Mostly a repair release, and an important one if you run Hamlib or a Kenwood. Three separate faults introduced in v1.9.23 are fixed here: dead S-meter, SWR and ALC readings on Hamlib rigs; the PTT and HALT buttons vanishing from ECHOCAT a few seconds after connecting; and VFO B being unusable on Kenwood radios. If you are on v1.9.23 and any of that sounds familiar, this is the update.
Hamlib: your meters are back, and so is PTT on the phone
One bug caused both. When a radio's Hamlib backend refuses a command — and plenty refuse the VFO query, the IC-706MKIIG and IC-7300 among them — that refusal was cancelling every other reading still on its way back from the radio. Two things followed.
S-meter, SWR and ALC went dead, because the meter reading arriving behind the refusal was thrown away (thanks N4RDX, whose log showed a poll running perfectly with nothing to show for it).
Worse, the discarded VFO reply was then mistaken for the radio's mode. The desktop told every client the rig was in a mode called "VFOA", and because both the phone and the web client only offer PTT and HALT on voice modes, those two buttons quietly disappeared — while FT8, which doesn't check, kept transmitting normally. LZ3AW reported it as "the PTT button disappears after a few seconds," and it took a full trace from the ECHOCAT side to connect the two symptoms. Both are fixed at the root: replies are now matched to the command that actually answered them, a query a radio doesn't support is asked once and then dropped for the session, and a VFO name can no longer be seated as a mode anywhere in the stack.
Kenwood: VFO B works, and Split reports for the first time
LZ3AW again, on a TS-480. Switching to VFO B showed B's frequency for about a second and then snapped back to A's while the radio stayed on B. The radio's status reply was being read one character off, so POTACAT was looking at the scan-status digit instead of the VFO digit and the tone digit instead of the split flag — both of which sit at zero on a normal radio. The result was a receiver that reported "VFO A, no split" no matter what the rig was doing.
VFO B now sticks, the A/B swap button goes the direction you expect rather than always toward B, and the Split indicator reflects the radio's real state instead of never lighting at all.
The VFO display follows a QSY immediately
N2FSM QSY'd through spots on an IC-7300 and watched the radio move, the callsign and park text update, and the VFO window sit on the first frequency of the session. On CAT connections the VFO display had exactly one source — the frequency poll echoing the new value back — so it was only ever as current as the poll. And the poll is not always running: it pauses after twenty minutes of inactivity, and it slows on a busy serial link. Tuning now updates the display straight away, with the poll still correcting it if the radio lands somewhere else.
JTCAT: the waterfall that went blank overnight
Leave POTACAT running for a few hours, open the JTCAT popout, and you could get perfect decodes with a permanently black waterfall. Opening the popout while the engine was already running let audio pile up faster than the new window could acknowledge it, and the flow-control that exists to protect a struggling window would deadlock instead of recovering. It now forgives a stall and resumes, so the waterfall comes back within a couple of seconds.
Two more audio fixes alongside it. Icom network and RS-BA1 users had no popout waterfall at all — audio was only ever reaching the main window. And WSPR-decode-while-idle was cancelling itself about a quarter of a second after starting: the radio dutifully moved to the WSPR frequency and decoded away with no window to show for it, which is why it looked like it had never run.
DX cluster: no more hammering the node
KD4D runs a DXSpider node and found POTACAT reconnecting to it every ten seconds, from more than one machine. Two separate places were resetting the backoff timer, so the delay between attempts never actually grew. Reconnects now back off properly, give up after repeated failures rather than retrying forever, and POTACAT no longer ships with anyone's server as a default — you choose the node.
Preamp and attenuator reach all their steps
KB2UXB's FT-710 could only ever reach the first position: preamp went off or AMP1, attenuator off or 6 dB, with 12 and 18 dB unreachable. Both were being treated as simple on/off switches when most radios offer a ladder. The buttons now step through the positions your radio actually has, and on Hamlib connections the list comes from the radio itself.
JTCAT additions
Waterfall speed. The waterfall runs at one line per frame, which holds barely a second of history and turns every signal into a long vertical worm. You can now slow it down in the popout's gear menu — at lower speeds each transmission becomes a compact block and whole periods fit on screen, much closer to what WSJT-X shows and far better for judging how busy a band is. Frames between drawn lines are averaged rather than dropped, so slowing it down doesn't lose weak signals.
S-meter and SWR in the popout, beside the RX and TX sliders, so you don't need the VFO window open to see what the radio is hearing or looking into.
An ATU button on the popout's bottom bar, next to Log.
RX level is remembered per band. The level that hears 17 m isn't the level that hears 20 m; switch bands and the slider returns to where you last had it there.
Band-activity filters persist. CQ/73, Wanted, Chase and sort-by-signal survive closing and reopening the popout.
A per-QSO transmission cap that counts correctly. Setting "give up after 5 tries" gave up after 3, because periods where nothing was transmitted were counting against it (thanks KQ4MHD).
Also: the mode chip no longer shows an "AP" badge on decodes, and the Skip button moved down to sit with the QSO's message ladder.
Logging: the Log QSO window gets the spot
If you keep the standalone Log QSO window open — Ctrl+L, the one W9TEF keeps on a second screen — clicking Log on a spot now fills that window and brings it forward, instead of opening a second log form on top of it. Spots carrying more than one park program, multiple park references, or a multi-operator selection still open the full in-window form, which is the only one that can hold them.
ULTRACAT
For those who have found it: Run now stops calling into a dead band. When nobody has answered for a while and there are no new stations calling CQ, it pauses and listens, then picks CQ back up the moment someone workable appears. It also no longer re-works stations already in your log on that band and mode, which is what made "worked out" a state it could recognise at all.
New alongside it, at KQ4MHD's request: Hunt can now call CQ for you when there's nothing to hunt, and hands straight back to hunting when the band wakes up. Off by default, in the popout's gear menu. The thirty-minute attended-operator limit still applies throughout — this is not unattended operation.
Also in this release
- Windows notifications and taskbar grouping now identify POTACAT properly.
- The memory diagnostic names each window, so a memory report is readable.
- FT4 replies to portable and special-event callsigns encode through the native path, and an encode that fails now says so instead of appearing to call forever.