Nexus 0.27.0
Released 2 August 2026. The current release is 1.24.0. Download it · Every version
Fixed: the log strip no longer invites an accidental log entry
On the Phone and CW log panel, the callbook button sat in the same row as Log — and that row wraps to fit whatever width the log column has, so Log did not stay in one place. Reaching for a lookup mid-QSO could commit the contact instead, and several contacts got logged that way.
Log and Spot now have a row of their own, directly above the caller card and immediately under the line that spells out what will be written ("logs to the shared logbook as SSB · 20m · 14.200 MHz"). Read it, then commit — and Spot stays right beside Log, where it has always been. Nothing else about logging changed: the same fields, the same Enter behaviour (Enter on a fresh call still pulls the callbook first, then logs), the same Clear button beside the callsign.
The callbook button is also now labelled Lookup rather than QRZ, and is styled as the quiet utility it is instead of a second action button. Lookup is the honest name: it asks QRZ first and falls through to HamQTH, so it has never been QRZ-only, and it works with a free HamQTH account and no QRZ subscription at all.
Fixed: a satellite uplink Nexus cannot verify is now refused instead of reported applied
Hold a linear bird on an IC-9700, pick "Main = downlink, Sub = uplink" — the layout the radio is actually built for — and the pass rail answered "the uplink was not written". It never said why, and the mapping the operator picked was the right one for their rig.
The missing piece was which CAT backend was serving. Nexus can drive an IC-9700's satellite mode two ways, and only one of them is wired: Native CI-V, where Nexus owns the CI-V port itself, engages the rig's satellite mode, writes the uplink into the Sub band and reads every step back off the radio before reporting it done. Native CI-V is off by default, so the usual IC-9700 station runs on Hamlib's rigctld instead — and this build has no verified satellite split there.
-
"Main = downlink, Sub = uplink" when Native CI-V is not what is serving.
Nexus now sends nothing at all — no split command, no transmit frequency — and the CAT status line says so in one sentence: Nexus drives that layout only through its own native CI-V backend, that backend is not what is serving this radio, nothing was written and nothing was transmitted, and your receive dial is still being corrected. It does not guess which backend is serving. Nexus may have started
rigctlditself, fallen back to it when its own daemon failed to start, or simply attached to arigctldsomeone else launched — and in that last case it has never even read the radio's model. Test CAT names the one in use. Before, the split went out, and arigctldthat answered "ok" was enough for the pass rail to display a transmit frequency the radio was not on. -
Every uplink mapping that rides VFO A/B on the IC-9700, IC-910, IC-9100 and IC-905 — a behaviour change.
Nexus has no verified cross-band VFO A/B satellite split for these four radios; sending one unverified is how 0.24.2's "uplink" went nowhere. The rule is now that the uplink is written only where Nexus can place both legs of the pass on the same band. A cross-band pass is refused — and so is a pass it cannot place on the band plan at all, because then it cannot prove the two legs share a band. That second case is not hypothetical: the band plan stops at 23 cm, so an IC-905 microwave bird is refused even when both legs really are on 10 GHz. "Transmit only" is refused on these four radios for the same reason — that mapping never writes a receive dial, so there is no second leg to compare against. V/V and U/U birds up to 23 cm keep the A/B split: both legs on one band is the case A/B carries, and it still carries it. Every other radio — the FT-847, TS-2000, TS-790 and FT-736R class, and every ordinary HF rig — is untouched and still takes a cross-band uplink on VFO B as before.
-
Settings no longer pre-fills a mapping it cannot drive.
The one-click "Main = downlink, Sub = uplink" offer now appears only where Nexus's own CI-V engine can actually serve the radio — an IC-9700 or IC-905 on a serial/USB port. The IC-910 and IC-9100 are outside that engine's coverage in this build, and so is any radio reached over the network rather than a serial port, so those stations are asked instead of pre-filled. "Serial" here means exactly what the CI-V daemon itself means by it, so a settings file written before Nexus had a connection setting counts as serial, the same as it does everywhere else in the app. Doppler keeps correcting your receive dial wherever it was already correcting one, so no refusal above costs you the downlink; work the uplink from the radio's own front panel for that pass. On the A/B side a refusal does cost something real, and it is worth being plain about: the IC-910, IC-9100 and IC-905 used to get an uplink attempt on a cross-band pass, and now they do not. That attempt was only ever acknowledged — the radio answered "ok" and nothing in this build read the frequency back off it — so what it bought was a number on the pass rail that may or may not have been where the radio was transmitting. Losing it is the safer half of the trade: on a linear bird, a split you believe and the radio did not take puts your carrier in the transponder's downlink passband, on top of everyone else working it. (On the IC-9700 that attempt is known not to land — it is the 0.24.2 bug.) Ordinary A/B splits — including every terrestrial pile-up "UP 5" — are untouched on every radio and on both backends. The Hamlib satellite recipe stays deliberately unwired rather than sent hopefully. Hamlib can answer a frequency read-back out of its own cache, so a read-back a stale value can satisfy cannot tell a landed uplink from an echo of what we just asked for, and confirming the difference needs a real radio on a real
rigctld— which neither this tree nor CI has. Guessing wrong there puts your carrier in the transponder's downlink passband, on top of everyone else working the bird.
Fixed: switching modes no longer forgets your frequency
Working a station on 20 m phone at 14.240, a stray click into CW and back reset the dial to 14.225 — the start of the phone segment — and the contact was gone. Nexus now remembers the frequency you were using, per band and per mode, for the whole session. Switch from Phone to CW and back and you are on your own frequency again, not the segment default.
Every band keeps its own memory. Leave 20 m phone for 40 m CW and your 14.240 waits: switch to Phone on 40 m and you get your last 40 m phone frequency (or the usual segment start if you have not used one yet), and picking 20 m from the band dropdown brings 14.240 back. Every mode remembers on a mode switch — Phone, CW, RTTY, and the digital Operate section — and the band dropdowns in the Phone and CW cockpits restore the band's memory too. Anything that names an exact frequency still goes exactly there, with no memory overlay — a typed MHz, a band-plan channel, a spot or Needed click carrying the spot's own frequency.
The first visit to any band/mode this session behaves as it always has, and a restart starts genuinely fresh: the memory is never saved to disk, and the frequency Nexus starts on is not counted as one you chose — neither the dial in the settings file nor the one read off the rig when the CAT link first opens. It cannot be: those say where the radio IS, never who put it there, and that could as easily have been APRS, a repeater, an ISS pass or a satellite as you. So closing Nexus parked on the national APRS channel and opening it the next morning will never turn 144.390 into "your 2 m phone frequency". Memory starts the moment you tune somewhere yourself. License privileges are still checked on every return — if your class changed mid-session, the dial falls back to the legal segment start.
Turning the VFO knob counts as choosing a frequency, with two exceptions. The first dial the CAT link reports is not a knob move at all — it only says where the rig already was, as above. And while Nexus is holding the dial on one of its own channels — APRS, an FM repeater or simplex channel, the ISS SSTV downlink, a satellite transponder — tuning around inside it is working that channel, not picking a frequency for the band, so it is not remembered. Move the knob to another band and the dial is yours again. The knob reports a frequency and nothing else, so a frequency you reach that way comes back on the band's usual sideband rather than the one you happened to be using on the band you left.
Two coarsenesses worth naming. An FM channel or FM-bird hold stays in force until you tune somewhere explicitly (that is what keeps the rig in FM instead of dropping it the moment you nudge the VFO), and while it is set any knob move at 29 MHz or above counts as the machinery's, whichever VHF/UHF band you are on; APRS is held to 2 m only. The channel mark Nexus sets when it tunes one of its own channels lasts the same way: it is dropped at the next frequency you choose yourself — a band pick, a typed MHz, a mode home, or a knob move onto another band — and until then a knob move inside that band is the channel's. It names the one band the app parked on, so anywhere else it says nothing at all.
Frequencies the app tunes for its own machinery are never mistaken for yours, and never the other way round either. Each part of Nexus that moves the dial says whose frequency it is at the moment it moves it, so the answer never depends on when the question is asked:
- Work a satellite from 20 m phone and your 14.240 is still waiting when you come back.
-
A pass that ends leaves the radio parked on the bird.
That Doppler-corrected frequency is never adopted as your frequency on 2 m, however long you stay there afterwards — and neither is a spot you hand-tuned to inside a linear bird's passband.
-
Tuning APRS from 146.520 keeps 146.520 as your 2 m phone frequency, even though 146.520 and 144.390 are the same band…
Tuning APRS from 146.520 keeps 146.520 as your 2 m phone frequency, even though 146.520 and 144.390 are the same band and nothing about the move looks like leaving it.
-
Switching radios banks the dial you were on before the handoff.
The memory is station-wide: a recalled frequency routes to the radio that owns that band and mode, exactly as any other band change does.
-
The ISS SSTV auto-arm's 145.800 is one of those channels too: your 2 m frequency banks on the way in, the 145.800…
The ISS SSTV auto-arm's 145.800 is one of those channels too: your 2 m frequency banks on the way in, the 145.800 itself is never remembered, and hand-chasing the pass's Doppler inside the armed window does not overwrite what you were on either. Two smaller things. Picking the band you are already on from the dropdown now keeps the frequency you are on, instead of snapping back to the segment start. And that dropdown's tooltip says what the control does: your last frequency on that band in that mode, else the start of your licensed segment.
Added: hear the pass — AOS and LOS alerts for the armed track
Arm a pass with "Work this pass" and walk away from the screen: the moment the bird rises, Nexus now plays a rising three-tone and shows a loud popup with the facts you need at the rotor — the bird, the azimuth it rises on, how high the pass peaks, and how long you have. At LOS a falling tone (tellable from the rising one blind) accompanies the pass-complete notice, which keeps saying exactly what was handed back — the dial, the uplink split, and whether the rotor is about to park or move to ready on its own.
The alerts fire wherever you are in the app, not only with the Satellites section open — the pass-complete notice used to be the section's alone, so working the Logbook at LOS meant missing it. They report the track's own state transitions, never a wall-clock timer: if the machine slept through the rise, waking mid-pass says "pass in progress" with the minutes left instead of a stale "starting now", and waking long after LOS says nothing. Alerts never touch the radio, the rotor, or the track — they only tell you what happened.
This is the armed track's own channel, separate from the per-pass ⏰ alarm on the schedule. An alarm lead ("wake me 15 minutes before") still fires on its own moment; at AOS the two channels coordinate so you never get two popups for the same rise.
The popup never steals keyboard focus and never blocks the screen. The tones are on by default — that was the ask — with a "Pass alert sounds" switch beside the satellite Doppler settings for operating from a quiet shack; the popups stay either way.
Changed: the pass strip says what it ranks — "Next" and "Best 24 h", over every bird
The strip above the Satellites schedule was labelled "Next up" but ranked by pass quality, so a spectacular pass hours away could sit above a workable one rising in ten minutes — entries "further away than what's in the schedule", as the field report put it. It also read only your ★ favorites, so a good pass on an unstarred bird could never surface there.
The strip is now two labelled pairs, drawn from every workable bird over your grid: "Next" is the two soonest passes in clock order — a pass already in progress leads, marked "already up" when it rose before the scan window — and "Best 24 h" is the two highest-quality passes in the next 24 hours, by the same elevation-first rank the old strip used. A pass that earns both spots shows once, under Next, and Best fills in with the next one down. ★ marks your favorites, the mode pill ([FM voice], [Linear SSB/CW], …) rides every classified row, and ▶ Work runs the same chain for any bird, starred or not. Needed-grid chips stay on ★ rows — they come from the favorites schedule, which is the only place that computes them.
The strip admits exactly the birds the "Other birds overhead" band admits — one rule, so a bird that band refuses (reported dead or re-entered, placeholder rideshare names) never appears in the strip either. The schedule below and the band are unchanged.
Added: hide the countries you have stopped chasing
On a busy evening most of a waterfall is a handful of countries you worked years ago. Band Activity gains a Countries button beside the filter chips: tick the ones you do not want to see and they leave Band Activity and the Call Roster.
The list is 18 — the high-density entities an experienced chaser actually wants gone, rather than the full DXCC table with a search box:
United States (K/W/N) · Canada (VE) · Mexico (XE) · Germany (DL) · Italy (I) · Spain (EA) · England (G) · France (F) · Japan (JA) · Brazil (PY) · Argentina (LU) · Poland (SP) · European Russia (UA) · Ukraine (UR) · Netherlands (PA) · Czech Republic (OK) · Slovenia (S5) · China (BY)
It is a view filter and only a view filter. Decoding, logging, the QSO sequencer and your needed-entity alerts never see it, so nothing you have switched off can cost you a contact:
- a station calling you always shows, whatever its country;
- the station you are working stays put mid-exchange;
-
needed outranks excluded
— a new entity or a new band slot from a country you have hidden still comes through. Matching is on the DXCC entity the callsign resolves to, not the letters it starts with, so
VE3XYZ/W1counts as United States andW1ABC/VE3as Canada, and hiding Germany catches DA, DB, DJ and DK as well as DL. While anything is hidden both panes carry a "3 countries hidden" chip with a Clear beside it, so a thinned band is never a mystery — and the "N heard" and roster counts state what is on screen, not what arrived. The Rx Frequency pane is deliberately left alone: it answers "what is on my frequency", and a station you are not chasing sitting on top of you is exactly what you need to see. Your choice is remembered per station, not per window — pop out a band map and it shows the same band as the main window.
Added: full FT8/FT4 decoder configuration — AP on/off, CQ-only AP, single decode
The Decoder section (Settings ▸ Weak-signal digital) grows from depth-and-passband to the full set of receive controls the built-in WSJT-X decoder actually takes. A-priori (AP) decoding — the hypothesis-assisted passes that pull marginal replies out of the noise using your call, the DX call and the QSO state, including the cross-cycle replay — now has its WSJT-X "Enable AP" switch, on by default exactly as before. A second control restricts AP to the bare CQ hypothesis, the guard WSJT-X applies by itself after five idle minutes; Nexus makes it an explicit choice. And Single decode narrows the search to ±25 Hz around your RX marker — the same one-station window WSJT-X uses for a double-click re-decode — to isolate one weak station instead of working the whole passband. Single decode is FT8 and FT4 only: 50 Hz is narrower than a single JT65, Q65 or MSK144 signal, so those modes keep the full passband whatever the switch says.
Every control is wired through to the decoder itself and proven there: tests assert at the decode-job boundary that flipping a switch changes what the Fortran is called with, and behavioural tests show AP-off really does silence the FT8 AP passes and the cross-cycle replay, and that CQ-only really does cost FT4 the deep-hypothesis recoveries. Nothing placebo. Honesty notes, stated in the UI too: the AP on/off switch is FT8-only — FT4's decoder has no such flag (its AP is part of Normal/Deep depth), though the CQ-only restriction applies to both; and on a WSJT-X UDP companion source, decodes arrive already made, so none of the Decoder settings act. Defaults are untouched — leave the section alone and the decoder behaves byte-for-byte as it did.
Fixed: the caller card's distance and bearing now agree with QRZ
Work a station on Phone or CW and the card that appears when the call resolves reported a bearing that did not match the one on the station's QRZ page. The maths was right — the inputs were coarse. QRZ measures between the two stations' exact coordinates; Nexus was measuring between the centres of grid squares, and a square is a box, not a point.
QRZ and HamQTH both hand us the station's real position in the same lookup that fills in the name and QTH, and Nexus was throwing it away. It now keeps it, and the caller card measures from it. Worked against a 4-character locator that is worth about a degree of bearing and a dozen miles — W1AW from EN52 read 823 mi · 89°, and now reads 835 mi · 88°, measured to the same pin QRZ computes its own figures from.
Nexus only accepts a position the callbook actually vouches for. QRZ returns coordinates for every record, but tags where they came from — a real pin, a geocoded address, or a fallback back-derived from the grid square or the DXCC entity. The last of those can sit hundreds of miles from the station, so Nexus uses only the first two and falls back to the locator otherwise. A station with no position on file is unchanged.
The larger half of the gap is your own grid. Everything is measured from where you say you are, and a 4-character locator says only "somewhere in this ~100-mile square" — so Nexus measures from the middle of it. On a DX path that costs about a degree, but on a station a couple of hours away it is worth up to ~30°, whatever the other end does, and it puts every distance out by up to ~60 miles at any range. The setup wizard used to tell you 4 characters was plenty. It no longer does, Settings says the same, and the caller card's tooltip now names whichever side is still a square instead of presenting the number as exact. Setting a 6-character grid closes it.
Distance is unchanged in method (great-circle, statute miles) and gains the same precision. The Operate roster's bearing column and the map are untouched. So is rotator pointing, which answers a deliberately different question — "point at this callsign's DXCC entity" — and still swings to the entity's nominal centre, not to the station. On DX that is within a beamwidth; for a stateside contact the two numbers are not meant to agree.
Changed: the displays answer the audio — waterfalls, scopes and meters respond faster
The voice/CW/RTTY/SSTV displays lagged the ear. You heard a signal and the screen answered late and rounded-off: the waterfall took a third of a second to bring a new signal to full brightness, the RTTY and SSTV waterfalls threw away five of every six spectrum rows, and both meters answered roughly half a second after the audio. Nothing here adds animation or invented motion — every change removes real delay between the antenna and the pixels.
-
The spectrum's analysis window is half as long.
A key-down or a voice syllable reaches full brightness in ~170 ms instead of ~340 ms, and short CW elements stop blurring together. Nothing visible is lost: the finer raw resolution the long window bought was below what the display can draw, and computing the shorter window costs less CPU.
-
The RTTY cockpit and the SSTV band waterfall scroll at 20 rows per second — the same cadence the Phone/CW rig scope already runs — instead of 8.
The FT8/FT4 waterfall is unchanged: those modes are slot-synchronous and their display is right as it is.
-
The RX audio meter has real instrument ballistics: it snaps up with the audio (90% of a step within ~60 ms) and falls smoothly, like a hardware S-meter.
It used to smooth the attack and the decay equally, so the needle registered a signal ~150–300 ms late.
-
Both meters read through a new fast lane, polled every 100 ms, that no longer rides the radio loop or the 300 ms status…
Both meters read through a new fast lane, polled every 100 ms, that no longer rides the radio loop or the 300 ms status poll — so a slow CAT read can no longer freeze the needles for seconds at a time, and a meter whose readings stop arriving falls back to rest within about two-thirds of a second instead of holding a stale needle. On a healthy link the CAT S-meter is also re-read every 360 ms instead of every 750 ms — twice as fresh, without crowding the link the radio dial shares. Slow serial links keep the gentler cadence; that limit is the radio's, not the app's.
-
The meter bars no longer ease their width in CSS.
The eased bar smeared every real reading by another 80–120 ms; color changes still ease, measurements do not.
Fixed: "TLE 26 days — STALE" on a catalog that is hours old
The Satellites screen carried an amber "TLE 26 days — STALE · refresh" chip, a warning chip in the status bar on every screen, and a Settings line reading "The element mirror is unreachable and your elements are 26 d old — import a fresh element file or retry later." None of it was true, and there was nothing to do about it.
Every one of those surfaces was reading the age of the single oldest satellite in the catalog and presenting it as the age of your elements. The catalog is 367 birds, and the typical one is a few hours old — but a handful are legitimately old. AO-7 launched in 1974 and gets fresh elements when someone observes it; a few dozen more are re-observed every few weeks rather than every day. One of those set the number for all of them.
Worse, the number could not come down. A satellite whose elements pass 30 days is dropped from the set, so the oldest one still counted is always just under 30 — as one aged out, the next in line inherited the badge. The chip was going to read stale forever no matter how fresh your elements were, and the calm "your elements are current" message the app already had could never appear.
The chip, the status-bar chip, the Settings line and the Connect Passes badge now report the median age of the elements they draw on: what the typical satellite in your catalog carries, not what the slowest-observed one does. On the catalog shipped with this release that reads a fraction of a day. A set that genuinely goes stale still says so — once more than half the birds pass the 14-day line the median goes with them.
Satellites held back by the 30-day ceiling are counted and reported in their own right, so "my catalog is current and a few birds sit out" no longer reads identically to "my whole set has gone stale". The Satellites header states the number beside the age ("367 birds · 30 sit out past 30 d"); Settings ▸ Radio ▸ Orbital elements carries it on the line that is always there, not only while a refresh has failed; and every refresh result — landed, blocked or failed — accounts for those birds.
Satellites in the 14-to-30-day band are counted too: still used, still drifting. A median can hide that band, and this is the shape it hides — half your birds sitting at 29 days while the other half arrived this morning reads as a current catalog, because the typical bird is one. The Satellites header carries both counts whenever they are not zero, and when most of the elements you hold are past the 14-day line the status bar and the Connect Passes pane say so as well. A slow-cadence tail on an otherwise current catalog stays quiet, which is what the shipped catalog looks like.
The Satellites section always has a way to refresh elements. The amber chip was also the section's refresh button, and it correctly disappears now that the reading is honest, so a quiet "⟳ refresh elements" chip sits in the header whenever the amber one does not.
Unchanged: the per-satellite rules. A satellite past 30 days is still refused by name and age when you arm it, arming a satellite with elements past 14 days still asks first, the Birds list still explains every excluded bird individually, and a fresh bird still arms cleanly no matter how old anything else in the catalog is.
Changed: the sky dome carries azimuth and elevation on the satellite
Working a pass with a manual rotator means reading the azimuth and elevation off the screen and turning the mast to match. Both numbers lived in a tooltip, so following the pass meant holding the mouse on a moving dot to see where to point next.
Azimuth and elevation now ride on the satellite and stay there for the whole pass. They sit on a plate that holds them legible over the pass track, and they move to the other side of the satellite near the edge of the dome or when the antenna marker is on that side — so they never run off the dome, and never cover the gap between where the antenna is pointed and where the satellite is.
The satellite is drawn as a spacecraft — body and solar panels — rather than a plain dot, and matches the mark the world map already uses for the same bird. The text readout under the dome and the hover tooltip are unchanged.
Changed: the rise and set marks on the sky dome carry their bearings
The two triangles on the rim of the dome are where a pass starts and ends, and they are what you pre-point a manual rotator by. Their bearings lived in a tooltip. The set mark is drawn as an outline rather than a solid triangle, and an outline answers the mouse only on the line itself — so on that one, catching the tooltip meant landing the pointer on a hairline.
Both marks now print their bearing on the dome: the rise or set bearing with its compass point, on a plate beside the triangle it belongs to. Each plate names which mark it is, in words and with the same up and down arrows the readout under the dome uses, so a pass that rises and sets close together never leaves two bare numbers you have to tell apart.
The plates are azimuth only. Both marks sit on the horizon, so an elevation there would be restating the picture rather than reporting anything.
A plate sits inside the dome next to its triangle and steps aside when that spot is taken — it never covers either triangle, the satellite, the satellite's own readout, or the antenna marker, and it stays inside the horizon at every bearing. The whole body of the set mark is now hoverable, so its tooltip, which also carries the time, is reachable without threading the mouse onto a line.
Added: the satellite surfaces say how high the bird is
The Birds list, the world map's satellite hover and the sky dome each show the bird's altitude — how far above the earth it is, right now.
It is the number that tells you what kind of pass you are about to work before you open anything: a 630 km LEO screams over in ten minutes with the Doppler running away from you, while an elliptical bird near apogee loiters for the better part of an hour with a shift that barely moves. Down the Birds list the figures line up in a column, so a low circular bird and a high elliptical one tell themselves apart at a glance.
Altitude is always labelled, and on the sky dome it sits directly under the range it is easily confused with — range is how far the bird is from you, altitude is how far it is from the ground. The dome's number is recomputed with the rest of the pass geometry every three seconds, because on an elliptical orbit it genuinely moves.
A bird nothing carries current elements for shows no altitude rather than a zero, and neither does an armed pass before the bird rises: there is no position computed yet, and 0 km would put a satellite on the ground.
Fixed: the starter favorites now cover the birds you can actually work
The one-time favorites seed ranked every bird on pass count, and pass count is mostly a statement about altitude. FO-29 flies an 800 × 1320 km ellipse at 13.5 orbits a day; a 480 km cubesat flies 15.3, so the cubesat shows more passes over your grid every day of the year. Run the app's own pass predictor over the catalog it ships with — 24-hour windows from six grids at four different start times — and the old ten came out 3 to 9 beacon-only telemetry cubesats out of ten, 6.5 on average: birds you can hear and cannot work. That is the shape of the catalog rather than bad luck. 305 of the 367 birds that carry elements have a downlink and no uplink, and 18 of the top 20 by orbital rate are among them, so a pass-count ranking fills its ten mostly from birds nothing can be worked on.
What happened to FO-29 is the other half of it. Across those same 24 runs it ranked anywhere from 2nd to 81st, and it reached the top ten in 6 of them; some SSB/CW transponder reached the ten in 14 of 24. Ranked on passes alone, whether you got a workable bird at all came down to the geometry of the day you first opened Satellites.
The catalog now records what each bird is worked WITH — an FM voice repeater, a linear (SSB/CW) transponder, a digital/packet channel, or a beacon downlink and nothing else — and the seed fills its ten by rotating through those three workable kinds, each one offering its own best bird under the same pass-count-and-elevation rank as before. A beacon-only bird takes a slot only after the three are exhausted. On today's catalog that is four linear, three FM and three digital — at every one of those 24 grid-and-window measurements, with no beacon-only bird in any of them.
Nothing else about the seed changed. It still runs once ever, still never seeds over anyone who has stars or ever had them, still requires a bird to be alive, to carry a live amateur transmitter, to hold current elements and to fly a workable pass over your grid, and still waits rather than spending itself when your grid or the catalog is not there yet. A bird can hold more than one kind at once, which 60 of today's 372 active amateur birds do — QO-100 carries a linear transponder, digital segments and a beacon — and each one counts wherever it belongs.
Existing stars are untouched. This affects a first run only.
Fixed: satellite SSTV puts the radio in FM
Clicking an SSTV row on the ISS — 145.800 "Mode V Imaging", 437.800 or 437.550 "Mode U - SSTV" — set the radio to USB. Every SSTV downlink in orbit is narrowband FM, so the picture arrived garbled. Satellite SSTV rows now set FM. HF SSTV is unaffected: it really is an SSB mode, and no satellite transmits SSTV below 30 MHz.
Known: a satellite that publishes an LSB downlink is still tuned USB
FO-82 (NORAD 40320) declares its transponder downlink as LSB. Picking it sets the radio to USB and the uplink is then derived from that, so both ends come out on the wrong sideband. It is one bird, and the fix belongs with the wider transponder-tuning work rather than here.
Added: other birds, without leaving your schedule
The Satellites schedule shows your ★ favorites, and the Birds list names every bird — but nothing connected the two. Finding a workable pass on a bird you had not starred meant a trip to an alphabetical name list that carries no pass information at all.
The schedule now carries a second, collapsed band under your favorites: "Other birds overhead · N workable · 24 h". The count is live. Expand it and you get one row per bird — its best pass in the next 24 hours, ranked by how workable it is — with the same star and ▶ Work controls your own rows have. Star a row and the bird moves up into your schedule, gaining the needed-grid chips, the pass alarm and the 48-hour view. Collapsed, the band adds nothing to the screen: your schedule is exactly what it was. Twelve rows show by default, with "show all N" for the rest; placeholder rideshare objects that share a name with their siblings are left to the Birds list, where starring one cannot star six.
With no favorites starred at all, the section no longer replaces the whole planning column with one sentence — the schedule renders with the band open, so the answer is on screen from the first visit.
Birds' mode class shows as a pill beside the name ([FM voice], [Linear SSB/CW], [Digital], [Beacon]) on schedule and discovery rows, wherever the catalog can say — no class, no pill, never a guess.
Added: a way back out of a bird's detail
Opening a bird pinned its detail to the side column for the rest of the session — no close, no Escape, and the Birds list starts more than a full screen below it. The detail heading is now sticky with a ✕, Escape closes it too, and the tracking badge in the header opens the tracked bird again with one click. Closing a detail never touches the track, the rotor or the dial.
Fixed: a pass already in progress reported a wrong rise time
The Satellites view's 24-hour pass scan started at "now", so a bird already above the horizon was reported as rising at the moment you looked, with its peak elevation understated to what remained — and the cached scan repeated that answer for up to ten minutes. The scan now looks back six hours, the same way the schedule always has: a pass in progress keeps its real rise time and its full peak. This also corrects the map view's pass pane and the first-run favorites seeding, which read the same scan.
Changed: the FT8 Classic view puts the decode window where you operate
Field feedback from an advanced DX operator, running Nexus beside WSJT-X on an ultrawide: the Rx Frequency pane — the stream you actually run a QSO from — was a small box at the bottom of the side rail, and too much of the screen went to chrome.
The Classic layout is now three columns: Band Activity on the left, a full-height Rx Frequency pane beside it with the Tx1–Tx6 message machine docked underneath — the same bottom-right geometry as WSJT-X — and the Stations roster on the right. Your own transmissions appear in the Rx Frequency stream in yellow, interleaved with the answers, so you click a call and watch the whole exchange line by line. Clicking a decode there does exactly what it does in Band Activity. A drag handle between the pane column and the roster lets you tune the balance; the split is remembered.
The chrome got out of the way. The status line, the QSO panel's three rows and the permanent TX-meters line are now ONE compact strip under the waterfall. Call CQ / S&P and TX On / Tune / Stop TX / Hold Tx come first, anchored at the strip's left edge, and they hold the same position whether the strip is idle, mid-QSO or transmitting — the DX call appearing or the rig keying never moves Stop TX. After the buttons: the TX state and sequencer readout with the DX call and report, what is being sent now, the free text box (sized to the 13 characters it can actually hold), and — moved into the strip from the old status line — TX AUTO / Skip Tx1 and the next-slot countdown, plus the SWR/ALC/Po/COMP meters in a fixed-width cell that shows live bars while transmitting and the dimmed last readings between overs, so nothing jumps when transmit starts. Every control that was on screen before is still on screen, always visible, just smaller and closer together. The header packs into one row on wide screens: the DXpedition mode is a dropdown, Spot and Pop out are icon buttons, Record QSO is a dot, and the memory chips keep to a single line.
At ultrawide half-screen the strip is a single row and the Rx Frequency pane grows from roughly 15 visible decode lines to over 60; Band Activity and the roster both gain height. On a 1366x768 laptop the three columns hold and the strip wraps at a fixed point into two stable rows — buttons and readout above, sending/free-text/meters below; on narrow windows everything stacks in one column as before.
Changed: Doppler corrects your downlink without being asked
Arming a pass and picking a transponder is the whole ask. Nexus corrects the receive dial for the pass from that moment, with nothing to switch on first.
Before this, two settings stood between an armed pass and a moving dial — Satellite Doppler, and the VFO mapping — and both shipped off. A station that had never opened Settings ▸ Radio armed a pass, held a transponder, and watched the dial sit still with nothing on screen saying why. Staying on an SSB signal as it walks several kHz across a 70 cm pass is most of what Doppler correction is for, and it was behind two switches.
The two switches were never the same kind of thing, and only one of them is kept as a precondition:
-
The downlink is automatic.
Correcting the receive dial cannot transmit. The worst case is that you do not hear the bird, so it asks for nothing.
-
The uplink is still confirmed, once per radio.
A wrong VFO mapping transmits on your own downlink — into the satellite's output passband, on top of everyone working the bird — so nothing reaches your transmit VFO until you have said which VFO carries the uplink on the radio in use. The confirmation now happens where you already are. On the pass readiness rail, the Doppler row reads what your radio can do and offers it in plain words — "Confirm the uplink and Doppler drives IC-9700 as Main = downlink, Sub = uplink" — with one button and the mapping list beside it if the derived answer is wrong. Confirm it once and that radio is never asked again. Where the answer is not certain, Nexus asks instead of guessing. An IC-9700, IC-910, IC-9100 or IC-905 running on Nexus's own CI-V connection has exactly one full-duplex layout it can express, and that is the one offered. A full-duplex radio Nexus drives through Hamlib, an FT-847, FT-736R, TS-2000 or TS-790, or a radio it cannot identify gets the question, not a pre-filled answer. A single-VFO radio is offered nothing at all: its downlink is corrected and the transmit dial stays yours. The confirmation is recorded per radio because the satellite path routes. A pass can hand the QSY to whichever rig covers the band, and peg-lock or a mid-pass handoff can change which radio is under the split — so the uplink is driven only on a radio you confirmed, re-checked every correction rather than assumed at arm time. If you were already correcting the uplink — Satellite Doppler on, a VFO mapping set — nothing changes for you. The mapping is kept exactly as written, it is not re-derived, and you are not asked to confirm it again on any radio your station had when you upgraded: the old station-wide grant is recorded as a confirmation for each of those radios, and it survives every save and relaunch. A radio you add later gets its own one-time confirmation, like any second radio. If you had picked a mapping but never turned the old Satellite Doppler switch on, that pair never tuned anything — so the mapping is kept, your downlink now corrects automatically, and the transmit VFO waits for the same one-per-radio confirmation a fresh install gets. Nothing reaches a transmit VFO on upgrade that was not already being driven before it. Satellite Doppler is still a switch you can turn off — Settings ▸ Radio ▸ Satellite Doppler — and off still means off, both legs, no dial and no split. What is gone is having it off by default. Non-satellite stations are untouched either way: the correction runs only inside a pass you armed, on a transponder you are holding, on a dial that pass owns. The pass rail now says which legs are actually being driven, separately — downlink corrected with the transmit VFO still yours reads as exactly that, and the Doppler readout shows an uplink frequency only when there is an uplink being written. That honesty extends to the birds with nothing to split: a one-channel (simplex) bird rides one dial and the rail says so instead of offering an uplink confirmation the pass cannot use, and a beacon — downlink only — never puts anything on the transmit VFO. A radio you remove takes its uplink confirmation with it (a replacement radio starts unconfirmed), config profiles carry neither the VFO mapping nor the confirmations between stations, and an operator who answers "Downlink only" is not asked again. The confirmation itself has one writer. Picking a mapping — on the pass rail or in Settings ▸ Radio — applies immediately and is recorded for the radio in play at that moment; a Settings window left open across radio changes cannot re-point the mapping or revive a removed radio's confirmation when you later press Save. When your chosen mapping is not yet confirmed for the radio under the split — a second rig, or an upgraded uplink-only station — the rail's confirm button offers exactly that mapping for exactly that radio; it never swaps in a derived one over your choice. And every surface that names who owns a frequency keys on what is actually driven: under an uplink-only mapping the tracking badge, the rotor strip and the end-of-pass notice say the transmit (split) VFO is Doppler's and the dial stays yours, instead of claiming a dial that never moved.
Known limitations
-
Multi-window per-radio profiles keep their files under a directory named by the radio's internal id, and ids are…
Multi-window per-radio profiles keep their files under a directory named by the radio's internal id, and ids are reused: remove a radio and add another, and the new radio can inherit the removed radio's window settings, geometry and journals. Pre-existing and unchanged in this release — recorded here so it is not mistaken for a new fault. The uplink confirmation is NOT affected: it is pruned when a radio is removed, so a replacement radio always starts unconfirmed.
Every version
- 1.24.0 11 Nov 2026
- 1.23.0 5 Nov 2026
- 1.22.0 1 Nov 2026
- 1.21.0 26 Oct 2026
- 1.20.0 22 Oct 2026
- 1.19.0 16 Oct 2026
- 1.18.0 12 Oct 2026
- 1.17.0 7 Oct 2026
- 1.16.0 3 Oct 2026
- 1.15.0 26 Sept 2026
- 1.14.0 20 Sept 2026
- 1.13.0 16 Sept 2026
- 1.12.0 14 Sept 2026
- 1.11.1 8 Sept 2026
- 1.10.3 4 Sept 2026
- 1.10.2 3 Sept 2026
- 1.10.1 1 Sept 2026
- 1.10.0 31 Aug 2026
- 1.9.2 29 Aug 2026
- 1.9.1 26 Aug 2026
- 1.9.0 25 Aug 2026
- 1.8.1 23 Aug 2026
- 1.8.0 23 Aug 2026
- 1.7.6 21 Aug 2026
- 1.7.5 20 Aug 2026
- 1.7.0 18 Aug 2026
- 1.6.1 17 Aug 2026
- 1.6.0 16 Aug 2026
- 1.5.0 16 Aug 2026
- 1.4.0 15 Aug 2026
- 1.3.0 13 Aug 2026
- 1.2.0 10 Aug 2026
- 1.1.0 10 Aug 2026
- 1.0.5 8 Aug 2026
- 1.0.2 6 Aug 2026
- 1.0.1 5 Aug 2026
- 1.0.0 4 Aug 2026
- 0.28.2 3 Aug 2026
- 0.28.0 3 Aug 2026
- 0.27.0 2 Aug 2026
- 0.25.0 1 Aug 2026
- 0.24.0 31 Jul 2026
- 0.21.5 29 Jul 2026
- 0.21.0 29 Jul 2026
- 0.20.0 28 Jul 2026
- 0.19.7 27 Jul 2026
- 0.19.6 26 Jul 2026
- 0.19.4 26 Jul 2026
- 0.18.0 25 Jul 2026
- 0.17.22 25 Jul 2026
- 0.17.21 25 Jul 2026
- 0.17.20 25 Jul 2026
- 0.17.19 25 Jul 2026
- 0.17.18 25 Jul 2026
- 0.17.17 25 Jul 2026
- 0.17.16 25 Jul 2026
- 0.17.15 25 Jul 2026
- 0.17.14 25 Jul 2026
- 0.17.13 25 Jul 2026
- 0.17.12 25 Jul 2026
- 0.17.11 25 Jul 2026
- 0.17.6 25 Jul 2026
- 0.17.5 25 Jul 2026
- 0.17.4 25 Jul 2026
- 0.17.3 25 Jul 2026
- 0.17.2 24 Jul 2026
- 0.17.1 24 Jul 2026
- 0.16.4 24 Jul 2026
- 0.16.3 24 Jul 2026
- 0.16.2 24 Jul 2026
- 0.16.1 24 Jul 2026
- 0.16.0 24 Jul 2026
- 0.15.24 24 Jul 2026
- 0.15.23 24 Jul 2026
- 0.15.22 24 Jul 2026
- 0.15.21 24 Jul 2026
- 0.15.20 24 Jul 2026
- 0.15.18 24 Jul 2026
- 0.15.17 24 Jul 2026
- 0.15.16 23 Jul 2026
- 0.15.15 23 Jul 2026
- 0.15.14 23 Jul 2026
- 0.15.1 22 Jul 2026
- 0.15.0 21 Jul 2026
- 0.14.0 21 Jul 2026
- 0.13.0 19 Jul 2026
- 0.12.0 18 Jul 2026
- 0.11.1 18 Jul 2026
- 0.11.0 18 Jul 2026
- 0.10.0 17 Jul 2026
- 0.9.7 17 Jul 2026
- 0.9.6 16 Jul 2026
- 0.9.5 16 Jul 2026
- 0.9.4 16 Jul 2026
- 0.9.3 16 Jul 2026
- 0.9.2 15 Jul 2026
- 0.9.1 15 Jul 2026
- 0.9.0 15 Jul 2026
- 0.8.9 15 Jul 2026
- 0.8.8 14 Jul 2026
- 0.8.7 14 Jul 2026
- 0.8.6 14 Jul 2026
- 0.8.5 14 Jul 2026
- 0.8.4 13 Jul 2026
- 0.8.3 13 Jul 2026
- 0.8.2 13 Jul 2026
- 0.8.1 12 Jul 2026
- 0.8.0 12 Jul 2026
- 0.7.1 12 Jul 2026
- 0.7.0 12 Jul 2026
- 0.6.0 11 Jul 2026
- 0.5.2 11 Jul 2026
- 0.5.1 10 Jul 2026
- 0.5.0 10 Jul 2026
- 0.4.1
- 0.4.0
- 0.3.0
- 0.2.0 3 Jun 2026
- 0.1.0