`loose.includes(entry)` ran once per LETTER inside the band loop — O(L²)
on a dense band of hundreds of loose letters, recomputed on every layer
re-render. splitYearLetters now also returns its `byEvent` map, so a
letter's disposition is `byEvent.has(linkedEventId)` and an event's card
is `byEvent.get(eventId)`, both O(1); `consumed` is a plain object. No
behavior change. Fixes review finding #3.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The bare "· {count}" spans in both the same-year and cross-year headers
announced as "· 2" with no context. Each now pairs the aria-hidden visible
count with an sr-only "{count} Briefe" via a new Paraglide key
(timeline_cluster_letter_count, present in de/en/es). Fixes review
finding #6 (count half).
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The cross-year card header emitted a bare aria-hidden ✉ with no sr-only
label, unlike the same-year header and LetterCard — a screen-reader user
heard only the title with no cue that this is a letter group. It now uses
the shared GlyphLabel (✉ + sr-only "Brief"). Fixes review finding #6
(glyph half).
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`showDate = !compact || !entry.title` dropped the date chip for ANY titled
compact letter. But titles are free-form OCR/import text — a letter titled
"Brief an Mutter" lost its month/day entirely, and inside an event card
the band frames only the year. The chip now drops only when the formatted
date actually appears in the title (e.g. "H-0023 – 6. Juli 1916"), so the
row-height win holds where valid and no information is lost otherwise.
The spec that asserted the date vanishes for any title is rewritten to the
correct contract, plus an inverse test. Fixes review finding #4.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A titleless or whitespace-only event stored `''` in the lookup, so its
letters still clustered and rendered a label-less `✉` mystery card. The
lookup now skips events whose trimmed title is empty — those letters stay
loose. Fixes review finding #8.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
buildEventLookup also collected `timeline.undated`, so an undated curated
event — which renders as a plain EventPill in the undated bucket, out of
clustering scope — still seeded clusters: its dated linked letters
scattered into year bands and each collapsed into a ✉ cross-year card
with no edit link and no spatial tie to the pill, showing the event title
twice with no relationship.
Only year-band events are collected now. Fixes review finding #7.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
buildEventLookup keyed on `kind === 'EVENT' && eventId` with no type
check, so a HISTORICAL curated event with ≥1 linked letter entered the
lookup and rendered as a mint EventCluster card — silently downgrading
from the full-width WorldBand that #779 REQ-009 mandates ("world-bands
render exactly as before").
The lookup now excludes `type === 'HISTORICAL'`, so a world event always
keeps its WorldBand and its letters stay loose chronological. Closes the
spec gap pinned as REQ-014. Fixes review finding #2.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
resolveLetterEventLinks iterated `eventRepository.findAll()` and touched
the lazy `documents` collection on every event — including events removed
by the type/person/generation/year filters and never shown. On a large
archive that hydrates join rows that are immediately discarded.
It now runs over the events that survived the filter (collected once in
the existing event loop). A letter whose only linking event was filtered
out links to nothing, which matches the frontend's filter-then-cluster
(the letter renders loose either way). Fixes review finding #10 (DevOps).
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A letter whose document is in more than one curated event's `documents`
set was linked by `putIfAbsent` over `eventRepository.findAll()`, whose
iteration order JPA does not guarantee — so the same letter could cluster
under a different event across re-seeds/VACUUM with no data change.
resolveLetterEventLinks now runs over a stably-ordered copy (earliest
event date, undated last, then lowest id), so the link is a deterministic
property of the data. Fixes review finding #9 (Architect-3).
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two spec gaps the Requirements Engineer flagged on PR #851 are now pinned
as requirements in the #850 issue body and traced here (Status Planned,
flipped to Done as the tasks land):
- REQ-014: a HISTORICAL curated event keeps its full-width WorldBand and
never clusters into a card, even with linked letters.
- REQ-015: a cross-year card sits at its earliest linked letter's
chronological position, never appended after later-dated loose letters.
Refs #850
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The grep gate fails if any lib/timeline component reaches for the raw-HTML
directive (CWE-79, REQ-010). The RTM gains thirteen rows tracing every #850
requirement to its implementation file(s) and test(s), Status Done.
Refs #850
TimelineView builds the event lookup once over the whole timeline and threads
it (plus canWrite) to every YearBand, so a curated event's letters cluster
under it inline. The /zeitstrahl meta-line drops its 'Gruppierung: Datum'
segment (toggle-free view, REQ-011); the now-unused timeline_grouping_date
key is removed from de/en/es and the messages parity guard, which now asserts
the new show-more/less keys.
Refs #850
A curated event with same-year linked letters renders as one EventCluster
card (no separate pill); a cluster whose event lives in another band renders
as a cross-year text-header card. Letterless/derived/world events stay plain
pills/world-bands. Loose letters keep the alternating left/right layout and
fold into one density strip past 12 — and the layout + strip count ONLY the
loose letters, so a clustered letter never re-appears loose.
Refs #850
A curated event with linked letters renders as one contained card: the
event is the header (accent glyph, title, date · provenance, count, and a
curator edit link), its letters sit inside as compact .lcard.ev cards. The
first CLUSTER_PREVIEW (5) show, then a keyboard-operable show-more/less
toggle reveals the rest. A cross-year card (no event prop) gets a plain
'✉ title' text header with no edit link. Titles render through default
{...} escaping. Adds timeline_bucket_show_more/less keys to de/en/es.
Refs #850
buildEventLookup maps each curated event in the (already layer-filtered)
timeline to its title; splitYearLetters partitions a year's letters into
event clusters (keyed by a linkedEventId present in the lookup) and the
loose chronological remainder. A letter linking to a filtered-out event
falls back to loose (filter-then-cluster); each letter appears once.
Refs #850
The event variant adds the .lcard.ev marker for letters living inside a
contained event card; the compact variant tightens the row (py-1, text-xs
title) and drops the redundant date chip when the title already embeds the
date. suppressTagChip lets a caller that already conveys the topic hide the
per-letter root-tag chip. Plain Datum letters are unchanged.
Refs #850
For each LETTER entry, resolve the curated event whose documents set
contains the letter, in one batched pass over the events already loaded
(no per-letter query, no new column). The DTO gains a nullable
linkedEventId; non-letter entries keep null.
Refs #850
vite 7.3.3 carries two high-severity advisories (GHSA-v6wh-96g9-6wx3
NTLMv2 UNC disclosure, GHSA-fx2h-pf6j-xcff server.fs.deny bypass), both
flagged by the CI gate `npm audit --audit-level=high --omit=dev`. 7.3.5
is in-range of the existing `^7.3.3` constraint, so this is a
lockfile-only patch with no package.json change. Gate now exits 0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fixes#848.
## Symptom
CI `Backend Unit Tests` goes red despite **all tests passing**: after the last test, the fork hangs at JVM shutdown and Surefire reports `There was a timeout in the fork` → `BUILD FAILURE`.
## Root cause (corrected after investigation)
My first theory (slow shutdown needs a bigger timeout) was **wrong** — raising `forkedProcessExitTimeoutInSeconds` 30→120 only delayed the kill by ~90s (total time 12:35 → 14:04), proving an *indefinite* hang, not slowness.
The real cause is **Testcontainers teardown with Ryuk disabled**:
- The job set `TESTCONTAINERS_RYUK_DISABLED: "true"` (carry-over from the old NAS runner).
- With Ryuk off, containers are reaped by the **in-JVM `JVMHookResourceReaper`** at shutdown. That reaper crashes (`NotFoundException`) and **leaks containers run-over-run**.
- The run boots ~30 per-context Spring contexts (`PostgresContainerConfig` is a per-context `@Bean`), so ~30 Postgres containers are torn down in-JVM at shutdown.
- As leaks accumulate on the runner, per-run teardown degrades until the fork hangs at shutdown → fork timeout. **The server had 21 orphaned `postgres:16-alpine`/`minio` containers up to 5 weeks old**; manually killing them is what restored CI before (a recurring pattern).
Environment confirmed via `ssh root@raddatz.cloud`: CI now runs on a root server with **Docker 29.4.3** (8 CPU, 62 GB, socket access) — so the original reason to disable Ryuk no longer applies, and Docker is *not* slow.
## Change
1. **Re-enable Ryuk** (remove `TESTCONTAINERS_RYUK_DISABLED`) — Ryuk reaps each run's containers out-of-process after the JVM exits, so they never accumulate. Automates the manual "kill all testcontainers."
2. Keep `forkedProcessExitTimeoutInSeconds=120` as a harmless backstop.
3. Drop the stale "NAS runner" comment on `DOCKER_API_VERSION`.
Operational: the 21 leaked containers were already removed from the server (by `org.testcontainers=true` label; real services untouched), giving immediate relief.
## Validation
Validated by this PR's CI run on the real runner (watching it). If Ryuk can't start in the runner's docker-outside-docker setup, the integration tests fail fast and I revert — fallback is a singleton Postgres container.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Marcel <marcel@familienarchiv>
Reviewed-on: #849
The person-add-event link wraps a single text label, so the flex gap
never applies — only the sibling edit link (icon + text) needs it.
Removing the dead utility per the UI/UX review nit.
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
REQ-011 covers direct nav to both /zeitstrahl/events/new and
/{id}/edit, but the row cited only the /new guard + test. The
[id]/edit route shares the same requireWriteAll helper and already
carries its own 403 gating test (shipped with #781); cite both so the
traceability matches the requirement. Closes the Tester/Security
review note (no new test needed — the guard test already exists).
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add REQ-001..011 rows for the timeline-curator-affordances feature (all
Done) and a messages.spec parity guard pinning timeline_add_event +
person_add_event across de/en/es. REQ-010/011 cite the existing #781
new-event route + its 403 guard test (no new production code).
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A curator on a person's page can now seed a timeline event from that
person: a gated link to #781's /zeitstrahl/events/new?personId={id},
which prefills the person and returns to /persons/{id} on save. Hidden
for a Reader. New i18n key person_add_event in de/en/es.
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wrap the heading in a flex-wrap header so the gated add-event link drops
below the title at narrow widths instead of overflowing; the CTA targets
#781's /zeitstrahl/events/new and only shows for a WRITE_ALL viewer. Also
thread canWrite into TimelineView so a curator sees the edit pencils on
the real page. New i18n key timeline_add_event in de/en/es.
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A curated HISTORICAL band had no edit affordance at all. Mirror the
EventPill pencil inline at the end of the band (data-testid="event-edit",
/edit href, aria-hidden glyph + sr-only Bearbeiten), gated on
canWrite && eventId != null. Thread canWrite to WorldBand through both
the year-band path and the undated bucket.
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Thread a gate-closed canWrite prop through TimelineView -> YearBand ->
EventPill and the undated bucket so a Reader never sees a dead-end edit
link. canEdit now also requires canWrite; the default false keeps an
un-threaded caller closed. The real boundary stays the #781 route guard
plus the backend permission -- this only hides the link.
Refs #842
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The /zeitstrahl header sub-line counted the unfiltered timeline, so a
hidden layer (e.g. Letters off) still showed its entries in the totals
("1 Brief" with no letters on screen) — the documented D1 limitation.
Derive the meta from filteredTimeline so the range and letter/event
counts always match what is actually rendered. hasContent stays on the
full timeline so the filter bar and meta line still appear whenever the
archive has content.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add the ten REQ-001…REQ-010 rows for the /zeitstrahl layer filter, each linked
to its implementation file(s) and test(s), all Done.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Playwright spec for /zeitstrahl: the primary journey (hide Letters → letter
cards vanish + trigger reports "1 aktiv" → reset restores) and a 375px axe pass
with the collapsible open in light and dark mode. Not skipped — #779 ships the
route. E2E is not wired into CI, so this runs locally only for now.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A static boundary gate (mirroring the project's no-`{@html}` greps) that reads
TimelineFilters.svelte and /zeitstrahl/+page.svelte and fails if either ever
reintroduces goto(, url.searchParams, api.GET, or fetch( — the filter must stay
presentation-only and fully client-side.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The route holds the three layer toggles in $state, binds them into
TimelineFilters, and derives a client-side filtered view of the SSR-loaded
timeline that it passes to TimelineView — no goto, no URL param, no extra
fetch. When the active toggles leave nothing visible it renders a calm
filtered-empty message plus a one-click reset below the still-open filter bar,
never a blank page and never the generic "no events" state. The meta-line keeps
counting the unfiltered timeline (D1 known limitation).
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A dumb, client-side layer-filter bar for /zeitstrahl: three $bindable layer
toggles (Personal/Historical/Letters) in a fieldset/legend, a sticky
"Filter (N aktiv)" trigger driven by hiddenLayerCount, and a reset text button
shown only when a layer is off. Toggles mirror the SearchFilterBar undated-toggle
markup (aria-pressed, ✓ glyph, 44px touch target, semantic tokens). The
collapsible slide honours prefers-reduced-motion by zeroing its duration. No
goto, no fetch — the route owns the state and the filtered view.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Eight Paraglide keys for the /zeitstrahl layer filter — layer labels, the
fieldset legend, the sticky trigger (distinct timeline_filter_trigger and
timeline_filter_trigger_active({count}) so it never reads "Filter (0 aktiv)"),
the reset button, and the filtered-empty message — added to all three locales.
messages.spec asserts their presence and the {count} signature.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pure helpers for the /zeitstrahl layer filter: isDefaultState and
hiddenLayerCount drive the "Filter (N active)" trigger, and filterTimeline
derives a client-side view that hides personal/historical/letter layers and
drops year bands left empty. Letters ride the Letters layer, HISTORICAL events
the Historical layer, and curated PERSONAL plus derived life-events the
Personal layer.
Refs #780
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes#837
Makes `PersonRelationship` fully editable (type, related person, dates, notes), migrates its dates from `Integer fromYear/toYear` to `LocalDate + DatePrecision` (mirroring the #773 person pattern, ADR-039 / V76), activates the previously-dead `notes` column, and gives the Zeitstrahl's derived **Heirat** events full date precision for free.
Both Open Decisions confirmed as adopted: **no `@Version`** (last-write-wins, single-writer archive) and **`DELETE` ownership-mismatch aligned 403 → 404** (anti-enumeration, matching the new `PUT`).
## What's in it
- **V78** migrates `person_relationships.from_year/to_year` → `from_date`/`to_date` + NOT-NULL `*_date_precision` (default `UNKNOWN`); pre-check abort on corrupt years, `YYYY-01-01`/`YEAR` backfill, 5 named CHECK constraints, year columns dropped.
- **`PUT /api/persons/{id}/relationships/{relId}`** (`@RequirePermission(WRITE_ALL)`) re-runs every create invariant (self / coherence / order / reverse-PARENT_OF / duplicate) and re-flags family membership; orientation preserved per viewpoint.
- New `ErrorCode.INVALID_RELATIONSHIP_DATES` registered in all four sites (§3.6).
- `TimelineEventService` sources the derived marriage date from `SPOUSE_OF.fromDate` + precision.
- Frontend: `RelationshipDateField` (DAY/MONTH/YEAR), upsert-capable `AddRelationshipForm` (pre-fill + notes + in-flight submit lock), `RelationshipChip` Edit affordance, `updateRelationship` server action, read-view date range + notes, `formatRelationshipDateRange` helper. `api.ts` regenerated.
- Docs: ADR-044, db-orm/db-relationships diagrams, DEPLOYMENT §5 deploy note, RTM REQ-001…REQ-019.
## Requirements
All 19 EARS requirements implemented red/green and marked `Done` in `.specify/rtm.md`.
## Test plan
- **Backend** (targeted, green): `RelationshipMigrationTest` (Testcontainers pg16, 8), `RelationshipServiceTest` (22), `RelationshipControllerTest` (15), `RelationshipServiceIntegrationTest` (real DB, 10), `DerivedEventsAssemblyTest` (17), `ArchitectureTest` (14); `clean package` builds.
- **Frontend** (green): `relationshipDates.spec.ts`, `AddRelationshipForm.svelte.spec.ts`, `RelationshipChip.svelte.spec.ts`, `PersonRelationshipsCard.svelte.test.ts`, `page.server.spec.ts`, `messages.spec.ts`. `npm run check` = 798 (below the ~834 baseline); `npm run lint` clean.
## Notes for reviewers
- **Spec deviation:** the edit form was built by making `AddRelationshipForm` upsert-capable rather than a duplicate `EditRelationshipForm` (DRY); RTM rows reference `AddRelationshipForm.svelte.spec.ts`.
- `api.ts` regenerated from the live spec; only relationship-relevant hunks remain (one springdoc `PageableObject` field-reorder pruned).
- **Deploy:** V78 is one-way and not rolling-deploy-safe — stop old JAR → start new JAR (Flyway runs first); targeted `pg_restore -t person_relationships` for rollback. No maintenance window.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Marcel <marcel@familienarchiv>
Reviewed-on: #841
The npm-audit job filed its tracking issue via `curl -sf`, which collapses
every HTTP >=400 into a bare "exit 22". When the NIGHTLY_AUDIT_TOKEN secret is
missing, expired, or under-scoped, the step failed with an opaque
`exitcode '22'` and no hint at the cause (run #6707).
Route all five API calls through an `api()` helper that reads the HTTP status
and, on >=400, emits an actionable `::error::` naming the status and the token
secret before failing — without ever echoing the token value. Extend the
in-workflow self-test (mocked curl) to cover both the success and HTTP-error
paths.
Closes#839
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two cleanups flagged in review, both behaviour-preserving:
- collapse the {#if color}/{:else} square-marker branches (identical but for the
neutral fill) into one element via class:bg-ink-3={!color}; squareStyle is
already empty when color is null, so no var(--c-tag-) leaks into the neutral
chip.
- drop the redundant `truncate` class from the name span — the inline
overflow/ellipsis trio (kept so it applies before the stylesheet loads,
REQ-008a) already expresses exactly what `truncate` would.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
resolveRoot silently falls back to returning the tag itself when no null-parent
ancestor surfaces — an orphaned parent_id or a chain deeper than the
findAncestorIds CTE depth guard. The chip then renders a non-root tag as if it
were the theme, with no trace. Log a warning (UUIDs only, per REQ-014) before
the fallback so the anomaly is diagnosable.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
mapDocument re-ran the alphabetical min() scan over the letter's tag set to
look up its already-resolved root, duplicating the work resolveLetterRootTags
had just done and leaving two independent definitions of "primary tag" that
could silently diverge. Key the resolved-root map by document id and compute
the primary tag exactly once per letter; drop the redundant resolvePrimaryRoot
helper.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add one row per requirement for the zeitstrahl-tag-chips feature, each mapped
to its implementation file(s) and the test(s) that prove it, Status=Done.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The doc comment described escaping by naming the raw-HTML directive literally,
which trips the lib/timeline grep gate that forbids that token. Reword it the
way LetterCard already does — behavior unchanged.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
LetterCard now renders a TagChip beneath the sender→receiver/date line
whenever the entry carries a rootTagName, mapping rootTagColor to the chip
(neutral when null). Because the chip lives on LetterCard it shows up wherever
a LetterCard does — the global timeline and the expanded YearLetterStrip — with
no per-surface special-casing; a tagless letter shows no chip. A long name
truncates inline so the card never overflows at 320px.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
TagChip renders a letter's primary root tag as a small rounded pill — a
decorative aria-hidden colored square (var(--c-tag-{token}), neutral when the
color is null) plus the escaped tag name, prefixed by the sr-only theme label
so color is never the only cue. Truncation is set inline so a long name
ellipsizes without forcing the card into horizontal scroll, and the full name
stays reachable via the chip title. Timeline-local by design — lib/timeline may
not import lib/tag (eslint boundary).
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
timeline_tag_chip_label (de "Thema" / en "Topic" / es "Tema") is the sr-only
prefix the /zeitstrahl letter tag chip reads out so color is never the only
cue. Pinned per locale in messages.spec.ts; the tag name itself is rendered as
data, never translated.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
openapi-typescript pickup of TimelineEntryDTO.rootTagId/rootTagName/
rootTagColor (all optional), so the SvelteKit timeline can read the new
letter chip fields. Regenerated from the live dev spec; only the additive
fields differ from the committed baseline.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
TimelineEntryDTO gains three nullable letter-only fields — rootTagId,
rootTagName, rootTagColor (token) — assembled in-transaction in TimelineService
(ADR-036): id + name + token only, never a serialized Tag entity. A letter's
primary tag is the root ancestor of its alphabetically-first assigned tag
(#827 Resolved Decision 3); roots are resolved through TagService in one
batched pass over the distinct primary tags (no per-letter N+1). The fields are
null for non-letter entries, untagged letters, and (color only) a colorless
root, so they are deliberately not @Schema(requiredMode = REQUIRED).
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
TagService.resolveRootTags(tags) maps each tag to its root ancestor as a
RootTag (id, name, color token), keyed by the input tag id. A root maps to
itself; a child is walked to the parentless ancestor via the existing
recursive-CTE findAncestorIds — one CTE per distinct non-root tag (memoized),
plus a single batched findAllById — so a timeline of many letters sharing few
tags costs O(distinct tags) queries, never O(letters). The color is read from
the resolved root's stored token (null when the root has none).
This is the shared enrichment the /zeitstrahl tag chip (#835) and, later, the
Thema buckets (#827) both consume. Unit-tested in TagServiceTest; the
DB-dependent ancestry walk is pinned against real Postgres in
TagServiceIntegrationTest.
Refs #835
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The spine offset (0.5rem phone / 50% desktop) was hard-coded in both
TimelineView's .timeline-axis::before and YearBand's .year-node/.letter-dot,
kept in sync only by a comment. Declare --spine-x once on .timeline-axis
and have the markers consume it by inheritance, so a change to the spine
position moves the markers with it. Add a test that the year-node tracks
the token.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The aria-hidden glyph + sr-only label markup was hand-copied in LetterCard
and YearLetterStrip. Extract a small GlyphLabel component and use it at
both sites so the accessibility idiom has a single owner.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the flatMap intermediate array plus two filter passes with one
walk over the year bands and the undated bucket. Same counts, no
throwaway allocation.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
