Files
OpsLog/internal
rouggy a81125eab1 fix(icom): a set whose acknowledgement is lost is sent once more
Extends to frequency and mode what PTT already had. A missing FB is not a
missing command: the rig acts on the frame as it decodes it, and what
expires is our wait for the answer, on a bus shared with the rig's own
transceive updates. JTDX in "Split: Fake It" moves the dial and the mode
immediately before every key-down, so those acks queue behind each other.

Losing one is fatal to the client rather than merely untidy: rigctld answers
RPRT -9, JTDX reads that as losing rig control and tears the connection down
mid-over. An operator's log shows three set_freq failures and one set_mode,
each followed within 300 ms by a fresh rigctld client -- and shows the PTT
resend rescuing an over that would otherwise have ended there.

Opt-in per caller rather than folded into exec: only a command that says
"be in this state" can be repeated safely, and a relative one must never
come through here.

The acknowledgement loss itself is still unexplained. Every failure in that
log is preceded by a state read reporting SSB on a rig in DATA, which points
at CI-V frame desync rather than a slow rig, and needs a trace to pin down.
2026-08-17 16:23:54 +02:00
..
2026-08-02 20:03:16 +02:00
2026-07-03 19:08:50 +02:00
2026-07-04 11:39:33 +02:00
2026-08-17 13:18:32 +02:00
2026-08-17 13:18:32 +02:00
2026-08-02 23:51:46 +02:00
up
2026-06-15 23:45:14 +02:00
2026-06-07 01:11:37 +02:00
2026-06-11 21:54:35 +02:00
up
2026-06-15 23:45:14 +02:00
2026-07-09 17:32:13 +02:00
2026-08-02 06:40:10 +02:00