fix(otasim): two-GUI handshake completes + sim-mode QoL fixes#29
Merged
Conversation
1. --monitor-audio flag for ultra_gui
Add `--monitor-audio` (+ optional `--monitor-device <name>`) to
ultra_gui. When -sim is on, opens an SDL output device and tees the
received OTASim RX audio into local playback. Restores the "feels
like a real sim" experience that the in-process loopback used to
provide before the OTASim refactor.
Queue depth is capped at ~500ms (24000 samples / 96000 bytes); on
overflow we clear and resync rather than letting latency build.
2. Short-circuit PTT/Hamlib in sim mode
In sim mode there is no radio, so the Hamlib serial-open spam ("port
open: serial_open(/dev/cu.usbserial-…) status=-6") that polled every
~1.5s when a CAT preset was saved in settings was pure noise.
ensurePttReadyLocked() and updateWaterfallFrequencyDisplay() now
bail early when simulation_enabled_ is set, releasing any existing
ptt_driver_ and skipping the rig open / frequency poll.
3. --log-file actually captures the whole session
initLog() in app.cpp was unconditionally calling
ultra::setLogFile(g_gui_log_file) after opening logs/gui.log, which
clobbered whatever --log-file had set in main_gui.cpp. The result
was 12 lines of modem init in the chosen log file, then everything
silently rerouted to logs/gui.log. initLog() now adopts ultra::g_log_file
if it is already set externally, so --log-file captures every
category for the entire session.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three independent bugs were compounding to block the OTASim two-GUI
connect handshake. With the fix, ALPHA + BRAVO ultra_gui instances
pointed at a single ota_simulator serve daemon now complete the full
PING / PONG / CONNECT / CONNECT_ACK / MODE_CHANGE / CONNECTED sequence
and exchange OFDM-CHIRP DQPSK R1/4 ACK traffic in-session.
1. OtaAudioBackend client RX cap was 20 s
kMaxRxBufferSamples = 960000 at 48 kHz allowed up to 20 seconds of
audio to silently accumulate when the GUI render loop briefly
stalled (waterfall spike, ImGui frame). Real audio frames then sat
behind multi-second silence, well past the ARQ window and the
modem's sync-search budget. Lowered to 23040 samples = 480 ms,
which mirrors a real soundcard's bounded driver buffer (continuous
in, consumer drains at real-time rate, oldest drops on stall).
~8x a 60 Hz render budget so typical jitter is absorbed without
building multi-second latency.
Also added opt-in `#ifdef ULTRA_OTASIM_AUDIO_DIAGNOSTICS` queue-
depth counters in OtaAudioBackend for future similar work. Off
by default.
2. Modem callsign defaulted to 8P9QC even in -sim mode
GUI's `Connect to <remote>` passes the OTASim --station-id as the
destination callsign in the frame header. The modem's local
callsign was loaded from settings.callsign (default "8P9QC") so
ALPHA's modem saw frames addressed to "ALPHA" with local_call=
"8P9QC", classified them as "different station", and dropped them
silently at TRACE level. LDPC reported 3/3 CWs decoded OK; the
frame never reached deliverFrame()'s INFO log and never reached
the protocol layer.
In -sim mode with a non-empty --station-id, force the modem's
local callsign and protocol callsign to the station id so peer-
addressed frames are accepted.
3. --log-file was overridden by initLog()
App::initLog() unconditionally called ultra::setLogFile(g_gui_log_file)
after opening logs/gui.log, clobbering whatever main_gui.cpp had
set from --log-file. The chosen per-station log file captured ~12
lines of modem init, then went silent for the rest of the session.
initLog() now adopts ultra::g_log_file if it is already set
externally, so --log-file captures every category for the full
session.
Test gate:
ctest --test-dir build -R "Otasim|UltraGuiOta|UltraTncSimAudio|SessionContext"
--output-on-failure -j1
→ 3/3 (or 4/4) pass; the new latency / soundcard-cap assertions in
test_ultra_gui_ota_client.cpp validate idle backlog and round-trip
latency on the in-process passthrough.
ACK diversity (each ACK sent ~2x ~440 ms apart) and CONNECT_ACK
rescue retry (one proactive resend) are intentional HF robustness
mechanisms. They are visible-but-harmless on OTASim's clean AWGN; the
receiver's SR-ARQ correctly de-duplicates at the base/bitmap level.
See selective_repeat_arq.cpp:1266 comment. Documented in CHANGELOG.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Two-station
ultra_gui -siminstances pointed at a singleota_simulator servedaemon now complete the fullPING → PONG → CONNECT → CONNECT_ACK → MODE_CHANGE → CONNECTED handshake
and exchange OFDM-CHIRP DQPSK R1/4 ACK traffic in-session.
Five bugs fixed across two commits:
`f0ad88c` — sim-mode quality-of-life:
local SDL output device so the simulator sounds like a real radio.
`port_open: serial_open(...) status=-6` spam when no radio is
attached.
(previously `initLog()` clobbered the destination after ~12 lines).
`1f215a7` — handshake completes end-to-end:
real soundcard's bounded driver buffer; consumer stall drops oldest
instead of accumulating multi-second latency.
inbound frames addressed to ALPHA/BRAVO aren't dropped by
`deliverFrame()` as "different station".
ACK diversity (each ACK repeated ~2x ~440 ms apart) and CONNECT_ACK
rescue retry (one proactive resend) are intentional HF robustness
features, visible-but-harmless on OTASim's clean AWGN. Documented in
CHANGELOG with link to the in-source rationale.
Test plan
state 3 (CONNECTED), MODE_CHANGE negotiated, in-session ACKs
decode in OFDM control profile
OTASim client backend cap and GUI sim-mode wiring changed
🤖 Generated with Claude Code