user saw.
- Icon: the Tuner Genius top-bar toggle used Zap, same as the CW keyer — changed
the tuner's icon (top bar + widget + card) to Gauge so the two are distinct.
Push the tuner control further so it mirrors the native 4O3A app and the way the
PowerGenius XL is surfaced.
Backend (internal/tunergenius):
- Status now carries both RF channels A and B (source/mode, band, frequency,
bound Flex nickname, per-channel bypass, antenna, PTT), the active channel,
the C1/L/C2 relay-network positions, and the 3-way-vs-SO2R hardware variant
(learned once from the `info` reply). Flat freq/antenna still mirror the active
channel for the compact widget.
- New TunerGeniusActivate(ch) binding → `activate ch=N` (or `ant=N` on 3-way).
Frontend:
- New shared TunerCard, styled exactly like AmpCard (PWR/SWR meter bars,
A/B channel selector with source+freq+antenna, Tune/Bypass/Operate). Used in
BOTH the FlexRadio panel (its own card, like the PGXL) and Station Control.
- Docked TunerGeniusPanel widget now shows the two channels A/B with their
source/frequency/antenna and lets you click to make one active — the missing
A/B state the user flagged.
- i18n EN/FR for the new labels (channels, antenna, in-line/bypassed, title).
Still UNTESTED on hardware — verify the per-channel field names/units and the
activate behaviour on the real box.
New internal/tunergenius client speaking the same "Genius Series" text API
as the Antenna Genius / PowerGenius XL: banner on connect (with optional
"AUTH" for remote access), "C<seq>|<cmd>\n" commands, "R<seq>|<code>|<msg>"
replies and the "S<seq>|status k=v …" snapshot; async "M|<msg>" info lines
are consumed inline. Polls status ~1.5s and exposes SWR (return-loss dB
converted to a VSWR ratio), forward power (dBm→W), and the operate/bypass/
tuning/active-channel state. Actions: autotune, global bypass, operate/standby.
Controlled directly (not via the radio) so OpsLog uses only one of the box's
four connection slots, per the device's protocol doc.
Wiring:
- app.go: tunergenius.* settings keys, Get/Save/start, GetTunerGeniusStatus,
TunerGeniusAutotune/SetBypass/SetOperate; started in the background at launch.
- Settings → Tuner Genius panel (enable + IP + optional remote code).
- Docked TunerGeniusPanel widget (SWR/power readouts + Tune/Bypass/Operate),
top-bar toggle, shown when enabled — mirrors the Antenna Genius widget.
- i18n EN/FR (sec.tunergenius, tg2.*, tgp.*).
Command verbs (operate/bypass/autotune/status/auth) come straight from the
4O3A "Tuner Genius XL — Protocol Description"; UNTESTED on hardware.
award_refs was only reset when the callsign was emptied, so swapping from one
call to another (e.g. clicking successive spots) kept the previous station's
references in the F3 Awards tab. Move the reset to just past the same-call guard
so any change clears them; the new call's spot POTA + live detection re-populate
right after. Changelog 0.21.1.
Note that entering address/city/state and per-band rig & antenna, then ticking
"Fill my station fields from my profile" on ADIF import, backfills the MY_* fields
automatically for a fully-described log.
In the QSO details Awards tab, the detected-refs line grew tall (DXCC/WAS/WAZ/
WAC/WPX/USA-CA… each with its full name), squishing the AwardRefSelector above so
you couldn't see the selected awards. Show just CODE@REF chips (full name on
hover), cap the block height with overflow scroll, and add a separator. Changelog 0.21.1.
Add a keyNav prop to BandMap: when set (only the docked Main-view map, not the
multi-band Band Map tab), Ctrl+ArrowUp / Ctrl+ArrowDown selects the next in-band
spot above / below the rig frequency and tunes to it via the existing spot-click
handler. Higher freq is up on the map (freqToY), so Up = next higher spot.
Ignored while typing in an input. Changelog 0.21.1.
flex-wrap with fixed-width cards started each new row below the TALLEST card of
the previous row, leaving big gaps under short panels (e.g. the amplifier alone
on a second line). Switch the dashboard to balanced CSS columns (column-width
430px, break-inside-avoid): cards flow top-to-bottom and pack tightly by height.
"Auto" fits as many columns as the window allows; 1-4 cap the column count via
the container max-width. Grip-drag reorder unchanged. Changelog 0.21.1.
GroupCard was defined inside ExportFieldsDialog, so it got a new component
identity every render and React remounted the whole grid on each sel change,
making the per-group All/None buttons (and checkboxes) feel unresponsive. Hoist
GroupCard to module scope with sel + handlers passed as props. Changelog 0.21.1.
The theme (and other per-profile UI prefs) are read via a.settings.Get, which is
scoped to the active profile. GetUIPref only guarded on a.settings==nil, so in
the startup window AFTER NewStore but BEFORE SetProfile(active) it resolved with
the wrong (empty) scope. The frontend theme self-heal treats a resolved "" as
"answered, unset → keep default, stop retrying", so a race landed on the light
default and stayed. Add a settingsScoped atomic flag set right after
SetProfile; GetUIPref/SetUIPref return "not ready" until then, so the frontend
keeps retrying and restores the real theme. Also prevents seeding prefs into the
wrong profile scope. Changelog entry added to 0.21.0 (EN+FR).
Big architecture change (settings/logbook DB split) warrants a minor bump. Rename
the unreleased 0.20.12 changelog block to 0.21.0 and prepend a clear ⚠️ warning
(EN+FR) explaining the automatic non-destructive split, that nothing is lost, and
advising a data-folder backup before updating.
Per feedback: New/Open/Rename/Save-copy/Reset made no sense under the settings
database — moved the meaningful DB actions to the logbook (New database / Open
existing / Rename-relocate). The settings section now shows just its path + Open
folder. Also fixed the "Logbook switched…" confirmation rendering twice (dropped
the duplicate render in the SQLite block; kept the shared one). The unused
settings-db handlers/bindings stay for a later re-add.
"Choose a dedicated file" points at a fresh/empty file; there was no way to
rename or move an existing logbook WITH its data. RenameLogbook VACUUM INTOs the
active profile's SQLite logbook to a new path, repoints the profile and switches
the live logbook (no restart). Deletes the old file only when it was a dedicated
per-profile file; the shared default logbook.db is left for other profiles.
Errors on a MySQL logbook. UI: "Rename / relocate…" button in the logbook section.
Design flaw: opslog.db held BOTH the config (settings + profiles) AND the SQLite
logbook, so the only way to a separate SQLite logbook — the whole-app "change
database location" pointer — swapped the config too, wiping the operator's setup
(reported: creating a new SQLite for a visiting op lost all profiles/settings).
Now the two are separate files:
- Settings database: settings + profiles. Fresh installs name it settings.db;
existing installs keep opslog.db.
- Logbook (QSOs): a dedicated logbook.db next to the settings db, or a
per-profile file (profile.ProfileDB.Path), or MySQL. connectLogbook opens the
logbook file (db.Open creates+migrates on first use); the settings db is used
as the logbook only if the split ever fails.
One-time, non-destructive migration at startup: if there's no logbook file yet
but the settings db already holds QSOs (legacy combined opslog.db), seed
logbook.db with a clean copy via VACUUM INTO — contacts move to the logbook, the
originals stay in the settings db as an untouched backup.
UI: the Database panel now shows the settings database and this profile's logbook
as two clearly labelled sections (+ "Open folder"), and the logbook selector is
SQLite (default logbook.db, or a dedicated file via "Choose a dedicated file…")
vs MySQL — no more confusing shared/separate choice. New binding RevealDataFolder.
Design flaw: the SQLite backend conflated the app/config database (opslog.db —
settings + profiles) with the QSO logbook. connectLogbook returned a.db for any
non-MySQL profile, so the ONLY way to get a separate SQLite logbook was the
whole-app "change database location" pointer — which swapped config + profiles
too, wiping the operator's setup (reported: creating Jerem.db lost all configs).
profile.ProfileDB gains a Path field: a non-empty path routes that profile's
logbook to its own SQLite file (db.Open creates + migrates it on first use),
while opslog.db keeps settings/profiles. connectLogbook opens the file; empty
path = shared db (unchanged default, backward compatible). MySQLSettings carries
sqlite_path; Get/Save wire it through.
UI: the Database panel gains a third backend option "SQLite — separate file"
with a file picker, and now clearly labels the shared application database
(settings + profiles) vs this profile's logbook, so the two are never confused.
Operators rarely write the canton in their address/QTH but do write the city.
Add a per-reference city regex to every one of the 26 cantons (same mechanism as
WAPC): the description match now fires on the canton NAME or any of its main
cities (Genève→GE, Lausanne→VD, Zürich→ZH, Bellinzona→TI, …), with accented and
ASCII/localised spellings. Scoped to DXCC 287 so cross-border city-name clashes
can't misfire. Catalog version bumped 1→2 so the update reaches anyone already
running H26 v1 (unless they've edited it).
The random per-install UUID lives in the LOCAL SQLite, so one operator running
OpsLog on several machines (laptop / desktop / Maestro) or after a reinstall
counted as several distinct "users" — making the user total unreliable. Use the
station callsign as distinct_id when set (public in ham radio, stable across all
of an op's machines); the install UUID stays as a fallback and rides along as an
install_id property so per-machine detail isn't lost.
A SteppIR only covers part of the spectrum (a standard one is 20 m–6 m) but
can't report its own range, so the follow loop's existing out-of-range guard
(FreqMin/FreqMax) never engaged for it. Tuning the rig to a band the antenna
can't reach (e.g. 30 m) made the loop keep trying — the immediate cause of the
stuck TX interlock addressed in the previous commit.
Add a Tunable range (MHz) setting to the Antenna panel, wired into the SteppIR
adapter's Status().FreqMin/FreqMax so the follow loop skips out-of-range
frequencies entirely: no tune command, no TX inhibit. Defaults to 13–54 MHz
(20 m–6 m, the common model) when unset; widen the low edge for a 40 m-equipped
SteppIR. Ultrabeam is unaffected (it reports its own per-band coverage).
Keeps the last-commanded-freq deadband from the previous commit as a universal
safety net for any flaky-status case within the valid range.
The motorized-antenna follow loop keyed its deadband off the antenna's reported
frequency. A SteppIR reports an intermittent/stale status frequency (flips
between the commanded freq and its home/6 m value), so the deadband tripped on
almost every 1.5 s poll and re-issued a tune command. Each command refreshed
noteMotorMoveCommanded(), keeping motorTXInhibitLoop's recentCmd window alive, so
the Flex transmit-inhibit ("Interlock is preventing transmission") never
released — confirmed in a real log (moving=0 throughout, recentCmd was the
driver).
Follow loop now keys the deadband off the LAST COMMANDED rig frequency, only
re-tuning when the radio actually QSYs beyond the step — immune to a flaky
antenna status. Falls back to the antenna's freq only until the first command.
Also dropped SetTXInhibit's `interlock set reason=` sends: `reason` is a
read-only field on the Flex interlock object, so writing it is rejected
(cmd error 5000002D on V1.4.0.0) and did nothing; `transmit set inhibit=` is the
real mechanism and is unchanged.
New built-in award file internal/award/catalog/h26.json (USKA "The Helvetia 26
Award HF"): 26-canton reference list, matches the canton from the QSO address
(OR-rule: QTH) by description, scoped to DXCC 287 (HB) on HF. Fixed a typo in
the description ("Swityerland"). Changelog entry added (EN+FR).
This reverts commit 6a28344. Restore the once-a-day anonymous PostHog heartbeat
and its opt-out toggle: telemetry.go (sendTelemetryHeartbeat, GetTelemetryEnabled/
SetTelemetryEnabled, PostHog host + API key), the startup goroutine, the Settings
→ General toggle + i18n keys, and the README/wiki mentions. version.go is dropped
again (appVersion moves back into telemetry.go). release.ps1 (untracked) pointed
back at telemetry.go by hand.
The matching "five rules" were already documented; add the parts that were only
in code: the per-QSO Explain trace (Test a callsign in the award editor), the
Missing-refs gap finder (in-scope-but-no-reference + bulk assign), how one QSO
credits several references (n-fer POTA / VUCC grid lines, comma/semicolon split),
code-vs-exact matching, and how built-in award updates skip awards you've edited.
The SteppIR client logged commands it sent but never the status frames it read
back, so a misread motor-status byte — which drives the app's "block TX while
the antenna is moving" interlock on a FlexRadio — was invisible. Log the raw
11-byte frame plus the decoded freq/dir/moving and the motor (buf[6]) and
direction (buf[7]) bytes, but only when the frame changes, to keep the 2 s poll
from flooding the log. Purely diagnostic; no behaviour change.
Drop the once-a-day anonymous "app_opened" heartbeat to PostHog and everything
around it: telemetry.go (sendTelemetryHeartbeat, GetTelemetryEnabled/
SetTelemetryEnabled, install-ID/last-sent settings keys, PostHog host + API key)
and the startup goroutine that fired it. The 'Send anonymous usage statistics'
toggle is removed from Settings → General (TelemetryToggle + i18n keys), and the
telemetry mentions are stripped from the README and wiki.
appVersion (still used by changelog/update/livestatus/log-email) moves to a new
version.go; release.ps1 now rewrites it there instead of telemetry.go.
writeRecord gained a fourth `allow map[string]bool` argument in the export
field-picker change; the round-trip test still called it with three. Pass nil
(no field filtering — the historical behaviour).
The RECEIVE card had grown tall after adding WNB and the SmartSDR v4 DSP rows
(NRL/NRS/NRF/ANFL/RNN/ANFT). Keep the everyday NB/NR/ANF visible and tuck the
supplementary controls behind a collapsible 'DSP' button. The button badges +
highlights when any hidden control is ON so an active filter is never hidden,
and its open/closed state persists in localStorage (opslog.flexDspOpen).
Global "Export to ADIF" gains a third option "Choose fields…" next to the
standard-only and full-OpsLog choices, and right-clicking selected QSOs adds
"Export selected — choose fields…". The picker separates official ADIF 3.1.7
fields (grouped by category) from OpsLog/non-standard tags actually present in
the log (discovered via DistinctExtraKeys over extras_json), with All/None per
group and reset-to-defaults; the selection is persisted per user.
Backend: adif.Exporter gains a Fields allow-set that gates every promoted and
extras field in writeRecord; app.go ExportADIF/ExportADIFFiltered/
ExportADIFSelected take a fields []string param, plus new ADIFExtraFields.
The External-services panel (HRDLOG, eQSL, LoTW, POTA) and the many input
placeholders across all tabs (QRZ/ClubLog included) were still hardcoded
English. Added ~40 es.* keys (EN + FR) and wired every label, placeholder,
hint, checkbox, button and the "coming soon" fallback through t(). Renamed the
tab-map param off `t` so it no longer shadows the i18n function.
The Password-encryption, ClubLog-exceptions and ClubLog-Most-Wanted blocks in
the General settings panel were still hardcoded English. Added ~30 gen.* keys
(EN + FR) and wired every heading, checkbox label, description, button and
status line through t(), plus the passphrase-mismatch error.
A QSO with a non-Latin character (Cyrillic Я = \xD0\xAF, Polish ł, …) failed to
save on the shared MySQL backend with "Error 1366: Incorrect string value …
for column 'qso'.'address'". Cause: the database had been pre-created by a
hosting panel (o2switch/cPanel) as latin1, so OpsLog's CREATE DATABASE IF NOT
EXISTS … utf8mb4 was a no-op and every table inherited latin1 — which can't
store those letters, even over a utf8mb4 connection.
OpenMySQL now runs ensureMySQLUTF8MB4 after migrations: ALTER DATABASE to
utf8mb4 (for future tables) + CONVERT any table with a non-utf8mb4 text column
via an information_schema probe. Idempotent and cheap on a healthy DB (the probe
returns nothing → no ALTERs); best-effort so a charset issue never blocks
logbook access. SQLite is unaffected.
The pass-order "#" is a rowIndex valueGetter, but AG-Grid keeps the same row
node across a reorder (getRowId by id) and caches the value, so a moved station
kept its old number. Force-refresh the __order column on every model update in
passOrder mode.
The on-air list was ordered by log time, but a net controller needs the order
the mic goes around. The on-air grid now renders in a controller-owned order:
- A pinned "#" column numbers the round; header sorting is disabled on this
grid (new passOrder prop on RecentQSOsGrid) so a stray click can't reshuffle
it, and any previously-saved sort is stripped on restore.
- ⬆⬇ buttons move the selected station up/down the queue.
- New check-ins append to the end; logged/cancelled stations drop out; the
order is reconciled against the live session list and persisted per net
(localStorage opslog.netOrder.<id>), so a tab switch keeps the round.
- "Log & end" now auto-advances to the NEXT station in pass order.
The first RX-audio CW decoder decoded poorly on real signals (vs SDC) and was
removed in cafade0. This rebuilds the DSP core from scratch in
internal/cwdecode, keeping v1's good ideas (pitch lock, Flex cw_pitch
targeting, tiered acquisition) and fixing its proven failures:
- Two-way DEBOUNCE (pending-commit ~0.3 dit): v1 rejected short marks but not
short spaces, so a one-hop fade inside a dah shattered it into dits — the
single worst real-signal failure.
- Per-character BATCH dit/dah classification at flush time, using the batch's
own bimodal boundary — the first letter of an over decodes at any speed; the
timing clusters (muDit/muDah) are updated with the same attribution, killing
the classify-then-learn feedback spiral ("all dits at 60 WPM").
- dB-domain envelope with separate floor/peak trackers, hysteresis slicer,
span CAP (30 dB — else quiet backgrounds put the off-threshold in the
analysis-window skirts and letters merge) and window de-bias on durations.
- Double squelch: span >= 6 dB AND peak >= bank-median + 9.5 dB (span alone
cannot reject pure noise).
- Hamming-windowed Goertzel, 16 ms window / 5 ms hop (a 20 ms window left
40 WPM inter-element gaps with no envelope dip).
- Hardened acquisition (overlapping windows made "3 stable hops" meaningless)
plus SUPERVISED RE-LOCK: a clearly stronger tone on another pitch sustained
~1 s takes over — heals noise locks and follows a QSY.
- Gap-fed dit-length tracking (inter-element gaps are timing evidence too).
Proven by a synthetic-signal test suite (all passing): clean 12-40 WPM, three
pitches, added noise, QSB fading 0.35-1.0, 10 ms dropouts inside dahs, QRM in
auto and targeted modes, noise-only squelch, mid-over speed change 25->15 WPM,
and jittered hand keying (+/-20-25%, 3 seeds).
Wiring and UI restored from git (app_cw.go, Ear header button, decoded-text
strip with WPM/pitch/level + pitch lock + click-a-word-to-fill-callsign, Tools
menu entry) — strip strings translated (cwd.* i18n keys, EN/FR) this time.
The whole "Manipulateur CW" settings panel was hardcoded in English, so it
stayed English even with the app in French. Added ~35 wk.* keys (EN + FR) and
wired every label/checkbox/select/placeholder through t(). Also removed the long
serial-engine explanation paragraph as requested.
Confirmed from HB9HBY's log: ptt=true, yet the FT-891 dropped to RX between
words during a CQ macro — while the same SCU-17 works in N1MM. Root cause:
go.bug.st/serial drives DTR/RTS through GetCommState/SetCommState, which on the
SCU-17's Silicon Labs CP210x drops the asserted RTS (PTT) the moment DTR (CW)
is toggled. N1MM/WSJT use EscapeCommFunction, which sets one line without
disturbing the other.
OpsLog now drives the lines the same way: a new lineCtl abstraction with a
Windows implementation (linectl_windows.go) that opens the COM port via
CreateFile and toggles DTR/RTS with windows.EscapeCommFunction (SETDTR/CLRDTR/
SETRTS/CLRRTS). linectl_other.go keeps go.bug.st/serial for non-Windows builds.
The serial keyer holds a lineCtl instead of a serial.Port. RTS (PTT) now stays
asserted for the whole macro, so the rig holds TX between words.
Promotes golang.org/x/sys to a direct dependency.
Reported: on a CQ macro the rig drops to RX between words instead of holding TX
for the whole macro. Root cause is PTT not being held (usePTT off, or the RTS
PTT line not wired/honoured). Improvements:
- "Key PTT line" now defaults ON when the Serial engine is selected — without it
the rig runs on semi-break-in and drops between words.
- The serial keyer's line options (PTT, key line, invert, lead/tail, speed) are
now atomic and live-updatable: SaveWinkeyerSettings -> ApplySerialConfig applies
them from the next character, no reconnect needed.
- Each transmission logs its PTT state / key line / lead/tail, so a rig that
still won't hold TX can be diagnosed from the app log.
The macro itself is sent as one string (not fragmented), so PTT is genuinely
held across word gaps when usePTT is on — pending on-air confirmation via the log.
Opt-in (Settings → General). Shows a "MW #rank" pill next to the DXCC entity
name in the entry-strip band/slot matrix — ClubLog's Most Wanted ranking
(1 = most wanted), personalised to the operator's callsign and refreshed daily.
- internal/clublog/mostwanted.go: fetch mostwanted.php?api=1&callsign=<CALL>
(rank→DXCC JSON, real User-Agent), invert to dxcc→rank, cache to
<dataDir>/clublog_mostwanted.json; NeedsRefresh invalidates on callsign change.
- app.go: keyClublogMostWanted setting, a.clublogMW, startup refresh goroutine,
activeCallsign() helper, and Get/Set/Download bindings mirroring the cty ones.
WorkedBefore now carries MWRank (qso.WorkedBefore.MWRank) when enabled.
- BandSlotGrid.tsx: MW pill next to the entity name, colour-tiered by rank.
- SettingsModal General: toggle + download button + status, mirroring the
ClubLog cty-exceptions block.
Also reorganises the changelog: the theme-persistence fix (landed after the
v0.20.11 tag) plus this feature go under a new 0.20.12 entry; 0.20.11 keeps only
what shipped in its binary.
The settings store is backed by the LOCAL SQLite config DB (wired before the
slow remote-MySQL logbook connect), so the "not ready" window is the brief
startup gap before OnStartup builds the store, not a MySQL delay. Fix behaviour
is unchanged; only the comments and changelog wording are corrected.
An update can clear the WebView's localStorage; the theme is then restored from
the portable DB pref, but GetUIPref returned ("", nil) while the settings store
was still starting (indistinguishable from a genuinely-unset pref), so the
self-heal gave up after ~2.4s and fell back to the light default.
GetUIPref now returns an error when the settings store isn't wired yet, and the
theme self-heal treats that as "not ready → keep retrying" (up to ~36s) vs a
real empty answer ("unset → keep default"). A dark theme survives the relaunch.
Confirmed from an FT-891 CAT log: OpsLog tuned via OmniRig's SetSimplexMode,
which returned OK on every spot click but never moved the VFO (readback stayed
put), while SetMode on the same .ini worked — so the CAT link was healthy and
only the frequency write was a no-op. SetSimplexMode was chosen because Icoms
accept direct FreqA writes but don't move; Yaesu/Kenwood are the opposite.
Now, for non-Icom rigs (or if SetSimplexMode errors), also write the VFO
frequency property (FreqA/Freq) directly, which does move them. Icom keeps the
SetSimplexMode-only path so the Main/Sub VFO isn't nudged by a direct write.
Some CW interfaces key on the opposite line level (transmit at rest / key
backwards). New "Invert keying (active-LOW)" toggle flips the asserted level for
both the CW and PTT lines, and the idle state drives both lines to their
inactive level accordingly so an active-low rig isn't left keyed at rest.
Default off = N1MM convention (line HIGH = active).
Adds a 4th CW keyer engine alongside K1EL WinKeyer, Icom CI-V and Flex CWX:
OpsLog bit-bangs the Morse itself on a serial control line (DTR keys CW, RTS
keys PTT — swappable), the hardware-keying method N1MM/WSJT use with a Yaesu
SCU-17 and generic interfaces. No WinKeyer chip needed.
- internal/winkeyer/serialcw.go: serialKeyer — worker goroutine, morse table,
WPM + Farnsworth timing (live speed changes), abortable, PTT lead-in/tail/hang.
- winkeyer.Config gains Type ("k1el"|"serial") + CWKeyLine ("dtr"|"rts");
Manager routes Connect/Send/Stop/SetSpeed/Backspace/Disconnect to it.
- app.go: keyWKCWLine setting; WinkeyerConnect sets Type from the engine.
- Settings -> CW Keyer: new "Serial port (DTR=CW/RTS=PTT)" engine with COM port,
key-line selector, speed/Farnsworth/lead-in/tail/PTT. Macros, auto-CQ and
<LOGQSO> reuse the WinKeyer path unchanged.
UNTESTED on hardware (no SCU-17 to hand): polarity assumes N1MM convention
(line HIGH = active). Weight not yet honoured; Farnsworth is.
The rig's 0x27 waveform frames carry the same leading main-scope selector byte
as the dual-scope IC-7610/9700 (`27 00 00 <seq> <total> …`), but the code keyed
the layout off the CI-V address and treated the 7300 as selector-less, reading
the sequence from Data[1] (always 0x00) so every frame hit the `seq == 0`
guard and was dropped — blank scope. The waveform parser now detects the
leading selector byte from the frame itself (address-independent), and the
scope config/set commands (mode, span, edges) include the selector the 7300
requires. The IC-7610/9700 path is byte-for-byte unchanged (main → idx 2, sub
frames skipped). Confirmed against a real IC-7300 CI-V capture.
0.20.10 was released before the "closing OpsLog switches off the radio's own
scope" fix landed, so its entry is reattributed to 0.20.11 (the fix code is
already committed in 1668455).
SetScope(false) sent both 0x27 0x10 00 (scope DISPLAY off) and 0x27 0x11 00
(CI-V data output off), so quitting OpsLog blanked a local IC-7300's own scope
screen. Only the display-on (0x27 0x10 01) is now sent, and only on enable;
disable stops the CI-V stream alone and never touches the radio's display.
The multi-frame waveform header (USB path) read the second frequency field
as an absolute high edge, but in center mode it is a span — so high < low and
the panadapter had no valid range to map the trace onto (blank scope). It now
applies the same center+span vs low+high disambiguation the IC-7610 single-
frame path already used. FIXED mode was unaffected.
Also stamps changelog 0.20.10 with today's date.
- Station Control: the amplifier panel is now the exact same card as the FlexRadio panel (extracted shared AmpCard: OPERATE/STANDBY, ON/OFF, power level, output-power bar, live FWD/ID/temp meters). Every configured amp gets its own card — no dropdown.
- Help → "Send log to F4BPO" (only when SMTP is configured): e-mails opslog.log + previous session + crash log to the developer. New email.SendFiles for multi-attachment.
- Bulk edit: new "Frequency (MHz)" field sets freq_hz AND recomputes band; out-of-band values rejected. Fixes a run logged on a stale/default freq after CAT dropped.
- Cluster: Time column sorted lexically on the HHMMZ string, so spots after 0000Z sorted below 2359Z and looked frozen at the UTC rollover. Now sorts by real arrival time.
- Also folds in the DVK auto-CQ/2-column layout and Station Control dashboard batch (changelog 0.20.10, EN+FR).
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
## What this is
OpsLog is a Windows ham-radio logger built with **Wails v2** — Go backend, React/TypeScript frontend, compiled into a single `.exe`. Author F4BPO. See `README.md` for the user-facing feature list and `wiki/` for end-user documentation.
**Pure Go, no CGO** — SQLite is `modernc.org/sqlite`, serial is `go.bug.st/serial`. Any dependency requiring cgo breaks the build; check before adding one.
## Commands
```bash
wails dev # hot-reload dev (Go methods also reachable at http://localhost:34115)
wails build # full build → build/bin/OpsLog.exe (~25 s)
go test ./internal/steppir/ -run TestApplyPendingDirHold -v # one test
gofmt -w <file>
cd frontend && npx tsc --noEmit # frontend typecheck alone (wails build does this too)
```
Release: `.vscode/release.ps1` (Ctrl+Shift+P → *Tasks: Run Task* → *Release OpsLog*) — bumps the version, pushes to Gitea, builds and publishes to Gitea + GitHub.
Prefer `wails build` for final validation: it is the only command that exercises Go, the TS typecheck, the bindings and the asset embed together.
## Navigating app.go
`app.go` is ~14 000 lines and holds nearly every Wails binding. **Do not read it linearly** — it is organised into banner-delimited sections:
```bash
grep -n "^// ──\|^// ---" app.go # table of contents
```
Sections map to features (`── Motorized antenna (Ultrabeam / SteppIR) ──`, `── NET Control ──`, `── DX-cluster spot alerts ──`, …). Find the banner, then read that range. The `App` struct (~line 448) is the other useful landmark: its fields document every subsystem and carry substantial explanatory comments.
Exported methods on `*App` are the entire frontend API. Adding or changing a signature requires `wails generate module`, which writes `frontend/wailsjs/go/main/App.d.ts` and `models.ts`. Parameters and return types must be Wails-serializable — the generator prints `Not found: time.Time` noise for unsupported types (that particular message is long-standing and harmless).
Backend→frontend push uses Wails events (`runtime.EventsEmit` / `EventsOn`), e.g. `qso:logged`, `update:progress`. Most hardware status is **polled** by the frontend on an interval rather than pushed.
### Two separate databases
This split is easy to get wrong and matters:
- **`a.db`** — settings database (`settings.db` / `opslog.db`). Always SQLite. Holds settings and profiles. Its location is chosen by the user and recorded in `data/config.json`.
- **`a.logDb`** — the logbook (QSOs). Either a per-profile SQLite file, the default `logbook.db`, or a **shared MySQL** so several operators log into one database. `a.dbBackend` says which.
QSOs never go to `a.db`. `internal/db` holds the connection helpers and the MySQL specifics; `internal/qso` is the repository.
Remote MySQL is the slow path: startup deliberately brings CAT/rig links up **before** connecting the logbook, and hot paths (cluster spot enrichment, alert matching) avoid per-row queries — hence the in-memory `wcbm` worked-index and the `clusterEvents` queue that keeps the telnet socket draining.
### Per-profile settings scoping
`internal/settings.Store` is a key/value store over the settings DB. Every key is transparently prefixed with the active profile (`p3.`), so each station profile has a complete independent set. Two consequences:
-`App.settingsScoped` (atomic) gates reads until the active profile is known. Anything reading settings early must respect it or it reads the wrong scope.
- Settings keys are declared as `key<Thing>` string constants near the top of `app.go` (~180 of them) — grep `key[A-Z].*= "` to find one.
Password-type keys are encrypted at rest when the secret vault is unlocked (`internal/secret`); a locked vault returns `""` rather than ciphertext, and callers treat that as "not configured".
### Hardware device pattern
The codebase talks to ~20 devices (rigs, amplifiers, antenna controllers, switches, keyers, rotators). They all follow the same shape — match it when adding one:
1.**`internal/<device>/`** — a self-contained package exposing a `Client` with `New(...)`, `Start()`, `Stop()`, `GetStatus()`. The client owns its own goroutine: a reconnecting poll loop, a mutex serialising the shared connection, and a cached last-known `Status`. Transports are usually TCP *or* serial behind one `io.ReadWriteCloser`.
2.**Settings** — `key<Thing>*` constants plus `Get<Thing>Settings()` / `Save<Thing>Settings()` bindings in `app.go`.
3.**Lifecycle** — a `start<Thing>()` method that tears down any existing client and rebuilds it from settings; called at startup and again whenever settings are saved.
4.**Status binding** — `Get<Thing>Status()`, polled by the frontend.
5.**UI** — a settings panel in `frontend/src/components/SettingsModal.tsx` and a live widget in `frontend/src/components/`.
`internal/steppir` and `internal/antgenius` are compact, well-commented references. Wire-protocol packages carry the byte layout in the package doc comment and pin it with table tests against **real captured frames** — keep that up, since a wrong byte silently mistunes an antenna.
Where two devices are interchangeable (Ultrabeam / SteppIR), `app.go` defines a small interface (`motorAntenna`) plus thin per-device adapters rather than branching everywhere.
### CAT
`internal/cat` (~7 000 lines, the largest package) is the rig abstraction: OmniRig, native FlexRadio/SmartSDR, native Icom CI-V (USB **and** remote-over-internet), and TCI (SunSDR / Expert Electronics). `cat.Manager` exposes a backend-agnostic `State()`; backend-specific features are reached through typed escapes such as `FlexDo(func(cat.FlexController) error)`. Check `State().Backend` before using one.
### Logging
`internal/applog` — Wails builds with the Windows GUI subsystem, so `fmt.Println` is discarded. Use `applog.Printf` (or plain `log.Printf` inside `internal/` packages) so output reaches the rotating log file in the user data dir. That file is usually the only evidence available when diagnosing a user's hardware problem.
## Conventions
**Changelog is mandatory.** Every user-visible change gets an entry in `changelog.json`, **in both `en` and `fr`**, in the same session as the change. Keep entries to one or two sentences — rationale belongs in the commit message. New work goes under the next version number; the release script bumps the version constants.
**Append new entries at the BOTTOM of a version block**, never at the top. A block is read top to bottom, so a feature must appear before the fixes made to it — prepending meant an operator read "the Yaesu meters are corrected" several entries before learning a Yaesu console existed at all.
**Version lives in two places** and must stay in lockstep: `appVersion` in `telemetry.go` and `APP_VERSION` in `frontend/src/version.ts`.
**Bilingual UI.** Every user-visible string goes through `t()` from `frontend/src/lib/i18n.tsx` (~700 keys), with both English and French provided. No hardcoded display strings.
**Commit messages must not include a `Co-Authored-By` line or any mention of the model.**
**Comments explain *why*, not *what*.** The existing code documents the reasoning behind non-obvious decisions — which protocol source was trusted, what field failure a workaround addresses, why an ordering matters. Match that: a comment that restates the code is noise, a comment recording the hard-won reason is why this codebase is navigable.
## Gotchas
- **Adding a promoted ADIF field touches five places in lockstep.** `internal/adif` promotes ~30 ADIF fields to real QSO columns; miss one and imports silently drop data. All of: the struct field, column list, insert args and scan in `internal/qso/qso.go`; the dictionary entry (`Promoted: true`) in `internal/adif/fields.go`; the `adifPromoted` list *and* the assignment in `internal/adif/import.go`; the writer in `internal/adif/export.go`; then the frontend column in `RecentQSOsGrid.tsx` with its EN+FR i18n keys. Trace an existing field (e.g. `ant_path`) across the repo as a template.
- **Generated files — don't hand-edit.** `internal/dxcc/dxcc_names_gen.go` (cty.dat joined to the ARRL/ADIF entity list) and `internal/awardref/uscounties_gen.go` (emitted by `cmd/cntygen` from the FIPS county CSV; a one-shot generator, not part of the build).
- **Single-instance guard** (`main.go`): a second process would open its own CAT and antenna-follow loops and the two would fight over the rig frequency.
- **Frontend controlled inputs**: several editors normalise a value on every keystroke (trim, split, filter). Binding an input directly to the normalised form makes Enter/Space appear dead — keep raw text in local state and derive the stored value from it.
title={acom.transport==='serial'?'Power on (DTR/RTS pulse — needs the power-on pins wired)':'Power-on needs the serial DTR/RTS lines — not available over a network bridge'}
constTYPE_LABEL: Record<string,string>={webswitch:'WebSwitch 1216H',kmtronic:'KMTronic 8-relay',denkovi:'Denkovi USB (FT245)',usbrelay:'Denkovi USB (serial)'};
// Dashboard geometry: a grid of EQUAL columns whose cards also share a common
// HEIGHT per row — every card stretches to the tallest one beside it. Masonry
// packed tighter but left the row edges ragged, which read as untidy on a wide
// screen; aligned rows are what an operator expects from a dashboard.
constCARD_MIN=430;// narrowest a card may get before "Auto" drops a column
constTYPE_LABEL: Record<string,string>={webswitch:'WebSwitch 1216H',kmtronic:'KMTronic 8-relay',denkovi:'Denkovi USB (FT245)',usbrelay:'Denkovi USB (serial)',dingtian:'Dingtian IOT relay'};
// Relay count for a configured device: fixed by type, except Denkovi (4/8) and
// the generic USB-serial board, whose channel count the user picks.
rouggy