Nexus 1.11.1
Released 8 September 2026. The current release is 1.24.0. Download it · Every version
There is no 1.11.0 release. The first beta of this work shipped as v1.11.0-beta.2 and, through a defect in the release workflow since fixed, its installer was stamped 1.11.0 rather than 1.11.0-beta.2. A prerelease sorts below its release, so those installs believed they were newer than any 1.11.0 that followed and would never have been offered it. Releasing as 1.11.1 reaches them. Operators who were on 1.10.3 are unaffected and were never offered a beta.
Added
-
JS8: CQ and the heartbeat repeat on their own, with the countdown on the button.
JS8Call's CQ and HB buttons are switches — arm one and it calls or beacons on a timer, with the seconds to the next transmission ticking down in the button's own label (
CQ (12),HB (42),HB (now)). Nexus's were one-shots. They aren't any more. Set Settings ▸ Digital ▸ JS8 ▸ CQ repeat interval; at 0 the CQ button stays the single-shot it always was, above 0 it becomes the switch. The heartbeat's countdown appears whenever its schedule is running. This is the POTA and beacon habit: set it going and walk away. It stops when it should, and every existing transmit guard still holds. It keys nothing until both acts are present — the session TX latch and the repeat switch itself — so arming one with TX off shows "on", never "armed", and sends nothing. A station answering you turns the CQ repeat off, exactly as JS8Call does. The idle watchdog (60 min by default) stands both down, and a scheduled call cannot reset that clock the way your own sends do — an unattended station has a bound. Stop TX, leaving JS8 and changing operating mode each cancel the schedule and drop anything already queued. It is never remembered across launches: the app can never come back calling CQ. -
The FT-710 can draw its own band scope — in source builds only.
It is not in the 1.12.0 installers: the scope needs FTDI's closed-source LibFT4222, so it is built only with the
yaesu-wfbuild option, which release builds leave off. The radio has a real spectrum display inside it and an internal USB bridge that will hand it over; until now Nexus could only show the sound card's 4 kHz slice. Turn it on per radio in Settings ▸ Radio (it appears only for an FT-710), and the Phone and CW cockpits gain span and position controls that drive the radio itself rather than cropping its picture. Two things it needs, and it tells you which one is missing instead of showing an empty panel: SCU-LAN10 enabled in the radio's EX menu, and FTDI's LibFT4222, which is not bundled — it is closed source and Nexus is GPL-3.0-only, so the app names it and links the download rather than shipping it. Everything else works without it; this affects the RF panadapter only. One caveat worth knowing: in FIX position the radio does not report where its own window starts, so Nexus derives it from the band edge. That is measured, not assumed — but check it against the rig's own scale, and the tooltip says so. -
JS8 is on out of the box.
The JS8 section shipped hidden behind a Settings ▸ Features toggle while it was new. It now sits in the Digital group of the rail, after APRS, with no toggle to find first. Nothing else about it changed: opening it tunes the rig to the band's JS8 watering hole and starts decoding all four speeds, and it still transmits nothing until you enable TX in the header and switch on whatever you want it to send. If you don't operate JS8, turn it off in Settings ▸ Appearance ▸ Features. Two things worth knowing, because they have not changed either: the four speeds have not been run side by side with JS8Call on the air yet — Fast and Turbo have only ever been decoded from generated audio, never off a real band — and at Normal, Nexus needs roughly 2 dB more signal than JS8Call to print the same message. Reports from the air are welcome.
-
JS8 transmits.
The JS8 section now keys every period like JS8Call: your messages, CQ, heartbeats on a random free 500–1000 Hz slot, and — behind the session TX latch plus the persisted switch — autoreplies, relay and HB-ack, each with a visible, cancellable countdown. Heartbeats are exempt from the 6-minute TX watchdog the way WSPR/FST4W beacons are and bounded by JS8Call's 60-minute idle watchdog; Stop TX clears the queue, the heartbeat schedule and any pending reply.
-
A POTA activity map in FT mode.
A new Map button beside Classic/Roster in the FT header opens the map in its own window — put it on a second monitor next to the roster, the way GridTracker is used for park hunting. It plots every spotted POTA activator from the live feed (the same source GridTracker reads, with no mode or age filter), coloured by new-vs-worked, with the park, activator, band, mode and spot age on hover. Double-click a park to QSY, set the mode and tag it as your hunt target — it never keys the transmitter. The window remembers its own layers and projection, separate from the Connect map.
-
An audible alert for new POTA activations.
Settings ▸ Spots & Alerts has a new New POTA activation switch (off by default). With it on, Nexus beeps once when a park that was not on the air at the last check appears on the feed — from now on, never for the backlog that was already active when you turned it on. It works whether or not the map window is open.
-
A per-radio switch to hold FM-D while you are receiving SSTV.
Nexus commands the FM data submode (FM-D / DATA-FM) only while a picture is queued or going out, and puts the radio back in plain FM in between — deliberate, because an SSTV send once keyed a data mode into an FM repeater input, but wrong if you park on an FM SSTV calling channel for the evening and expect the rig to stay in FM-D. Settings ▸ Radio ▸ Rig & CAT ▸ Advanced ▸ "Hold FM-D while SSTV is receiving", off by default, set per radio because it depends on how that rig is cabled and what you use it for. With it on, the radio is held in the data submode for as long as the SSTV receiver is running. Stop the receiver before you go back to voice — the receiver keeps running after you leave the SSTV screen, and while it does, transmit audio comes from the data port and your microphone modulates nothing. The switch's own hint says so. ⚠️ ** What is not proven is a radio's on-air behaviour when it is held in FM-D between pictures. (#130) This one was reported as already fixed, twice. It was not — see below.
-
The manual is illustrated, and corrected.
Every one of the 22 chapters now carries screenshots — 137 of them, up from 11 — and the pages an operator reaches first were the ones most in need: the first-run wizard, the waterfall, the settings reference. An audit against the running app also turned up a stack of things the manual said that were no longer true: the wizard is four steps and not three, the waterfall takes three mouse gestures and not two, the grid wants all six characters, and the mode chapters described transmit capabilities that several modes do not have. Those are fixed. Where a picture would have had to be staged to exist, there is no picture and the text says what is missing instead.
-
The JS8 station list carries what a DX operator needs, and band activity is listed by offset.
The stations pane gained distance, beam heading, a worked-before tick, the name and comment from your last contact, and a pin for the stations you are watching — JS8Call's own column set, and its own rule that worked-before means any band and any mode, not the band/mode dupe scope. Alongside it, a new pane lists band activity ordered by audio offset with its time delta, so finding a clear slot is a glance rather than a hunt. Both panes are ⊞-hideable like their siblings.
Fixed
-
The JS8 screen was drawing the Tempo screen underneath it.
Opening JS8 rendered the whole Tempo workspace below the cockpit: its Fast and Deep tier buttons, its station roster, its conversation pane, and a second waterfall beside JS8's own. Every other digital cockpit is wired to leave that slot empty and JS8 had been missed, so it fell through to the chat screen's layout. JS8 now shows JS8 and nothing else.
-
Nothing could clear the callsign card in the FT cockpit, and F4 did nothing while you were typing.
Two separate faults behind one report. The card follows whichever station you have open plus whoever the sequencer is working, and no control anywhere put it back to empty — the answer that said F4 did it was wrong, because F4 cleared the DX Call and Grid boxes in the Tx Messages panel, which is a different block. F4 now clears both, and the card comes back on its own the moment it would be about a different station. Separately, F4 was disarmed whenever the cursor sat in a text box, which is exactly when you reach for it; it now fires while you type, the way WSJT-X does. Alt+F4 still closes the window and clears nothing. (#204)
-
The 60 m FT8 dial, and what changed underneath it.
The band button tunes 5.3715, the US channel centred on 5373.0 kHz, and that was reported as the wrong frequency — which it was when the report was filed, because 60 m FT8 lived on 5.357 worldwide. It is not wrong now: on 13 February 2026 the FCC split US 60 m into four 100 W ERP channels (5332.0 / 5348.0 / 5373.0 / 5405.0 kHz) plus the worldwide 5351.5–5366.5 kHz segment at 9.15 W ERP, and eliminated the 5358.5 kHz channel that 5.357 dialled. US FT8 moved to 5.3715 to keep the power. So the band button stays where it is — moving it to 5.357 would drop a US station's legal ceiling by about 10 dB with nothing on screen saying so — and the entry now says which dial is which, and why. Outside the US, and for QRP, 5.357 is still the one you want, and it is a Memories preset alongside it. A 60 m spot on either dial is now recognised as FT8 rather than a bare "Digital", which it was not before. The 60 m notes on the Memories presets were also wrong after the rule change — they described a 100 W channel at 5358.5 that no longer exists — and are corrected. (#175)
-
The diagnostic log's "Always on" now says why, and there is no
--debug.A reply implied the log could be switched off and that a
--debugcommand-line flag existed. Neither is true, and neither is going to be: a log you can turn off is missing on exactly the launch that needed it, and the switch would live in a settings file a failing launch may never reach. The Settings entry now states that rather than leaving "Always on" as a bare assertion, and the troubleshooting guide names the one switch that is yours — Settings ▸ Logging & Connectors ▸ "Extra detail in the diagnostic log", which applies live with no restart — and states plainly that--profileis the only argument Nexus takes. (#101) -
Hound is one click, on the FT8 screen.
The DXpedition Hound mode was a dropdown in the Operate header behind a setting only Settings otherwise wrote; it is now a single Hound button you click on and off while you work. That matters because Hound is a per-DXpedition mode, not a station setting — Nexus drops it at every restart for exactly that reason — and you should be able to enter and leave it in the middle of a session without going anywhere. Toggling it mid-QSO is safe: a contact already on the air keeps the rules it started under, and the button governs the next one. The retired SuperHound option is not offered as a choice, because it never did anything a plain Hound did not.
-
Nexus tells you a DXpedition is running SuperFox before you call it.
SuperFox is a transmission format this version has no decoder for, so such an operation never appears in the decode list and Hound mode does not change that — which, found out mid-pileup, looks exactly like bad propagation. When the DXpedition calendar shows a SuperFox operation on the air, the Operate header now names it beside the Hound button and its card on the DXpeditions board says the same thing: work that one in WSJT-X.
-
Turning Hound off during a QSO with a Fox could strand the contact.
A DXpedition Fox packs two replies into one transmission, and the half confirming you arrives without a sender; Nexus put the Fox's callsign back on it so the exchange could close. That repair was keyed to the Hound switch as it stood right then rather than to the contact, so switching Hound off part way through stopped it: the Fox's confirmation no longer read as one, and you kept calling a station that had already rogered you. The reverse could happen too — switching Hound on during an ordinary contact let a passing Fox's confirmation be read as your partner's. A contact now keeps the rules it started under from end to end.
-
The satellite catalog was publishing years-old orbits for birds that had stopped being tracked.
The mirror took the freshest elements it could find for each active bird — but for a bird that has dropped out of Celestrak's amateur groups the only source left is a SatNOGS cache that never expires, so it kept serving that bird's last recorded orbit, one day staler every day. Forty birds were affected, the oldest carrying an orbit from 2014, and four of them had already re-entered: pointing an antenna at one of those was pointing it at nothing. An element more than 30 days old is no longer published. Those birds keep their row in the satellite list with the status chip that says why, exactly like the ones that have gone silent — a row that reads "no current elements" is honest, a track twelve years out of date is not. Every bird anyone actually works is unaffected: ISS, SO-50, RS-44, AO-91, AO-7, QO-100, PO-101, AO-73 and FO-29 all carry elements hours old.
-
The RTTY auto-sequencer could not log a single Winter Field Day contact.
Winter Field Day uses its own station classes — H (Home), I (Indoor), O (Outdoor), M (Mobile) — but the sequencer matched the ARRL Field Day letters A–F for both events. None of the Winter classes is in that set, so a station sending a perfectly legal
2M EPAnever satisfied the required Class slot: the exchange stalled, the QSO never auto-logged, and the only way through was to log it by hand. The sequencer now copies against the class set of whichever event is selected, so2M EPAcompletes at Winter Field Day and2A WIstill completes at ARRL Field Day. An ARRL Field Day sequencer continues to refuse a Winter class, and vice versa — the two sets are not interchangeable. Scoring, Cabrillo and ADIF output are untouched. -
The Chase pane stopped updating once you selected a station.
Its "open now / best 1400Z" column and the colour accent on each row come from the modelled band outlook, and that was deliberately not refreshed while a station was selected — correct for the map, which switches to a per-station prediction, but it also starved Chase, the Chase feed and the band-outlook heatmap, which use it all the time. Pick a station and their propagation column froze at whatever it last held. It was most visible in a detached Connect window left open on a second monitor, but it happened in the main window too. The outlook now refreshes on its own every minute regardless of what is selected; the map is unchanged.
-
CW and SSB park activations were missing from the roster for digital-only stations.
If the CW or Phone features were off (the wizard's FT-only choice), a POTA or SOTA activator running CW or SSB was dropped as if it were a CW/Phone need — so the Nexus roster showed fewer parks than GridTracker's. A park activation is someone to hunt, not a mode you have to operate: it now stays visible regardless of which mode features are on. Genuine CW/Phone needs are still hidden exactly as before, and a park activator who also happens to be a new entity shows as a park, not as a CW award row at the top of a digital-only board.
-
Your log never recorded which callsign made a contact, so a special event batch uploaded afterwards was signed under the wrong call.
Nexus has always transmitted correctly under a special event or club call — a 1×1 like
W6Ris an entirely ordinary callsign to FT8 — but nothing wrote down which call was in force, and the LoTW and HRD Logbook upload paths filled that in from whatever callsign happened to be set at the moment you pressed upload. Work a weekend asW6R, put your own call back on Monday, then upload, and every contact from the event reached ARRL under your home call: a confirmation the other operator will never get and cannot diagnose. Every contact now carriesSTATION_CALLSIGNin the log and in every ADIF export, stamped with the call that made it, and both upload paths sign from the record rather than from the current setting. Contacts logged by earlier versions carry no station call and upload exactly as they did before. -
The QRZ callbook row in Settings ▸ Connections could never say it was working.
It was wired to nothing, so a paid subscription with a working password and a subscription that had lapsed months ago showed the same grey "not verified yet" row, and the only way anyone found to clear it was to push a QSO through. Every callbook lookup now updates that row, and so does the Test connection button on the QRZ Logbook row — which also could not clear its own dot before. Because the row is a lookup and not an upload it now says "last lookup" rather than "last upload". A lookup QRZ refuses is kept apart from a callsign QRZ simply does not have. QRZ answers both the same way — on a live session, with an error and no record — so a refused lookup was being counted as a completed one: the row went green, and green does not just fail to warn, it clears a red that was already there. A match and a genuine miss now read as a working subscription; a refusal and a session QRZ will not accept read as failures, each naming what to go and check. Telling those two apart reads QRZ's own not-found wording, and it now has to be QRZ's not-found reply rather than any refusal that happens to use the words — a subscription-level refusal worded "callsign not found at your subscription level" was being read as a plain miss, which turned the row green.
-
Cloudlog and Wavelog threw away the reason your upload was rejected.
The instance sends back what it actually refused — a read-only API key, a station profile id not linked to that key, a missing ADIF field — and Nexus replaced all of it with "auth rejected — check the API key", which sent people to check the one thing that was usually fine. The instance's own words now lead in the upload result, and where it said nothing the guess names the station profile id as well as the key. A failure to reach the instance at all is no longer blamed on the URL either: an antivirus or company proxy inspecting HTTPS traffic is named as the likely cause, which is what it usually is.
-
A logbook service's own error text is no longer kept in your config directory.
The Connections panel's failure line was stored word for word in a file in your config directory, and for Cloudlog and Wavelog that text can contain your API key — the key travels inside the upload request, and an instance running in debug mode, or a proxy or firewall page in front of it, answers by quoting the request straight back. What gets stored is now always Nexus's own sentence, naming the failure and what to check. The service's exact words are still shown to you, in the upload result and in this session's connection log; they are simply not kept. The same rule now covers HRDLog.net and World Radio League, whose replies were being stored with no length limit at all. It covers the diagnostic output as well. On Linux the desktop session captures a program's console output into a world-readable file in your home directory, and the connection log was echoing every service reply there verbatim — so the same API key was landing in a second file, kept until your next login. The console line now names only the connector and whether it succeeded, and the ClubLog most-wanted fetcher, whose request carries your ClubLog key in its URL, no longer echoes ClubLog's refusal there either. And a file that already had a key in it from an earlier version is cleaned the next time Nexus starts: a stored failure line that this version could not have written is dropped, and the row keeps the time it failed.
-
A logbook service's own error text is no longer written into your log file — or uploaded to ARRL with it.
When a QRZ Logbook or ClubLog upload bounced, the service's reply was stored on the contact itself, inside
log.adi. Both services answer a rejected upload by quoting the request back, and the request carries your API key — so the key was written into the logbook, and from there it rode out through every export built on it: the batch TQSL signs with your callsign certificate and uploads to LoTW, the per-contact eQSL upload, and any range or operator export you make. That is a secret leaving your machine under your own signature, and it cannot be recalled. What the contact now records is Nexus's own description of why the upload failed — the credentials were refused, the record was refused, TQSL signed only part of the batch — and never the service's words. LoTW's own tool is covered by the same rule. You still see exactly what the service said: it is in the upload result and in this session's connection log. A log file that already carries one of these replies is cleaned the moment Nexus reads it, which is before any export can quote it — the contact keeps the fact that its upload bounced, and when, and loses only the text. And so are the safety copies of it. Nexus keeps alog.adi.bakbeside your log — taken the first time it opens a log and never overwritten — plus dated snapshots in abackupsfolder. Those copies are made from the raw file, so on any machine that ran an affected version they hold the same replies, and nothing ever rewrote them. Upgrading now cleans them too: the copies keep every contact, they lose only the reply text, and any copy that held one is rewritten so only your own account can read it. Your log itself is cleaned once, in place, the first time this version opens it. A log with nothing to clean is not rewritten at all, so a large log still opens as fast as it did. -
The automatic QRZ Logbook sync no longer prints QRZ's reply where the desktop can keep it.
When the hourly sync failed, Nexus printed QRZ's own refusal line — which on Linux the desktop session saves to a world-readable file until your next login, and QRZ quotes your failing request, API key included. The failure now says what went wrong in Nexus's own words: no key stored, the sync never reached QRZ, or QRZ refused it. QRZ's own wording still reaches you, in Settings ▸ Connections, and it goes no further than that session.
-
A LoTW upload that TQSL turns down now says what TQSL said, and which of two problems it was.
The reason on the contact is Nexus's own sentence — the service's wording is not written into your log — and that sentence sends you to this session's connection log to read what TQSL actually printed. For LoTW it was not there: every other connector wrote its line, LoTW wrote none, so a failed sign left the real message only in the toast of the click that produced it. It is now recorded for every failing upload, automatic or manual. And a batch that fails to sign no longer reports "check your credentials" for two problems with nothing in common: a missing or expired Callsign Certificate (request or renew it at ARRL, load the .p12 into TQSL) is now told apart from an unusable Station Location (create it in TQSL, or correct the name in Settings).
-
A QRZ callbook row no longer claims a working subscription it cannot prove.
The row went green on far too little. QRZ answers "no such callsign" and "your subscription will not do that" in exactly the same shape, so several versions of this check tried to tell them apart by reading QRZ's wording, and each was beaten by the next wording — one of them still read "Not found: your subscription does not cover this record" as a plain miss and marked the connector green, clearing a real failure. The wording was never the real problem. When your XML subscription lapses, QRZ does not send an error at all — it sends a successful lookup with the free-tier fields in it: the callsign, a name, a country, and no grid square and no state, because those are what the subscription buys. Every version of this check saw a record come back and called the subscription healthy, so the row read green for operators who were no longer subscribed to anything. So the check was turned around. The row goes green only on positive proof — a lookup that comes back carrying a subscriber-only field, read from the returned record itself and not from QRZ's surrounding prose — and it now also believes QRZ's own subscription field: a reply that says
non-subscriber, or whose subscription expired, reads red no matter what else it carries. Every other answer, including shapes QRZ has not sent before, reads as not confirmed. Two trades, both deliberate: looking up a callsign that genuinely does not exist marks the row as failing until your next successful lookup, and so does a lookup whose record happens to carry neither a grid nor a state. A row that reads red until your next lookup costs you a glance; a row that reads green over a subscription that has quietly lapsed is the bug this was reported as. <!-- ℹ️ NICE-TO-KNOW, no longer load-bearing (round 7 pass 2): green would rest on QRZ withholding <state> from a non-subscriber (<state> is FCC-ULS-derivable, so QRZ MIGHT return it to a free US account) — EXCEPT the predicate now also reads QRZ's own <SubExp> and refuses an explicitnon-subscriber/expired reply regardless of which fields it carries. So if QRZ ever does return <state> free, the SubExp disqualifier is what catches it (the reply carries <SubExp>non-subscriber</SubExp>), and #245 cannot return through that door. Proven bya_non_subscriber_body_carrying_a_real_state_is_disqualified_by_subexpin qrz.rs. The only residual — a free reply with <state> and NO SubExp marker at all — is worth confirming, not the single assumption the fix rests on. CHECK when convenient: one lookup from a lapsed/free QRZ XML subscription — does the <Callsign> carry <state>, and does the <Session> carry a <SubExp>? Do NOT use live credentials in code/CI. --> -
A manual HRDLog.net or World Radio League push that never reached the service now records it.
Pressing the per-contact upload button and getting a network failure left the Connections row saying "stored — not verified yet" forever, while the automatic upload recorded the identical failure — so the button an operator presses because the row has never been verified was the one that could not change it.
-
A Cloudlog failure now says which failure, after a restart too.
Every way an upload could fail collapsed into one stored sentence, so once the connection log had gone with the session the Connections row could not tell a station profile id that is not linked to your key from a URL that is not a Cloudlog instance from the instance being down. Each of those now has its own line on the row, naming what to go and check — the same way the HRDLog.net and World Radio League rows already worked.
-
A connector row could read green in the second it failed.
The panel compares when a connector last worked with when it last failed, and those are stamped to the whole second — so a QSO that uploaded and then failed inside the same second read as working, and stayed that way until the next failure. The auto-upload worker pushes to every service back to back with no spacing, and a Cloudlog instance on your own network answers in milliseconds, so this was reachable. A tie now reads as failing.
-
A LoTW report that failed to download said which step failed, not what went wrong.
Every failure reading the response became "could not read the response body", so a request that died after a minute and one that never started looked identical. It now says which happened and how long it waited — and where it genuinely cannot tell (LoTW still assembling the report, or a report too large to arrive inside the deadline) it says both instead of guessing, and tells you to narrow the date range if it keeps happening.
-
An upload code pasted with a stray space or newline made every upload fail, silently.
Copying an HRDLog.net upload code, a QRZ Logbook API key or a World Radio League key out of a web page usually brings a trailing newline with it. Nexus stored it exactly as pasted and sent it that way, so the service rejected every QSO and nothing on screen explained why — the code looked right because it was right, apart from a character you cannot see. Codes are now trimmed when you save them. Existing stored codes are unaffected until you re-enter one.
-
settings.json is no longer world-readable.
It holds your ClubLog API key (and, briefly, a Cloudlog key on its way into the OS keychain), but was written so any account on the machine could read it. It is now owner-only. Existing files are tightened the next time Nexus saves them.
-
Upgrading no longer destroys a stored Cloudlog key when the OS keychain is unavailable.
The one-time move of a legacy Cloudlog/Wavelog key from settings.json into the keychain cleared the key from the file whether or not it had actually been stored — so on a Linux box with no Secret Service running, the key was silently deleted at launch. It is now cleared only once it is safely in the keychain, and the migration retries on a later launch otherwise.
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