The conditionally-required match-mode and sortDir radios placed
`defaultValue` inside `props`, which Formly ignores — its core extension
reads `field.defaultValue`. So opening any board with >=2 included or
excluded tags left those required fields undefined and the Save button
disabled. Even with the default lifted, Formly skips defaults for
fields hidden at init and on hidden→visible transitions unless
`resetOnHide: true` is set, so picking a sortBy re-locked the form.
Lift `defaultValue` to field level and add `resetOnHide: true` on
`includedTagsMatch`, `excludedTagsMatch`, and `sortDir`. Add a unit
spec exercising the panel fieldGroup and a Playwright e2e that drives
the dialog end-to-end.
* feat:Add a keyboard shortcut to collapse/expand groups
* fix(collapsible): scope group behavior to isGroup and address review feedback
* fix(collapsible): scope group behavior, fix a11y focus, clean CSS, and update docs
* feat(work-view): apply isGroup to Overdue/Later/Done/Recurring sections
Aligns implementation with the wiki keyboard-shortcuts docs added in this
PR — those sections are described as participating in the expand/collapse-all
group shortcuts but were missing the [isGroup]="true" input.
* feat(tasks): ArrowLeft on top-level task collapses parent group
When a top-level task (no parent) is focused and the existing arrow
handling would otherwise call focusPrevious, hand off to the closest
ancestor collapsible group instead: collapse it and move focus to the
group header. Makes the new group shortcuts reachable from the normal
task-driven keyboard flow rather than requiring a Tab to the header.
Sub-tasks and selected-task hide-detail behavior are unchanged.
* feat(collapsible): ArrowUp/Down on group header focuses adjacent task
When a group header is focused, ArrowDown moves focus into the next
<task> element in document order (the first task of the group when
expanded; the next group's first task when collapsed). ArrowUp moves
focus to the previous task. Pairs with the ArrowLeft hand-off so the
group shortcut and task navigation form a single continuous keyboard
flow.
* feat(tasks): ArrowUp/Down on group-edge task focuses the header
ArrowUp on the first task of a group focuses the current group's
header; ArrowDown on the last task focuses the next group's header.
Routed through new handleArrowUp/handleArrowDown methods so other
focusPrevious/focusNext callsites (move-to-backlog, etc.) keep their
existing semantics.
* refactor(util): extract findAdjacentFocusable for arrow nav
Three places had near-identical DOM-walking logic for "find the next/prev
focusable element across tasks and group headers". Consolidates them
into a single utility with 9 unit tests, plus a shared GROUP_NAV_SELECTOR
constant exported from collapsible.component.
Behavior change: ArrowUp/Down on a group header now considers other
collapsed group headers as next-focusable (previously walked only across
<task> elements). For an all-collapsed list, ArrowDown on a header now
moves to the next header instead of skipping to a far-away task.
* refactor: trim findAdjacentFocusable + tighten collapsible API surface
Multi-agent review consensus:
- Drop the compareDocumentPosition fallback in findAdjacentFocusable.
Both real callers pass an element matching the selector; the fallback
branch only existed to satisfy a synthetic test. Util shrinks to ~10
lines of obvious code.
- Re-privatize setAllGroupExpanded and expandIfCollapsed — neither is
called from outside CollapsibleComponent.
- Scope focusHeader's selector to ':scope > .collapsible-header' so a
nested collapsible projected through ng-content can't steal focus.
---------
Co-authored-by: Johannes Millan <johannes.millan@gmail.com>
* feat(mentions): match substrings instead of only prefixes
The default mention autocomplete filter (used by #tags, @due dates,
and +projects in task titles) only matched items whose label started
with the search string. Switch to substring matching so typing 'aa'
matches a tag named 'xxaaxx'.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* feat(mentions): rank prefix matches above mid-string matches
Sort filtered results by indexOf(searchString) ascending so items whose
label starts with the query appear above items where the match is in the
middle. The pre-existing alphabetical ordering from addConfig() is
preserved as a tiebreaker via the stable sort.
Refactors the filter to share a single getLabel() helper between the
filter and sort steps.
---------
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-authored-by: Johannes Millan <johannes.millan@gmail.com>
Mirrors the issue-panel tab-drag pattern: cdkDropList on the mat-tab-group
with cdkDrag on each user-board label. Drop dispatches a new BoardsActions.sortBoards
action persisted via op-log (OpType.Move, isBulk).
Closes#4163
Non-default locale data is registered lazily via requestIdleCallback
(commit d989c064f3). If the schedule view renders before the idle
callback fires — e.g. cold start with zh-cn during initial sync replay —
formatDate(date, fmt, locale) throws NG0701 "Missing locale data".
Wrap the four schedule formatDate call sites in a safeFormatDate helper
that falls back to DEFAULT_LOCALE on throw. After the idle callback
registers the user's locale, any signal change re-runs the computed
and the user's locale takes effect.
Refs #7383
Make explicit in the backlog-query help text that the default behavior
without a token imports every open issue, not just the user's, so anyone
enabling auto-import without credentials understands the scope.
The default backlog query 'sort:updated state:open assignee:@me' fails with
HTTP 422 ("The listed users cannot be searched...") when the provider has
no token configured, because GitHub's Search API can't resolve @me without
an authenticated request. This broke auto-import for public repos, which
the docs and UI advertise as not requiring a token.
Fall back to 'sort:updated state:open' when no token is set, and update
the help text to describe both defaults.
Closes#7381
When 'sync focus sessions with time tracking' was enabled, selecting a
task auto-started the session and bypassed the preparation screen even
if 'skip preparation screen' was off. The two settings should be
independent.
syncTrackingStartToSession$ now early-returns when isSkipPreparation is
off, leaving the user on the preparation screen (already shown by
autoShowOverlay$) so they must click 'Start' to trigger the rocket
animation — matching the manual-start flow in focus-mode-main.
Replace the misleading "downloaded to documents folder" toast with a
dialog that displays the logs in a textarea and offers Copy and (on
native) Share file actions, so users can always retrieve logs even
when the system share sheet fails silently or app-private storage
is invisible to file managers on Android 11+.
Copy uses ShareService.copyToClipboard so the execCommand fallback
covers older Android WebViews — the same class of silent-failure
platform that motivated the dialog.
The edit dialog dispatched updateTaskRepeatCfg with the entire
TaskRepeatCfg object as the changes payload. rescheduleTaskOnRepeatCfgUpdate$
filters by `field in changes`, so every save matched the filter — even when
only the time changed. The effect then asked getNextRepeatOccurrence for the
next slot with lastTaskCreationDay === today, which always returns tomorrow,
shifting today's instance forward by one day per save.
Compute and dispatch a true delta. Reuse distinctUntilChangedObject for
deep array comparison. Derive isRelevantChangesForUpdateAllTasks from the
delta so we don't diff twice.
Fixes#7373
Real-time drag tracking with velocity-based dismiss decision for the
mobile bottom sheet:
- Custom pointer-event handler on the panel header drags the sheet
height in 1:1 with the finger; touch-action: none on the header lets
pointermove be passive (no main-thread arbitration per event).
- Pointer listeners run outside the Angular zone — drag state is not
template-bound, so per-event change detection was pure overhead.
- Single-pointer gesture with setPointerCapture; multi-touch and
cross-pointer events are rejected.
- pointercancel resets state without invoking the close decision so
OS-interrupted gestures do not dismiss the sheet.
- Release decision: upward fling expands; downward close requires
velocity floor AND projected end-position past the close zone
(Apple WWDC18 / Vaul / Android BottomSheetBehavior pattern). Held
finger and slow drags never trigger close.
- Close animation uses translateY for compositor-only paint, with
velocity-coupled duration and easing for natural inertia. Reads
offsetHeight directly so the slide-off matches the rendered height.
- A .closing class lifts min-height: 20vh through the close animation;
hand-off from .dragging to .closing is one step to avoid a one-frame
snap to 20vh between dragend and slide-off.
- Regrab during a close/expand animation cancels the in-flight timers
and clears residual styles before starting a new drag.
* feat(gitea): add label-based include and exclude filters
Adds two optional config fields to the Gitea issue provider. filterLabels
is a comma-separated allowlist forwarded to the issues endpoint as the
`labels` query param (AND-matched server-side). excludeLabels is a
comma-separated denylist applied client-side, since Gitea/Forgejo's
issues endpoint has no negation syntax. Scoped labels (scope/name) work
with both.
This lets users point multiple Gitea integrations at the same repository
and route disjoint subsets of issues into different Super Productivity
projects, e.g. one project filtering `project/foo` and another excluding
it.
Filters apply only to ingestion (search suggestions and auto-import of
open issues). Already-imported tasks keep refreshing regardless of label
changes, matching how the existing `scope` filter behaves.
* docs(gitea): document label include/exclude filtering
Update the issue-integration comparison page with the new label
allowlist and denylist filters, including scoped-label support.
* feat(gitea): enforce filterLabels AND match client-side
Gitea's two issue endpoints (`/repos/{o}/{r}/issues` and
`/repos/issues/search`) parse the `labels=` query param into different
internal fields and produce different semantics (AND vs OR), so the
same `filterLabels` config behaved differently on auto-import vs
manual search. The AND on the auto-import path is also a Gitea bug
(go-gitea/gitea#33509) that may flip to OR upstream at any time.
Apply AND-matching on the client (mirroring how `excludeLabels`
already filters client-side) so behavior is consistent across both
endpoints and independent of server version. The `labels=` query is
still sent as a coarse pre-filter (always a superset of the right
answer regardless of server semantics). Forgejo inherits the same
endpoints from Gitea, so the same fix covers it.
- Synced pt-br.json with the latest en.json keys
- Fixed outdated or incorrect translations
- Improved grammar, clarity, and natural phrasing
- Standardized recurring UI terms for consistency
- Ensured placeholders and interpolation tokens remain correct
This update brings pt-BR translation closer to the source and improves overall UX for Brazilian users.
* test(e2e): open attachment dialog via detail panel after #7314
The attach-dialog entry point moved out of the task context menu in
PR #7314 and now lives in the detail panel. Update the WebDAV sync
attachment test to use openTaskDetailPanel() and click the attachment
input-item instead of right-clicking and looking for an "Attach" menu
item that no longer exists.
* refactor(tasks): drop dead addAttachment from task context menu
Leftover from #7314, which moved the attach dialog into the detail
panel. The method, its TaskAttachmentService injection, and the
DialogEditTaskAttachmentComponent import are unreachable from the
template.
* fix(api): reject inherited fields when creating subtask via REST
POST /tasks with parentId silently dropped any supplied projectId/tagIds
(the reducer forces tagIds=[] and projectId=parent.projectId), so callers
got 201 with values different from what they sent. Reject the request
with 400 UNSUPPORTED_FIELD instead, symmetric with how subTaskIds and
parentId-on-PATCH are handled.
* refactor(simple-counter): inline countdown wrapper, document set invariant
Inline the single-call-site `_hasStartedRepeatedCountdown` private wrapper.
Promote the `setCountdownRemaining` invariant note to JSDoc so it surfaces
in IDE tooltips at every call site — paused-display is gated by
`hasStartedCountdown`, and writing through `setCountdownRemaining` without
a prior `startCountdown` call would silently leave the value invisible.
* ci(plugins): run unit tests when plugin code changes
New workflow runs each plugin's `npm test` in a parallel matrix when
its directory or a shared package (plugin-api, vite-plugin) changes
on a PR. Detects per-plugin changes via three-dot `git diff` against
the PR base.
* fix(sync): break iOS WebDAV conflict-dialog loop (#7339)
Two compounding bugs trapped iOS WebDAV users in a per-minute conflict
dialog that no button could resolve:
1. FileBasedSyncAdapter's snapshotReplacement heuristic re-fires
gapDetected on every sync from a non-writing client (clientId \!=
excludeClient is true forever), so the download keeps coming back
with snapshotState and OperationLogSyncService keeps throwing
LocalDataConflictError despite the local clock already dominating
the remote snapshot. Skip hydration and conflict when
compareVectorClocks(local, remote) is EQUAL or GREATER_THAN, gated
on both clocks being non-empty so a fresh client still hydrates a
legacy/clockless snapshot.
2. SyncWrapperService._openConflictDialog$ filtered undefined out of
the afterClosed() stream, so a programmatic close (iOS WebView
lifecycle, re-entry) collapsed the observable and firstValueFrom
threw EmptyError — the user's Use Local/Use Remote/Cancel click
never reached the resolution branches. Drop the filter so undefined
flows through to the existing cancellation path.
The dominate-skip deliberately does NOT append result.newOps to the
op log: VectorClockService.getEntityFrontier is last-write-wins by
seq, and writing historical remote ops at the current tail would
regress per-entity frontiers and let future LWW resolution overwrite
local data. Trade-off documented inline.
Adds an adapter-level integration reproducer that asserts gapDetected
re-fires forever for a non-writing client (the upstream loop trigger),
a 3-client WebDAV e2e that reproduces the loop end-to-end against a
real provider, plus service-level tests for the dominate-skip, the
empty-clock guard, the concurrent-clock conservative path, and
consecutive-sync loop prevention.
* fix(sync): improve Dropbox auth dialog UX on sandboxed Linux (#7139)
The "Get Authorization Code" button silently failed on Flatpak because
shell.openExternal rejects without renderer feedback when the
org.freedesktop.portal.Desktop talk-name isn't granted. After the user
gave up and reopened the dialog, the second attempt failed again with
`invalid_grant: invalid code verifier` because each call to
Dropbox.getAuthHelper() generated a fresh PKCE verifier that no longer
matched the originally-shown URL's challenge.
Three changes:
- Cache the in-flight PKCE Promise on the Dropbox provider so concurrent
callers and consecutive dialog opens share one verifier+URL pair.
Cleared on successful exchange, on clearAuthCredentials(), and on a
rejected generation (so a one-time crypto failure doesn't poison the
session). Five regression tests cover reuse, success-clear, explicit
clear, concurrent calls, and rejection-recovery.
- Render the auth URL as user-selectable text under a <details>
disclosure. Escape hatch when both shell.openExternal and the
clipboard portal are denied — the user can triple-click to select and
Ctrl+C the URL into a manually-opened browser. Adds
D_AUTH_CODE.MANUAL_URL_HINT translation key.
- Pipe shell.openExternal rejections through
errorHandlerWithFrontendInform so the existing IPC.ERROR snack
channel surfaces a "Could not open the link in your browser" message
instead of swallowing the failure to electron-log. Wrapped in a
try/catch since errorHandlerWithFrontendInform throws synchronously
if the renderer isn't ready.
The Flathub manifest also needs --talk-name=org.freedesktop.portal.Desktop
and --socket=wayland to fully fix the user-reported issue, but that
change lives in the flathub repo.
* refactor(sync): rename dialog-sync-initial-cfg to dialog-sync-cfg
Pure rename — no behavior change. The component is the canonical sync
config dialog, used both for first-time setup and editing. The "Initial"
qualifier was misleading once it became the only sync-config surface
(see #7362).
* refactor(sync): consolidate sync actions into the sync dialog
Move Re-authenticate, Force overwrite, and Restore from history into
DialogSyncCfgComponent so the dialog is the canonical surface for sync
configuration and management.
- Re-authenticate (OAuth providers, currently Dropbox) shown inline
when the active provider has getAuthHelper() and is already authed.
Bypasses the dirty-form gate so credentials can refresh without
losing in-progress edits.
- Force overwrite and Restore from history live in a new "Advanced"
section gated on isWasEnabled() (existing config) and disabled when
the form is dirty with a "Save changes first" tooltip. Restore stays
SuperSync-only.
- Force overwrite reuses the existing native confirm in
SyncWrapperService.forceUpload() — no extra confirmation dialog to
avoid double-confirm on the other 5 internal callers.
The legacy buttons on the settings page are removed in a follow-up
commit.
Refs #7362
* refactor(sync): move WebDAV Test Connection button into the sync dialog
Test Connection was previously injected into the inline sync form on
the settings page. It needs to live wherever the sync form is
rendered. The next commit removes the inline form, so move the
injection into DialogSyncCfgComponent first.
No user-visible change: the button still appears in the WebDAV
section of the sync form, just sourced from the dialog component now.
* refactor(config-page): replace inline sync form with status row + buttons
The Sync & Backup tab now hosts a compact status surface, not a full
form clone:
- When sync is disabled or unconfigured: empty-state hint + "Set up
sync" primary button.
- When configured: provider name, optional "Authentication required"
pill (OAuth providers) and "Encrypted" pill, then "Sync now" primary
button + "Configure" secondary button.
The dialog is the canonical sync config surface — reachable from this
tab, the header sync icon (right-click / long-press), or
SyncWrapperService when first-time setup is needed. Force overwrite
and Restore from history live inside the dialog now (see prior commit).
Removes _buildSyncFormConfig (~250 lines of formly button injection),
the empty authStatus placeholder field in SYNC_FORM, and the unused
tour-syncSection class.
Refs #7362
* refactor(sync): consolidate BTN_FORCE_OVERWRITE translation key
The SUPER_SYNC.BTN_FORCE_OVERWRITE key was a duplicate with no
consumers — its English copy ("Force Overwrite Server") even drifted
from the canonical F.SYNC.S.BTN_FORCE_OVERWRITE ("Force Overwrite").
The third occurrence under D_ENTER_PASSWORD ("Use Local Data") stays
— that's a different conflict-resolution action, not a duplicate.
Refs #7362
* refactor(sync): simplify sync dialog — collapse advanced fields, kebab menu
The previous "Advanced" zone (hr + heading + hint + button stack) plus
always-visible interval/manual-only/compression toggles made the dialog
visually noisy. Two changes:
- Move syncInterval and isManualSyncOnly into the existing
"Advanced Config" collapsible alongside compression. The collapsible
is hidden for SuperSync (which uses fixed settings) — most users
never need to expand it.
- Move Force overwrite + Restore from history out of the dialog body
into a kebab (more_vert) menu in the dialog title row. They act on
saved config — the kebab placement makes that scope clear and removes
the dirty-form gate / "Save changes first" ambiguity.
The kebab is hidden during first-time setup (gated on isWasEnabled).
The Re-authenticate button stays inline as before.
Drops three translation keys added in the prior commit (ADVANCED,
ADVANCED_HINT, SAVE_FIRST_HINT) — no longer referenced.
Refs #7362
* fix(sync): drop primary color from Cancel button in sync dialog
Cancel is a secondary action; only the affirmative button (Save)
should carry the primary color.
* fix(sync): clarify Dropbox info text for the new dialog flow
The previous copy said "Click the button below to authenticate" — but
in the consolidated dialog there is no inline Authenticate button for
first-time setup; OAuth runs as part of Save. Make the copy describe
the actual flow instead of pointing at a button that no longer exists.
* refactor(sync): move Force overwrite + Restore into the Advanced collapsible
The kebab menu introduced earlier hid Force overwrite and Restore from
history behind a small icon and required users to know that "more_vert"
in the dialog title contained sync actions. Fold them into the existing
top-level collapsible instead — the same one that already hosts sync
interval, manual-only, and compression toggles.
- Rename the collapsible's label from "Advanced Config" to "Advanced"
via a new sync-specific translation key (the global ADVANCED_CFG key
is shared with issue providers, so we don't touch it).
- The dialog component appends Force overwrite (warn) and Restore from
history (SuperSync only) to the collapsible's fieldGroup in edit
mode. First-time setup keeps the original SuperSync hide so the
collapsible never appears empty.
Re-authenticate stays as an inline button below the form because its
visibility depends on an async provider.isReady() check that doesn't
fit Formly's sync hideExpression.
Refs #7362
* refactor(sync): single Advanced collapsible per provider, all actions inside
Each provider now has exactly one "Advanced" collapsible:
- non-SuperSync: top-level (interval, manual-only, compression,
enable-encryption, plus injected Re-authenticate / Force overwrite
in edit mode)
- SuperSync: nested in the SuperSync section (server URL, plus injected
Force overwrite / Restore from history in edit mode)
The inner SuperSync collapsible's label switches from "Advanced Config"
to "Advanced" so both surfaces read the same. Re-authenticate moves
from a button below the form into the non-SuperSync Advanced collapsible
as a Formly button gated on `syncProvider === Dropbox`. Drops the async
provider.isReady() check — re-auth is shown for Dropbox in edit mode
regardless, since `force=true` works for both stale-token and switching
accounts.
Honors the rule that no buttons sit below the Advanced collapsible
inside the dialog.
Refs #7362
* refactor(sync): use stroked style for all dialog buttons + add Nextcloud test
- Every action button inside the sync dialog now uses btnStyle: 'stroked'
for a consistent outlined appearance: Re-authenticate, Force overwrite,
Restore from history, WebDAV/Nextcloud Test connection, file-based and
SuperSync Enable encryption, SuperSync Get token, LocalFile folder
pickers.
- Force overwrite keeps btnType: 'warn' so it stays warn-coloured but as
an outline rather than a filled warn button.
- The conditional stroked-when-token-set toggle on Get token is dropped
in favour of always-stroked.
- Re-authenticate and Restore previously misused btnType: 'stroked' (a
color slot) instead of btnStyle: 'stroked' (the actual style flag) —
the formly-btn template ignored the unrecognised color, so they
rendered as filled buttons. Fixed.
- Adds a Test Connection button to the Nextcloud section, matching the
WebDAV one — Nextcloud's serverUrl + userName are translated into the
WebDAV-shaped baseUrl client-side before invoking WebdavApi.testConnection.
Refs #7362
* refactor(sync): apply multi-review fixes — fragility, races, dedup, lifecycle
Cross-validated findings from the multi-agent review:
**Correctness / lifecycle**
- `fields` becomes a `computed` signal of `_getFields(isWasEnabled())`,
removing the manual `fields.set()` re-sync after the enabled flag
flips. Single source of truth.
- Reset `_tmpUpdatedCfg._isInitialSetup = false` when the dialog opens
in edit mode so SuperSync encryption-warning hideExpressions stop
treating returning users as first-timers.
- Replace ad-hoc `Subscription` aggregator with `takeUntilDestroyed` on
both subscriptions in the dialog component.
**Race conditions**
- `syncSettingsForm$` subscription on the settings page now goes through
`switchMap` so a fresh emission cancels any in-flight `provider.isReady()`
probe — late callbacks can no longer overwrite newer state.
- Drop the redundant `_cd.markForCheck()` after `signal.set()` — signals
integrate with OnPush automatically.
- Use `??` instead of `||` when preserving `globalCfg` flags during
provider switch, so an explicit `false` is honoured.
**De-duplication**
- Promote `NextcloudProvider._buildNextcloudBaseUrl` to a public static
`buildBaseUrl()`. Dialog's `_testNextcloudConnection` now imports it
rather than duplicating the URL-building logic.
- New `OAUTH_SYNC_PROVIDERS` set in `provider.const.ts`. Re-auth button's
hideExpression and the settings-page `requiresAuth` derivation can both
reference it instead of hard-coding `Dropbox`.
**Simplicity**
- Settings page `syncStatus` collapses from 5 fields to 3 (providerId,
needsAuth, isEncrypted) — empty state derives from `providerId === null`.
- Drop redundant `D_INITIAL_CFG.ADVANCED` translation key in favour of
the existing `T.G.ADVANCED_CFG`.
**Security / logging**
- Re-auth catch path now logs `{ name: e.name }` instead of the raw error
object — log history is exportable, no auth detail leakage.
Refs #7362
* refactor(sync): apply pass-2 review fixes — error paths, marker, factory
Cross-validated findings from the second multi-agent review:
**Correctness — keep the syncStatus stream alive on probe failure**
- Wrap the inner async block in try/catch. Previously, a rejected
`provider.isReady()` would propagate as an observable error through
switchMap, killing the outer subscription and freezing the status
pill. On failure, default to `needsAuth: true` so the row stays
meaningful.
- Replace the imperative `subscribe + .set` bridge with `toSignal()`.
**Security — finish the redaction**
- The first pass only redacted SyncLog.err; the snack `translateParams`
still passed raw `e.message` into a `[innerHtml]` sink. Snack copy
now uses the same `_redactErrorName(e)` helper so neither surface
carries token/URL/stack details. Helper handles `null`/`undefined`
thrown values explicitly.
**Architecture — structural marker decouples routing from i18n**
- Both Advanced collapsibles now carry `props.syncRole: 'advanced'`.
The dialog routes on this stable identifier instead of the shared
`T.G.ADVANCED_CFG` translation key (also used by Jira/Azure DevOps
forms — a global rename would have silently broken sync).
**Simplicity — collapse 5 button factories into one**
- New `_actionBtn({ text, onClick, btnType?, hideExpression?, className? })`
helper. Each per-action factory becomes a 3-4 line call site. ~60
lines net reduction in the dialog component.
- Drop the orphan `props: { dropboxAuth: true }` marker on the Dropbox
fieldGroup — its consumer (`_buildSyncFormConfig`) was deleted.
Refs #7362
* refactor(sync): apply pass-3 review fixes — auth label asymmetry, dead snack param
Pass-3 review surfaced two real findings on top of polish:
**Correctness — non-OAuth providers no longer mislabelled "needs auth"**
- The pass-2 try/catch returned `needsAuth: true` unconditionally on
probe failure, which would surface "Authentication required" for
WebDAV/Nextcloud/LocalFile — providers that don't have an auth
helper. Now we capture `requiresAuth` BEFORE the throwable
`isReady()` call and reuse it in the catch, so non-OAuth providers
fall through to `needsAuth: false` even on transient failures.
**Simplicity — drop the dead snack translateParams**
- The `INCOMPLETE_CFG` translation key has no `{{error}}` placeholder,
so the redacted error name was silently dropped by the translation
pipe. Removing the param documents the actual contract: the snack
shows static credentials-missing copy; the discriminator goes only
to the (redacted) log.
**Polish — cleaner types, tighter helper**
- `_redactErrorName` collapses to two-arm helper: `Error.name` for
Error instances, `'UnknownError'` for everything else. The
null/undefined/typeof branches were exporting internal jargon to
the user.
- The `props.syncRole` marker now uses a typed `SyncCollapsibleProps
extends FormlyFieldProps` interface (exported from sync-form.const)
instead of a `Record<string, unknown>` cast. Eliminates the cast at
both write and read sites.
Refs #7362
* refactor(sync): polish round — shareReplay, subscription cleanup, doc tightening
Three deferred items from prior reviews now applied:
- shareReplay({bufferSize:1, refCount:true}) on syncSettingsForm$. The
config-page subscribes long-term and the dialog re-subscribes per
open; previously the no-provider branch re-fetched the
sync-config-default-override.json asset on every fresh subscription.
Cold-start cost is now amortised across both consumers.
- Drop the manual Subscription aggregator on ConfigPageComponent. The
remaining cfg$ + queryParams subscriptions now use
takeUntilDestroyed(this._destroyRef); ngOnDestroy + OnDestroy go
away.
- Tighten dialog _isInitialSetup handling. Pass-3 review noted the
pre-set + Object.assign pattern reads as if order matters when it
doesn't. Move the flag into the spread payload so the intent is
obvious at the call site, and use \!v.isEnabled directly so the
semantics are explicit (true ⇔ first-time setup).
- JSDoc on forceOverwrite() documenting why no Material confirm is
added — SyncWrapperService.forceUpload already gates the action with
a native confirmDialog, and that gate also serves the 5 internal
snackbar callers (LockPresentError / EmptyRemoteBody / JsonParse /
LegacySyncFormatDetected / retry). Wrapping here would double-confirm.
Refs #7362
* refactor(sync): pass-4 polish — refCount:false, widen updateTmpCfg, trim docs
Cross-validated findings from the fourth review pass:
- shareReplay flips to refCount:false. Pass-4 reviewer noted that
refCount:true only deduplicated the rare case where the dialog
opens *while the settings page is mounted*. The far more common
path — header-icon → dialog with no settings page open — saw the
dialog become the sole subscriber, complete via .pipe(first()),
and drop refCount to 0, so the next open re-fetched the JSON
default-override anyway. With refCount:false the cached value is
retained for the lifetime of SyncConfigService, matching the
comment's stated intent at negligible memory cost.
- updateTmpCfg signature widens to `SyncConfig & { _isInitialSetup?:
boolean }`. Drops the call-site cast (a code smell that pretended
the type was wider than it was) and documents that _isInitialSetup
is a real, expected payload field.
- Trim the misleading comment around `_isInitialSetup: \!v.isEnabled`.
The "Spread last so Object.assign honours this" line wasn't true —
it's an object literal property, not a spread. New comment states
the actual semantics: first-time setup ⇔ sync was previously
disabled.
- Trim forceOverwrite() JSDoc from 6 lines to 2 — same content,
matches the project's preference for terse method comments.
Refs #7362
Adds a new termination provision (§9(4) DE/EN, §7 HTML) stating that
accounts inactive for 3 months may be deleted along with their synced
data. Bumps "Last updated" to 25.04.2026 in all three ToS files.
The "Get Authorization Code" button silently failed on Flatpak because
shell.openExternal rejects without renderer feedback when the
org.freedesktop.portal.Desktop talk-name isn't granted. After the user
gave up and reopened the dialog, the second attempt failed again with
`invalid_grant: invalid code verifier` because each call to
Dropbox.getAuthHelper() generated a fresh PKCE verifier that no longer
matched the originally-shown URL's challenge.
Three changes:
- Cache the in-flight PKCE Promise on the Dropbox provider so concurrent
callers and consecutive dialog opens share one verifier+URL pair.
Cleared on successful exchange, on clearAuthCredentials(), and on a
rejected generation (so a one-time crypto failure doesn't poison the
session). Five regression tests cover reuse, success-clear, explicit
clear, concurrent calls, and rejection-recovery.
- Render the auth URL as user-selectable text under a <details>
disclosure. Escape hatch when both shell.openExternal and the
clipboard portal are denied — the user can triple-click to select and
Ctrl+C the URL into a manually-opened browser. Adds
D_AUTH_CODE.MANUAL_URL_HINT translation key.
- Pipe shell.openExternal rejections through
errorHandlerWithFrontendInform so the existing IPC.ERROR snack
channel surfaces a "Could not open the link in your browser" message
instead of swallowing the failure to electron-log. Wrapped in a
try/catch since errorHandlerWithFrontendInform throws synchronously
if the renderer isn't ready.
The Flathub manifest also needs --talk-name=org.freedesktop.portal.Desktop
and --socket=wayland to fully fix the user-reported issue, but that
change lives in the flathub repo.
Two compounding bugs trapped iOS WebDAV users in a per-minute conflict
dialog that no button could resolve:
1. FileBasedSyncAdapter's snapshotReplacement heuristic re-fires
gapDetected on every sync from a non-writing client (clientId \!=
excludeClient is true forever), so the download keeps coming back
with snapshotState and OperationLogSyncService keeps throwing
LocalDataConflictError despite the local clock already dominating
the remote snapshot. Skip hydration and conflict when
compareVectorClocks(local, remote) is EQUAL or GREATER_THAN, gated
on both clocks being non-empty so a fresh client still hydrates a
legacy/clockless snapshot.
2. SyncWrapperService._openConflictDialog$ filtered undefined out of
the afterClosed() stream, so a programmatic close (iOS WebView
lifecycle, re-entry) collapsed the observable and firstValueFrom
threw EmptyError — the user's Use Local/Use Remote/Cancel click
never reached the resolution branches. Drop the filter so undefined
flows through to the existing cancellation path.
The dominate-skip deliberately does NOT append result.newOps to the
op log: VectorClockService.getEntityFrontier is last-write-wins by
seq, and writing historical remote ops at the current tail would
regress per-entity frontiers and let future LWW resolution overwrite
local data. Trade-off documented inline.
Adds an adapter-level integration reproducer that asserts gapDetected
re-fires forever for a non-writing client (the upstream loop trigger),
a 3-client WebDAV e2e that reproduces the loop end-to-end against a
real provider, plus service-level tests for the dominate-skip, the
empty-clock guard, the concurrent-clock conservative path, and
consecutive-sync loop prevention.
New workflow runs each plugin's `npm test` in a parallel matrix when
its directory or a shared package (plugin-api, vite-plugin) changes
on a PR. Detects per-plugin changes via three-dot `git diff` against
the PR base.
Inline the single-call-site `_hasStartedRepeatedCountdown` private wrapper.
Promote the `setCountdownRemaining` invariant note to JSDoc so it surfaces
in IDE tooltips at every call site — paused-display is gated by
`hasStartedCountdown`, and writing through `setCountdownRemaining` without
a prior `startCountdown` call would silently leave the value invisible.
POST /tasks with parentId silently dropped any supplied projectId/tagIds
(the reducer forces tagIds=[] and projectId=parent.projectId), so callers
got 201 with values different from what they sent. Reject the request
with 400 UNSUPPORTED_FIELD instead, symmetric with how subTaskIds and
parentId-on-PATCH are handled.
Leftover from #7314, which moved the attach dialog into the detail
panel. The method, its TaskAttachmentService injection, and the
DialogEditTaskAttachmentComponent import are unreachable from the
template.
The attach-dialog entry point moved out of the task context menu in
PR #7314 and now lives in the detail panel. Update the WebDAV sync
attachment test to use openTaskDetailPanel() and click the attachment
input-item instead of right-clicking and looking for an "Attach" menu
item that no longer exists.
* fix:google calendar text exceeding background
* Chore: removed comments
* fix(issue-panel): wrap long provider names instead of clipping
Replace the clip-and-tooltip approach with white-space: normal on the
button label so long provider names (e.g. "Google Calendar (iCal)")
wrap to a second line rather than overflowing or being silently cut
off. Also reverts unrelated package-lock, source-map, and lint-comment
changes that snuck into the original PR.
Refs #7331
---------
Co-authored-by: Johannes Millan <johannes.millan@gmail.com>
The plugin-api contract for fieldMapping.toTaskValue/toIssueValue declares
ctx as { issueId: string; issueNumber?: number }, but the host adapter
was constructing ctx with issueId only at both call sites
(_extractTaskFieldsFromIssueWithSyncValues and _applyFieldMappingPull).
This caused mappings that interpolated ctx.issueNumber, like the GitHub
plugin's title mapping (`#${ctx.issueNumber} ${title}`), to render
"#undefined ..." whenever the title was pulled via the update path.
The initial add-task path was unaffected because titles came pre-formatted
from mapSearchResult.
Reads issue.number from the PluginIssue record at both sites and
includes it in ctx, falling back to undefined when absent so non-GitHub
plugins are unaffected.
Adds an advanced checkbox on the GitHub issue provider that, when enabled,
drops the hardcoded `is:issue` filter in both backlog auto-import and
in-provider search. Lets users sync PRs (assigned, authored, or
review-requested) into their backlog with a query like
`state:open is:pull-request (assignee:@me OR review-requested:@me)`.
Default is off, existing behavior is unchanged.
The JSDoc promised "first occurrence on or after startDate," but the
implementation anchored on `max(startDate, today)` — so a past startDate
was silently advanced forward. This was the root cause of #7344 and
forced the introduction of `getFirstOccurrenceAnchor` as a call-site
workaround.
- Refactor `getFirstRepeatOccurrence` to always anchor on startDate.
Drop the unused `today` parameter. Return null when startDate is
missing instead of silently falling back to 1970-01-01; existing
callers already handle null.
- Delete `getFirstOccurrenceAnchor` + spec — the workaround is obsolete
now that the util honors its contract.
- Switch `rescheduleTaskOnRepeatCfgUpdate$` to `getNextRepeatOccurrence`.
Editing an existing recurring cfg needs the next occurrence from
today, not retro-anchoring on startDate.
- Rebase deterministic spec tests whose `startDate` was past-dated but
expected today-anchored results under the old contract. Intent is
preserved; only the setup now matches the contract explicitly.
Follow-up to #7344. When a timed task with a past dueWithTime was converted
to recurring, the preserved first-occurrence date produced a past remindAt
and the reminder module flagged it as overdue, firing a "missed reminder"
popup immediately on save.
Clamp the anchor to today when cfg.startTime is set and the planned day is
past. Non-timed past-day preservation (the #7344 fix) is untouched.
The paste field sat below the "Get Auth Code" button, so on iPhone the
on-screen keyboard covered it and users could not paste the code. After
the user taps the link the field now moves to the top of the dialog and
receives focus synchronously with the click, so iOS opens the keyboard
on a visible target. Adds autocomplete/autocorrect hardening to avoid
mangling pasted tokens.
Fixes discussion #7340.
- Rename util getAnchorDateForRepeatCfgConversion → getFirstOccurrenceAnchor
to match neighboring "what it returns" naming convention
(getEffectiveRepeatStartDate, getNewestPossibleDueDate, etc.).
- Collapse the duplicated multi-line "why anchor" comments at both call
sites into single-line pointers; full rationale lives in the util JSDoc.
- JSDoc on the util now explicitly documents why rescheduleTaskOnRepeatCfgUpdate$
does NOT use it (different semantics for cfg edits vs. initial conversion).
No behavior change. Follow-up polish on the #7344 fix.
The "X starts at {{start}}" countdown banner was rendering remindAt (the
reminder fire time, e.g. 2:30 for a 30-min-before reminder) instead of
the task's actual scheduled start (dueWithTime, e.g. 3:00), confusing
users about when the task actually begins. Fall back to remindAt for
reminder-only tasks that have no dueWithTime.
Also extend the "leave banner as is" early-return guard to include
dueWithTime, otherwise rescheduling a task while keeping the same
reminder offset would leave a stale time in the banner.
Closes#7343
When an existing task with a past dueDay was converted to a recurring
task, both conversion effects called getFirstRepeatOccurrence with
`new Date()` as the anchor, which silently advanced the first occurrence
to the next future date matching the recurrence pattern — shifting the
task out of the user's planned date without consent.
Anchor on the task's existing dueDay only when it equals cfg.startDate
(the dialog default). If the user overrode startDate, fall back to
today — preserving existing behavior for all other flows.
Closes#7348.
Sizing
- Clock face scales 175/225/280/110px across breakpoints; digits scale
44/58/72/28px via --clock-time-size. Tokens hoisted to .clock-display.
- Breathing dot (flow mode) tracks the clock face at 80% of diameter.
- Duration slider fills the clock face in countdown prep and its
editable text scales with the circle (~16% of diameter).
- tabular-nums + line-height: 1 on .clock-time so countdown digits
don't shift width/metrics while ticking.
Layout
- Pomodoro settings cog moved inside the clock face, centered at
bottom; fades in on hover like the other in-circle controls.
- --control-offset derived from clock-face/clock-time/icon-button so
pause/done icons sit ~4px from the digits at every breakpoint,
clamping to --s on the smallest circle.
- Duration slider hides the underlying clock digits via an
is-duration-slider-active class bound from isShowDurationSlider().
Component API
- input-duration-slider now accepts --input-duration-slider-size and
--input-duration-slider-value-size (defaults 130px / 16px preserved
via var() fallbacks — the three existing dialog consumers are
unchanged).
- breathing-dot accepts --breathing-dot-size (default 144px preserved).
POST /tasks now accepts `parentId` and creates a subtask via
TaskService.addSubTaskTo (inheriting the parent's projectId). Invalid
parent references return 404 PARENT_NOT_FOUND; nested parents return
400 INVALID_PARENT. `subTaskIds` on POST and `parentId`/`subTaskIds`
on PATCH are now rejected with 400 UNSUPPORTED_FIELD instead of being
silently dropped.
Fixes#7346
The existing copy ("Data Damage Detected" / "Your data appears to be
damaged" / "This should not normally happen") implied a defect when the
dialog actually surfaces routine Checkpoint-D cleanup of orphan
references after Last-Write-Wins resolution of concurrent cross-device
edits — working as designed, just communicated as a bug.
Reword D_DATA_REPAIR_CONFIRM, D_DATA_REPAIRED, and the DATA_REPAIRED
toast to describe the behavior accurately while preserving the "report
if frequent" invitation. No behavioral change.
Catalogs the six post-sync corruption shapes that can trigger the
"data was automatically corrected" alert. Calls isRelatedModelDataValid
directly to bypass Typia and isolate cross-model checks.
Attach the console listener in the legacy-migration helper after
page.unroute('**/*.js') so the benign `Failed to load resource:
net::ERR_FAILED` messages from the intentional seeding-phase aborts
stop surfacing as console errors. pageerror stays attached early
since no JS runs while bundle loads are aborted.
Chromium's StatusIconLinuxDbus writes the tray PNG to TMPDIR and hands
gnome-shell the path via DBus IconThemePath. Inside a strict snap, /tmp
is bind-mounted to /tmp/snap-private-tmp/snap.<name>/tmp with mode 0700
on the parent, so gnome-shell (outside the mount namespace) cannot
stat the file and falls back to the three-dot placeholder icon.
Redirecting TMPDIR to \$XDG_RUNTIME_DIR (/run/user/\$UID) puts the
write where both sides can read it. Matches Signal Desktop's snap and
the official Electron snapcraft guide. electron-builder's SnapOptions
docstring still claims this is a default but app-builder-lib 26.8.1's
snap target no longer injects it.
Refs #7298
Template binds `countdownTime$ | async` which is `number | null`. The
function already tolerated `null` for `total`; widen `remaining` the
same way so strict template type checking passes. Also pick up the
prettier indentation fix on the progress-circle element.
The non-native describe block stubbed cfg$ as { subscribe: () => {} }
which worked while the effect only subscribed to cfg$ lazily. After
6614dfa4 introduced a _reminderCfg$ class-field initializer that pipes
on cfg$ at construction time, the stub fails with
"cfg$.pipe is not a function" for all 4 non-native tests.
Swap to NEVER — a real Observable that never emits. Matches the
stub's intent and has .pipe().
Follow-up to 6466ddc0 addressing multi-agent review feedback.
- Extract _reminderCfg$ = cfg$.pipe(map(c => c?.reminder),
distinctUntilChanged()). Both scheduling effects now subscribe to
this narrow slice, so unrelated global-config edits (theme, sync,
pomodoro, etc.) no longer trigger reschedule loops and serial
native bridge calls.
- Tighten the flip-disabled test to assert the notification IDs
passed to cancelReminder, not just the call count.
- Replace `any` in the new spec blocks with ReminderConfig,
TaskWithReminder, and a small TestCfg shape.
- Update SYNC-SAFE JSDoc to note the cfg reactivity.
Refs #6980
The "Disable all reminders" setting was only enforced at the in-app
dialog layer. On Android, native notifications scheduled via
AlarmManager kept firing regardless of the toggle, because the mobile
NgRx effects did not observe the config.
- scheduleNotifications$ now subscribes to cfg$ and, when
disableReminders is true, cancels every tracked reminder and
short-circuits further scheduling.
- scheduleDueDateNotifications$ extends its existing disabled branch
to respect the master disableReminders flag in addition to
notifyOnDueDate.
Known remaining gap: SyncReminderWorker (WorkManager) can still
schedule reminders from the server every 15 min for users with
background sync configured. Tracked as a follow-up.
Refs #6980
The sync conflict dialog title and body still hardcoded "Dropbox" in 9
locales even when users sync via WebDAV, making the prompt misleading.
English was corrected in 2020 but the translations never got updated.
Fixes#7339
