hamradiotools

Nexus 1.1.0

Released 10 August 2026. The current release is 1.24.0. Download it · Every version

Fixed

  • Picking an SSTV frequency now puts the rig in the right mode, whatever screen you came from.

    The SSTV view deliberately doesn't touch the radio when you open it (it decodes wherever you're tuned) — but that meant picking a frequency from its band strip left the previous screen's mode policy in charge, so arriving from RTTY could land 20 m SSTV in DATA-L instead of USB. An SSTV pick now claims the Phone section and tunes in one step: you listen (and talk) in plain USB/LSB, and the DATA mode is only commanded while an image is actually transmitting — then it hands the mic back. The ISS 145.800 downlink and the 2 m FM calling channel now correctly command FM too.

  • Double-clicking a station in the Call Roster now moves your RX/TX to where they were heard

    — exactly like a Band Activity double-click. The roster remembers each station's last decode offset; Hold Tx is respected as always (RX follows, TX stays put when held).

  • Sharing your radio with VarAC, WSJT-X or a logger is now solid — and on by default.

    The address under Settings ▸ Radio ▸ Share this radio is now answered by Nexus itself from its live radio state, instantly. Until now it pointed at the underlying rigctld: on a busy serial link its reply could arrive late enough that the other program asked again and then read two answers stuck together — VarAC's "Input string was not in a correct format" — and the link dropped whenever Nexus tested or reconfigured CAT. None of that can happen against the new address: shared programs stay connected through Test CAT and every settings save, and their connections show in the connection log. Long command spellings (\get_freq and friends, which VarAC sends) are now understood too. Shared programs can still key the radio — turn that off in the same block if you want them read-only — and every Nexus transmit safeguard applies to them regardless. Upgrading from 1.0.4 or earlier turns sharing on automatically; if you installed 1.0.5, flip it on once in Settings ▸ Radio.

  • Two identical radios no longer get pointed at the same sound card.

    With two rigs whose USB audio shows up under the same name (two Yaesus both presenting "USB Audio CODEC"), Detect used to suggest the first sound card for both — so radio 2 transmitted into radio 1's audio, silently. Sound cards are now handed out one-per-radio: the worst case is the two swapped (visible, and one question to fix), never shared. Auto-test also skips ports that belong to your other configured radios instead of spending its time probing ports it can never open.

  • The radio-control helpers no longer listen on your whole network.

    The rigctld and rotctld servers Nexus starts accepted connections from any machine on your LAN, while the share UI promised "this computer only". Both now bind localhost, matching the promise — sharing between computers was never a supported path, and rig control is not something to leave open on hotel Wi-Fi.

  • Clicking a spot no longer bounces the rig off a default frequency on the way.

    Working an FT4 spot while on FT8 (or the other way round) used to switch the mode first — briefly commanding the mode's standard calling frequency — and then jump to the spot. On a slow CAT link the two commands could land out of order, intermittently leaving the radio on the wrong frequency or mode after the click. The mode switch and the QSY are now one atomic command: the rig hears only the spot's exact frequency. The same click from the torn-off Needed board also now switches FT8/FT4 to match the spot — it previously left the decoder on the old mode.

  • The hourly QRZ logbook sync no longer wipes out a QSO in progress.

    With auto-sync on, the timer's bookkeeping ran through the same heavyweight path as a full Settings save, which resets the operating state and clears anything queued to transmit — once an hour, mid-QSO, with nothing on screen saying why. The sync now records its progress without touching the operating state.

  • Band changes no longer let the radio pick the wrong mode.

    The big one from the bench: selecting 12 m or 30 m in the CW section landed the rig in DATA-U; picking 20 m back in the FT8 section landed it in CW-U. The radio itself was doing it — modern rigs keep a per-band last-used-mode memory, and when Nexus commands the mode before the dial (required since 1.0.1 to stop the FTdx10's 650 Hz CW-pitch walk), a band-crossing dial write let the rig's own memory override the mode Nexus had just set — and Nexus, believing its command stood, never checked. Now every band-crossing retune reads the rig's real mode back and re-asserts once if the rig overrode it, then re-seats the dial. Your click wins, on every band, in every section — and the pitch-walk fix stays intact. (This was in 1.0.1 through 1.0.5 too: it is the "intermittent wrong mode" some operators reported, deterministic per band.)

  • A Settings save can no longer flip you out of your operating section's mode.

    The operating mode is live section state now, like the dial: a Settings form saved after you switched sections carried the old mode and silently reverted the rig. Every stale-save door is closed at once. Awards — a tester put their 31,000-QSO log next to their official ARRL account, and every difference below is chased to the award's actual rules:

  • 5-Band DXCC now computes the award ARRL actually grants.

    It used to count entities confirmed on all five classic bands at once; ARRL wants 100 on each band, and they don't have to be the same 100. The tile now shows your weakest band — the number that decides the award — and says so. 5-Band WAS follows the same per-band rule.

  • Satellite contacts no longer inflate your band DXCC.

    A QSO tagged as satellite counted toward 2 m DXCC and Mixed, which ARRL restricts to the satellite awards — one bird QSO was enough to make Nexus disagree with your LoTW account by one on 2 m. Satellite entities now stand in their own Satellite DXCC count, shown on the Sat VUCC card, matching LoTW's Satellite column.

  • VUCC is now the real award, not a grand total.

    The old tile divided your all-band grid count (mostly HF FT8 grid exchange) by 100 and called it VUCC. The award is 50 MHz and up, judged per band with per-band thresholds — 100 on 6 m or 2 m, 50 on 222/432 MHz, 25 above. The tile now shows your best real standing ("VUCC ✓ 6m") and keeps the all-band count visible as the grid tracker it is. Terrestrial 70 cm and 1.25 m grids — real VUCC bands that previously counted for nothing — now count.

  • IOTA's checkmark now follows IOTA's own rules.

    The programme credits QSL cards and Club Log matching, never LoTW — so a LoTW-only confirmation still tracks your progress but no longer lights the award. The card shows your card-confirmed count beside it.

  • Clearer sentences where the numbers were right but the words weren't: Honor Roll now says "1 more confirmed needed —…

    Clearer sentences where the numbers were right but the words weren't: Honor Roll now says "1 more confirmed needed — entry at 331" instead of the ambiguous "1 confirmed to Honor Roll"; the Confirmed tile says it counts LoTW-or-card only; the Sat VUCC card now states on screen that it counts PROP_MODE=SAT-tagged contacts — which Nexus now writes automatically when you log on a bird's downlink (see below).

  • What is not changed, on purpose: Nexus computes from your log, so it will honestly differ from an ARRL account that…

    What is not changed, on purpose: Nexus computes from your log, so it will honestly differ from an ARRL account that includes decades of paper-card credits applied at ARRL or QSOs from merged callsigns that aren't in the imported file. Import your full history and your LoTW report to close most of that gap. Installing:

  • The installer no longer looks frozen, and the docs no longer send you to install WebView2 by hand.

    Everything Nexus needs ships inside the installer — WebView2, Hamlib, the DSP stack — and always has. But the 209 MB WebView2 step ran silently (minutes of apparent hang), and an old troubleshooting page still said to go download the runtime yourself, which is exactly what a new operator did (IK6HQL's report). The WebView2 step now shows its progress, both troubleshooting pages say plainly that nothing needs installing by hand, and the release pipeline now proves the WebView2 payload is inside every installer it publishes.

  • The installer speaks Italian

    (and German, French, Spanish, Portuguese, Dutch and Japanese) — it asks your language up front instead of assuming English.

  • Download docs corrected: GitHub Releases listed alongside SourceForge (a generator slip had listed SourceForge twice…

    Download docs corrected: GitHub Releases listed alongside SourceForge (a generator slip had listed SourceForge twice and GitHub never), the documented size matches the real ~250 MB, and the README's file table no longer hardcodes an old version number.

  • Answering a station right after a mode switch could transmit on their own cycle.

    Switching FT8 to FT4 (or back) changes the length of the transmit periods, so everything Nexus knew about who transmits in which half is discarded — correctly. But the Stations list survived the switch, still offering every station from the old mode, and answering one silently kept whatever cycle you were already on. If that was the DX's own cycle, you transmitted every time they did: they could never hear you, and nothing anywhere said so. From the same POTA field report as Field mode. Three things changed. The Stations list now clears on a mode switch, exactly as the decode panes always have — it refills with the first decodes in the new mode. Clicking a leftover decode line that can no longer be placed on the right cycle is refused with a message telling you to wait for the station's next transmission, instead of calling blind — the wait is one decode long. And typing a callsign in by hand still works exactly as before, on whatever cycle you have set, which is what WSJT-X does with a typed call too.

  • You can now see which transmit cycle the app picked for you.

    Double-clicking a station has always set your cycle opposite to theirs — that part worked — but the big Tx 1st / Tx 2nd buttons only light when you lock a cycle by hand, so the automatic flip showed only as tiny text and looked like nothing happened. The side the app is currently on now carries a clear underline, distinct from the lit look of a manual lock, so a working flip finally looks like one. From the same field report as Field mode.

  • Contacts broadcast to N1MM+ or N3FJP now say which radio made them.

    Every packet claimed radio 1, so a two-radio station had its whole session attributed to one rig — and silently, because a logger receiving the feed has no way to tell a wrong radio number from a right one. Nothing looks broken until someone reads the log and the bands make no sense. Both the ordinary log broadcast and the Field Day feed now carry the radio that actually made the contact. A single-radio station is unchanged: still radio 1.

  • Changing band no longer leaves the radio on the wrong sideband.

    Picking 20m while you were on 40m carried LSB across with you: the menu offered 14.230 USB and the rig was commanded to 14.230 LSB. Typing a frequency by hand did the same thing. Transmitting on the wrong sideband is as bad as transmitting on the wrong frequency — nobody hears you, and nothing tells you why. Crossing 10 MHz now follows the usual convention, LSB below and USB above. Moving around within one side still keeps whatever sideband you are on, so if you have deliberately put the rig on USB down on 40m for digital work, an ordinary retune leaves it alone. FM, AM and the data submodes are untouched. Reported by patpell.

  • KG4 callsigns are no longer all credited to Guantanamo Bay.

    Only a KG4 call with exactly two characters after it — KG4AB — is Guantanamo; KG4ABC and KG4A are ordinary United States calls sharing the same block. Every KG4 was being filed as Guantanamo, which is a rare entity, so it showed up as a burst of activity from a place that was not on the air and could put a wrong entity in your log and on the Needed board. The country file names some of these exceptions individually and those are still honoured — a few KG4 holders really are in Alaska or Hawaii — but it can only list the calls it already knows, so the rule is now applied by shape as well. Reported by graafpeter-web.

  • A callsign with a stroke in it resolves in the callbook again.

    Looking up W1AW/1 found nothing, while clicking through to QRZ loaded the record straight away — which made it look like Nexus could not find a call QRZ plainly has. The two are genuinely different: QRZ's website redirects a stroked call to its base, and the data service Nexus queries does not. Nexus now asks for the call exactly as you typed it and, only if no callbook has it, asks again for the base call. That order matters — a portable or club call can hold a record of its own, and looking up the base first would hand you a different operator's details without saying so. Reported by kr4fqg.

  • Your memory channels, watchlist and chase lists were being kept somewhere that does not survive a reinstall.

    They lived in the browser storage inside the app window, along with your profile list, your armed satellite and DXpedition alarms, and your UI scale — which is an accessibility setting, so losing it is not a cosmetic matter. None of it sat beside your settings file, none of it was covered by a backup of that file, and none of it was per-profile. Reinstalling rather than upgrading in place could clear the lot, with no warning and no way back. All of it now lives in ui-state.json, next to your settings, and is carried across automatically the first time you run this version. Nothing is deleted from the old location, so going back a version loses nothing either. Genuinely cosmetic things — a collapsed panel, the tab you were last on — stay where they were, which is what that storage is for.

  • Some rigs were keyed the moment Nexus connected, whatever you set PTT to.

    Reported on a TS-2000 with a Digirig, and it is not specific to either: about fifty radios tell Hamlib they use hardware flow control on the serial port — the Kenwood TS-2000, TS-590 and TS-990S, the Yaesu FTDX10, FT-991 and FT-891 among them. On those, the RTS line belongs to flow control and sits raised for the whole session, and Hamlib refuses to let anything hold it down. On the very common interfaces that key from RTS — Digirig and most homebrew cables — a raised RTS is transmit. So the radio went into transmit as the port opened and stayed there, which is also why changing the PTT setting made no difference: flow control is not keying, so no PTT choice ever touched it. Nexus now turns that flow control off — but only when RTS on that port is genuinely wired to keying: a recognised keying interface (the Digirig class) or a serial PTT/CW line you have configured there. A rig on its own plain USB CAT cable keeps its hardware handshake untouched. Hardware flow control is close to meaningless for the short messages CAT exchanges — WSJT-X drives these same radios without it — and a radio that transmits unattended is the worse outcome by a distance. Reported by vk6mo.

  • The Linux .deb installed cleanly and then would not start.

    The package named Hamlib, WebKit and GTK as its dependencies but never mentioned the Fortran runtime or single-precision FFTW that the modem is linked against — so apt reported success, the menu entry appeared, and the binary then died on libgfortran.so.5: cannot open shared object file the moment you clicked it. Nothing in the install said anything was missing, because as far as dpkg was concerned nothing was. It only ever worked by luck: on a machine that happened to have both libraries already, for some other reason. The package now declares libgfortran5 and libfftw3-single3, so apt installs them with Nexus — and on a system that genuinely cannot supply them, the install refuses up front instead of handing you an application that cannot open. Reported and fixed by Justin Johnson (G0KSC).

  • macOS: RX audio could go permanently silent with no error shown.

    The bundle shipped with no NSMicrophoneUsageDescription in Info.plist, so macOS could never show the audio-input permission prompt at all. Once the OS records a denial for the app's bundle ID, CoreAudio does not surface that as an error — the input stream opens normally and the audio pipeline runs, it just delivers silence, which looked identical to a device-selection or gain problem. Added src-tauri/Info.plist with the usage description so macOS can actually ask; anyone hitting this on an existing install also needs tccutil reset Microphone com.kd9taw.tempo once to clear the stuck denial before the next launch will prompt again. Reported and fixed by g0fqb. Applies to a build from source — macOS is not a shipped platform.

Added

  • Satellite contacts tag themselves now.

    Log a QSO while a transponder is held and your dial is on that bird's downlink, and the record gets PROP_MODE=SAT plus the satellite's LoTW designator (SO-50, not the catalog name) — written as a pair, which is what TQSL demands. Your pass QSOs now count toward Satellite VUCC and the new Satellite DXCC figure in the app, and upload to LoTW as creditable satellite contacts, with nothing to edit by hand. The passband check means an HF contact made while a bird is still held from an earlier pass is never mistagged. One honest exception: ISS contacts stay untagged (there is no LoTW designator Nexus can safely derive from the catalog name) — add the two fields yourself for those.

  • The wizard no longer asks what you want to do — you get everything.

    The goals step (pick DX chasing / POTA / contesting…) is gone: every mode and every section starts ON, because Nexus is one program instead of six and there is nothing to unlock. The final step is now just your license class, the optional starter channels, and the walkthrough offer. The goal profiles still exist in Settings ▸ Appearance for anyone who wants a leaner app.

  • The top bar got its space back, and Light/Dark moved home to Settings.

    On an ultrawide the chips used to pile against the right edge with a dead zone after your callsign; in a snapped window they wrapped to a stray row. The chip cluster (operator, Help, Field) now sits with the mode pills on the left, ahead of the TX clock, so the slack lands at the end of the bar instead of inside it — and the buttons themselves slimmed down. The Light/Dark theme switch is a set-once preference, not an operating control, so it now lives in Settings ▸ Appearance instead of spending top-bar space on every screen.

  • The setup wizard now finds your radio instead of waiting to be told about it.

    Open the rig step and the scan is already running — USB radios and FlexRadios on the network appear without pressing anything. A radio behind a generic USB cable (no model in its USB name) is identified by probing the common rigs at their real baud rates, with a plain elapsed counter while it works. When the probe finds the port but has to guess the model — an FT-991A answers the same probe as an FTdx10 — the wizard says so and asks which radio is this? instead of quietly saving the guess; picking a fixed-rate rig sets its one true baud automatically. Radios configured here now key via CAT instead of the old silent VOX default that made the wizard's own Test CAT report failure. And the step ends with the same live Setup-health strip Settings has — Rig / RX audio / TX, with Prove TX — so you leave the wizard on evidence, not hope.

  • A second radio is one button in the same step.

    "I have a second radio" adds it, probes the remaining ports (your first radio's port is skipped — it's busy being your first radio), and saves the port, speed and its own sound card to the new radio's profile. If both rigs use identical USB sound cards, they are shared out one each and the wizard tells you the one thing to check — and gives you a single swap them button if the wrong rig's meters move. It also finally says out loud: run both radios at once by opening Nexus twice.

  • Field mode — one tap for operating outdoors.

    A POTA activator reported the screen was very difficult to read outdoors even shaded, in either theme, and he is right about "either": daylight on the panel washes out low-contrast text first, and about two-thirds of the text in Nexus is deliberately de-emphasised grey — that, not the type size, is what disappears in the sun. The new Field chip in the top bar turns the de-emphasised text near-solid, pushes the surfaces to full black-and-white (in whichever theme you are in), and steps the whole interface up a size or two to match what your window can hold. Tap it again and you are back exactly where you were — it never touches your theme or your saved scale. Pop-out windows follow it. The tooltip will tell you the Light theme reads best in daylight, but the choice stays yours.

  • You can skip the burned-in callsign for a picture that already shows it.

    A pre-made QSO card carries your call as part of the artwork, and the burn-in was covering it to repeat it. Tick My picture already shows my callsign under the transmit preview and the plate is left off — for that picture only. The tick clears every time you load a new image, on purpose: it is a statement about one picture, and identification is your responsibility when you make it — Nexus cannot read your artwork. With the box unticked nothing changes, and a station with no callsign set still cannot transmit at all. Asked for by akhepcat.

  • Back up your whole setup to a file, and restore it on another computer.

    Settings ▸ Radio now has Back up and Restore. One file holds your radios, operating preferences, memory channels, watchlist and chase sets — for a new laptop before a contest, a rebuild after a disk failure, or a second machine that should match the first. Until now there was no way to keep a copy of any of it: it lives in a configuration folder most operators never open. It holds no passwords or API keys. Those stay in your operating system's keychain, so a restore asks for them again and the file is safe to carry on a USB stick or email to yourself. Your contact log is separate and is not included — export that from the Logbook. Restoring replaces your current setup and says so before it does anything.

  • The operator at the key is shown in the top bar, and swaps in one click.

    Set an operator and it appears beside Help, in the one group no cockpit hides — because a wrong operator is silent: nothing misbehaves and you find out at submission, when the log is already wrong. Click it to hand over to anyone already in the log, or to go back to single-operator when the activation ends. Nothing shows at all if you have not set one, which is the normal single-op case.

  • Export one ADIF per operator.

    With more than one operator in your log, the Logbook grows an Export per operator button: one file per operator, named with their callsign, plus the combined log — which is still the one carrying any contacts logged with no operator set. POTA and Field Day both want each operator to submit their own, and these get uploaded from a phone in a car park, so the filenames say who they belong to at a glance. A single-op station never sees the button, because the file it produced would be a copy of the one beside it.

  • Two of you on one radio? Your contacts can now say who made them.

    Set Operator at the key in Settings ▸ Station and every contact you log carries it, so a shared POTA activation or a Field Day shift can be split by operator afterwards instead of hand-edited — both programs want each operator to submit their own log. It is the ADIF OPERATOR field, which is a different question from your station callsign: one is who was operating, the other is whose station it was, and that distinction is exactly what the two fields exist for. Leave it blank if it is just you and nothing is stamped — an operator equal to the station call in every record says nothing. This setting existed before but only inside Field Day, where a POTA pair would never have found it, and it only ever reached an N3FJP feed rather than your own log. Asked for by a POTA operator running two ops on one laptop.

  • The band map tunes now — click it, or scroll on it.

    It has always been a frequency scale with your dial marked on it, but it was the one place showing frequencies that you could not act on: the readout digits and the waterfall both tune by wheel, so the map staying inert read as something broken rather than something missing. Click anywhere on the track to go there, or scroll to tune the way every other dial in Nexus does — same step, same sensitivity setting. Clicking a spot still works that station rather than tuning to it, because a spot's label is nudged aside from its true frequency to keep a crowded band readable, so the label is not where the signal is. The map stays read-only while CAT is down or you are transmitting. Asked for by kr4fqg.

Every version