Repository navigation
Capture and persist four circuit gates safely on device #108
Description
Parent: #105
Objective
Configure all four gate lines from validated GNSS positions while stationary and persist a dedicated versioned JSON track file.
Scope
- Select circuit layout, gate, and left/right endpoint in Setup.
- Capture the current accepted GNSS fix only while stationary and outside active timing.
- Show accuracy, fix age, captured coordinate, completion state, and a line preview/summary.
- Require both endpoints and semantic validation before Save.
- Atomically write one JSON file per layout through a storage adapter outside the UI/parser.
- Support Cancel, overwrite confirmation, storage failure, and interrupted-write recovery.
- Keep prebuilt build-time JSON deployment and host import/validation supported.
- Do not require or offer interaction while driving.
Acceptance criteria
- A user can capture, validate, save, reload, and select all four lines without rebuilding firmware.
- Invalid/partial capture never replaces a valid definition.
- Active or moving capture is rejected visibly.
- Simulator fixtures cover success and every failure state.
Related: #37.
Activity
PurpleSentinel commentedon Oct 16, 2026
Reopening this question rather than the issue, for whoever decides.
This closed in August on the criterion "a user can capture, validate, save, reload, and
select all four lines without rebuilding firmware". That is not met on the device and
never has been:
There is no menu entry, no screen root and no routing on the hardware, so the screen cannot
be opened at all. Everything the issue specified was built and tested; the device just
cannot reach it - the same shape as #18, which was closed while nothing in the tree called
the runtime it had delivered.
The capture logic itself is sound, including the two parts that matter most: it sets
geometry_status = device_capturedon save, andSdTrackStore::write_atomicwrites intothe separate
device-capturedpack. So this is wiring, not rework.#204 now carries the remainder. Whether this reopens or stays closed with #204 owning it is
a judgement call - flagging it rather than making it, since the delivered work here was real.
PurpleSentinel commentedon Oct 21, 2026
Staying closed, deliberately, with #204 carrying the remainder.
For the record: this issue's acceptance criterion — "a user can capture, validate, save, reload, and select all four lines without rebuilding firmware" — was not met when it was closed in August and was not met on the device until #205 made the capture screen reachable at all. It is still not fully met: a capture has yet to be made at a circuit.
Rather than reopen and have two issues describing the same unverified capture, #204 holds the remaining criteria, the wiring (#205) and the screen corrections (#232).