BUILDING.md never mentioned the floret-kit submodule, so its instructions
could not produce a working build — a plain clone fails to configure. It
also understated the SDK requirement: compileSdk is 37, not the "36+" it
claimed. Both matter now that the Codeberg move is meant to make an
outside clone viable.
The docs map advertised superpowers/specs/ and superpowers/plans/, neither
of which exists any more, while omitting BUILDING.md and RELEASING.md.
Dropped the dead rows and listed what is actually there. Purged the
remaining superpowers references in .planning/ and ARCHITECTURE.md's
dangling "see plan 03" the same way — a pointer to a deleted file is worse
than no pointer.
ARCHITECTURE.md still called OptionCard "the app's only sanctioned
selection-dialog style", which the 2026-07-07 flip to full-screen pickers
reversed, and described no layer for floret-kit despite it being a
compiled-from-source composite build. Its CI note credited Gitea Actions;
contributor CI runs on Codeberg now.
PROJECT.md's Source section described the old world outright ("hosted on
self-hosted Gitea") and now records the split: Codeberg canonical, Gitea
build infrastructure.
Documentation pass after the 2.0 milestone:
- docs/ARCHITECTURE.md — principles (provider as single source of truth,
observer-driven UI, JVM-first tests, no network), layer + reminder
mermaid diagrams, navigation (overlay/held-key, no nav lib), and the
provider lessons (recurring-write invariants, conflict snapshots)
- docs/README.md — map of what documentation lives where, incl. the
convention that superpowers/ plans are historical artifacts while
.planning/ stays current
- README.md — showcase layout (centered header, badges, screenshot
gallery from the fastlane assets, grouped features, install/build/
architecture/roadmap sections); renders on Gitea
- .planning/{PROJECT,REQUIREMENTS,STATE}.md unstaled: read-only-V1 talk
removed, V1/V2 checklists marked shipped, state points at v3 + the
Locations & People go/no-go
release.yaml gains a gitea-release job: on every tag push it extracts the
tag's CHANGELOG section and creates a Gitea release with it as the notes.
No APK assets — distribution stays with the F-Droid repo. Idempotent
(skips an existing release), gated on the test job only so notes appear
even when the F-Droid upload hiccups.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The create form (v1.2) now edits: a pencil on the detail screen (writable
calendars only, contextual WRITE upgrade like delete) opens it prefilled via
EventDetail.toEditForm; populated sections always show, the calendar is
fixed, and a dirty-check writes only changed columns (pristine saves are
no-ops). Saving a dirty recurring event parks in SaveUiState.AwaitingScope
and asks how far the change reaches (Google model): "only this event" =
modified-occurrence exception via CONTENT_EXCEPTION_URI (empty optionals as
explicit NULLs since the provider clones the parent row), "this and all
following" = series split (insert new event first, then truncate), "all
events" = series-row update with the time delta applied to the series
DTSTART. A changed rule drops the exception option. Delete gained the same
middle scope.
Recurrence: EventForm.rrule + SimpleRecurrence (FREQ/INTERVAL/UNTIL/COUNT +
weekly BYDAY with locale-ordered weekday toggles) behind a picker on create
and edit; unrepresentable rules render humanized (shared ui/common
RecurrenceText) and survive verbatim. UNTIL validation flags rules ending
before the event starts.
Provider lessons baked in (verified on-device via adb probes): instance
caches regenerate only from an update's own values, so truncation sends the
full time-column set (truncateSeries) — RRULE-only updates left a stale
duplicate occurrence on the split day; UNTIL is written as the local end of
day in UTC (toRRule(zone), previousLocalDayEndUtcMillis) so UTC+x zones
can't leak an extra day. Reminder edits reconcile against actual provider
rows, keeping untouched rows' methods.
Tests: RecurrenceTest (parse/render/round-trip, truncation), update/exception
mapper paths, repository pass-throughs, prefill + populatedFields, raw-title
mapper.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The date-picker half of M2 is dropped entirely; the "Today" half already
shipped in v0.5. V1 is now feature-complete and only a polish/QA pass
remains before v1.0.
Updated the living planning docs (ROADMAP, STATE, REQUIREMENTS) and the
design spec; corrected the v0.5.0 CHANGELOG note that promised M2 would
return in v1.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- ROADMAP: mark v0.1 (Foundation & CI) as complete
- REQUIREMENTS: move Foundation & CI from Active to Validated (shipped)
- AndroidManifest: drop redundant android:label and android:theme on
MainActivity - both inherited from <application>
- build.gradle.kts: move ui-tooling-preview to debugImplementation
(@Preview annotations are dev-only; release APK stays smaller)
All foundation verification (lint + test + assembleDebug) still green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code review on Task 2 flagged that 'Kotlin 2.3' could mislead a future
reader into bumping to 2.4.0, which has no KSP release yet. Spell out
the full versions and the KSP pairing constraint in both PROJECT.md
(Stack section) and REQUIREMENTS.md (Constraints section).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
