hamradiotools

Nexus 1.22.0

Released 1 November 2026. The current release is 1.24.0. Download it · Every version

Added

  • A solar flare as NOAA sees it, on the map and the globes.

    During an M- or X-class flare, the absorption on the sunlit side is now drawn from NOAA's own D-RAP absorption map, which NOAA updates every minute. Between updates, or when NOAA's map cannot be reached, Nexus's own estimate from the X-ray reading stands in, and the flare chip says which one you are looking at and when it is from. Coloured rings round the point under the sun show, band by band, where each HF band is weakened (dashed) and where it is out (solid), on the Flat map, the Globe and the 3D globe. The flare chip gives NOAA's own start and peak times (when NOAA's flare list is silent, or has not yet listed a new flare, it says the trend is Nexus's estimate and never repeats the previous flare's times; while a flare fades after NOAA calls it ended, it keeps NOAA's "ended"; its own rising or falling comes from the last half hour of X-ray readings, and without recent readings it says neither), which bands are out under the sun and at your QTH, and how long until HF recovers: NOAA's own estimate when it gives one, otherwise NOAA's rule of thumb from the X-ray level. Click the chip for those details and a six-hour X-ray graph. The flare heads-up quotes the same "HF below" figure and the same recovery time as the chip, and says where each comes from (NOAA's absorption map, or estimated from the X-ray reading; NOAA's own recovery estimate, or NOAA's rule of thumb); the Insights flare line says its figure is estimated from the X-ray reading. Below M class the map stays quiet and only the sun turns from yellow toward orange, with the X-ray class on hover. Nexus asks NOAA for the absorption map only during a flare, at most once a minute, and waits longer each time NOAA does not answer. A flare's X-rays arrive at the speed of light, so a flare is drawn the moment it is measured, never as "coming".

  • Search Band Activity.

    FT8 and FT4's Band Activity list has a search box beside its title, in Classic and Roster. Type a call, a grid, a report, a CQ modifier such as DX or POTA, or a country, in any case, and only the lines that contain it stay on screen; with several words, every one must match, and * and ? work as wildcards. While the box has text the list says how many lines match, for example "12 of 87 match", so a filtered list never looks like a quiet band, and new decodes are filtered as they arrive. ✕ or Esc in the box clears it and everything shows again; Esc there still stops transmitting, as it does everywhere. The search only changes what the list shows: decoding, alerts, the Rx Frequency pane, auto-sequencing and transmitting are unchanged, and double-clicking a line works that station exactly as it does without the search. The search is not kept when Nexus restarts.

  • Recent contacts in the FT8/FT4 cockpit.

    The Recent contacts pane that Phone, CW, RTTY and PSK have is now in the FT cockpit too: tick it in ⊞ Panels and it stands at the foot of the side rail, under the callsign card, the Stats pane and the rail's lists, in both Classic and Roster. It shows the last five contacts in the log, newest first, and a click on one opens it in the Logbook editor. Ticking it changes nothing else on the FT screen: the Tx1–Tx6 messages and their divider stay where they are. The lists in the side rail keep a usable height and the side rail scrolls when it runs out of room. It starts hidden, so nobody's FT screen changes on update.

  • Start fresh in Getting Out, for trying an antenna.

    Connect's Getting Out box ("Am I getting out?") has a Start fresh button. Press it and the box counts only the reception reports that arrive from then on: the number hearing you, the furthest, the compass rose and the list, and the map's "hearing you" dots follow. The count's line ends "since 14:05:30 UTC" (the time you pressed it), and Show all counts every report again. Nothing is deleted. The box shows the last 30 minutes of reports, as before, so Start fresh ends by itself 30 minutes after the press and the box goes back to its usual last 30 minutes. A restart inside those 30 minutes keeps it. Each Connect window keeps its own, and a Remote screen keeps its own without changing the shack's.

  • The Path card says why a path works, in one sentence.

    Select a station and the Path card (Conditions ▸ Path to the station) now says how the signal is modelled to get there right now and what may get in its way, for example "20m short path, 2 F2 hops, crosses the auroral zone — an aurora (Kp 5 forecast at 1800Z) may weaken it." The same sentence shows on the selected station's card in the Stations list of the Classic Operate layout; the Roster layout does not show it. It names the best band open now, the number of hops and the layer (E or F2, with the ITU-R P.533 engine), whether the path crosses the auroral zone, the polar cap or the live auroral oval, and whether either end is on the grey line. A storm warning appears only when the space-weather forecast pushes the auroral zone over that path in the next 24 hours, and it is worded as a forecast. When Nexus holds no space-weather data (just after start-up, or when the space-weather service did not answer), the sentence says Kp is unknown and makes no claim that depends on it. Click the sentence on the Path card (or hover it on the station card) for the numbers behind it: hops and take-off angle, how much of the path lies poleward of the auroral-zone edge and at what Kp, whether that Kp was measured, forecast or is unknown, the highest geomagnetic latitude, and how much is on the grey line or in daylight. Up to 7000 km the hops and take-off angle are the model's monthly median; beyond 7000 km they are an estimate the model makes from the distance alone, the same every hour, and are labelled so. The storm, auroral-zone and aurora parts come from space weather, which the model does not include. The sentence describes the short path; the long path has its own line, below.

  • Short path or long path, and which way to beam.

    With the ITU-R P.533 engine, Nexus now models both ways round the world to the station and says which one the model favours on the best band open now (or the band in focus): for example "Beam 80° for 20m: long path (possible) — the short path is modelled closed now." The line shows in two places: the Path card (Conditions ▸ Path to the station), and the selected station's card in the Stations list of the Classic Operate layout (the Roster layout does not show it). When both routes work it says so ("works either way", with both headings); a route the model calls closed is never offered as one of the two. Near the antipode it makes no call, because every heading is then about the same distance. A chip names a route that crosses the polar cap ("Polar path · Kp 5 forecast from 1800Z") or the auroral zone, or that a forecast storm would bring into the auroral zone ("Auroral zone if Kp 5 (forecast from 1800Z)"); its hover says whether that hour's Kp was measured, forecast or the latest reading, and with no Kp reading there is no chip. Hover for the numbers: both headings and their modelled reliability, and the difference in median signal-to-noise. The long path is the same published method run along the other arc of the great circle; the reference model it was checked against covers the short path only, so wherever the long path appears it is labelled "modelled, unvalidated for long path". Until it is validated, the model never calls the long path more than "possible"; the short path can still be "likely". On the map and the 3-D globe the selected station now shows both routes at once, the short path solid and the long path dashed, each coloured by how well the model expects it to work now (green open, amber marginal, grey closed), with the stretch past the auroral-zone edge in violet (none when Kp is unknown) and the stretch on the grey line in gold, and its heading written at your end. When the model favours the long path the LP button on the map is ringed. All of this is advice: nothing turns the rotator.

  • Path over the day.

    Two new Conditions boxes show when a path opens and which way round. Path over the day draws the selected station's path as a strip of bands against the 24 UTC hours, each hour coloured by how well the model expects the better route to work and marked ▸ for the short path or ◂ for the long path (nothing when the model puts them within 6 dB of each other). A line over the strip says the best of it, for example "Best: 20m 12Z–16Z by the long path; 40m 20Z–00Z by the short path." SP and LP show either route on its own. A violet underline marks the hours a route crosses the auroral zone or the polar cap at that hour's Kp (the hover says which, and an hour only a forecast storm puts it there reads "Auroral zone if Kp 5 (forecast from 1800Z)"), a line marks the highest band under the MUF, ◐ marks the hours either end is on the grey line, and a Kp ribbon runs under the hours: solid where Kp was measured, hollow where it is forecast, dashed where no forecast covers the hour and SWPC's latest reading is carried forward, and dotted where there is no space-weather data and Kp is unknown. Click an hour and the 2-D map moves its grey line, sun and the station's path to that hour ("Map at 14Z", with Now to go back); the 3-D globe stays at now. Paths to the regions shows the same strip for ten DX regions (Europe, Japan & East Asia, Australia & New Zealand, Pacific, South America, Africa, Caribbean, Middle East, South Asia, Antarctica), each on its best band, with ▸ to open every band. Regions where the Needed board hears stations come first, then those you have logged most, so with an empty log they read in that order. North America is not one of the ten. With the ITU-R P.533 engine the regions take a couple of seconds to work out, once every six hours or when your grid, power or the sunspot number changes, and are remembered over a restart. Every figure is the model's monthly median, not a measurement, and the long path stays labelled "modelled, unvalidated for long path".

  • Which way your contacts went.

    The Logbook has a new Path column (turn it on in Columns) and a Path filter (Long path, Across the polar cap). For each contact the column shows whether the signal most likely went the short way round (▸) or the long way (◂), how sure that is, and ⊕ when the route crosses the polar cap; hover for why. A contact made with the Path card open on that station is graded as you log it: Nexus keeps the grade in your own log and its ADIF exports, and sends it to no service (LoTW, QRZ, Club Log, eQSL, HRDLog, Cloudlog and the rest get the contact as before). Nexus never fills in a contact's antenna heading or antenna path (ADIF ANT_AZ, ANT_PATH) by itself; set your antenna path yourself on the contact card (Short path / Long path, written when you save), and it travels with the contact like any field you type. Other receivers along one route and a station whose decodes run consistently late are mentioned as reasons and never make a grade surer. A contact logged with a propagation mode other than F2 reflection or ionoscatter (satellite, EME, a repeater, sporadic E, TEP, aurora and the like) is never graded. Contacts already in your log are graded in the background for display only, a few minutes after start, and nothing is written to them: those grades are kept beside your settings and drawn dimmer, as the model's monthly median at the solar flux the contact logged (or today's for a contact from the last three months); an older contact with no flux logged is not graded. Double-click a contact and its card shows the path in one sentence and the path over that day, with the contact's hour marked. Hover the Path column's header for how well the model has matched your contacts over the last 30 days: how many fell in hours it calls open, polar against other paths, and, for contacts that logged a beam heading, how often it agreed with the model.

  • Delete several SSTV pictures at once.

    Each picture in the SSTV gallery now has a tick box, and above the gallery are Select all, Clear selection and Delete selected (N). Delete selected asks once, naming how many, then deletes every ticked picture. If some cannot be deleted, a message says how many and why, and those stay ticked.

  • CW audio scope: Filter, Std and Full span buttons.

    The CW cockpit's audio scope still opens on Filter, your radio's filter centered on your pitch, for zero-beating. Std shows 200–3000 Hz of receive audio and Full 200–4000 Hz, as the waterfalls do, so you can see the rest of the passband. A click tunes the same frequency at every span. The scope remembers your pick, each window its own.

  • The RF scope fits the band, and draws it at the finest resolution your scope can give.

    On the Phone and CW screens the radio's panadapter has a new Band span, now the first choice: it shows the band you are tuned to, edge to edge with a little margin, so 30 m shows 10.095–10.155 MHz instead of an SDR's whole 2 MHz. A span you picked before is kept. Band and the ±25, ±10 and ±5 kHz spans are now asked of the source instead of cut out of a wider picture: an Icom over CI-V is set to the narrowest span that covers the window, and Nexus's FlexRadio panadapter to the bandwidth that covers it, once each time you change band, span or radio, and again when your tuning takes the band outside the span Nexus set. A span you set yourself, on the radio, with the rig scope's own span buttons or with the FlexRadio pan width, stays as you set it while you tune (on an Icom, until the band needs a different span step; on a FlexRadio, until the band no longer fits the pan); an SDR through TCI or SDRconnect is read more closely (about 30 Hz per point on a 2 MHz SDRplay, where ±25 kHz used to be some fifty points stretched across the screen), and its own sample rate is left alone. Nothing is sent to the radio while you transmit; a change made then goes out when you unkey. Where the radio cannot sweep the whole band (a 2 m band on an IC-9700, a 48 kHz TCI stream) the scope draws the part it has and says so under the span buttons. The FT-710 keeps its own span ladder, whose Fix position already fits the band. Not yet checked on a radio.

  • Test buttons for LoTW, eQSL and ClubLog.

    Settings ▸ Services ▸ Connections now has a Test button on the LoTW, eQSL and ClubLog rows, like the one QRZ Logbook already had. Test signs in with the login you saved and only reads: it uploads nothing and changes nothing on the service. It shows what the service said ("Username/password incorrect", "Error: No such Callsign found") or, when the service could not be reached, says so. The LoTW Test checks your LoTW website login, which downloads your confirmations; uploads are signed separately, by TQSL with your certificate. ClubLog's Test checks the email, Application Password and callsign your uploads use; if ClubLog refuses them, Nexus stops sending them, uploads included, until you enter the password again, as ClubLog asks. Needs a check against real LoTW, eQSL and ClubLog accounts.

  • A storm on its way: NOAA's watch, counted down on the map.

    When NOAA issues a geomagnetic storm watch (a coronal mass ejection or a fast solar-wind stream on its way), the map and the 3-D globe show a chip beside the Layers panel: "🧲 G1 watch · NOAA expects 31 Oct–1 Nov (in 21 h)", then "in force" while it runs, then nothing once it is over. Under it, a second line gives the WSA-Enlil model's view (the fastest wind it expects and when, or the window in which it expects the stream to arrive), labelled as a model. When the solar-wind monitor at L1, about an hour upstream of Earth, measures a sudden jump, the chip says "Arrived at L1 02:14Z · reaching Earth ~03:10Z (measured)"; it is the only line that ever says "arrived". Hover the chip for where each line comes from and when NOAA issued it, NOAA's own comment on the watch, and the solar wind at L1 now. Each new watch raises one quiet notification, once: not again on the next check or after a restart, and again only when NOAA issues a new watch. A model run with no NOAA watch shows no chip. Hover the lines of the Kp outlook box for NOAA's forecasters' reasoning behind the three-day outlook (in NOAA's English), with the time it was issued. Nexus asks NOAA for each of these on its own schedule (the L1 wind every 5 minutes, the watch every 15, the model hourly) and waits longer after each failure. If Nexus has not heard from NOAA for six hours, the watch is no longer shown as in force, and a model run more than a day old is not shown.

  • APRS path quick picks.

    Beside the beacon's Path box, a menu offers Direct (no path), WIDE1-1, WIDE2-1, WIDE1-1,WIDE2-1 and ARISS. Picking one fills the box and saves it, as typing it would; you can still type any path.

  • The ISS on the APRS channel menu.

    The APRS screen's channel menu now has the International Space Station's APRS digipeater, 145.825 MHz FM, after the regional channels, and a computer driving the station over Remote can pick it too. With the ISS chosen and a path other than ARISS, a line under the beacon says so, with a Use ARISS button; Nexus never changes your path by itself. Transmitting there is held to your licence and the TX switch exactly as on any 2 m channel.

  • Doppler correction on the ISS APRS channel.

    During a pass the ISS is up to about 3.6 kHz off 145.825 MHz, high while it approaches and low as it leaves, and what it hears from you is shifted the other way. With the ISS channel chosen and the ISS above the horizon, Nexus now keeps the radio on the corrected receive frequency between packets (moving it when it has drifted 500 Hz or more, at most once a second), and before each packet you send moves it to the corrected transmit frequency, keys only once the radio reads that frequency back, holds it for the whole packet and goes back to receive afterwards. It never retunes while the radio is transmitting, and the licence check judges the frequency actually transmitted on. Tune the radio off the channel yourself (by the knob, or to another band) and the correction stops and says so; Nexus never pulls the radio back. A radio that rounds the corrected frequency to its own tuning step has the correction turned off for that pass, so a packet is never sent off frequency, and a packet the radio is not on frequency for within 5 seconds is dropped with a notice, never sent later. Below the horizon, without current orbital elements for the ISS, without a grid square or without CAT, the radio stays on 145.825 and the APRS screen says why. A switch under the beacon, Correct the ISS Doppler shift, turns it off; it is on by default and only ever acts on the ISS channel. The health chip accepts a dial a few kHz off 145.825 on the ISS channel, so a pass you correct by hand does not read as the wrong frequency.

  • Talk through this computer's microphone in Phone.

    A new per-radio switch, Settings ▸ Radio ▸ Audio ▸ Use this computer's microphone, sends a microphone plugged into the computer to the radio while you hold PTT in Phone: through the sound card's output on a sound-card radio, and over TCI audio on a TCI radio. Pick the microphone and set its gain beside it. It is off by default, so a station that talks into the radio's own microphone, or routes a microphone through the operating system, works as before. The microphone is open only while you hold PTT, and only once every transmit check passes (TX on, your licence, the TX watchdog); Stop TX or TX Off cuts the audio and unkeys at once, and so does a microphone that is unplugged mid-over. Releasing PTT lets the end of your last word finish before the radio unkeys: what the microphone had already picked up plays out, at most 0.2 seconds of it by the clock, and nothing it hears after the release is sent. Every stop unkeys first and only then closes the microphone. The digital modes never take it, nothing you say is played back to you, and a voice-keyer message and the microphone never share the transmitter. A microphone that could put the band or this computer's own audio back on the air is refused, with the reason in the status line: any radio's receive input under any of its names, the input receive is using, a Flex DAX receive channel, a loopback or virtual cable, or a sound server's default input such as default, pulse or pipewire on Linux. The microphone is judged by the same device check the JS8 sounds use, so on Linux it must be a sound card picked by name, and receive must be on its own, different card by name: with receive on the system default, Nexus cannot prove the microphone is not the band, and says so. A microphone that sends only digital silence never keys the radio; once you are on the air, a microphone that gates to silence between words keeps the over up, and four seconds of nothing but digital silence ends it. On a Flex that Nexus's own Flex client serves with native audio on, where the radio keeps its own microphone in Phone, a press switches the radio's transmit audio to DAX for as long as you hold PTT: Nexus keys only once the radio reports DAX, sends nothing the microphone heard before then, and gives the radio its microphone back once the over has ended and the radio has let go of the transmitter. If the radio has not switched within 2 seconds, nothing is sent and the status line says why; if it leaves DAX during the over (SmartSDR's own DAX button), the over ends at once. Every stop above ends it there too. From a computer driving the station, the three settings show read-only, Set at the station. Needs a check on the air with a sound-card radio, a TCI radio and a Flex.

  • Upload one contact to LoTW.

    Each Logbook row's ⋯ menu has Upload to LoTW. It signs that one contact with TQSL and uploads it, the same way Upload to LoTW does for the whole log: your Station Location from Settings, the same marks on the contact, and the same messages if TQSL is missing, the Station Location is not set or LoTW refuses it. A contact already uploaded, already confirmed, or with no time of day is not sent, and Nexus says why.

  • APRS can use FM-D on the Icom IC-7100, IC-9700, IC-705 and IC-905.

    The per-radio "APRS uses DATA-FM" switch in Settings ▸ Radio ▸ Rig & CAT ▸ Advanced is offered on these Icoms as well as the FT-991/991A. Turned on, APRS puts the radio in FM-D (FM with DATA on) instead of plain FM, so the radio takes its APRS audio from the USB sound card through its DATA MOD input (set DATA MOD to USB in the radio's menu). It works over Hamlib and over the built-in Icom connection, which now also reads FM-D back as FM-D. Over Hamlib, FM-D on an Icom (APRS with this switch, or an SSTV picture on an FM channel) also sets the radio's FM filter to its default, FIL1. Off by default: an APRS setup that works in plain FM is unchanged, and repeater voice stays in plain FM either way.

  • Nexus for Intel Macs.

    Each release now has Nexus_<version>_x64.dmg beside the Apple Silicon Nexus_<version>_aarch64.dmg: the same app built for Intel, for macOS 12 or later, signed and notarized so it opens without a Gatekeeper warning, with Hamlib inside so CAT needs nothing installed, and updating itself from inside Nexus like the Apple Silicon build. A Mac takes the file for its chip (Apple menu ▸ About This Mac). Needs a check on Intel Macs, on the newest macOS and on macOS 12.

  • The first-run setup takes you down your own road.

    Its first step now also asks how many radios Nexus will run. Answering "Two or more" opens the radio setup (which radio for which bands and modes) as soon as the second radio is saved on the "Your rig" step. Answering "Experienced", for someone coming from another program, the log step shows the import preview before anything is written (records, dates, new and already present, problems; Import or Import new only) and lists where your contacts go: each logbook service's state as Settings ▸ Connections shows it, QRZ Logbook's Test, and an offer to open Settings ▸ Confirmations when you finish. Nothing is uploaded and no login is asked for in the setup; the line under the services says what saving a login in Settings sends. New to the hobby still ends with the first-contact walk. Every step stays skippable, and a station that updates sees none of it.

  • Choose the audio waterfall on an Icom

    (#417). Settings ▸ Radio ▸ Rig & CAT ▸ Advanced has a Waterfall source for an Icom whose own scope Nexus shows (native CI-V, or the Icom network connection): Radio's scope, as before, or Audio, the sound card's waterfall, which is finer close to the dial than an IC-7300's scope at ±25 kHz. With Audio the radio stops sending its scope, CI-V stays connected for CAT, and every screen draws the audio waterfall, CW's HiDef included. Each radio keeps its own choice; Radio's scope stays the default.

  • HiDef on the CW scope.

    A steady carrier on the CW screen's audio waterfall was drawn about 12 Hz wide however far the scope zoomed in, because the picture came from the same 171 ms analysis as every other waterfall. The CW scope now analyses its own passband: a carrier draws about 4–5 Hz wide, two signals 10 Hz apart are two lines, and 25 wpm keying still shows the gaps between dits and dahs. It is on by default. The picture runs about a quarter of a second behind the audio; the Fast/HiDef button in the scope's header switches back to today's picture, which follows the audio closely, and remembers your choice. Every other waterfall and the zero-beat indicator are unchanged. Display only: nothing here tunes the radio.

  • The APRS map has the Connect map's tools.

    Above the APRS map: the same map choice as Connect (Globe, Flat, Beam, and Street once a street map is downloaded), zoom buttons, Centre to come back to your station, and a Layers panel (relief, coastlines, US states, the grid and its labels, satellites, the stations). Moving stations draw their track: where each has been while it stayed on the map, its own colour, the selected one heavier with a dot at each report. Every station is labelled, and a label with no room beside its symbol is left out instead of lettered over another; turn Labels off to name only the selected and moving stations. Two filters, by how recently a station was heard and by kind (vehicles, fixed, weather, digis and iGates, objects, aircraft, boats), narrow the list and the map together. The station card says how far a moving station has come and over how long, and Follow keeps it in the middle of the map as it reports; Centre or the Following chip stops it. Your choices are kept for each window. Full screen is left out here so Stop TX and the TX switch stay on screen. Nothing on the map transmits.

  • Move the Logbook's columns.

    Drag a column's header left or right to put the column where you want it, or press Alt+Shift+← or Alt+Shift+→ on a focused header to move it one place (a screen reader says where it went). A click on a header still sorts, and a drag never does. The order is kept with the rest of the column choice in your settings, a choice saved before keeps the order it had, and Reset to the default columns puts the default order back. ✎ Edit and ⋯ stay at the end.

  • Set up radios: a walk-through for a station with more than one radio.

    Settings ▸ Radio ▸ Radios has a "Set up radios…" button (and the first-run setup offers it once a second radio is added). It takes you through your radios, which radio for which band, which radio for each mode where two radios share a band, the radio for everything else, what happens when Nexus needs the other radio, each radio's receive and transmit sound cards and keying method (and, for Serial RTS or DTR, the port the keying line is on), which you can change there with the same checks Settings makes when it saves (CAT keying on a radio with no rig model is refused, and a radio Hamlib cannot key over CAT says so), with a Test CAT for each one, and a "Where would this go?" check that answers from the same routing Nexus uses on the air. Finish saves the band lists, rules, default, sound cards and keying it showed, only where you changed them, and nothing else: a mode you change on a shared band becomes a rule above your existing rules, which all stay, so a band and mode you did not change routes as before. If your radios or their routing change elsewhere while it is open, Finish says so and saves nothing; from the first-run setup, a sound card or keying change you make there to the radio being set up is what first run then saves; Cancel, at any step, saves nothing; a station that never runs it is unchanged. A later Save in Settings keeps what you changed here, since it sends only what you changed in the radio open for Edit. Test CAT on a radio that is not in use only reads it through the connection Nexus already keeps to it: it does not switch to it or key it.

  • Setup asks how much you have done, once.

    On a new install the setup wizard now starts with one question: new to the hobby, new to digital, or experienced. New to the hobby starts with FT8, the log and the maps on, and offers the Getting started guide and a short walk through a first FT8 contact after Finish; new to digital adds the Phone and CW screens; experienced starts with everything on, as before. For the two "new" answers, each Settings tab opens on the sections needed first, with the rest folded to their headings until Show all, a search or a link opens them (Accessibility and Features never fold). Nothing is hidden for good: every screen is in Settings ▸ App ▸ Features and every setting is found by Settings search, including one inside a folded section. Change the answer any time in Settings ▸ Station ▸ Your experience, which asks first, naming them, whenever the change would turn screens off; re-running the setup wizard from Settings keeps the screens you have on. The walk-through is also under Help ▸ Your first FT8 contact. It changes nothing about transmitting: TX still starts off. If you are updating, nothing changes: no wizard, no question (even after the app's browser data is cleared or Nexus is reinstalled, since the setup wizard's record now sits with your settings), and your screens and settings stay as they are (Your experience reads Not set until you pick one).

  • Every laptop at a club shows the club log.

    The club's contacts, newest first, with the time, call, exchange, band, mode and the position that logged each, and another position's repeat marked as a dupe, used to be on the host only. Every joined laptop now shows the same list under its club board, on the Contest screen and in its pop-out: the newest 100 and how many the club holds, as the host lists them and as Club ADIF and Club Cabrillo are written. It follows the host as contacts arrive, a take-back and its Restore included, and a laptop back from a dropout is sent the list as it stands when it rejoins. It is view only: each laptop still takes back only its own contacts. Club sync moves to a new version for this, so every laptop in the club needs this release: an older Nexus is turned away when it joins, and told to update.

  • New IOTA on the Needed board, the decode lists, Spots and Pounce.

    A station on an IOTA island group you have not worked now carries an IOTA chip (in its own coral colour) and ranks just below a new one. Its group comes from a DX-cluster spot whose comment names exactly one reference ("IOTA EU-131", "OC-146 QSL via …"), or from a callbook lookup this session for exactly that call (never a portable call such as W1ABC/P, whose callbook entry is the licensee's home). The award counts each group once, so a group worked on any band clears it, and one worked but confirmed by neither card nor LoTW asks for a confirmation instead. The Needed board has a New IOTA filter, Spots shows the reference a comment names (in the chip's colour while it is a need), and Pounce's widest setting now also alerts for a new IOTA group.

  • IOTA island names.

    IOTA Ltd's island-group list ships with Nexus, by IOTA Ltd's permission, so from the first launch the Logbook's edit form and the contact card show the group's name beside a reference, with IOTA Ltd's credit and a link, and say when a reference is not an island group. Nexus checks iota-world.org for a newer list at most once a month and uses it in place of the one it shipped with; a check that fails keeps the list you have. Settings ▸ Logbook ▸ IOTA island list turns the names off: none are shown, the monthly check stops, a downloaded copy is deleted, and a reference is checked by its shape only, as before. Turning it off during a check also drops what that check downloads. A copy of Nexus built without the list says so in Settings until its first check, and the names show as soon as that check brings the list, with no restart.

  • An IOTA reference offered when you log.

    When a DX-cluster spot comment for the station names an island group, or your callbook lookup for that call does, the log strip, the Logbook's Add form and the confirm-before-log popup show it as a chip: "IOTA EU-131", the group's name if you have the island list, and where it came from. Press Use to log it with the contact, or × to leave it out. Nothing is filled in by itself, a reference you typed is never replaced, and an FT contact logged by itself with no popup logs without one. Use is for that one contact, logged or let go of with Clear or a band change: the next contact starts with nothing taken, even with the same station on another band. When two spot comments for the call in the last 20 minutes name different groups, none is offered. A reference taken from a spot or the callbook stays in your log and its exports but goes to no upload service, and not to DXKeeper or HRD Logbook (both can upload it themselves), until you confirm it on the contact card or edit it. Confirming it later does not send it on to DXKeeper or HRD Logbook, which keep the contact as they first got it; and once confirmed, a LoTW confirmation no longer replaces it.

  • QSL labels and a cards-to-send list.

    A contact's QSL menu in the Logbook can queue a paper card for it (ADIF QSL_SENT=Q), and take it out again; a log imported with cards queued (Q) or requested by the other station (R) keeps them, where they were dropped before. The Logbook's Cards to send lists those contacts to tick and prints their address labels on Avery 5160 (US Letter, 30 per sheet), Avery L7160 (A4, 21 per sheet) or a Dymo 30252 / Brother DK-11201 roll: the station's call, its QSL manager when the log has one, the date, UTC time, band, mode and report sent of each contact, your call and "TNX QSO 73", sorted by call, with several contacts with one station on a label if you like and a start position for a part-used sheet. The labels open in your browser to print at actual size. Nexus then asks whether they printed, and marks the cards sent (bureau or direct, dated today) only when you say yes; nothing else about a contact changes.

  • Award status: Submitted and Granted, and what is ready to submit.

    Each contact now carries an award status you can set: on the contact card under Award status (DXCC, WAS, WAZ, IOTA and USA-CA, each Not submitted, Submitted or Granted), and for many contacts at once in Bulk edit. It is written to the ADIF fields made for it, CREDIT_SUBMITTED and CREDIT_GRANTED, under the award's own credit code; setting one clears the other, and nothing else on the contact changes and nothing is sent. The Awards view has an Award status panel: pick an award to see how many entities are worked, confirmed, submitted and granted, and the contacts ready to submit, one per entity confirmed by that award's own rule and not yet submitted or granted (its oldest confirmed contact, or its newest). Export the list as CSV, save it as a page to print, or mark the whole list Submitted: Nexus says how many contacts it will change and changes them only when you confirm, and Undo takes it back. Granted credit LoTW sends for DXCC counts as before.

  • US Counties award (USA-CA).

    The Awards view counts CQ's United States of America Counties Award: the 3,077 counties, from each contact's county (CNTY), with the 500 to 3,077 levels. The county is read as ADIF writes it (IN,Marion) and as other logs write it (Marion, IN; IN, Marion County; a county alone beside the contact's state; Saint or St.); Alaska counts by Judicial District, and a Virginia independent city, which counts for a bordering county you choose, is left uncounted, as is any county Nexus cannot place (the card says how many). A county is confirmed by a QSL card or an eQSL from an Authenticity-Guaranteed member, as CQ's rules say; CQ names no LoTW credit for this award. The county list comes from the US Census Bureau.

  • Logbook ▸ Clean up ▸ Blank grids.

    Lists the contacts logged with no grid for which Nexus knows the station's grid: the grid it sent on the air, heard by Nexus in ALL.TXT or this session within 24 hours of the contact (the one in effect at the contact's time; a grid heard further from it is not offered as heard, since the station may have been elsewhere), or else its home QTH from a callbook lookup you already made, labelled "home QTH (callbook)" (a call with no stroke and no park, on a contact from the last 12 months). Nothing changes until you tick contacts and confirm; the fill is one change that Undo takes back, and a contact given a grid in the meantime is left as it is. A heard grid counts like any logged grid. A callbook grid is kept for the distance and the map only, as on an FT8/FT4 contact logged with no grid: it counts toward no grid award, fills no GRID need and is sent to no logging service. The fill also writes the distance to the station where the contact has your own grid, as logging a contact does. A contact already uploaded keeps its upload marks, as with every Clean up change: nothing is sent again, and each service keeps what it was sent; the list marks those contacts. Tick all ticks the contacts shown (up to 200), never ones you have not seen, and the confirm says how many of the ticked grids are heard (counted) and how many are from the callbook (for the map only).

  • JS8 sounds from any screen, a WAV file for each, on the output you pick.

    The seven JS8 sounds (a CQ, a heartbeat, an ACK for you, a message to you, a message in your Inbox, a new callsign, one you have worked) now play whichever screen is showing while JS8 is the mode, as JS8Call's do, so a CQ sounds while the Logbook is open. Each plays its own tone or a WAV file you choose in Settings ▸ Digital ▸ JS8 ▸ Sounds (Choose…, Reset to built-in, Test); a file that cannot be played plays the tone and Nexus says so once. A file on a network share (\\server\share, or a link to one) is never opened, since opening it can send your Windows sign-in to that computer: the tone plays and Nexus asks you to copy the file to this computer. A drive letter mapped to a share (or a subst drive) is not recognised: a file there is read like one on this computer, so keep sound files on a local drive. Play the sounds on picks the output, the system default unless you pick one. Nexus never plays a sound on a device your radio transmits from, where it could key a rig on VOX: it checks at every sound, so a sound is refused even if a radio's TX audio device is changed to that output later, and says so. The device the radio actually opened for TX audio is refused too, while the radio holds it: with TX audio set to the system default, a sound is not played on that device after the system default moves to another one. Two identically named sound cards count as one device, whatever number Windows or Linux gives each ("Speakers (2- USB Audio CODEC)", CODEC_1), since the numbers move when one is unplugged. On Linux, the sound server's default output (default, PulseAudio, PipeWire) may be the radio's card, so the sounds play only on a sound card picked by name, and only while the radio's TX audio is set to its own, different card by name. A device whose route is set in a sound configuration file (one of your own from ~/.asoundrc, or plug:, oss, a bare dmix) may reach any card, so a sound is refused against it either way round. With a radio's TX audio set to the system default (or a radio profile with none), pick a specific output for the sounds. A sound plays at most ten seconds, one at a time, and Test plays on the output chosen in the form before Save. All sounds stay off by default. The sound files and Play the sounds on are set at the station. On a computer driving the station over Remote, the sounds play their built-in tones on that computer, where you hear the station, and Settings says so. None is chosen there, and the station's sound files are not sent to that computer: not in Settings, not in a saved profile, not in a settings backup made from it. Such a backup also leaves out where the TQSL program and the voice keyer's recordings are, and restoring it at the station keeps the station's own.

  • JS8: the new-message notice from any screen.

    A message filed in your Inbox puts JS8Call's "New Message Received" notice up wherever you are while JS8 is the mode; Show opens the JS8 screen on that station's messages.

  • JS8's waterfall darkens the audio outside the decode filter

    , as JS8Call's does: half-dark on each side, with a yellow line at each edge, while the filter is on. Every other mode's waterfall is unchanged.

  • JS8 screen options, as JS8Call has them.

    Minimal column labels in Band activity's and the call list's ⋯ ▸ Show columns (short headings, no units in the cells). Reset Band activity, the call list and the RX pane at startup in Settings ▸ Digital ▸ JS8 ▸ Lists and marks (off by default; display only, so HEARING? and the other answers still know stations heard before). Colours and fonts for the RX pane, the two lists and the compose box, each colour kept for the theme you picked it in, with a warning when text would be hard to read and a Reset per pane. The RX pane's Clear and Clear all lists now empty the compose box too, as JS8Call's do, except while its message is going out.

  • Nexus in Russian.

    Русский язык. Pick Русский in Settings ▸ Appearance ▸ Language, or Nexus picks it up from a Russian system language. Translated by loloka; completed and reviewed for this release, so every screen is in Russian, with Russian plural forms (1 контакт, 2 контакта, 5 контактов). Frequencies, callsigns, reports, band and mode names stay exactly as they are in every language. The kit for corrections is in translations/ru/.

  • Every Logbook column sorts.

    Click any column's header to sort by it, as the first sixteen already did: End time, your grid and rig, Name, QTH, Power, Operator, CQ and ITU zone, Continent, Distance, the QSL sent and received dates, each service's upload mark, the contest exchange sent and received, and Notes. Times and dates sort in time order and open newest first, power, zones and distance by value, and text without regard to capitals; a number in an exchange sorts by its value, so serial 9 comes before 10 and 100, and "NY 3" before "NY 12"; an upload mark sorts failed, waiting, then uploaded, and moves as soon as an upload or a card marked sent changes it. A zone or distance that is not a number shows as it was logged and sorts after the numbers. Notes sort by their first 64 characters, and contacts whose notes begin alike keep their time order. A contact with nothing in the column comes after the rest, whichever way it is sorted. Shift-click adds a column as the next key, as before. On a 100,000-contact log each of them sorts in a third of a second or less, as the columns that sorted before do.

  • Nexus says when it switches radio by itself.

    With two or more radios, a band or mode your routing gives to another radio switches to it as before, and now a line says so: "Switched to IC-9700 for 2m · Weak-signal digital", with Undo (back to the radio you were on, at its own frequency) and Why? (Settings ▸ Radio ▸ Radios, where the rules are). The line closes as soon as another radio is in use, whoever chose it, and Undo then does nothing; Undo is refused while you are transmitting, and says so if the radio you were on has since been disabled or removed. The switch itself is unchanged, and so is what happens to a contact in progress. A new install is told by default; an existing station keeps the silent switch it has today until it picks "Tell me, with Undo" under When Nexus switches radio by itself in Settings ▸ Radio ▸ Radios.

  • Each radio's routing in plain words.

    With two or more radios, every card in Settings ▸ Radio ▸ Radios says what Nexus uses that radio for (its rules, its bands, "any band no other radio lists", "everything else"), and the same lines, one per radio, stand above the rules list.

  • A contact card in the Logbook.

    Double-click a contact in the Logbook (or press Enter on the one the keyboard is on) and a card opens with every field it holds to correct: call, date and time on and off in UTC, band, frequency, mode, RST sent and received, power, name, QTH, grid, state, park, summit and WWFF references (theirs and yours), IOTA, your grid and rig, QSL sent and card received, Comment (sharable) in its own full-width box above Notes, and Notes. More fields lists every other ADIF field the contact carries, to correct, remove or add one; the satellite, DXCC entity, station callsign and operator are shown there and changed where they always were. On a contest contact the exchange's own fields (a county, section or zone) are shown there too and are not changed from the card, because exports, uploads and awards read them from the exchange. Your own station's fields (your grid, MY_STATE, MY_CNTY, OWNER_CALLSIGN and the like) can be corrected too; they are not sent to LoTW again, and on a contact LoTW already has the card says so. More fields also shows, read only, what the old contact view showed: when the QSL card was sent, whether the contact is confirmed (and for awards), when each service took its upload, and the county, zones, continent and distance under their own names. The card shows the callbook photo when a lookup has brought one (initials otherwise, with the reason on hover), Look up, a link to the call's QRZ page, the last 10 contacts with the call (UTC date and time, band, frequency, mode, RST and confirmations; click one to open it on the card), and the contact's history with its Undo. Save is the Logbook's own edit: one change, one Undo, refused if the contact changed elsewhere meanwhile, and a field you did not touch is written back exactly as it was. A box the log reads as "leave it" when empty (state, end time, the parks, IOTA, your grid and rig) cannot be emptied here, and the card says so rather than saving nothing. Closing it or opening another contact with unsaved changes asks first; after an Undo in its history it shows the contact as the undo left it. A multi-line field keeps its line endings exactly as stored unless you edit its text. A single click still selects, and the pencil still opens the edit form. Nothing is looked up unless you press Look up.

  • Download confirmations in the Logbook.

    One button downloads new confirmations from LoTW and eQSL, where you have set them up, one after another, through the same downloads their own button and Settings run, and says what each one brought or why it failed. The QRZ sync keeps its own button, since it can also add contacts logged elsewhere. While a download runs, the other download buttons wait for it; Settings' LoTW and eQSL downloads also wait for each other.

  • A Cancel button on the Logbook's edit form

    , beside Save. It closes the form without saving, as Esc does.

  • ATU ON (#344).

    The ATU button now reads ATU ON while the radio reports its built-in tuner in line: after a tune-up, and within about ten seconds of switching it at the radio. A click on the button runs a tune-up as it always did, ATU or ATU ON. To switch the tuner out, right-click the button (or press and hold on a touch screen) and choose Switch tuner out; Nexus sends the same CAT command it starts a tune-up with, set to off (AC000 on a Yaesu). Switching out puts nothing on the air, is refused while anything is transmitting, and a switch-out the radio refuses or that could not be sent in time is said on the CAT line. The menu is offered only where Nexus can switch the tuner out: not on a Flex run by Nexus's own connection, and not on the built-in Icom connection, which has no tuner command. Needs a check on the radios.

  • JTTY: your own F1 to F8, and WSJT-X's FT8-style set.

    Edit keys turns the JTTY F-keys into their messages to edit, as in WSJT-X, and Messages picks WSJT-X's Contest set or its FT8-style set (%H %M %G, %H %M %R, %H %M R%R, %H %M RRR, %H %M 73, CQ %M %G); each set keeps its own messages between sessions and Defaults puts a set back. While the keys are open for editing, F1 to F8 send nothing, pressed or clicked, and say so; they send again when you finish editing. A message in one of WSJT-X's contest forms goes in JTTY's compact form, as WSJT-X sends it, and any other goes as text. %R is now the SNR of the newest decode naming the call you clicked, as in WSJT-X, rather than always -10, and it goes only with the call it was taken for.

  • JTTY asks to log the contact after TU.

    When a message starting TU goes out, as WSJT-X opens its Log QSO window, the log pane asks "Log the contact with W1AW?", naming the station the TU thanked: the station your last messages went to, even with keys such as F4 and F8 that send no call (after F6, the one before the station it calls). Asking changes nothing on screen. Log QSO logs through the strip's own Log only when the strip holds that call; an empty strip is filled for you to check first, and another call is left alone with nothing logged. The first answer closes the question on every screen, Remote included, and so does logging that contact any other way, such as with the strip's own Log, so a contact is never logged twice. If the log line then cannot log it, the question comes back. A second TU to another station before you answer replaces the question and says whose it was, and an unanswered question goes after two minutes; when Log QSO finds the question already closed, it says why. Nothing is logged unless you press it, and nothing more is sent.

  • JTTY: decode again.

    Decode in the All decodes pane decodes the last three minutes of receive audio again, as WSJT-X's Decode button does, and a double-click on a waterfall row decodes 40 seconds from just before that row. What one press finds shows under one JTTY review heading in both panes, after the live lines, and never pushes live lines out of the window; it is not written to ALL.TXT, spotted or counted as hearing a station.

  • JTTY: WSJT-X's smaller controls.

    Ctrl+H types both calls (%H %M ) into the compose bar, Ctrl+Shift+K empties it and stops what is still going, Rx→Tx and Tx→Rx copy one frequency to the other, and a Frames count (sent/total) shows beside Send while a transmission runs. PSK Reporter spots of a JTTY message that carries a grid (CQ W1AW FN31) now include the grid, as WSJT-X's do.

  • ARRL RTTY Roundup.

    Settings ▸ Contesting and the contest setup now offer the ARRL RTTY Roundup (first full weekend of January, never on January 1; 1800Z Saturday to 2359Z Sunday), from ARRL's current rules. A station in the 48 states, DC or Canada sends RST and its state or province; everyone else sends RST and a serial from 1. Each station counts once per band for one point, and the multipliers (states, Canadian provinces and territories with Labrador, and DXCC entities other than the USA and Canada, Alaska and Hawaii included) count once per log. The Cabrillo file is ARRL-RTTY with every contact as RY, and its LOCATION is your ARRL section (DX outside the USA and Canada), as ARRL's log upload asks; the contest setup asks for the section. The contest is RTTY only: a contact in JTTY, FT8, PSK31, CW or any other mode is logged and kept but scores nothing, and the screen says so; nothing is refused. The setup and the Contest screen offer the RTTY screen to work it from, and the contest setup no longer asks for a county for CQ WW RTTY or this contest, neither of which sends one. Operating time counts breaks of 30 minutes or more as off time, as the rules do.

  • Finals temperature beside the transmit meters.

    The TX meters (SWR, ALC, PO, COMP) now show a TEMP row where the radio reports its power-amplifier temperature: a Flex on its own connection, in degrees, and the FTDX101D/MP as its own TEMP meter (Yaesu publishes no degree scale for it, so it reads as a share of the meter, not degrees). Nexus asks the Yaesu for its ID first and reads the meter only when the radio says it is an FTDX101D or MP, whatever radio is chosen in Settings. Degrees follow Settings ▸ Units: °C, or °F with Imperial. With an amplifier set up, its temperature shows too, labelled AMP beside the radio's RIG; an Elecraft KPA's is converted like the radio's, and an SPE's is shown as the amplifier sends it, since its protocol does not say °C or °F. The row is read while receiving and stays between overs; a reading the radio stops giving goes from the screen. Radios that report no temperature over CAT (the other Yaesus, every Icom, the Kenwoods through Hamlib) show no row and are asked nothing new.

  • Set the speech processor from the COMP meter, and step the SSB transmit bandwidth.

    In the Phone cockpit the COMP bar still shows compression, and a line on it now marks the processor level: drag the bar, or use the arrow keys, to set it. This works wherever the processor-level control does, Icom and Yaesu alike. While the radio is transmitting, a new level waits and goes to the radio once, right after the unkey, on any radio controlled through Hamlib or another connection: a radio that stops answering mid-over would otherwise hold the CAT link, and your release behind it, for several seconds. On Nexus's built-in Icom connection it goes while you talk, at most four times a second; there a radio that stops answering can hold the release for one short request (about a third of a second), and no more level goes until you unkey. During a voice or CW message or a tune it goes once that ends. The line keeps the new level until the radio reports it. On an Icom on Nexus's built-in Icom connection (IC-7300, IC-7300MK2, IC-7610, IC-705, IC-9700, IC-905, IC-7760), PROC beside the bar switches the processor, and a chip steps the SSB transmit bandwidth through WIDE, MID and NAR; the chip shows what the radio reports back, is held while transmitting, and a change is made between overs. On a Yaesu there is no PROC button or bandwidth chip on the meter yet: through Hamlib, the radio's own menu write holds the CAT link for a second or two, and the PROC switch needs checking on a radio first. The processor switch in the transmit chain is unchanged.

  • A note about DAX for a Flex over SmartSDR CAT.

    In SSB a Flex transmits from its own mic unless DAX is on for voice, so a voice-keyer message keys the radio with no RF. The first time a voice message plays on a Flex controlled through SmartSDR CAT, a note in the voice keyer says to turn on DAX in SmartSDR's P/CW panel; Got it hides it for good. It never holds up the message. The Flex page of the manual says the same.

  • Save the settings backup where you like.

    Settings ▸ App ▸ Backup & reset now has Save as… beside Back up: it opens your system's save dialog in Downloads with the usual dated name filled in, so you can pick a USB stick or any other folder. Back up still saves straight to Downloads.

  • The Icom IC-7850 and IC-7851 in the radio list.

    Pick "Icom IC-7850/7851" in Settings or the setup wizard and Nexus drives it through Hamlib (model 3075) over its serial CI-V port. It has not been on a bench yet. Nexus's own Icom connections (native CI-V and the network port) do not cover this radio.

  • Add ticked memories to a group in one go.

    In Memories, the bar that appears when you tick rows now has Add to group…: pick a group and every ticked memory in view joins it, keeping the groups it was already in. Rows ticked but narrowed out of view by a search or filter are left alone, as with Delete.

  • Import a laptop's file into the club log.

    The host's Contest screen has a new button in the club section, Import a laptop's file…. Choose a laptop's contest journal (the file named fieldday_backup_ and 8 characters, .adi, in that laptop's Nexus settings folder) and its contacts go into the club's log exactly as club sync would have sent them: the same dupe rules, and the same contacts kept out (another contest, an earlier running such as a rehearsal, a contact logged on another call than the club's, one taken back on that laptop). The host then says how many it added, how many were duplicates or already in the club log, and how many it kept out and why. Importing the same file twice adds nothing, and nothing is ever taken out of the club log. The import adds the contacts a few hundred at a time without holding up the radio; if the club event changes partway through, it stops and says how many it added, and importing the file again adds the rest. A file larger than 16 MB is refused. This covers a host PC that dies with contacts only on its own disk, and a laptop that never synced. The Field Day manual's club checklist has the steps.

Changed

  • Space weather is logged with each contact on a new installation.

    Log space weather on each contact (Settings ▸ Services ▸ Integrations & Feeds ▸ Propagation) now starts on for a new installation, so every contact carries the K index, A index and solar flux it was made under, which the Path column uses to model that contact. If you already use Nexus, nothing changes: it stays as you had it.

  • Settings is regrouped so each thing sits where you would look for it.

    Satellite Doppler and the orbital elements have a Satellites tab of their own (they were on Radio). The logbook services, uploads and the programs that log through Nexus are on Services, the new name for Logging & Connectors. Features, App updates, Start at sign-in, This computer, Performance, the data folder and Backup & reset are together on App, which takes in the old Config tab and the computer settings that were on Appearance; Appearance keeps how Nexus looks. On Digital, JT65, MSK144, the WSPR and FST4W beacons, FST4 and Q65 sit under one Weak-signal and beacons heading. Remote keeps its own tab. Nothing you set is changed or lost: every setting keeps its value, search finds each one where it is now (and "services", "app" and "satellite" find the new tabs), and a link or pointer to an old place lands on the new one. For a station that answered "new to the hobby" or "new to digital", the weak-signal heading starts folded like the other sections a first contact does not need; Features and Accessibility never fold.

  • Logbook rows: Edit and a ⋯ menu, with room for the contact.

    Each contact now shows two buttons, always: ✎ Edit and ⋯. Everything else a row could do is in the ⋯ menu, one choice at a time: spot the station, upload it to QRZ Logbook, Club Log, HRDLog, World Radio League or LoTW (each by its full name, with whether this contact was already sent there, for example "Uploaded, 2026-10-28 14:06" or "Not uploaded"), correct it on QRZ Logbook, mark a QSL card sent or received, tag a satellite, or delete it (which still asks first). The menu works from the keyboard (arrow keys, Enter, Escape). Until now a row's buttons were invisible until you pointed at the row but still took clicks, so a click or a touchscreen tap on the empty-looking right end of a row could upload a contact to QRZ without you seeing a button; nothing in a row is clickable any more unless you can see it. The buttons are 32 by 32 pixels with 8 pixels between them, and the narrow button column leaves the width to the contact: the callsign, date and time now show in full at 1024 by 768, and Country and Notes get more room. Choosing Compact under Density (Settings, Appearance, Workspace) makes the rows shorter.

  • FT8: rover (/R) calls and "TU;" messages show only in Contest mode or as a Hound.

    With neither on, FT8 no longer shows decodes of rover (/R) calls or "TU;" messages, as WSJT-X does: away from a contest such decodes are most often false, and this should remove the odd /R station that seemed to be calling you. To work rovers, or to give and take "TU;" exchanges in a contest, turn Contest mode on; as a Hound they show too. FT4 is unchanged, as in WSJT-X.

  • The Phone audio scope remembers its span.

    Auto (your radio's filter width) is still where it starts; the Full, Voice, 1.5k or 800 button you pick is now kept for the next time, each window its own.

  • Each mode's waterfall remembers its own span.

    RTTY, PSK, JS8, SSTV, JTTY, Tempo and the FT modes each open their waterfall at Std (200–3000 Hz) until you pick a span in that mode, and a span you pick in one mode stays in that mode. Before, all of them shared one span, so choosing 1 kHz on the RTTY waterfall made FT8, PSK and JS8 open on a 1 kHz window too. A span you had already chosen keeps opening as it did in every mode until you pick another there. WSPR still shows its 200 Hz sub-band.

  • JTTY uses its published dials.

    The JTTY band picker now serves JTTY's published (preliminary) dial table: 160 m 1.838, 80 m 3.575, 40 m 7.090, 30 m 10.140, 20 m 14.090 (the main meeting place), 17 m 18.100, 15 m 21.090, 12 m 24.920, 10 m 28.090, 6 m 50.160 and 2 m 144.160 MHz, USB. 160 m, 80 m, 40 m and 17 m moved (from 1.839, 3.572, 7.078 and 18.104); 17 m is now the same dial as FT8, and 80 m the same as FT4, so expect their signals in the passband.

  • Saving your first QRZ Logbook, ClubLog or eQSL login asks before sending your old log.

    Saving a login turns that service's uploads on and sends every contact not yet sent there, and contacts brought in by Import ADIF are never marked as sent, so saving a QRZ key after importing your log from another program started sending the whole old log. The first login you save for one of these services now asks "Send N contacts not yet sent to QRZ Logbook?" (or ClubLog, or eQSL). With the default, "Only contacts made from now on", no contact already in your log at that moment is sent there automatically: not by saving the login again, not after a restart, a factory reset or restoring a backup, and not from any other Nexus window that shares the log (each radio's window when you run radios at the same time, or the main window afterwards). Older contacts can still be sent from the Logbook whenever you choose. "Send them" sends them as before. With nothing waiting nothing is asked, and uploads start from that moment, so a log you import later is not sent by itself either. If you turned that service's upload on before saving its login, nothing is asked and the contacts waiting for the login are sent, as before. Clearing a login and saving one again asks again, and answering "Only contacts made from now on" again keeps the first starting point. A station that already has its logins saved sees no change, and saving one of them again asks nothing; so does a save when the system keychain or the log cannot be read, which then sends as before.

  • A callbook lookup no longer fills in the IOTA reference by itself.

    A callbook's island group describes the licensee's record, not necessarily where the station is today, so it is now offered when you log (see Added) instead of being written into the contact; Logbook ▸ Update from callbook no longer fills it either. Contacts already in your log are unchanged, and their references are sent as before.

  • FT8/FT4: "Prompt before logging" opens the popup by itself with auto-log off too.

    As WSJT-X's "Prompt me to log QSO" does, the confirm-before-log popup now opens by itself at the end of every completed contact whenever that setting is on, whether auto-log is on or off; before, with auto-log off it did nothing until you pressed Log. The contact is offered once, so Log then has nothing more to log, and a CQ run calls CQ again while the contact waits in the popup, as it already did with auto-log on. If 64 contacts are left waiting in the popup, the oldest finished contact is logged without your confirm, also as with auto-log on; a contact you opened by clicking the station's RR73 or RRR never is. With auto-log on and the popup off a contact still logs itself, and with both off nothing is logged until you press Log, as before. The Settings hints for both now say exactly this, in every language.

  • RTTY: "Prompt before logging" opens the popup by itself with auto-log off too.

    A contact the RTTY Auto sequencer completes now follows the same rule as FT8/FT4: with "Prompt before logging" on the confirm-before-log popup opens by itself, whether auto-log is on or off; with auto-log on and the popup off it logs itself, and with both off nothing is logged, as before. What RTTY transmits, and when, is unchanged.

  • A contact with nothing in a Logbook column sorts last, whichever way you sort.

    In Country, Mode, Sent, Rcvd, Park, Grid, State, IOTA, Contest and County, an ascending sort put the contacts with that column empty at the top; they now come after the rest in both directions, as in every other column. A value with a space in front of it now sorts by its first letter. Contacts are not changed, only their order on screen.

  • The prompt-to-log popup has a propagation mode, and your power, operator and grid.

    With "Prompt me to log QSO" on, the popup now has WSJT-X's Prop mode picker (EME, Sporadic E, Meteor scatter, Tropospheric ducting, F2, Aurora and the rest of WSJT-X's list), and the mode you pick is offered again for the next contact until you change it, as WSJT-X's Retain does, so an EME or 6 m Es run is picked once. Power (W), Operator and My grid are in the popup too, filled in as Nexus fills them today; change one and the contact is logged with your value, clear one and Nexus fills it in as before. A grid you change there goes to LoTW, as one changed in the Logbook editor does; one left as Nexus filled it in (typing only its start, EN52 over EN52ab, counts as leaving it) stays out of the LoTW upload, so TQSL signs from your Station Location. A contact made on a satellite opens on Satellite with its satellite name, as the satellite strip sets it, whatever mode you picked last, and the mode you picked is offered again for your next contact off the satellite.

  • ALL.TXT is written by default on a new install

    , as WSJT-X writes it. An existing install keeps its setting: if you had Settings ▸ Services ▸ "Write ALL.TXT decode log" off (the old default), it stays off until you turn it on. JS8's DIRECTED.TXT follows the same switch, as before.

  • Every mode's waterfall starts at the same size.

    FT8/FT4, Phone, CW, JS8, JTTY, PSK, RTTY and Tempo now open with a waterfall the full width of the window and a quarter of the cockpit's height. CW's was much shorter (13 %) and Phone's and FT's a little shorter (22 %). Tempo's waterfall leaves the narrow right-hand column for a full-width strip across the top, with a divider to set its height like FT's; Band Activity and the link pill stay in the right column. With Band Activity, Spots or Needed on Phone's left side, the scope and the transmit strip now run the full width above the side instead of beside it, and PTT, Stop TX and Tune stay where they were. A height you have already dragged is kept as you left it. On a small window a cockpit whose own panes need the room (CW, PSK and RTTY at 1024×768) starts its waterfall as tall as that room allows. SSTV keeps its larger picture area.

  • RTTY's tuning scope moves beside the waterfall.

    The crossed-ellipse tuning display now stands to the right of the waterfall, in the same strip, with ◀ ▶ and the words beside the square, so RTTY's waterfall is as tall as every other mode's and the tuning square keeps its size. The waterfall is that much narrower on RTTY. The decoded text and the log no longer give up room for it. Hide Tuning and the waterfall takes the full width; hide the waterfall and the tuning display keeps the strip. On a short screen the decoded text, the log strip and the waterfall give up room, so RTTY fits a 1080p screen at 150 % UI scale.

  • The Logbook uses the whole window.

    The log sat in a centred column with empty space on both sides (1100 px of a 1920 screen, 1600 px of a 2560 one), and a table with added columns scrolled sideways inside that column. It now spans the window at every size from 1024×768 up; where the columns you chose still do not fit, the table scrolls sideways inside itself, header and rows together, and the page never does.

  • PSK Reporter shows your Nexus version.

    Spots you upload to PSK Reporter now name the reporting software as "Nexus v" and the version you run (for example "Nexus v1.21.0"), as WSJT-X and other programs send theirs, instead of "Nexus" alone. The WSJT-X UDP Heartbeat that loggers read now gives the same version (it said 0.2.0).

  • Play receive audio here stays off on an output that might be your radio's TX audio.

    It already refused the radio's TX audio device by name. It now also refuses one it cannot tell apart from it, as the JS8 sounds do: a second identically named sound card, whatever number Windows or Linux gives it ("USB Audio CODEC #2", "Speakers (2- USB Audio CODEC)", CODEC_1), and, on Linux, the system default, PulseAudio or PipeWire output against a radio on a sound card (or the other way round), since either may be the radio's card. On Linux, pick Headphones or speakers by its card name; Settings ▸ Radio ▸ Receive audio on this computer says so. Two different sound cards monitor as before. When it refuses, the notice names the radio and the device it clashed with (or "the system default output") and what to change. Every radio's TX audio counts, the radio in use or not, but a radio whose transmit audio never goes through this computer's sound card (the Icom network connection, TCI with its audio over TCI, Nexus's own Flex client with native audio) and that has no TX audio device set no longer counts as the system default, for the monitor and the JS8 sounds alike.

  • Tune is in the same place on the FT8/FT4 screen as everywhere else.

    The FT8/FT4 strip now opens with TX On/Off, Tune, ATU, Stop TX and Hold Tx, then Call CQ and S&P, so Tune sits right after TX On/Off exactly where every other operating screen's TX strip has it. What each button does is unchanged.

  • Pounce follows the Spots view's "Spotted from" chips.

    Narrow the Spots view to spots heard from a continent or a country and the Pounce alert listens only to spots heard from there too: with France chosen, a new one heard only by a US skimmer stays quiet, and the same station heard from France then fires. With no chip chosen Pounce hears every spot, as before. The chips of a Spots pane on another screen do not change Pounce.

  • Sharing the radio steps aside when its port is taken.

    If another program, such as your own rigctld, is already listening on the sharing port (4532 by default) when Nexus starts, sharing no longer stops with a complaint: it stays on and moves to the next free port, leaving the other program's port alone. Settings ▸ Radio shows the address sharing really uses, its Copy button copies that address, and a note says which port was taken. Your "Sharing port" setting is not changed, and the next start tries it first again. A NET rigctl radio pointed at that other program's rigctld (for example your own rigctld on 127.0.0.1:4532) now connects once sharing has moved, instead of being refused as "Nexus talking to itself"; a radio pointed at the port sharing really uses is still refused. Sharing also never takes the address of such a radio's rigctld when that rigctld is not running yet: it moves to the next free port, so you can start your rigctld after Nexus and the radio connects to it, and the note says the port is your radio's rigctld address.

Fixed

  • Restoring a settings backup keeps this computer's TQSL program, JS8 sound files and voice-keyer recordings.

    A backup made elsewhere could point Nexus at a different program for LoTW signing, or at different sound or recording files; a restore now always keeps the paths set on this computer, in your settings and in each config profile. The backup's voice-keyer slot labels still restore; restored onto a new computer, those slots have no recording until you record or import one there. To use a different TQSL program, choose it in Settings.

  • Settings backups never carry a profile's Club Log or Cloudlog key.

    A settings backup left out the Club Log and Cloudlog keys in your own settings but kept the ones saved in each config profile, so a backup you mailed yourself or shared for help could carry them. It now leaves them out of every profile too. Restoring a backup keeps the keys already saved on this computer, its own and each profile's of the same name; before, a restore cleared the Club Log key. A backup made from a computer driving the station over Remote now holds only the station's own part of the screen state (memories, watch list, chase lists, profiles), not that computer's layout or UI scale, and restoring it keeps the station's own layout.

  • Restoring a backup that points Cloudlog or an Icom radio at a different address asks you to enter that key again.

    If the backup's Cloudlog/Wavelog URL, or an Icom network radio's address or port, is not the one this computer uses, Nexus removes the key or network password saved here for it, and the Connections log says which to enter again in Settings. The same address, written with different capitals, a trailing slash or the usual port, keeps the key. Restoring onto a fresh install or reset settings, where this computer has no address of its own yet, keeps the key saved here but does not use it until you say so: a notice at the top of the window asks "Use your saved Cloudlog key with …?" or "Use your saved password for … with …?", with Use and Remove. Until you answer, Cloudlog uploads wait in the queue and the radio does not sign in with that password. A restore that leaves a radio out of the roster removes that radio's saved network password too, as removing the radio does, so a radio you add later never sends it to another address. An older backup that still carries its own Cloudlog key puts that key in use right away, where before it waited for a restart, and the Connections log no longer says a key was removed when none was saved. Removing a radio when the keychain cannot delete its saved network password keeps that password in the notice, never used, and a radio you add never takes its place or signs in with a password left behind. A restore is refused while you are transmitting, with "Finish or stop the transmission first, then restore.", so a slow keychain can never hold the transmitter on past the end of an over, or while TX is enabled between the overs of a digital mode, so it can never make the next over late either. Remove in the notice never holds up the end of an over while the keychain waits for you to unlock it, and deletes only the key the notice asked about: a key you type in Settings, or a backup restores, while the notice is up stays saved. A key Nexus kept only because the keychain would not remove it is shown as such, with Remove only: it was saved for another address, so it is never used with the new one. The notice asks about a radio's network password only once that radio has a network address, and its refusals are in your language. While the Cloudlog key waits for your answer, older contacts held back for Cloudlog do not slow down the catch-up uploads to your other logbooks; once you answer, they go one at a time. Use in the notice answers only for the address it showed: if a restore or a Save changed that address since, nothing is let go and the notice asks again.

  • Two radios in two windows no longer fight over each other's radio.

    With Simultaneous radios on, each radio runs in its own window, but every window also opened a read-only connection to the other window's radio, starting a second rigctld on that radio's port. One window ended up sharing the other's rigctld and lost its radio every time the other window reconnected, while the other kept trying to open a COM port that was already in use, and the radio was polled by both windows at once. A radio window now connects only to its own radio. The single window that follows the band (Simultaneous radios off) still watches every radio as before.

  • PSK Reporter stays live in both radio windows.

    Two radio windows under the same callsign connected to PSK Reporter under the same name, so each knocked the other off in turn and the live spots in both kept dropping. Each radio window now connects under its own name. The DX cluster can do the same thing when both windows log in with the same callsign: give each window a different Cluster login SSID (for example 1 in one window and 2 in the other) in its Settings.

  • The geomagnetic "tomorrow" forecast showed today's.

    The G↗ tmrw chip beside the space-weather scales was reading NOAA's forecast for today, not tomorrow. It could say a storm was coming tomorrow when only today's forecast had one, or show nothing when tomorrow's forecast had a storm. It now reads the forecast NOAA dates for tomorrow. If NOAA's file has no forecast for tomorrow, the chip is not shown, rather than showing today's in its place. A forecast kept from before midnight UTC (Nexus keeps its last copy for a few minutes, and while offline) is no longer shown as tomorrow's once that day has begun.

  • Split Operation never sends an over on the wrong frequency.

    With Split Operation set to Rig or Fake It, Nexus moves the over's audio into 1500–2000 Hz and the transmit frequency by the same amount. If the radio refused that frequency change, or did not answer it in time, the over was sent anyway from the receive frequency, 500 Hz or more from where you put it, and nothing said so (seen on a FLEX-6600M, and on any radio connection that does not do split). Now, as WSJT-X does, nothing is sent, TX turns off, and the status bar says the radio did not take the split frequency and suggests Fake It or None.

  • Rig split stays on between overs, as in WSJT-X.

    Nexus turned the radio's split on before every over and off after it. On a FlexRadio through SmartSDR CAT that can open a new slice every over. Split is now turned on once and left on while you operate; each over still sets the transmit frequency and mode. It is turned off (or your own split put back) when you change Split Operation, change band or radio, turn TX off, leave the digital screens, or quit Nexus. If the radio does not take that, Nexus tries again every couple of seconds while nothing is transmitting, sends no FT over until the radio has taken it, and never mistakes that leftover split for one you set yourself. A radio that reads back its split as off also counts as done. If Nexus still cannot confirm SPLIT is off, the status bar says so: turn SPLIT off on the radio, then disconnect and reconnect the radio in Nexus (Rig Model None, Save, then your radio again, Save). A radio that refuses the first split-on of an over (an Elecraft K3 with VFO B in a mode it will not pair) now gets VFO B's frequency and mode set and is asked again, as before.

  • Split Operation shifts a TX offset of exactly 2000 Hz, as WSJT-X does

    (+500 Hz on the dial, audio at 1500 Hz). Only 1500–1999 Hz is sent without a shift.

  • A voice-keyer message on a Flex never hands the transmission to the shack's microphone.

    With the Flex client's native audio on, a recorded message goes to the radio over DAX. If the radio was switched back to its own mic while a message played (SmartSDR's DAX button), the radio stayed keyed to the message's end and the shack room went out in its place. The message now stops and the radio unkeys as soon as the radio leaves Nexus's DAX. A message also keys only while the radio is taking Nexus's DAX, including during a Remote session, when the radio is already on DAX for the Remote operator; otherwise it is dropped and the keyer says why. Beside SmartSDR's own DAX, and with native audio off, messages play as before.

  • On a Flex, an APRS beacon or SSTV picture right after a voice message no longer goes out on the shack's microphone.

    With the Flex client's native audio on and the radio on its own mic for Phone, APRS and SSTV refuse to send, because the radio would put the mic on the air instead of the packet or picture. For a moment at the end of every voice-keyer message (longer on a busy computer), Nexus briefly thought the radio no longer had its mic, so a beacon or picture started just then could key the radio on its own mic, and a voice message pressed just then could be dropped as if the radio still had the mic. Nexus now reads the radio's mic setting in the same instant as everything else it checks, so the refusal holds all the way through. Needs a check on a Flex.

  • A Remote operator talking on a Flex never puts the shack's microphone on the air.

    With the Flex client's native audio on, a Remote operator's voice goes to the radio over DAX, but their press keyed the radio whatever it was taking its transmit audio from: if the radio had not switched to DAX (the switch refused, not yet answered, or SmartSDR's own DAX button set back to MIC), the shack room went out in their place. A press now keys only while the radio takes Nexus's DAX; otherwise nothing is transmitted and the Remote page says the station's audio changed and to press again. An over already on the air ends at once if the radio leaves DAX under it. With native audio off, a Remote press keys as before. Needs a check on a Flex.

  • The proton-event chip on the map has its own violet edge again.

    It was drawn with the flare chip's orange edge. It now also sits under the flare chip in one column, so an unfolded flare chip never covers it.

  • Space weather shows only what NOAA measured, and when.

    When NOAA's space-weather service did not answer but the spot feeds did, the Space Wx gauges and the top bar of the Connect window and TV page showed made-up values (SFI 120, Kp 2, A 8, X-ray B1.0) as if they were readings, in green, and the band strip called bands open or closed from them. Each number now carries the time NOAA measured it. One NOAA could not give shows "—" and "no reading from NOAA", and one that is out of date shows how old it is, dimmed. Neither colours the gauge, gives band advice, or raises a storm or flare alert, and an out-of-date number is dimmed too. With no current solar flux or Kp reading, a band nobody has heard reads "—" (unknown) in the band strip, the Band Advisor, the band menu, the band tiles and the Basic lines, instead of "Open". The Band Advisor no longer gives the model's reason for such a band ("open per model") or ranks bands by the model, and when nothing has been heard the headline says "Nothing heard yet." instead of suggesting a band. The same holds on the TV page's summary view, which also shows a dash for any index that is not current. The storm and flare lines in Insights, the map and Remote, and the polar-cap edge on the map, also speak only from a current Kp or X-ray reading, and so do the path outlook, the band outlook and the DXpedition windows. Hovering a gauge shows where its number came from and when. One NOAA product failing no longer blanks the others, and a missing X-ray reading is shown as missing instead of "no significant flares". The SFI is now the day's noon reading, the same number as the 30-day trend beside it.

  • The A gauge is NOAA's daily A index.

    It showed an estimate worked out from the minute-by-minute Kp, labelled as the daily A. It now shows NOAA's daily planetary A with the day it is for; the estimate is on hover. With "log space weather" turned on, a contact gets an A index only for its own UT day. NOAA publishes a day's A after that day ends, so a contact logged live usually has no A yet and A_INDEX is simply left off; it is never given the day before's A. A contact you type in later for an earlier day gets that day's A, while NOAA's daily list (the last 30 days) still has it. The K index goes on only if NOAA's Kp reading was taken within three hours of the contact, and the solar flux only if it was measured within two days of it. Any index NOAA did not give is left off the contact instead of being filled with a guess. Contacts already in your log are not changed.

  • The aurora oval is labelled as a forecast.

    NOAA's aurora oval is a forecast for half an hour to an hour and a half ahead. The map's Aurora row now says which time it is a forecast for, and when the solar wind it was made from was measured (on hover); the globe's row says Forecast. A forecast whose time has long passed says how old it is. The space-weather scales say how old they are when NOAA has not answered for a while, and the ionosonde MUF readings kept through an outage age as they should.

  • Flare recovery estimates for the biggest flares.

    The fade-recovery estimate stopped growing above an X5 flare, so an X10 read as recovering as fast as an X5. It now follows NOAA's estimate all the way up (about 150 minutes for an X10). The flare alert is also translated now.

  • Tune and PTT no longer reach the radio after you let go.

    When rigctld held a command for many seconds (for example a rigctld-compatible bridge or network rigctld that answers slowly), Nexus stopped waiting for the key after 0.7 seconds and sent the release on a new connection. rigctld could then carry out the release before the key, so the radio keyed after you had released Tune or PTT, and could stay keyed. The release now always goes out behind the key on the same connection, so the radio gets the key first and the release after it, and until the key's answer arrives Nexus sends the radio nothing but the release, a further key, or the stops of a Stop TX, each queued behind it in order. If a rigctld-compatible program loses a key's answer altogether, Nexus gives that connection up once the release has gone out behind the key and the answer has been missing for 30 seconds (counted from when the key first went unanswered, not from the release), and starts again on a new one. Tune now plays its tone only after the radio confirms the key. If the radio does not answer the key in time, the tune stops, no tone is played, and the release is sent behind the key. The delayed key can still reach the radio when rigctld frees up, so the radio may key for a moment then, with no tone, and it is released straight after. This includes a radio so slow that rigctld itself reports it did not answer. The status bar now says PTT NOT CONFIRMED, as a warning, when the radio did not answer the key in time, and keeps PTT NOT ACCEPTED for a radio that refused it. Voice messages, SSTV pictures and the Remote microphone still go out on a key that was not answered in time, as before, now with the PTT NOT CONFIRMED warning. FT8, FT4 and the other slot modes are unchanged. On a network radio, a key now waits up to 2.5 seconds for the radio's answer, so a key that works but answers slowly is no longer reported as a problem. Not yet checked on a radio.

  • The 3-D globe rests while you cannot see it.

    With Nexus minimised, behind another tab, or the globe scrolled out of view, the 3-D globe on the Conditions screen kept drawing: once a second while satellites were shown, and all the time while it was spinning on some computers. Each picture is a lot of work for a computer without a graphics card. It now draws nothing until you can see it again, then catches up at once and carries on spinning if it was. What it shows is unchanged.

  • A sound card that is not plugged in no longer keeps Nexus busy.

    When the sound card chosen in Settings is not there (the rig switched off, the USB cable out), Nexus used to look for it every 2 seconds for as long as it ran, asking the computer for its whole list of audio devices each time and writing the same line to the diagnostic log. It now looks again after 2, 4, 8 and 16 seconds, then every 30 seconds, asks for the list once per look, and writes the log line only when the wait gets longer. Plugging the radio back in is still noticed within half a minute, and choosing a different sound card, or switching radios, tries at once. The no-audio warning reads as before.

  • Your own transmission lists after the message it answered.

    In FT8 and FT4's Band Activity and Rx Frequency lists, the line for something you sent could appear above the station's message it was a reply to: answer a station's RR73 or RRR with 73, and your 73 sometimes showed above their RR73. It happened most when the reply was keyed at the start of a period before that period's decodes had come in, which in Q65, FST4 and MSK144 is every time, and when the message switched mid-transmission (both lines are now shown). What was sent and when was always right; only the list order was wrong. Your transmission now follows the messages heard in the period before it, and comes before the messages heard while you were sending, oldest first or with Newest on top.

  • Nexus asks ClubLog for DXpedition data far less often.

    The DXpedition board's list of operations on the air now is fetched from ClubLog at most once a day, and ClubLog's Most Wanted list at most once a month. Both are kept on disk, so restarting Nexus does not fetch them again. If ClubLog does not answer, Nexus waits an hour before asking again, then longer each time, and keeps showing the last list it had. Before, an NG3K DXpedition calendar that would not load made Nexus ask ClubLog again every few minutes; now the calendar is retried on its own, also with a growing wait. The Most Wanted list no longer sends Nexus's ClubLog key, which it never needed. Rarer DXpeditions now rank higher on the DXpedition board and in the chase feed: ClubLog's Most Wanted list was never being read, and Nexus asked for it again each time it updated.

  • Band Activity and Rx Frequency use far less of the computer on a busy band.

    Several times a second, every time the radio's state came in, both decode lists filtered and sorted their whole history (up to 3,000 lines) and redrew every line on screen, twice, even when nothing new had been heard and even while the FT screen was not showing. They now do that work only when a decode arrives or something about a line changes (worked, a need, a filter you pick), and redraw only the lines that changed. What the lists show, their order, colours, need tags, search count and scrolling, and what a click or double-click does, are unchanged.

  • Less CPU while nothing is happening on screen.

    A red alarm in the top bar, such as RADIO ENGINE STOPPED when the sound card is missing, now pulses three times and then stays lit at full strength. It used to pulse for as long as it was up, which made Nexus redraw the window 60 times a second on an otherwise still screen. A different alarm pulses again when it appears. The panes on the operating screens no longer blur what is behind them, a blur you could hardly see behind their solid background that cost work on every frame the waterfall or scope drew. The panels floating over the map and the globe keep their blur.

  • The waterfall no longer keeps the computer busy when it is not on screen.

    A waterfall kept ready on a screen you have left (the FT8 screen while you are on Phone, CW or the Logbook, or a waterfall popped out into its own window) still woke 60 times a second to do nothing, so with the FT8 screen's waterfalls that was up to 180 times a second on every screen, and the computer never got to rest. A waterfall not on screen now does nothing until you come back to it, and it picks up where it left off, as before. A paused waterfall also redrew its PAUSED label and time scale 60 times a second; it now redraws them only when a time on them changes, or at once when you scroll back. After you resume, the PAUSED label and time scale no longer linger on the live waterfall. Otherwise what the waterfall shows is unchanged.

  • The Phone and CW scope does less work between sweeps.

    The dial line, the filter passband, the notch and the frequency scale over the scope are now drawn again only when one of them moves, rather than about 30 times a second, and with "Row per sweep" the picture is no longer redrawn while nothing in it has changed. The scope looks and behaves as before.

  • Band Activity keeps its rows in a short pane.

    In the FT8 and FT4 Roster layout on a 1024×768 screen, the filter chips wrapped onto three lines and left Band Activity's list one decode row, and with Recent contacts ticked the "heard" count at the foot of the pane was cut off. When the pane is too short for both, the chips now fold behind one Filters button in the title row, beside the search box: click it (or Tab to it and press Enter) to open them, and click it again to fold them away. It is lit and shows a count while a filter is on, so a filtered list never looks like a quiet band. The list shows about four rows there, or three and a half with Recent contacts. Taller panes, Classic at 1024×768 and every layout at 1920×1080 keep their chips on screen as before.

  • Receive decoders stop when you leave their screen.

    Opening RTTY, PSK, JTTY, SSTV or APRS starts that mode's decoder, so the screen is never dead, but nothing stopped it again: look at each one and go back to FT8, and all five kept listening for the rest of the session, a noticeable share of the CPU on a slower computer. Now a decoder that started because you opened its screen stops when you go to another operating screen (FT8, Tempo, JS8, Phone, CW or another of the five), and starts again when you come back, with its text, received stations and saved pictures still there. An SSTV picture still coming in when you go to another operating screen is lost, as it would be if you pressed Stop. A look at the map, the logbook, Settings, Satellites or the contest screen stops nothing. A decoder you started yourself with its Arm or Monitor button keeps running wherever you go, as does SSTV armed for an ISS pass or with a manual receive start, SSTV while the picture viewer window is open, RTTY once you have turned Auto on, and APRS while the receive-only iGate is on. Turning RTTY Auto on also starts the RTTY decoder if it is off, unless you stopped it yourself. While another computer is connected and driving the station (watch-only included), or a Remote session has control, leaving a screen stops nothing; the decoders whose screens were left meanwhile stop about 10 to 15 seconds after it disconnects (10 seconds after Nexus notices it has gone, which for a computer that drops off without signing out can take up to 5 seconds; a shorter dropout stops nothing), and opening a screen again starts its decoder again. The RTTY tuning scope also does no work while nothing is showing it. Nothing about transmitting changes.

  • The operating screens you are not looking at no longer redraw a few times a second.

    Nexus keeps the FT8, RTTY, PSK, JTTY, SSTV, APRS and JS8 screens running behind the one you are on, so their decodes keep collecting while you are away. Each of them used to redraw itself in full every time Nexus checked the radio, about three times a second, even with nothing new to show. Now a screen you are not looking at redraws only when something it shows has changed, such as a new decode, a new station or a new frequency. What you see and when it updates is the same as before.

  • A Flex or TCI connection that fails inside Nexus no longer keeps a processor core busy.

    If the Flex native client's or a TCI connection's link to the radio stopped because of a fault inside Nexus, rather than closing normally, a part of Nexus kept checking it in a tight loop, using a whole processor core, while Nexus went on showing the radio as connected. Nexus now sends the unkey that a normal disconnect sends, shows the radio as disconnected, and, if a transmission was on the air, says the radio may still be transmitting, as it does when the link to the radio is lost mid-over. An idle Flex or TCI connection also uses a little less processor time: it no longer copies everything the radio has reported fifty times a second when nothing has changed.

  • Closing Nexus right after logging waits for the log.adi copy.

    If you quit in the moment between saving a contact and its line reaching the log.adi copy, the quit could report that copy as having nothing left to save and close before writing it. The contact itself was always saved in the logbook; now the quit waits for the copy too.

  • The diagnostic log no longer reports a transmit underrun at the end of every over.

    Each FT8 or FT4 over added one to tx_underruns in the "sound-card health" line, written just before the unkey, with nothing wrong: the last piece of every over runs out part-way through a sound-card buffer, and that was counted as a gap. Soundcard CW did the same at every word space. Nothing was ever cut or delayed on the air, and what is transmitted is unchanged. The end of a slot over (FT8, FT4, JS8, WSPR and the other slot modes), a CW word, a PSK or RTTY message, a voice-keyer message, an APRS frame or an SSTV picture is no longer counted. Audio that runs out anywhere else still counts at once, however long the gap, including a Tune carrier the program stopped feeding. The end of a remote-microphone, TCI-audio or JTTY transmission still adds one, as before.

  • The Spot dialog keeps the keyboard while it is open.

    In the Spot to the DX cluster dialog (from the Logbook's ⋯ menu, the CW and Phone log, Operate's and Phone's 📢 button, and the Spots board), Tab and Shift+Tab now go round the dialog's own boxes and buttons instead of walking out onto the screen behind it, and Escape closes it from any of them, Cancel and Spot included (it used to work only in the three boxes). When it closes, the keyboard goes back to the button that opened it, or, if that contact has scrolled out of the Logbook meanwhile, to the list it was in, never to nowhere. The dialog looks and sits as before; the screen behind it is now dimmed like every other dialog's, without the slight blur.

  • Clicking a call shown as <...> explains itself in your language.

    When a station's call shows as <...> (Nexus has not yet heard it in full) and you click it, Nexus refuses, as WSJT-X does, and says to click the station's next message once its call shows. That explanation was always in English, after the start of the message in your own language. It now appears in German, Spanish, French, Japanese or Russian too.

  • WSPR and FST4W beacons from a compound call go out as WSJT-X sends them.

    A beacon from a call such as PJ4/K1ABC or K1ABC/P went out as "PJ4/K1ABC FK52 37" on every transmission, which WSPR receivers read as power 0 and FST4W could not send at all. Nexus now alternates WSJT-X's two messages, one per transmission, starting with the call and power ("PJ4/K1ABC 37") and then the call with your full grid ("<PJ4/K1ABC> FK52UD 37"). As in WSJT-X, the message alternates on every transmission Nexus starts, even one the radio then refuses to key; an interval where Nexus starts no transmission (no power set, or TX off) does not skip one. Pressing Tune in WSPR also moves it on to the other message once, as WSJT-X's Tune does; a Tune in FST4W does not, as in WSJT-X. WSPR with a 4-character grid sends the call and power every time, since its other message cannot carry a 4-character grid. A plain call beacons as before.

  • An SSTV picture you delete stays deleted, and received pictures keep their details.

    If a picture finished decoding, or its callsign ID arrived, at the moment you deleted another picture, the gallery's saved list could still name the deleted picture, and in rare cases the list could be damaged so that every picture lost its frequency, line count and callsign ID at the next start. The two now take turns, so the saved list always matches the gallery. Two copies of Nexus sharing one gallery folder also no longer damage each other's list. A symbolic link placed in the gallery folder is no longer shown as a picture, and deleting through one is refused. Saving a received picture or deleting pictures no longer holds up the radio when the Pictures folder is slow, such as a network or OneDrive folder, so Stop TX and the end of an over are never kept waiting on the gallery. A picture that finished decoding while you were deleting another one no longer shows up twice in the gallery.

  • Deleting an SSTV picture can only delete a picture.

    A delete request for the SSTV gallery, one picture or several, was checked only for being inside the gallery folder, so a request naming another file there (from a Remote client, for example) would delete it: a file of your own kept in that folder, or the gallery's own list of pictures. Nexus now deletes only pictures the gallery shows and refuses anything else with "That file is not a picture in the SSTV gallery". Deleting pictures works as before.

  • The VFO A/B indicator asks each radio for itself.

    After you changed the radio model in Settings ▸ Radio without restarting Nexus, and the new radio answered at once, Nexus kept the last radio's answer to whether it can report its own VFO. A new radio that cannot report it then had Hamlib's copy of Nexus's last A/B command shown as the radio's own selection, and a new radio that can was never asked, so an A/B press on its front panel did not show. Nexus now asks the new radio.

  • Received audio is no longer lost while Nexus reconnects to a radio that is not answering CAT.

    With the radio's CAT silent (switched off, a cable out, CAT turned off in its menu), Nexus retries the connection every minute or so, and each retry held up the receive path for several seconds. On a sound card at 44.1 kHz that threw away the newest two seconds of the band every time, so a decode was torn on each retry and the "sound-card health" line in the diagnostic log climbed by tens of thousands of dropped samples. Nexus now keeps up to about 20 seconds of received audio while it waits, so a retry no longer throws any away. A retry that runs across the end of an FT8 or FT4 period can still cost that one period's decode.

  • A wrong LoTW password is called one.

    When a LoTW confirmation download (Fetch LoTW, Download confirmations) was sent with a LoTW username or password that LoTW turns down, Nexus said your username and password were fine and blamed LoTW. LoTW answers a wrong login with its sign-in page rather than an error, and Nexus took that page for any other page that isn't a report. It now says LoTW did not accept your username or password, so you know to check them in Settings. Any other page LoTW sends in place of the report is still reported as one, without claiming anything about your login.

  • A network radio that is switched off or behind a firewall no longer freezes Nexus.

    When the rigctld address in Settings ▸ Radio points at another computer that does not answer at all, Nexus waited for the operating system to give up on the connection, about two minutes on Linux and twenty seconds on Windows, and the radio, the slot clock and the transmit controls stalled meanwhile, again on every retry. It now gives up after three seconds, shows the radio as down with "rigctld at … is not reachable", and keeps trying in the background as it does for any radio that stops answering. A rigctld on your own computer connects exactly as before.

  • A cluster spot on an FT4 dial now says FT4.

    A spot on 80 m 3.575, 17 m 18.104 or 12 m 24.919 was labelled FT8, and on 6 m 50.318 and 50.323 the two labels were swapped, so the Needed board and Pounce named the wrong mode and working the spot opened the FT screen on FT8 at an FT4 dial. Each spot now takes the mode of the dial it is on.

  • eQSL download says what eQSL actually said.

    When eQSL did not hand over its download, Download confirmations and Check confirmations told you to check your eQSL username and password, whatever the reason. Now a login eQSL refuses says "eQSL did not accept your username or password", and any other page (an error page, maintenance) says so without blaming your login. A download with nothing new since the last one, which eQSL answers with "You have no log entries", is no longer an error at all: it ends like any download that found nothing, with no new confirmations, and the next download carries on from it. Check confirmations on an eQSL account with no cards reads as an empty InBox the same way. Needs a check against a real eQSL account.

  • A network radio whose computer is switched off reads as not answering, not as refusing.

    When the rigctld address in Settings ▸ Radio points at another computer that is switched off or has left the network, the connection fails with "No route to host" (or "Host is down", or "Network is unreachable"), and Nexus took that as the radio saying no: it kept sending frequency and mode changes to it and could give up on the band as if the radio did not cover it. It is now treated like any radio that stops answering: changes wait until the radio is back, and the messages say the radio's CAT link can't be reached instead of "no reply from the rig" or "rig rejected". The split and RF power notes say "CAT link unreachable" too, as they now do whenever Nexus has no connection to the radio at all. A radio that really refuses a setting is still reported as refusing it.

  • Mac CAT on macOS 12 and 13.

    The rig-control programs inside the Mac app (Hamlib's rigctld and the others) were built for macOS 14, so on an Apple Silicon Mac running macOS 12 or 13 macOS refused to start them, and CAT only worked where Homebrew's Hamlib was installed. Everything inside the app is now built for macOS 12, the oldest macOS Nexus supports, and the app itself says so to macOS (it said 10.13). Needs a check on a Mac running macOS 12 or 13.

  • Update from callbook runs when other programs use your QRZ account.

    Logbook ▸ Clean up ▸ Update from callbook counted QRZ's own tally of your account's lookups against its daily limit. That tally includes every program that uses your QRZ login, so a station whose other software looks calls up all day found every run stopping before its first lookup, with nothing changed, and no setting high enough to get past it. The daily limit now counts only the lookups Nexus makes. QRZ's count is still shown, beside today's lookups, as information. If QRZ does turn lookups down, the update still stops and says so.

  • Running the setup again from Settings keeps your licence class.

    Re-running the setup wizard (Settings ▸ Radio ▸ Re-run setup wizard…, or the link on Settings ▸ App) always showed "Outside the US" on its licence step, and pressing Finish without touching that step saved it: a Technician, General, Advanced, Extra or Novice station lost its class, and with it the transmit lockout that keeps you inside your licence privileges, without a word. The licence step now opens on the class you have saved, and setup saves a class only when you pick a different one. A class you pick is shown the next time you run setup, without restarting Nexus. If you re-ran setup on an earlier version, check Settings ▸ Station ▸ License Class.

  • Callbook answers and the IOTA island list are read only up to a sane size.

    A callbook answer (QRZ, HamQTH, callook.info) larger than 256 KB, a QRZ Logbook download (Logbook sync, Check confirmations) larger than 256 MB, or an IOTA island list download larger than 2 MB, is refused with a message instead of being read to the end. The IOTA download no longer follows a redirect away from IOTA's address, and an island group name longer than 200 characters is shortened. Normal answers and the real list read as before.

  • The Spots window works at its smallest size again.

    Since each row got its Spot button, the board's columns no longer fit a Spots window dragged down to its minimum: the list scrolled sideways and the last column was cut off. The columns fit again, the spotter and comment giving up the room. On a busy day, when the filter bar has more spotted-from countries than a short window can hold, the list now keeps its first rows and the filter bar scrolls, where before the list was squeezed away and the last filter chips were out of reach.

  • A status message no longer pushes the top bar onto a second line.

    When a message such as "PTT OFF FAILED — TX STOPPED" came up in the top bar, the bar could wrap: on the FT8 screen Auto / Tx 1st / Tx 2nd dropped to a second row and the whole screen moved down a row until the message went. The message now takes only the room the bar has, shortened with "…" when it must be (its tooltip still has the full text), and the bar keeps its height.

  • No empty band in the cockpit header.

    Where a CW, Phone or digital screen's controls wrap (a 1080p screen above 100 %, or a narrower window), ⊞ Panels and CAT took a line of their own under the last row of controls, leaving a band of empty header above the waterfall. They now sit at the right end of that last row when it has room.

  • The main window opens on the screen, borders and all.

    A window reopened un-maximised could run its borders off both sides of the screen and its bottom edge under the taskbar: the saved size was fitted to the screen without its title bar and borders, and a window whose corner was still on the screen kept its place even when its size carried it past the right or bottom edge. The whole window, title bar and borders included, is now fitted to the screen it opens on and pulled back from the right and bottom edges. Needs a check on Windows.

  • IOTA counts LoTW confirmations.

    The IOTA programme has credited contacts matched through LoTW since May 2020, as well as QSL cards, but the Awards IOTA card gave its ✓ only for 100 island groups confirmed on paper. The ✓ now follows the groups confirmed by LoTW or card, the count it already showed, and the note says so; how many are on cards is still shown beside it. eQSL confirmations still do not count.

  • "(custom)" on the band pill says what it means

    (#334, #184). The pill reads "80m (custom)" when the dial is not on a listed channel for the mode, and that is all it means; its tooltip and the top of its menu now say so: pick a channel there, or type a frequency.

  • Signposts that pointed nowhere.

    "Not spotted: no DX cluster is connected" named a "Spot Sources" page; it now names Settings ▸ Services ▸ Integrations & Feeds ▸ Spot Sources, where the cluster is set. The Spots board's license filter named "Settings ▸ license"; it now names Settings ▸ Station. The APRS beacon box now says the map centres on your grid, with a link to Settings ▸ Station, and that Lat and Lon set only where your beacon says you are (#330). A search in Settings for meters, SWR or ALC now says the TX meters are a panel: ⊞ Panels ▸ TX Meters on the FT8, Phone and CW screens (#343).

  • FT8/FT4: a contact a band change set aside logs WSJT-X's start time.

    A contact you changed band in the middle of, kept for Log QSO, was logged as starting at the moment of the band change when Tx2 and Tx3 had never been selected in it (for example after clicking a station's R-report). It now logs the start time WSJT-X estimates for such a contact, a period or two before, as the same contact logged before the band change already did.

  • FT8/FT4: a 73 sent after clicking a station's RR73 or RRR is offered for the log.

    If you clicked a station just after it sent you RR73 or RRR, Nexus sent your 73 but never logged the contact, because no signal report was exchanged in it. Now, as WSJT-X opens its Log QSO dialog, the contact goes to the confirm-before-log popup for you to log or discard: with auto-log or "Prompt before logging" on it appears by itself, and with both off Log QSO opens it. It is never logged without your confirm: not when the popup's queue is full, where it keeps waiting while the oldest other contact is logged, and not after a restart, which brings it back to the popup still waiting for your confirm. As in WSJT-X, it is offered on your first 73 only: when your last transmission, to any station on any band, was already a 73 or RR73, the click sends 73 again without offering the contact, so clicking a repeated RR73 does not offer it a second time. Changing band right after your 73 no longer loses it: it is offered then, on the band it was made on (or kept for Log QSO with auto-log and "Prompt before logging" both off). A logger program's Reply on the station's RR73 or RRR counts as a click, as in WSJT-X, so its 73 is offered the same way. Clicking (or a logger's Reply on) the RR73 of a contact whose 73 is already waiting to go out is that same 73, as in WSJT-X: the contact is logged once, with its reports, and no copy without them is offered. The popup never holds a second such offer for a station, band and mode already waiting in it, so a logger replying by itself cannot fill it; a finished contact with signal reports is always kept, whatever is waiting, and if you press Log QSO on an offer that is already waiting, the message says so instead of "Logged QSO". What is transmitted, and when, is unchanged.

  • FT8/FT4: a click no longer loses a contact waiting for Log QSO.

    With auto-log and "Prompt before logging" both off, a finished contact waits for you to press Log QSO. Clicking a station before you did, even the same station's repeated RR73, replaced it, and Log QSO then had nothing to log, or only a copy without the signal reports. The contact is now kept, and Log QSO logs it with its reports while the new contact has nothing to log yet; the Logged QSO message names the station logged. With auto-log or "Prompt before logging" on, a click right after your RR73 (on a caller decoded the period before, say) no longer loses the contact you just finished: it is logged, or waits in the popup, as the next period would have done. Pressing Call CQ, changing mode (FT8 to FT4 included) or clicking a CQ in a running contest in that period does the same, and with both settings off the contact is kept for Log QSO. With both off, switching FT8 to FT4 leaves the finished contact on screen as before, and a CQ run waiting for Log QSO keeps waiting, calling no CQ until you press it; Log QSO then logs the contact as FT8, on the frequency it was made on (it was logged as FT4). Each contact kept for Log QSO stays until Log QSO takes it, however many there are and whatever else is logged meanwhile, and nothing opens by itself. A count beside Log says how many wait ("2 waiting for Log"), in the main window, a popped-out Operate window and on a computer driving the station, and Log stays usable while any wait, even with no station on screen, such as after Call CQ. They are kept across a restart, and come back waiting for Log QSO. A finished contact a band change ends joins them too, and so does one a mode change ends (leaving Q65 or another mode with long overs while TX is on): counted, kept across a restart and logged on the band and in the mode it was made in. Log QSO pressed on a contact that is already logged takes only one of these: an unfinished contact a band change set aside waits, as before, for a press before the next contact's reports, and is no longer logged by that press; nor does Log QSO in a contest reach it. With no station on screen a press takes a contact the count names before such an unfinished one. The count is announced to a screen reader when it appears, Log's spoken name and its tooltip include it, and it reads at full contrast in the light theme too. Clicking the same station's R-report again right after your RR73 sends RR73 again, as before, and the contact is logged once, as in WSJT-X, however long it ran.

  • Updating from inside Nexus works.

    On the stable channel it never had: since 0.18.0, whenever Nexus found a new version, the download failed at once with "Cannot read properties of undefined (reading 'rid')", so the update never arrived and only the download page could install it. The beta channel was not affected. The download now runs, and Install restarts into the new version. An update waiting for Install is downloaded once, not again at every hourly check, and stays on offer when a later check fails (offline, say). Versions before this one still fail the same way: install this one from the download page once, and later updates come through Nexus.

  • A JTTY contact in the contest logger window is logged as JTTY.

    With the JTTY screen as the station's mode, the contest logger window logged its contacts as PSK31, and the contest screen, the contest setup's mode warning, the station card's new-mode check and the Spots board's Spot me named the mode PSK31 too. Each now names JTTY (and QPSK31 when the PSK screen is on QPSK31), and a POTA self-spot from JTTY is refused as a mode pota.app does not list instead of going out as PSK31. Contacts already logged keep the mode they were logged with.

  • A JTTY contact is logged on JTTY's frequency.

    A contact logged from the JTTY screen's log strip was logged on the dial alone, and one logged in a contest (the contest logger window, the log strip in a contest) at the dial plus PSK's audio centre. Both are now the dial plus JTTY's Tx frequency, as WSJT-X logs a JTTY contact; a PSK contact is logged as before. This holds on a computer driving the station as well. Contacts already logged keep the frequency they were logged with.

  • ARRL VHF Cabrillo LOCATION is your ARRL section.

    The January, June and September VHF contests headed the Cabrillo file with your state; LOCATION is now your ARRL section, which ARRL's log upload asks for, and the contest setup asks for it. A blank section still writes the state.

  • A PSK contact is logged on the frequency it was worked on.

    A PSK31 or QPSK31 contact logged from the PSK screen's log strip outside a contest was logged on the dial alone; it is now the dial plus the audio centre the decoder is netted on (where PSK transmits), as a contest contact already was, so 14.070 netted on 1200 Hz logs 14.0712. A contact drafted from a computer driving the station now keeps that same frequency (and JTTY's, the dial plus its Tx) instead of the bare dial. Contacts already logged keep the frequency they were logged with.

  • A computer driving a station on JTTY shows the JTTY screen.

    A computer connected to a station in JTTY followed it to the PSK screen; it now lands on JTTY, and on PSK when the station is on PSK. It still sends nothing to the station by following.

  • A band change ends the contact you were working.

    In FT8, FT4, FT2, Q65, MSK144, FST4, JT65, FT1 and DX1, a band change already stopped TX and turned Enable Tx off, but the contact stayed loaded: the standard messages and the TX line still carried the old station, and TX On called it on the new band. Now the contact or CQ run ends and the DX call and grid clear (WSJT-X stops TX the same way; it keeps the DX call, Nexus clears it). In the Contest screen's FT modes the contest contact ends the same way, and Running or S&P stays as it was. A contact already complete when the band changes (your RR73 or 73 sent) is logged at once with auto-log on, on the band and frequency it was made on. Any other contact whose reports had crossed stays loggable with Log QSO, with the band, frequency, mode and end time it had at the band change, until it is logged or a later band change sets another contact aside. Logging other contacts meanwhile does not drop it, and neither does a CQ run giving up on a caller who went quiet: Log QSO then logs one contact per press, the caller the run gave up on first. A contact made before a listen off the bands (WWV) logs on the band and frequency it was made on when the band then changes. Logging a set-aside contact while a CQ run is calling no longer stops the next caller's contact from being logged. A band change also clears the JS8 To box and the JTTY Call box, and sets aside an unlogged entry in the Phone, CW, RTTY, PSK, JS8 and JTTY log line, every box of it, including one with a name, report or comment typed but no call yet. A note offers Restore, which puts the whole entry back to log on the band, frequency and mode it was made on, also after a listen off the bands before the band change (a contest entry once the radio is back on that band), until you restore it, dismiss the note, log that same call on that band, or a later band change sets another entry aside; logging other calls meanwhile does not drop it. This applies to every way the band changes: the band buttons, a frequency typed or clicked on another band, the radio's own dial, band hopping, a Remote band pick and switching to a radio on another band. A move within the band, and a listen off the bands (WWV) and back, change nothing.

  • Clicking a station always starts the FT8/FT4 contact fresh, as WSJT-X does.

    A contact that stopped part-way could be picked up again hours later by a click on the same station (in the Call Roster, a station card, the Needed board or an alert): Nexus resent the old reply with the old signal report, and the contact logged with the start time of the first attempt, hours earlier. Now every click works the station from what it sent last: a new CQ gets your call with your grid (Tx1), or your report (Tx2) with Skip Tx1 on; a report to you gets your roger with the report you hear now; an RR73 or RRR to you gets your 73. Clear DX (F4 or the Clear button), or changing the call in the DX Call box, ends the contact as in WSJT-X: TX goes off and nothing more is sent to that station. In a CQ run a cleared caller is not answered again if it keeps calling, even after you clear the next caller too, until you turn TX back on, call CQ or click it; a contact with it kept for the Log button goes too. A call you type is then worked fresh. The start time also follows WSJT-X: a click, Clear DX or Tx1 clears it, and it is set when your report or roger (Tx2 or Tx3) is chosen; a contact that never had either (a click on an R-report) logs WSJT-X's estimate, two or three periods before your RR73 or 73. Turning TX on, or pressing a Tx button, still carries on with the contact as it stands, as in WSJT-X.

  • A contact's power is the power the radio measured.

    1.21.0 stamped every new contact with Settings ▸ Station power, whatever the radio was doing. A contact now carries the forward power the radio reported during its own over (the over's peak, in watts), on every path that reads the radio's power in watts: Hamlib, TCI, a Flex on its own connection and Nexus's built-in Icom connection. Settings ▸ Station power is used only when the radio reports nothing; with an amplifier set up on the radio, whose meter then reads the amplifier's drive; and during a tune. No new reading is asked of the radio, and contacts already in the log are not changed.

  • Icom forward power in watts on each radio's own scale.

    On the built-in Icom connection the power meter used one 100 W scale for every radio, so an IC-705 or IC-905 read about ten times its output and an IC-9700 read wrong on 70 cm and 23 cm. The meter now reads Icom's percentage of the radio's rated output on the band it transmits on (IC-705 10 W; IC-905 10 W up to 23 cm and 2 W on 2.4 and 5.6 GHz; IC-9700 100, 75 and 10 W; IC-7300 100 W, 50 W on 4 m; IC-7610 100 W; IC-7760 200 W). Where Icom gives no figure (the IC-905 on 10 GHz, the IC-9700 transmitting on the other band of a satellite split) the power row is hidden rather than wrong.

  • The shared comment has room.

    In the log strip and the Logbook's edit form, Comments (sharable) now has a full-width box of its own directly above Notes, and long text wraps over several lines instead of scrolling sideways in a narrow box. It is still one line in the log, as ADIF's COMMENT is: a line break you type becomes a space, and Enter still logs. Editing any other field of a contact whose stored comment holds line breaks (an imported one) no longer runs its words together.

  • The station card uses its width and shows the callbook photo.

    The card's text was squeezed to about the width of the callsign; it now uses the card's width, with Look up and QRZ on a line of their own. A card opened after the log strip had looked the call up showed initials and a Look up button; it now shows that answer, photo included, straight away. When the callbook sent no photo, hovering the circle says why (QRZ sends photos only to XML subscribers, and only when the station posted one).

  • Editing a contact keeps its stored country.

    Saving a correction from the Logbook's edit form rewrote the contact's country in Nexus's own spelling on every save, so a country imported or downloaded as "Germany" became "Fed. Rep. of Germany" when you only fixed the name. The country is now written back exactly as it was, unless it names a different DXCC entity than the DXCC number the awards count the contact under (LoTW's, where LoTW confirmed it as another entity, else the contact's own): then it becomes the name of that number's entity (a K1N contact with DXCC 182 stored as "United States" becomes "Navassa Island", and one LoTW confirmed as Navassa keeps "Navassa Island"). A country that already names that entity, in any capitals, is never respelled. A contact with no country gets the name of that same entity, so a later edit does not change it again. A LoTW DXCC number that names no entity is passed over for the contact's own, as the awards do. The same rules apply to an edit made from a Remote browser. Correcting a contact's callsign used to leave it with no country at all; it now takes the new call's country, from the edit form, the contact card and Remote alike, and never keeps the old call's.

  • Correcting a contact's callsign clears what belonged to the old call.

    A contact LoTW had confirmed as another DXCC entity kept counting toward that entity in Awards, the Needed board and the statistics after its call was corrected, with nothing on screen to say why (a K1N contact confirmed as Navassa and corrected to K1NX still counted as Navassa). A callsign correction now also takes off the old call's LoTW DXCC, CQ and ITU zones, continent and county, the paper card's, LoTW's and eQSL's received dates, and eQSL's Authenticity-Guaranteed mark, beside the confirmations it already took off; the contact goes out again to the logbook services under the new call as before. A value you change in the same save is kept, a State included. Undo puts all of it back. On a contact already uploaded, Undo decides service by service: each logging service the new call has not been sent to gets its upload mark back (LoTW's with its sent mark) and its waiting re-send is cancelled; each that already has the new call, or may have it, leaves the contact marked not uploaded there, and the Undo message names those services. Nothing queued under the new call is sent after the Undo. After a restart Nexus no longer knows what was sent, so Undo puts the contact back marked not uploaded everywhere and says so. On the contact card, saving a corrected call first says which of the values shown it will take off (the zones, continent and county, the paper card marked received and its date, and the LoTW, eQSL and QRZ confirmations), and saves only when you confirm. An ordinary edit (a name, a grid, a band) keeps every confirmation and its date. A callsign corrected on the contact card together with changes under More fields now goes to QRZ, Club Log, eQSL and the other services with those changes, not without them.

  • An Icom keeps its filter when Nexus retunes it (#417).

    On the built-in Icom connection every frequency change re-sent the mode without a filter number, and the radio answers that by selecting the mode's default filter, so an IC-7300 set to FIL1 jumped to FIL2 on each QSY. Nexus now reads the radio's mode first and sends nothing when it is already right, the way WSJT-X does, so FIL1, FIL2 or FIL3 stays as you set it. A real mode change is sent as before. Applies to the IC-7300, IC-7300MK2, IC-705, IC-905, IC-9700 and IC-7610; the IC-7760 is unchanged for now. Needs a check on the radios.

  • Digital modes leave your radio's filter alone (#82, #107).

    Going into FT8, FT4 or any other data mode, Nexus asked the radio for a 3 kHz filter, read back what it took and asked again, and said so on the CAT line when the radio kept its own. Some radios answered with a wider default (a Flex over SmartSDR CAT landed on 6 kHz), and an operator who wanted a 4 kHz DATA filter was held to 3. Nexus no longer sets a filter width on digital modes, the way WSJT-X does: the radio's own DATA filter is used, whatever you set there, including a narrow one the radio remembers for a band. The filter-width control in the CW and Phone screens still sets a width when you use it. Needs a check on the radios.

  • APRS can use DATA-FM on an FT-991/991A (#393).

    Entering APRS always put the radio in plain FM, so an operator whose radio sends USB sound-card audio only in DATA-FM had it put back to FM each time. A new per-radio switch, Settings ▸ Radio ▸ Rig & CAT ▸ Advanced ▸ "APRS uses DATA-FM", shown for the FT-991 and FT-991A, makes APRS select DATA-FM on that radio. It is off by default, so an APRS setup that works in plain FM today is unchanged. Turn it on only with the radio's menu 077 FM PKT PORT SELECT set to USB, or beacons go out silent. Repeater voice on the same radio stays plain FM, and with "Data modes use plain SSB" on, APRS stays plain FM. Needs a check on the radio.

  • Stop TX on an FT-710 or FTX-1 unkeys at once.

    Every Stop TX sends these radios Yaesu's CW playback stop, and Hamlib held the radio for its whole reply timeout after each one, because a Yaesu never answers a setting. The unkey waited behind them and reached the radio about 2 seconds late on an FT-710 and about 4 seconds late on an FTX-1. Changing the FT-710's scope span or mode held the next command, a key included, for about a second the same way. Nexus now sends these commands in a form Hamlib does not wait on, and the unkey reaches the radio within a few tens of milliseconds.

  • The Tx panel's dot follows the FT8/FT4 contact.

    In Classic, the dot beside the next message sometimes stayed put while the contact went on normally: when the report Nexus sends (from the decode it answered) differed from the station's latest one in the list, when Tx5 was edited, or when another decode was clicked mid-contact. The dot now follows the message actually going out.

  • A contest contact reaches your logger complete over the WSJT-X UDP link.

    A contact logged in any contest (ARRL Field Day, the Digi and VHF contests, FT Roundup, the QSO parties) went out as a bare QSO Logged message: no grid, no reports, no power, the dial without your audio offset, the same time as start and end, your station call as the operator, and outside Field Day a blank received exchange. No Logged ADIF message was sent at all, so a logger reading only that message got nothing. Each contest contact now goes out as WSJT-X sends it: QSO Logged with the grid (from the exchange, or the CQ the station was heard with), the reports the exchange carried, your power, the frequency on the air (dial plus audio offset), when the contact began and when it was logged, the operator at the key, your grid (the park's while you activate one) and both exchanges as they went over the air; then Logged ADIF with the full record.

  • WSJT-X UDP QSO Logged carries the satellite, satellite mode and Rx frequency, and your park's grid.

    WSJT-X sends three more fields after the propagation mode (satellite name, satellite mode and Rx frequency); Nexus stopped at the propagation mode, so a logger reading only that message lost a satellite contact's satellite and a split contact's receive frequency. They are now sent, in WSJT-X's order. During a park activation the message's own grid is now the park's, as the contact's ADIF already was, instead of Settings' grid.

  • ALL.TXT reads as the conversation happened.

    Your own FT8/FT4 transmission was written to ALL.TXT when Nexus prepared it, a moment before it went out, so the reply you were answering (decoded just after) landed below your answer with an earlier time. Your Tx line is now filed once its transmission has ended, in time order, so the file reads top to bottom as WSJT-X's does: their message, then yours. The Tx line itself is unchanged (stamped with the period it went out in), and nothing about when or what Nexus transmits has changed. Q65 lines now name the period and submode as WSJT-X writes them (Q65-30A) instead of plain Q65. Quitting now writes ALL.TXT's last lines, your own Tx line among them even when the quit cuts its over short; they used to be lost when Nexus closed before writing them.

  • Play Receive Audio Here works with Audio over TCI and with native Flex audio (#372).

    With a radio on a TCI connection and Audio over TCI on, or a Flex with native audio on, Settings ▸ Radio ▸ Play receive audio here played the computer's idle sound card instead of the radio, so it was silent while the decoders heard the band. It now plays the audio arriving from the radio over the network, the same audio the decoders hear, at the RX gain and listening level you set, and the sound card again when that audio is off. About a tenth of a second is buffered so network timing does not come through as clicks, and switching to another radio, however the switch is made, stops the first one's audio in the headphones and plays the new one's. It still goes quiet while your PTT is down and still refuses to open on the radio's TX audio device, and it also refuses SmartSDR's DAX TX device, which would send the band to a Flex's transmitter. Needs a check on a TCI radio and on a Flex.

  • The Repeaters note points at the right place in Settings.

    When a directory looks thin, the note on the Repeaters screen said to add a RepeaterBook API token in "Settings ▸ Integrations", which does not exist. It now says Settings ▸ Services ▸ Confirmations, where the token is, and clicking it opens that page.

  • FT8/FT4: a contact logged with no grid takes the grid the station sent, and a callbook grid no longer counts.

    A station that answered your CQ with a bare report logged with no grid when its CQ had been heard on another band or before a mode change, since the list of stations heard is cleared there. The contact now takes the grid that station sent on the air earlier in the session (its CQ, or a call to anyone with its grid), on any band and in any mode, if it was heard within the last 24 hours (a station heard days ago in a session left running may be portable somewhere else now), and it counts like any logged grid. Otherwise, for a call with no stroke and no park, a callbook lookup you already made fills it, as before, but on such an FT8/FT4 contact now for the distance and the map only, since a callbook gives the station's home QTH, not where it was: that grid does not count toward VUCC or the grid totals, does not clear the GRID pill or mark a grid worked in the decode lists, and is not sent to any logging service, to a logger listening for WSJT-X's logged-QSO message, in N1MM's broadcast, or to DXKeeper or HRD Logbook (both would count it, and could upload it); your own log and exports keep it. Correcting the call in the confirm popup drops a grid Nexus took for the miscopied call, as before, and a callbook grid for the corrected call is then for the distance and the map only too. Contacts already in your log are unchanged: a grid a callbook filled before this version (from a callsign lookup, or Update from callbook) still counts and is still sent exactly as it was, so no total drops, and a contact you log from the log form still takes a callbook grid as it always has. A grid the station sent, one LoTW confirms, or one you type or edit counts as always.

  • Split Operation set to Rig works with the Flex native client (Beta).

    With Nexus's own Flex client, Rig split was never carried out: the radio stayed on the receive frequency, so an FT8, FT4 or FT2 over with the TX offset outside 1500 to 2000 Hz went out 500 Hz or more below where you put it, and nothing said so. Nexus now does it the way SmartSDR does: it opens a second slice where your receive slice is, mutes it, moves the TX flag to it and tunes it to the transmit frequency, and the receive slice stays your dial. While split stays on, every over uses that same slice: a run of overs never opens a slice each. Turning split off moves the TX flag back to your receive slice and closes the extra slice. If the radio cannot make the slice (all its slices are in use) or refuses to move the TX flag, split is not turned on and no slice is left behind. On a radio shared with another station (multiFLEX), Nexus never takes the TX flag from the other station's slice. Nothing about the TX flag or the transmit frequency is changed while the radio is transmitting: if turning split off cannot finish at once, Nexus finishes it as soon as the radio is idle and keys nothing on the extra slice until then (an over, CW, the radio's own Tune or an ATU start), and changing band turns split off before the dial moves, so the transmitter never stays on the old band. If your receive slice is closed in SmartSDR or on another client while split is on, split turns off and the extra slice becomes your dial, its audio back on, so band changes and overs keep working. Needs a check on a Flex.

  • The Logbook's Card received, LoTW received and eQSL received columns show the date.

    Marking a paper card received put the C in the QSL column but left Card received empty, and the LoTW and eQSL columns stayed empty after a sync. A card you mark received is now dated with today's UTC date (ADIF QSLRDATE), unless the contact already has a card date, for example from an imported log, which is kept. Taking the card back takes its date off with it, and Undo in the contact's history puts both back as they were. A LoTW download now brings the date LoTW confirmed each contact (LOTW_QSLRDATE), and an eQSL download the date of each card (EQSL_QSLRDATE), onto contacts that have none; a date a contact already has is never replaced, and a download never changes a paper card's date. All three dates go out in ADIF exports. A contact confirmed before this version gets its LoTW or eQSL date only when a download brings that confirmation again (a regular sync brings only new confirmations), and a card marked before this version stays without a date.

  • Older eQSL cards in your InBox now confirm their contacts.

    eQSL marks a card in your InBox as received, and dates it, only when it stored the day the card was added, which older cards do not have. Nexus counted a card only by that mark, so a download or Check confirmations passed over an older card: its contact stayed unconfirmed, and a card for a contact missing from your log was not reported either. Every card in your InBox now confirms its contact by eQSL. A regular Download confirmations asks eQSL only for cards added since your last download, so it never brings these older cards again: to have the ones already in your InBox counted, open Logbook ▸ Check confirmations, press Check (it reads your whole eQSL InBox), and press Change. The check lists them as contacts that gain the confirmation eQSL holds for them, and Change adds it; nothing is taken off a contact unless you tick its line. Your first eQSL download, and the first one after you change your eQSL username or QTH nickname, read the whole InBox too. A card without a date leaves the eQSL received column empty; Nexus never puts the contact's own date there instead. A listener's report still confirms nothing, and only a card from an Authenticity-Guaranteed member is marked as one.

Every version