fy59/trainlog
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Packages Wiki Activity Actions

docs: record sleep persistence and agenda evidence

Browse source
This commit is contained in:
fy59 2026-09-24 06:52:13 +02:00
parent a60e1aa056
commit 4cb911d5c4
4 changed files with 50 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
CHANGELOG.md
View file

@ -45,6 +45,20 @@ also refined.
`trainlog.perf` deployment and read-only Firefox smoke validate the latest
real HR night and the latest real night containing medications from code
commit `2ce458d`.
- Reused that same read-only fallback projection in the multi-night Sleep
agenda, with a visibly estimated hatched interval and no Sleep write. The
agenda now derives an exact 18:00-to-18:00 hourly grid, presents at least
two-hour labels on ordinary desktop widths, expands to every hour when wide,
and reduces to six-hour anchors on narrow screens. A read-only real-device
investigation of the 2026-09-23 night found no medication intake in any
retained local revision and established that later synchronization only
published the already-empty current revision. Android regression coverage
now proves two pre-bed intakes, including quantity two, through repeated
repository reopen, bedtime, wake, final get-up, generation, desktop ACK,
replay, stale-ancestor input and a subsequent revision. The private service
serves Web code commit `a60e1aa`; a non-mutating Firefox smoke on the real
night confirms the identical 22:50–04:04 agenda/detail projection and the
responsive 13/25/5-label axis densities.
## 0.1.5 — 2026-09-23

18
docs/current_state.md
View file

@ -169,7 +169,7 @@ parallel implementation of its rules.
| Sleep tonight readiness delta | `TRAINLOG_SLEEP_TONIGHT_CLOSEOUT_V1=PASS`: Meson 111/111, import checks 6/6, Android `testDebugUnitTest` plus `assembleDebug`, Web 191/191 plus typecheck/build, exact V2 quantity replay, private vc47 schema-v32 migration, fresh CYCPLUS BPM and non-mutating Android/Web visual review. |
| 0.1.5 validation | `TRAINLOG_0_1_5_VALIDATION_V1=PASS`: native 111/111; Android unit-test task and APK build; Web 191/191 plus typecheck/build; JSON/import, C-format and diff checks pass. Private vc47 preserves real data through schema v32, renders pre-bed medication quantity controls, reconnects the CYCPLUS and presents the multi-night Web Sleep agenda. |
| Tomorrow session readiness | `TRAINLOG_TOMORROW_SESSION_CARDIO_READINESS_V1=PASS`: the private vc46 is clean on schema v31 with fresh CYCPLUS BPM and the next Program session is unambiguous; one combined regression proves two ordered occurrences loaded from the synchronized Program, Program provenance, factual timeline/HR ownership, finalization, full generation, desktop import, replay and correlated ACK |
| Current operational cursor | 0.1.6 Sleep graph overlays and post-review Web corrections implemented, fully validated, and served by the private `trainlog.perf` service from development code commit `2ce458d`; immutable stable v0.1.5 remains unchanged. |
| Current operational cursor | 0.1.6 Sleep graph overlays, shared agenda estimate and denser responsive agenda axis validated and privately served from Web code commit `a60e1aa`; immutable stable v0.1.5 remains unchanged. |
| Complete synchronization gap contract | Frozen dependency contract; operational USB/Drive slices required by v0.1.2 are delivered |
| Isolated synchronization test environment | `TRAINLOG_SYNC_TEST_ENV_V1=PASS/FROZEN` |
| Synchronization characterization | `TRAINLOG_SYNC_CHARACTERIZATION_V1=PASS/FROZEN` |
@ -195,8 +195,11 @@ full-hour ticks between exact visible bounds. Medication presentation retains
unit dose and quantity as separate facts. A missing declared sleep interval
uses only a labeled, non-medical Web estimate from Couché plus 45 minutes to
Levé minus 10 minutes when both bounds exist; it never changes canonical Sleep
data. The overlays remain available without an HR capture and do not fabricate
BPM.
data. The selected-night graph and the period agenda consume the same
projection. The agenda renders an estimated interval with distinct hatching,
never substitutes it for a factual interval, and retains the exact shared
18:00-to-18:00 geometry. The overlays remain available without an HR capture
and do not fabricate BPM.
Language is device-local presentation state. Android keeps it in dedicated
SharedPreferences; desktop keeps it in the XDG configuration file
@ -341,9 +344,12 @@ cluster. Program progress uses the shared success color while retaining its
factual completed/total text. Responsive charts preserve their viewBox aspect
ratio, and Analyse explicitly distinguishes unavailable exercise timing from
partial measured coverage instead of deriving time from sets or repetitions.
The shared Sleep agenda owns one label/track/facts grid. Its five clock labels
and every event, interval and medication marker use the same coordinates:
18:00 at 0%, 00:00 at 25%, 06:00 at 50%, 12:00 at 75%, and 18:00 at 100%.
The shared Sleep agenda owns one label/track/facts grid. Its 25 exact hourly
positions and every event, interval and medication marker use the same
coordinates: 18:00 at 0%, 00:00 at 25%, 06:00 at 50%, 12:00 at 75%, and 18:00
at 100%. Ordinary desktop widths show the 13 two-hour labels, wide layouts show
all hours, and narrow layouts retain the five six-hour anchors without changing
the underlying geometry.
The Equipment route projects supplied and custom definitions with exercise,
history, preparation and Program reference counts. Its guarded merge accepts a

9
docs/roadmap.md
View file

@ -65,11 +65,14 @@ synchronization transport.
The 0.1.6 Web refinement now also layers factual Sleep intervals, awakenings,
medication intakes and complete hourly ticks on the selected night's HR chart.
Its explicitly labeled fallback estimate is presentation-only; release work
remains the next separate operational step.
Its explicitly labeled fallback estimate is presentation-only and is shared by
the period agenda, whose responsive labels retain an exact hourly
18:00-to-18:00 geometry. Android lifecycle/ACK regressions now explicitly guard
pre-bed medication snapshots against restart, replay and stale ancestry.
Release work remains the next separate operational step.
```text
CURRENT_OPERATIONAL_CURSOR=0.1.6 Sleep graph overlays and post-review Web corrections validated and privately deployed; stable release work remains separate.
CURRENT_OPERATIONAL_CURSOR=0.1.6 Sleep graph/agenda projection and Android medication persistence proof validated and privately deployed; stable release work remains separate.
```
Sleep Diary V1 stays frozen/readable while active capable peers require

24
docs/tests.md
View file

@ -25,27 +25,39 @@ still conflicts. The equipment client test removes `crypto.randomUUID`, uses
deterministic `getRandomValues` bytes, and completes the protected merge HTTP
request; native merge tests retain their transaction/rollback coverage.
The real Sleep Bluetooth recovery regression starts with no Android session,
records a pre-bed medication with `quantity=2`, Couché, two BPM/RR samples
The real Sleep Bluetooth recovery regression starts with no Android session
and an empty pending Sleep revision, records two pre-bed medication intakes
(including `quantity=2`), reopens the repository after each meaningful
lifecycle transition, then records Couché, two BPM/RR samples
separated by Réveil, and Levé. It captures and publishes the production V2
Sleep and V1 heart-rate artifacts, imports them with the desktop consumer,
accepts and replays the ACK without duplication, then admits a second
Sleep-only revision. Separate tests prove connection-arrival auto-request
Sleep-only revision. It also applies the original intake-free ancestor after
ACK and proves that causal reconciliation cannot erase the current intakes.
Separate tests prove connection-arrival auto-request
policy with its cooldown, cross-run reuse of an unacknowledged immutable
generation and recognition of its original-run exact generation/digest ACK,
ledger-backed superseded-branch capacity, fresh-peer non-root Sleep ancestry,
service-installer upgrade behavior, and ACK-recovery progress frames over the
Bluetooth-selected orchestrator path.
UI regressions assert the Sleep labels at exactly 0/25/50/75/100%, persisted
UI regressions assert the Sleep agenda's 25 exact hourly positions, 13 visible
two-hour desktop labels and five narrow six-hour anchors, including the exact
0/25/50/75/100% coordinates. They also prove that both Sleep views reuse the
same presentation-only Couché +45 minutes to Levé -10 minutes estimate, that a
factual sleep interval wins, and that a missing bound produces no estimate.
Other UI regressions cover persisted
Dashboard sizing and reset behavior, English preference persistence across a
remount, centralized Dashboard English labels, non-stretched chart aspect
ratios, and distinct missing/partial exercise-timing messages.
The closeout run passes Web **31 files / 196 tests**, TypeScript typecheck and
The closeout run passes Web **32 files / 205 tests**, TypeScript typecheck and
the production build; native Meson **111/111**; Android `testDebugUnitTest` and
`assembleDebug`; JSON/import validators; changed-C formatting; and diff checks.
The restarted private service serves the rebuilt executable byte-for-byte from
the development checkout and reports healthy Trainlog `0.1.6`.
the development checkout and reports healthy Trainlog `0.1.6`. A non-mutating
Firefox smoke on the real 2026-09-23 night confirms the agenda and detail both
show 22:50–04:04 as estimated, with 13 ordinary-desktop labels, 25 wide labels
and five narrow anchors.
Sleep Diary coverage includes schema creation/migration, midnight and DST
offsets, revision conflict, logical deletion/reopen, companion replay and