makiolaj/clockula
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 3 Wiki Activity

docs: point the reboot repair at the milestone that actually owns it

Browse Source 
The roadmap puts the BOOT_COMPLETED and TIME_SET receivers, the ringing
service and the full-screen intent in M3; ARCHITECTURE.md credited them
to M6 and M7, which are Timers and Stopwatch.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
makiolaj
2026-09-11 13:53:11 +02:00
co-authored by Claude Opus 5
parent efd88e3060
commit 971ee4f7a3
1 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/ARCHITECTURE.md
+7 -6
View File
@@ -267,7 +267,7 @@ consulting the wall clock to decide staleness — would let a user moving the
system clock warp a running timer, which `PLAN.md` §5 forbids outright. system clock warp a running timer, which `PLAN.md` §5 forbids outright.
Closing it properly needs a persisted boot identifier, which is not in this Closing it properly needs a persisted boot identifier, which is not in this
schema. **M6's `BOOT_COMPLETED` receiver is the authoritative repair:** it runs schema. **M3's `BOOT_COMPLETED` receiver is the authoritative repair:** it runs
at the top of every boot, before the new uptime can climb past any stored at the top of every boot, before the new uptime can climb past any stored
anchor, and rewrites every running timer from `endsAtWallClock`. anchor, and rewrites every running timer from `endsAtWallClock`.
@@ -299,9 +299,10 @@ timer, and inherits the same blind spot: once the new boot's uptime passes the
stored `startedAtElapsedRealtime`, the reboot goes unnoticed and the run reports stored `startedAtElapsedRealtime`, the reboot goes unnoticed and the run reports
`accumulated + (elapsedRealtime - startedAtElapsedRealtime)` as **live and not `accumulated + (elapsedRealtime - startedAtElapsedRealtime)` as **live and not
stale** — a segment it never actually ran. So the discarding above is what stale** — a segment it never actually ran. So the discarding above is what
happens when the reboot *is* detected, not a guarantee; until M6's happens when the reboot *is* detected, not a guarantee; until a
`BOOT_COMPLETED` receiver pauses the run at boot, a stopwatch that spanned a `BOOT_COMPLETED` receiver pauses the run at boot — the receiver itself arrives
reboot can show a fabricated segment. in M3, the stopwatch's own repair in M7 — a stopwatch that spanned a reboot can
show a fabricated segment.
--- ---
@@ -444,8 +445,8 @@ metadata holder service.
| `AlarmScheduler`, the exact-alarm plumbing, snooze state (schema **v2**, with a tested migration) | M3 | | `AlarmScheduler`, the exact-alarm plumbing, snooze state (schema **v2**, with a tested migration) | M3 |
| Any UI, ViewModel or navigation beyond the theme | M4+ | | Any UI, ViewModel or navigation beyond the theme | M4+ |
| Alarm edit surface, ringtone picker, per-alarm override UI | M5 | | Alarm edit surface, ringtone picker, per-alarm override UI | M5 |
| `BOOT_COMPLETED` / `TIME_SET` receivers — the authoritative repair after a reboot | M6 | | `BOOT_COMPLETED` / `TIME_SET` receivers — the authoritative repair after a reboot | M3 |
| The ringing foreground service, full-screen intent, notifications | M6/M7 | | The ringing foreground service, full-screen intent, notifications | M3 (timers reuse its audio path in M6) |
| Stopwatch presentation, best/worst lap analysis | M7 | | Stopwatch presentation, best/worst lap analysis | M7 |
| ICU city and zone display names, offsets, day differences | M8 | | ICU city and zone display names, offsets, day differences | M8 |
| The `android.provider.AlarmClock` intent surface and its hostile-extra validation (`TimeOfDay.clamped` and `Zones.normalise` exist so M9 has something to call) | M9 | | The `android.provider.AlarmClock` intent surface and its hostile-extra validation (`TimeOfDay.clamped` and `Zones.normalise` exist so M9 has something to call) | M9 |