Review follow-up. spanFirstDay/spanLastDay resolved every event in the
device zone, but all-day events live at UTC midnights with an exclusive
end — east of UTC (e.g. Europe/Berlin) that pushed spanLastDay onto the
next day, so a single-day all-day event reported spansMultipleDays and
leaked onto a second agenda day. Resolve all-day dates in UTC, matching
the Week view and detail card. Adds eastern-zone regression tests that
the prior UTC-only tests could not catch.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The "→" arrow convention read as unclear. Spell each day out instead:
the first day names the start ("Starts 14:00"), the last day the end
("Ends 10:00"), and whole days in between read as "All day". All-day
multi-day events stay "All day" on every day. Single-day rows unchanged.
Adds agenda_span_starts / agenda_span_ends (owes Weblate backfill).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A multi-day event repeated the same "start – end" on every day it spans
("14:00 – 14:00"), which reads as meaningless. Show only the part
relevant to each day, with a "→" marking that it carries past the day's
boundary: the first day shows the start ("14:00 →"), the last day the
end ("→ 10:00"), and whole days in between an all-day span arriving from
and continuing into their neighbours ("→ All day →"). Single-day rows are
unchanged.
Factors the span first/last-day resolution into shared EventInstance
helpers reused by groupAgendaDays and the label.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A multi-day event now appears under every day it spans, so keying its
row by instanceId alone repeated the key across days and crashed the
LazyColumn ("Key already used") on scroll. Scope the key by day.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Search matched city, id, and the long localized name but not the
abbreviation, so typing "CEST" found nothing. Match it too: an exact
abbreviation hit ranks just under a city prefix, so typing an abbreviation
gathers every zone that shows it (all the CEST zones at once). It matches
the region-resolved abbreviation — i.e. exactly what the row displays.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
CalendarRepositoryImpl gained a SettingsPrefs parameter, but this
instrumented smoke test still called the three-arg constructor — so the
whole androidTest source set failed to compile. Pass a SettingsPrefs
built on the same DataStore.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
groupAgendaDays keyed each instance by its start day alone, so a
multi-day event surfaced only on its first day and vanished from the
rest of its span in both the Agenda screen and the agenda widget.
Expand each instance across every day from its start (clamped to the
anchor for ongoing events) through its last occupied day, bounded by the
visible window end. An event ending exactly at midnight — including the
exclusive next-midnight all-day events end at — does not reach that
boundary day.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
java.util gave abbreviations only for zones in the display locale's region:
an en-DE phone saw "CEST" for Berlin but "GMT-4" for New York, and an
en-US phone the reverse — ICU only surfaces the short names commonly used
where the reader is. Verified on-device (en-DE): US zones fell back to the
offset, EU ones resolved.
Ask ICU in the display language but the *zone's* region instead — New York
in en-US, Berlin in en-DE — using android.icu's zone→region map. On-device
that lights up EDT/PDT/CDT, plus BST, AEST, IST, JST that showed only the
offset before. Where a region still has no name (Athens in en-GR) the
device locale sometimes does, so fall back to it, then to the offset.
The region lookup needs android.icu, which domain/ can't import, so it's
injected: timeZoneOptions takes a regionOf lambda (default none, keeping
the module pure and JVM-tested), and the UI passes icuTimeZoneRegion.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The abbreviation came from java.time's "zzz" formatter, which on Android
has no short specific-zone names and silently degrades every zone to a
"GMT-4" form — so on device New York showed "GMT-04:00" instead of "EDT".
Desktop couldn't catch it: there "zzz" and java.util.TimeZone agree, and
that agreement is the whole trap.
Resolve through java.util.TimeZone.getDisplayName(inDst, SHORT, locale)
instead — ICU-backed, so it returns the real abbreviation on both Android
and the JVM. Zones with no named abbreviation still fall back to a GMT
form, which zoneDescriptor already collapses to a single token.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
"Central European Time (Europe/Berlin)" is too wide — it made the zone
field wrap and stretch. Lead with the id ("Europe/Berlin") and follow it
with the abbreviation instead: "CEST · GMT+02:00" in the picker and edit
card, "CET · 8:00 AM – 9:00 AM" on the detail card (abbreviation + the
event's own-zone time).
The abbreviation is resolved DST-aware at the same instant as the offset
(CET vs CEST) via "zzz"; zones with no named abbreviation fall back to a
"GMT+05:30" form, and zoneDescriptor drops the separate offset in that
case so it isn't stated twice. The long localized name is kept on the
option for search only — typing "pacific" still works — but no longer
shown.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The When card reads date-large, time-small-beneath. The zone card led
with the original time at titleMedium and dropped the zone label to
bodyMedium, inverting that — which made the foreign time the loudest
thing on the screen and pulled attention off the local time the reader
actually acts on.
Put the label back on top and the original time small beneath it, so both
cards read the same way and the original stays available without
competing. The label already names the zone, so the range drops its "in
New York" tail (and the string with it).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two review fixes.
The zone picker's search box was a raw Material OutlinedTextField — the
only one left in the app, and against the convention DialogControls
states outright ("the family's InlineTextField over a tonal surface, not
Material's outlined field"). Rebuild it on InlineTextField over a tonal
surface, with the clear button inside the surface since the picker's top
bar is the title rather than a search field.
Showing a pinned event only in its own zone answered "what was it set
to?" while dropping "when is it for me?" — the user had to do the offset
arithmetic. Show both whenever they differ:
- the edit form keeps editing the event in its own zone (that's the time
it was set at) and captions it with the local equivalent;
- the detail screen keeps local times primary and now leads the zone card
with the original ("8:00 AM – 9:00 AM in New York") instead of naming
the zone and nothing else.
EventForm.timesIn is pure, so the conversion — including crossing the
date line and each zone's own DST, which don't move together — is a
plain JUnit test rather than something only reviewable on a phone.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The time-zone picker needs the scaffold's `scrollable` opt-out, which
landed in 0.2.1. Pins the tag rather than the branch commit the work was
developed against, keeping the tag-pinning convention.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Surface the zone as an optional form field, right after the time fields
it qualifies: hidden on ordinary events, revealed automatically when the
event already carries a foreign zone, and withheld entirely while all-day
is on (a date-anchored event has no zone to show).
The picker is a full-screen one per the app's convention, but it can't
reuse OptionPicker: that composes every option eagerly, and ~600 zones
would all compose on open. It drives its own LazyColumn instead, which
needs the kit's new `scrollable = false` — the scaffold's own
verticalScroll would otherwise throw on a nested same-axis scrollable.
The device zone and recently-picked zones pin to the top so the common
case needs no typing; search is accent- and case-insensitive.
Recents persist in DataStore, capped at five, dropping ids the tz
database no longer knows.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Every write sampled ZoneId.systemDefault() and stamped it into
EVENT_TIMEZONE, so the column was real but only ever held the device's
zone: an event synced from elsewhere could be read in its zone, never
authored in one.
Give EventForm a nullable `timezone`, where null keeps meaning "the
device zone at save time" — so every existing call site behaves exactly
as before — and a non-null value pins the event to a zone it then tracks
across DST. toWriteTimes resolves the form's zone ahead of the device's;
toEditForm pins only when the stored zone differs from the device's, and
prefills such an event in its own zone so the form shows the wall-clock
the event actually means.
Two provider-contract bugs fall out of this:
- Editing the time of a foreign-zone event rewrote EVENT_TIMEZONE to the
device's. The instants stayed right, so nothing looked wrong, but the
event silently stopped tracking its zone and would drift an hour at the
next DST boundary. Only the timesChanged gate spared title-only edits.
- A zone change with an untouched wall-clock is still a time change (the
same 09:00 elsewhere is a different instant), so it now trips
timesChanged and rewrites DTSTART instead of being dropped.
All-day events keep carrying no zone at all: they're date-anchored, and
the UTC midnights they normalise to are an anchor rather than a location.
TimeZoneCatalog is pure JVM so the search ranking and DST-aware offsets
stay plain JUnit tests.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Tracks the tagged kit main rather than the feature branch the merge left it on.
A release branch pointing at an unmerged branch is the failure mode main carried
since 2.15.0 (its pointer lived only on feat/custom-snooze-duration): the release
pipeline builds from the tagged source tree, so a rebased or deleted branch would
break the published release, not just a local checkout.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Locale-aware calendar titles (#60): Month, Week and Day move onto the shared
formatter Agenda already used, the year drops out while you're in the current
one, and the Week title names its month rather than restating the day numbers
printed directly below it.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Month, Week and Day each hand-rolled their title from a German-style template —
"$weekday, ${date.day}. $monthName ${date.year}" — so every language got a
trailing ordinal dot and day-before-month order, including ones that write the
month first. English read "Fri, 17. Jul 2026" where it should read
"Fri, Jul 17, 2026".
396a561 already fixed exactly this, but only for Agenda: it added
localizedDateFormatter and migrated the four agenda files, leaving Month, Week,
Day and MonthWidget on their originals. Day and Agenda therefore disagreed about
the same date. All four now route through formatCalendarTitle, so there is one
place left that decides how a title reads.
Two behaviour changes fall out of the reporter's point that the bar wastes space:
The year is dropped while you are in the current year. The title sits above a
grid that already says which year it is; the year's absence is itself the signal
that you are in the current one, and it reappears when you page out — the moment
it starts carrying information. Since a skeleton is a field list, this is just
appending "y", and the locale still places it.
The Week title names a month instead of a day range. "24. Jun – 31. Jun" restated
the day numbers printed in the column headers directly below, in the widest
string in the bar. Naming weekStart's month keeps a straddling week on the
outgoing month until it is fully gone — a week is seven contiguous days, so the
earlier month has a day in it exactly while weekStart is inside it. No straddle
conditional, and the title depends on nothing but weekStart, so it cannot drift
with the direction you paged in from.
currentLocale/localizedDateFormatter move to floret-kit's core-locale (neither is
calendar-specific); LocaleSupport.kt goes away and its 11 callers repoint.
MonthWidget keeps Locale.getDefault() — Glance has no LocalConfiguration and the
widget re-renders on a configuration change, matching AgendaWidget.
Does not touch the FAB stack the issue opens with: the buttons overlaying content
is intentional and matches Google Calendar.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Move the Unreleased changelog under [2.15.0] — 2026-07-15, bump
versionName/versionCode to 2.15.0/21500, and sync the F-Droid
per-version changelog.
Milestone 2.15.0 (all integrated): #21#35#36#39#40#46#52.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a "Soften calendar colours" setting (Settings → Design, default on).
Turning it off paints calendar and event colours raw, exactly as the sync
source (DAVx5/CalDAV) publishes them, instead of the theme-fitting pastels
(#36).
Event/calendar colours now flow through shared eventFill()/eventInk()
helpers gated by a LocalSoftenColors composition local (widgets read the
pref directly). Event titles pick black or white text by the fill's WCAG
relative luminance, so a dark colour stays legible whether softened or raw
(#21) — previously always near-black.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Both landed as community translations through Weblate (French ~33%, Polish
~57%), above the bar already shipped for zh-CN. Add them to locales_config.xml —
the single source of truth for the in-app language picker and Android's per-app
language settings.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Also correct the #35 entry — the empty-today marker is a card in the app but a
plain line in the widget, so call it a "note" rather than a card.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The day view's 48dp edge-to-edge hour gutter centred its labels at 24dp, 4dp
left of the hamburger. Give the gutter content the same 8dp start inset as the
week view so the labels centre on the hamburger (4dp bar inset + 24dp half icon).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The month grid was inset 4dp while its weekday header (and the loading grid) used
8dp, so the week-number column — and the day cells under their labels — sat 4dp
left of where they should. Bring the grid to 8dp: the gutter centre now lands on
the hamburger (4dp bar inset + 24dp half icon), and day cells sit under their
weekday labels.
The week view's header badge and hour labels had the same drift (a 48dp edge-to-
edge gutter centres its content at 24dp, not 28dp). Its top section background
bleeds full-width when scrolled, so instead of insetting the whole content, give
just the gutter content an 8dp start inset to centre it on the hamburger too.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The agenda widget's event rows are plain (a colour stripe + text, no card), so
the rounded "No more events today" surface looked out of place. Render it as a
muted line indented to the event titles instead, matching the widget's style.
The in-app agenda keeps its coffee-cup card, where event rows are cards too.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a Duplicate action to the event detail top bar that opens the shared
event form seeded with a copy of the event as a new, unsaved event, so a
non-recurring event can be re-created with just the day and time changed
instead of re-entering every field.
The copy reuses the existing prefilled-create overlay (createEvent), so it
becomes an independent event with the default reminder applied. Recurrence
is dropped — a duplicate is a single event; the edit form still exposes a
recurrence picker for anyone who wants a series. The occurrence's own times
carry over unchanged. Duplicate is offered for any loaded event, including
read-only ones (WebCal, birthdays): the source calendar is kept only when
it's writable, otherwise the copy resolves to the first writable calendar.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Localize every agenda date via a new localizedDateFormatter helper that lays a
field skeleton out in the locale's own order (Android best-pattern), instead of
a hardcoded day-month-year layout: the range-window summary, the screen's day
headers, and the widget's day headers. This also fixes the window mixing two
orders (e.g. "15 Jul – Aug 13, 2026").
Refine the range bar: the banner drops the range name (it already sits on the
selector button beside it) and shows just the concrete dates; the selector keeps
a subtle neutral surface tint — distinct from the top-bar view switcher's
secondary container so the two don't compete — and its right edge lines up with
the switcher. The anchored empty-today card takes a single event row's resting
corner radius.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add an "Always show today" setting (Settings → Agenda, on by default) that
keeps today as the first entry in both the Agenda screen and its home-screen
widget even once nothing is left today. Under today's normal header a small
"No more events today" card appears — the coffee-cup empty-state motif in the
app, a rounded surface in the widget — so the first rows you see are clearly
today's rather than a future day's.
The anchor is a pure, JVM-tested helper (anchorTodayIfMissing) applied after
past-event filtering; in-app it only kicks in when the window starts on today,
never on a jumped-to date. The widget reads the pref reactively via per-instance
Glance state, mirroring the range/past-event settings.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round out the calendar-intent surface toward AOSP/Etar parity — the app
already handled VIEW (date + event), INSERT, and .ics open/share, but was
missing the edit action and the alternate .ics MIME labels.
- ACTION_EDIT on content://com.android.calendar/events/<id> now opens the
event in the edit form (previously only VIEW → read-only detail existed).
An assistant, task app, or widget can hand an event to Calendula to edit.
A bare EDIT URI with no occurrence extras falls back to the event row's
own DTSTART/DTEND, mirroring the #48 view-event fallback.
- ACTION_EDIT with no event id (AOSP's "edit a new event") maps to the same
prefilled create form as ACTION_INSERT.
- The .ics VIEW/SEND filters now also accept text/x-vcalendar (vCalendar
1.0 / .vcs) and application/ics — the alternate labels the same calendar
data arrives under from some file/mail apps (matches Etar's ImportActivity).
Deliberately excluded: webcal:// / http(s) remote-calendar subscription
(needs INTERNET, which the app doesn't have) and the Google-web-link handler
(Google-specific + network).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A reminder fired ahead of an event on a different day showed only the
event's time (e.g. "09:30 – 10:00"), making it look like it was
happening today. reminderTimeText now prefixes timed events with a
relative day: "Tomorrow"/"Yesterday", the short weekday for another day
this week, or the exact date for anything further out.
The this-week boundary honours the user's "week starts on" setting: the
resolved first day of week is threaded through from ReminderNotifier, so
e.g. a Sunday reads as next week under a Sunday-start locale. All-day
events keep their explicit date (never ambiguous), and cross-midnight
timed events keep both explicit dates.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
