Conversation
…rtSpin2k service UUID remains present in the legacy scan response.
|
@codex review |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: a460d31175
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| int logBufLength = ss2k_log_hex_to_buffer(pData, length, logBuf, 0, kLogBufCapacity); | ||
| int port = 0; | ||
|
|
||
| ergMode->onFTMSCommand(static_cast<uint8_t>(rxValue[0])); |
There was a problem hiding this comment.
Calling onFTMSCommand() for every opcode causes StartOrResume to clear resumeErgOnCadence, even though that command only updates training/status state and does not select another control mode. If an app sends Start/Resume while the rider is still below the cadence threshold, subsequent cadence recovery leaves the trainer in simulation mode instead of restoring the latest ERG target; restrict cancellation to commands that actually change the control mode, or explicitly re-arm recovery on Start/Resume.
AGENTS.md reference: AGENTS.md:L300-L307
Useful? React with 👍 / 👎.
…e and update related logging; adjust minimum cadence threshold.
…ast set point for improved responsiveness.
|
@codex review |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 71ceed871e
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| } | ||
|
|
||
| if (rtConfig->cad.getValue()) { | ||
| if (rtConfig->cad.getValue() > MIN_ERG_CADENCE / 2) { |
There was a problem hiding this comment.
When a cadence sensor reports 0–10 RPM at a stop, this gate skips computeErg(), so the new fallback changes only watts.target and never updates targetIncline; the stepper therefore remains at the previous high-load position. If the first resumed sample is already above 20 RPM, lines 87–90 restore the old watt target before any low-load position is commanded, leaving the rider in the ERG “black hole” this change is intended to prevent. Apply the low-cadence position adjustment even for stopped-cadence samples rather than limiting it to the 11–19 RPM window.
AGENTS.md reference: AGENTS.md:L405-L412
Useful? React with 👍 / 👎.
| if (rtConfig->getFTMSMode() == FitnessMachineControlPointProcedure::SetTargetPower && rtConfig->cad.getValue() <= MIN_ERG_CADENCE) { | ||
| if (rtConfig->watts.getTarget() != userConfig->getMinWatts()) { | ||
| SS2K_LOG(ERG_MODE_LOG_TAG, "Cadence below ERG minimum; lowering target to %dw", userConfig->getMinWatts()); | ||
| lastSetPoint = rtConfig->watts.getTarget(); |
There was a problem hiding this comment.
When the rider shifts while cadence is below the threshold, FTMSModeShiftModifier() applies the shift to the temporary minimum target; on the next loop this assignment replaces the original saved setpoint. For example, a 200 W target falls back to 50 W, one shift raises the public target to 60 W, and this stores 60 W so recovery resumes at 60 W rather than the expected 210 W. Low-cadence ERG shifts need to update the deferred target instead of overwriting it from the temporary fallback value.
AGENTS.md reference: AGENTS.md:L558-L564
Useful? React with 👍 / 👎.
1 participant
See Commits.