OBD-II signals renamed obd_*, plus timing, trims, O2; GNSS signals; elm_connected; reboot command
/ crosscheck (push) Successful in 1s
Rust consumers (L4T) / Clippy, test, build, glibc for every service that compiles contracts in (push) Failing after 6s

Decision 30: CAN feeds displays, OBD-II (K-line via the ELM) calibrates, so every
obd2 signal carries the obd_ prefix and the plain names are left for decoded ones.
yukart/gnss/<signal> gives displays the fix as single values; gnss/fix stays for
navigation. Boundary: reboot beside poweroff, same guards.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DDwfd6Ki8UfG4hK6W5YpJz
This commit is contained in:
2026-10-13 22:50:03 +09:00
parent 5feba8b354
commit 229d796cfd
4 changed files with 221 additions and 90 deletions
+1 -1
View File
@@ -9,7 +9,7 @@ type: guidelines
- ⚠ **The unit appears in both the topic name and the payload, so they can
contradict each other.** Generating both from one entry is what stops that; a
hand-written `"unit"` reintroduces the failure it was meant to prevent.
- Signal names carry their unit (`vehicle_speed_kph`, `coolant_temp_c`), so a unit
- Signal names carry their unit (`gnss_speed_kph`, `obd_coolant_temp_c`), so a unit
mismatch is visible at the call site rather than one registry lookup away.
## 2. Two clocks, and they answer different questions