- Seed items with ranks 'a0' and 'a1', drag second to top via rank 'Zz'
- Assert 'Zz' < 'a0' is true in JS (documents uppercase-before-lowercase intent)
- GET /api/lists/:id/items must return Zz-ranked item at index 0
- Fails now because MariaDB utf8mb4_uca1400_ai_ci sorts 'Zz' after 'a0'
- Will pass once rank column gets COLLATE utf8mb4_bin via migration
src/lib/rank.test.ts was the last co-located API test. The API tsconfig excludes
tests/ from the build, so all test files belong there; a test in src/ gets compiled
into dist/ and vitest then runs the stale compiled copy (the source of phantom
'22 todo' and dist sourcemap warnings). Moving it to tests/lib/ matches convention
and leaves zero test files in src/, so the production build no longer emits test
artifacts. Updated rank.test.ts path refs in phase-04 plan docs.
- Wire publishListEvent fan-out in lists.ts after every write mutation (item:added/updated/deleted, list:updated/deleted)
- Add GET /api/sse/lists scoped endpoint in sse.ts: resolveUserId → 401 on null; getAccessibleListIds → subscribe only to accessible channels; 30s heartbeat; cleanup on disconnect (D-04/T-04-01/T-04-02)
- Create useListSSE.ts: bounded-backoff EventSource wrapper (250ms→500ms→1s→2s→4s→cap 8s); MAX_ATTEMPTS=6; withCredentials:true; close-before-retry prevents reconnect storm (Pitfall 3); invalidates ['list', listId] on open (D-10) and on each event; onStateChange('disconnected') after exhaustion (D-11)
- Create LiveSyncIndicator.tsx: connected=green dot; reconnecting=pulsing muted dot + label; disconnected=red dot + 'Updates paused' (role=alert); correct ARIA per UI-SPEC
- Wire useListSSE + LiveSyncIndicator into ListDetail header; retain refetchInterval:30000 polling fallback (D-12)
- All 8 useListSSE tests pass; all 54 API tests pass; both typechecks pass
- playwright-cli: live update confirmed (eggs item added via API appeared in browser without manual refresh)
- rank.test.ts: 100-iteration zipper mid-point insert precision test (Pitfall 2);
rank-between-neighbors contract test; total 10 tests (was 8)
- lists.test.ts: 5 new LIST-03 ordering tests — PATCH position updates only rank
and GET returns new ASC order; one-row write asserts other items unchanged;
LWW (D-15): second PATCH overwrites first; T-04-07 two-field position PATCH → 400
- Note: tests use a0–a5 rank range (avoids uppercase ranks that sort differently
under MariaDB utf8mb4_unicode_ci vs JS lexicographic order)
- GET /: scoped access (owner + list_shares); activeCount/doneCount per list
- POST /: auto-populates list_shares for all other members when isShared=true (D-01/D-02)
- PATCH /🆔 rename + isShared toggle; reconciles list_shares on visibility change
- DELETE /🆔 owner-only; cascade handles items/shares via FK onDelete cascade
- resolveUserId helper copied verbatim from events.ts per project convention
- zod createListSchema (name 1..255, isShared default true) + patchListSchema
- T-04-02 / T-04-05 / T-04-07 / T-04-08 mitigations applied
- listsRouter mounted at /api/lists in index.ts (after sseRouter)
- Plan 06 SSE seam comments left at every mutation handler
- [Rule 1 - Fix] zValidator returns 400 (not 422); tests corrected to match convention
- All 23 tests green; full API suite 140 passed no regressions
Replaced singleFork:true + sequence config with the simpler fileParallelism:false
which correctly serializes test file execution. The previous singleFork approach
ran tests from multiple files concurrently within one process, allowing the global
afterEach cleanup (test/setup.ts: truncates lists/listShares) to delete rows mid-test
in another file, causing intermittent FK violations (ER_DUP_ENTRY, ER_NO_REFERENCED_ROW).
fileParallelism:false runs one test file at a time so afterEach cleanup for file A
never races with insertions from file B.
listEmitter.ts:
- Module-level EventEmitter singleton; setMaxListeners(200) (T-04-04)
- publishListEvent(listId, event): emits on list:${listId} channel
- subscribeListEvents(listId, handler): registers listener, returns unsub closure
- ListEvent type union: item:added/updated/deleted, list:updated/deleted
- D-04 isolation guaranteed by per-list channel keying
listAccess.ts:
- getAccessibleListIds(userId): two SELECT queries (owned + shared), Set dedupe
- Satisfies T-04-02/T-04-03: over-returning proven impossible by Test 7
listAccess.test.ts fix:
- Use randomUUID() suffix in seedUser to avoid oidc_sub unique-key collisions
across test re-runs (users table not truncated by global afterEach)
vitest.config.ts:
- pool: 'forks' + singleFork: true to prevent FK violations from concurrent
DB workers racing against the shared-state global afterEach cleanup
- sequence.concurrent: false as belt-and-suspenders
ioredis NOT introduced (D-18 abstraction boundary satisfied)
The two Wave-0 RED stubs (lists.test.ts, listEmitter.test.ts) were co-located in
src/ but all existing API tests live in apps/api/tests/. Move them to tests/routes/
and tests/lib/, add explicit vitest imports to match the tests/ convention, and
update path references in downstream plans 04-02..04-06. PWA tests keep co-location
(that IS the PWA convention).
- Add react-router@7, @dnd-kit/core, @dnd-kit/sortable, fractional-indexing to PWA
- Add fractional-indexing to API (rank generation server-side)
- ioredis NOT added (in-memory EventEmitter per RESEARCH Plan 02 justification)
- Create apps/api/test/setup.ts with afterEach DB cleanup for list tables
- Wire test.setupFiles in apps/api/vitest.config.ts
- Add 4 Wave-0 RED stub test files (LIST-01/02/03/04, D-04, D-11, D-07)
- All stubs run as todo, not import-error
The CR-01 fix appended .orderBy().limit(1) to the edit/delete event lookups
and CR-02 added .innerJoin(calendars).limit(1) to the freshest-etag re-read.
The existing test doubles terminated the mock chain at .where(), so the new
chain calls hit undefined methods → handlers caught the throw and returned 503
(events.test.ts) and the worker skipped the PUT (outboxWorker.test.ts).
Extend the mocks to match the corrected production chains. Behaviour-preserving:
mockWhereCalEvents stays the awaited terminal so etag override assertions still drive.
8 failing tests now green; full suite: api 103, pwa 141.
All-day: a single-day all-day event displayed across two days. iCal all-day
DTEND is EXCLUSIVE (1-day event = DTSTART:24/DTEND:25) and the server occurrence
carries that exclusive end, but Schedule-X treats all-day end as INCLUSIVE.
hydrateEvents now subtracts one day (clamped to >= start) so a 1-day event shows
on one day and an N-day event spans N days. Write path was already correct
(verified against stored VEVENTs). +regression test.
Color: a member's coral (#E8734A) was mistaken for the shared-family rose
(#F25C7A). Reorder COLOR_PALETTE so warm near-rose hues (amber, coral) are
assigned LAST; early members get cool, clearly-distinct colors (blue/green/teal).
Gate 2 A3 fail: a second member (amelia) got the same color as the first (luc),
both #E8734A. Color was assigned by COUNT(*) % palette; a deleted spike user
shifted the count so two live members landed on the same slot. Replace with
'first palette color not already in use by another user' (fall back to count
round-robin only once the palette is exhausted) — guarantees distinct, stable
colors for up to palette length members. +1 regression test (deletion frees a
slot → next member fills it, no collision).
The calendar flashed whenever the event popup/form closed or a post-write events
refetch landed. Root cause: CalendarContent was a function component DEFINED
INSIDE CalendarShell's render and used as <CalendarContent />. A nested component
has a new identity every render, so React unmounted+remounted its whole subtree
— including <ScheduleXCalendar> — on ANY CalendarShell re-render. The earlier
Bug B work only minimized re-renders (Zustand selectors) to dodge this; the
resync-before-done fix made the post-write ['events'] refetch deliver changed
data again, so the remount/flash returned.
Fix: render the content as a plain JSX element value (const calendarContent)
referenced at both layout sites instead of a nested component type. Element
values reconcile in place across re-renders — no remount, no flash.
The legend showed 'Member 972be1a3' because Authelia does not emit
name/preferred_username/email in the ID TOKEN (only at the userinfo endpoint),
and @hono/oidc-auth reads ID-token claims only. The real fix is an Authelia
claims_policy adding those claims to id_token for the familysync client.
App-side robustness so it self-heals once Authelia is fixed (no DB surgery):
- deriveDisplayName now returns null (not a synthetic 'Member <sub>') when no
real claim is present, so we never persist an ugly sub string; the UI degrades
to a generic 'Member'.
- upsertUser now tracks the IdP name authoritatively: a non-null displayName that
differs from the stored value updates the row (blank/stale 'Member …'/email →
real name on next login). A null value never overwrites a good stored name.
Two write-path cache bugs surfaced during Gate 2 live testing:
P1 (delete didn't work / ghost event): syncCalendar only UPSERTED events
present on Fastmail and never removed cache rows for events that disappeared.
A successful CalDAV delete left the row in calendar_events forever, so
GET /api/events kept returning it and the UI showed a ghost that 'wouldn't
delete' (even after refresh). Add a prune step: delete calendar_events rows for
this calendar whose uid is absent from the server response (scoped to cal.id so
it never touches another calendar or the other member's rows — BUG B). Empty
server result prunes the whole calendar's cache.
P2 (edit needed a manual refresh): the outbox worker marked a row 'done' BEFORE
triggerTargetedResync refreshed the cache. The PWA's SyncStateToast invalidates
['events'] the instant sync-status flips to 'done', so it refetched stale cache.
Re-sync first, then mark done — 'done' now guarantees the cache reflects the write.
Tests: +2 prune regressions (present-subset prune, empty-server prune-all).
Internal Server Error after returning from Authelia: processOAuthCallback threw
OAUTH_INVALID_RESPONSE ("unexpected state parameter") because the OIDC state
cookie no longer matched the state returned to /callback.
Root cause: eventsQuery (fetchEvents, redirect:'follow', retry:2) ran
concurrently with fetchMe on load. While unauthenticated, every /api/* request
hits the OIDC guard, which 302-redirects to Authelia AND sets a fresh state
cookie. fetchEvents could not follow the cross-origin redirect, so React Query
retried it up to 3x over ~3s — each retry overwriting the state cookie mid-login,
racing the single /api/login navigation that owns the real flow.
Fix: enabled: meQuery.isSuccess. Only fetchMe (redirect:'manual', retry:false)
touches a guarded endpoint while unauthenticated, so the top-level /api/login
navigation owns the state cookie uncontested. Realizes the documented design
intent that only fetchMe drives the login redirect.
The displayName claim-preference logic (name → preferred_username → email →
sub fallback) was duplicated verbatim in me.ts and events.ts resolveUserId.
Extract it to auth/user.ts as deriveDisplayName and use it in both call sites,
so the rule has one definition. Update the events.test.ts user.js mock to keep
the real helper (spread importActual) while stubbing only upsertUser.
The original toSQL() regression test hand-built the joined query inside the
test body and asserted the SQL contained a join — tautological: it never
exercised the handler, so removing .innerJoin from events.ts left it green.
Replace it with two tests that issue real PATCH/DELETE requests against the
mocked select-chain (from → innerJoin → where) and assert the handler returns
202 (not 503) AND invokes the innerJoin spy. Verified RED: removing the
edit+delete joins fails both tests; GREEN with the joins present.
BUG 3: GET / had no ownership predicate — it returned all users' events.
Second household member would see other member's private events.
- Resolve currentUserId at top of GET handler (same resolveUserId helper
as write endpoints); return 401 if unauthenticated.
- Add ownership predicate to WHERE: AND (calendars.userId = currentUserId
OR calendars.isShared = true). Combined with and() around the existing
date-window or() block. Mirrors the /writable-calendars idiom (D-03).
BUG 2: Both me.ts and events.ts resolveUserId were passing email (often
absent) as displayName to upsertUser, resulting in blank legend names.
Also, upsertUser returned existing rows unchanged even when displayName
was null and a better value was now available.
- me.ts: derive displayName via name → preferred_username → email →
"Member <sub-prefix>" fallback, checked defensively. Updated JSDoc.
- events.ts resolveUserId: same derivation so write-path upserts don't
re-blank a correctly-set displayName.
- user.ts: when existing row has null displayName and caller supplies one,
issue an UPDATE so already-existing blank rows are corrected on next login.
Authelia-side emission of name/preferred_username is an operator concern
(claim mappings / userinfo scope in authelia config) — out of scope here.
The code now reads whatever claims are present and falls back sensibly.
Lucas Berger