- Create docs/ai/today-tag-architecture.md documenting the virtual tag
pattern, explaining why TODAY_TAG must never be in task.tagIds and
how membership is determined by task.dueDay instead
- Add reference section to operation-log.const.ts pointing to related
timing constants in other domain files (sync.const.ts, meta-sync)
This addresses documentation gaps identified in the maintainability review.
Add safeguards to prevent silent failures when new entity types are
added without proper registration:
- Add canary test that fails if EntityType union grows but test arrays
aren't updated (verifies count matches expected 20 types)
- Create developer checklist at docs/ai/adding-new-entity-type-checklist.md
documenting all required changes when adding new entity types
Create a quick reference document with ASCII diagrams summarizing:
- Area 1: Write Path
- Area 2: Read Path (Hydration)
- Area 3: Server Sync (SuperSync)
- Area 4: Conflict Detection
- Area 5: Conflict Resolution (LWW)
- Area 6: SYNC_IMPORT Filtering
- Area 7: Archive Handling
Includes decision tables, key invariants, and file references.
Areas 8-12 placeholders for future additions.
- C.5: Correct "Current timestamp" to "Preserved maximum timestamp
from local ops" - critical for correct LWW semantics
- C.7: Rewrite "Late-Joiner Replay" as "SYNC_IMPORT Filtering" to
reflect actual "Clean Slate Semantics" implementation:
- CONCURRENT ops are now DROPPED, not replayed
- Uses SyncImportFilterService, not removed _replayLocalSyncedOpsAfterImport()
- Vector clock comparison determines causality, not UUIDv7 timestamps
- Fix CLAUDE.md path reference from docs/op-log/ to docs/sync-and-op-log/
- Rewrite section 2c in architecture diagrams to reflect actual implementation:
- Was: UUIDv7 timestamp replay (removed feature)
- Now: Vector clock filtering with clean slate semantics
- Update service reference from removed _replayLocalSyncedOpsAfterImport()
to SyncImportFilterService.filterOpsInvalidatedBySyncImport()
- Clarify that CONCURRENT ops are dropped, not replayed
- Update README.md with correct links and implementation status
- Add E2EE to implementation status (now complete)
- Update e2e-encryption-plan.md to mark as implemented
- Update operation-log-architecture.md:
- Add missing service files to file reference section
- Add E2EE to Part C and "Recently Completed" sections
- Fix broken references, update last modified date
- Remove reference to non-existent tiered-archive-proposal.md
Add comprehensive tests for operation applier testing gaps:
- Partial archive failure: verifies that when archive handling fails
midway through a batch, previously processed ops are reported
as successful while remaining ops are not applied
- Effects isolation: confirms that only bulkApplyOperations action
is dispatched (not individual action types), which is the key
architectural benefit preventing effects from firing on remote ops
- Multiple archive-affecting ops: ensures remoteArchiveDataApplied
is dispatched exactly once when batch contains multiple
archive-affecting operations
Total: 4 new test cases, 24 tests now passing
Move scattered architecture docs into centralized locations:
- Move operation-log docs from src/app/core/persistence/operation-log/docs/
to docs/op-log/
- Flatten docs/sync/sync/ nested structure to docs/sync/
- Move supersync-encryption-architecture.md from docs/ai/ to docs/sync/
- Copy pfapi sync README to docs/sync/pfapi-sync-overview.md
- Update all cross-references to use new paths
This improves discoverability and keeps architecture documentation
separate from source code.
Disable automatic data repair during sync/hydration to:
- Expose validation failures instead of silently "fixing" them
- Prevent REPAIR operations from propagating potentially corrupted data
- Help identify the root cause of archived tasks with subtasks being lost
When validation fails, an alert is now shown with error details.
The original repair logic is preserved in comments for easy re-enabling.
When syncing 100+ operations, the server's piggyback limit (100 ops)
was causing Client B to miss operations. The server returned 100 piggybacked
ops but latestSeq was set to the actual server sequence (e.g., 199).
Client B then updated lastServerSeq to 199, so subsequent download got 0 ops.
Changes:
- Server: Add hasMorePiggyback flag to UploadOpsResponse when piggyback limit
is reached and more ops exist
- Client: When hasMorePiggyback is true, store lastServerSeq as the max
piggybacked op's serverSeq instead of latestSeq, ensuring subsequent
download fetches remaining ops
- Effects: Change switchMap to mergeMap in autoAddTodayTagOnMarkAsDone to
ensure ALL mark-as-done actions trigger planTasksForToday
- Flush service: Implement two-phase wait strategy (poll queue + acquire lock)
to ensure all pending writes complete before upload
- Add diagnostic logging for operation counts at key stages
Test: High volume sync with 50 tasks + 49 mark-as-done (197 ops total)
now correctly syncs all done states via piggyback (100) + download (97).
Implements the ability to change the encryption password by deleting
all server data and uploading a fresh snapshot with the new password.
Server changes:
- Add DELETE /api/sync/data endpoint to delete all user sync data
- Add deleteAllUserData() method to SyncService
Client changes:
- Add deleteAllData() to OperationSyncCapable interface
- Implement deleteAllData() in SuperSync provider
- Add EncryptionPasswordChangeService to orchestrate password change
- Add DialogChangeEncryptionPasswordComponent with validation
- Add "Change Encryption Password" button to sync settings (visible
when encryption is enabled)
- Add translations for all new UI strings
Testing:
- Add 10 unit tests for EncryptionPasswordChangeService
- Add 14 unit tests for DialogChangeEncryptionPasswordComponent
- Add 5 E2E tests for complete password change flow
- Add changeEncryptionPassword() helper to SuperSyncPage
Also fixes:
- Add missing deleteAllData() to MockOperationSyncProvider
- Fix typo S_FINISH_DAY_SYNC_ERROR -> FINISH_DAY_SYNC_ERROR
- Delete docs/ai/sync/server-sync-architecture.md which incorrectly
stated "Status: Not Started" when server sync is fully implemented
- Delete deprecated src/app/features/time-tracking/store/archive.effects.ts
which was empty and marked for removal
- Update local-actions.token.ts comment to reference the correct
archive-operation-handler.effects.ts file
- Update hybrid-manifest-architecture.md status to Implemented
- Add documentation index (README.md) for operation-log docs
- Add index for docs/ai/sync/ with document categorization
- Replace duplicate docs with redirect stubs to canonical locations
- Add effect rules, multi-entity rules, and config constants to operation-rules.md
- Update vector-clocks.md with operation log integration section
- Expand super-sync-server README with API details and security features
- Update pfapi sync README with operation log integration info
- Add Section 2c to architecture diagrams with mermaid diagrams
explaining the late-joiner problem and solution
- Add Section C.7 to main architecture doc with code examples
- Add Example 4 to vector-clocks.md explaining dominance filtering
- Update Last Updated dates to December 12, 2025
Add class-level documentation explaining:
- Only SuperSync uses operation log sync (API-based)
- Legacy providers skip operation log sync entirely
- File-based methods exist for future extensibility but are never called
- If file-based sync is ever enabled, encryption/decryption must be added
The original review incorrectly identified file-based operation log sync
as having "critical gaps". This update corrects the analysis:
- File-based operation log sync is dead code, never called
- Only SuperSync uses operation log sync (API-based)
- Legacy providers skip operation log sync entirely and use pfapi LWW
Status changed from "NEEDS FIXES" to "PRODUCTION READY"
Mark all previously identified test gaps as resolved:
- Race conditions: 6 new tests in compaction service
- Schema migration: 6 version mismatch tests in hydrator service
- Download retry: 4 tests now passing (was pending())
- All critical and missing scenarios now have test coverage
- Add December 11 security review results (all issues verified as addressed)
- Document newly added tests (3-way conflict, delete vs update, large conflict sets)
- Update test coverage status table with completed items
- Mark supersync E2E test plan as IMPLEMENTED
- List all test files and their coverage areas
Update documentation to reflect recent simplifications:
- OperationCaptureService now uses simple FIFO queue
- OperationApplierService uses fail-fast approach
- ArchiveOperationHandler is unified for local and remote operations
- Update diagrams and file references
- Add "Recently Completed" items for December 2025
Consolidate all archive-affecting operations into ArchiveOperationHandler
to eliminate duplicate code and create a single source of truth.
Changes:
- Add ArchiveOperationHandlerEffects for local operations
- Rename handleRemoteOperation to handleOperation
- Add isArchiveAffectingAction helper for action filtering
- Make isIgnoreDBLock conditional on action.meta.isRemote
- Remove archive logic from individual effect files
- Update feature-stores.module.ts to use new effect
- Add archive write points documentation
This ensures archive operations (moveToArchive, restoreTask,
flushYoungToOld, deleteProject, deleteTag, deleteTaskRepeatCfg,
deleteIssueProvider) all go through one handler for both local
and remote operations.
Analysis shows the alleged race conditions are intentional defensive patterns:
- Double sync wait ensures stability during date changes
- Effects are complementary with staggered timing, not competing
- Infinite loop scenario explicitly handled in code
Note: Tests exist but are disabled due to Dropbox SDK mocking issues.
Re-evaluated the "missing subtask cascade in tag deletion" issue.
The current behavior is correct:
- A subtask with a surviving parent is NOT orphaned - it's accessible
through the parent in the UI
- If a parent is orphaned, its subtasks ARE correctly deleted
- Subtasks always have parentId, so they're never directly orphaned
This is intentional behavior, not a bug.
The snapshot was being saved before validation ran. If validation found
corruption and repaired it, the snapshot was stale/wrong. This fix
moves validation (CHECKPOINT C) BEFORE saving snapshot in both:
- Tail replay path (after replaying ops following a snapshot)
- Full replay path (when no snapshot exists)
Now the snapshot contains validated/repaired state.
Includes regression tests that verify validation happens before snapshot
save in both code paths.
The incrementCompactionCounter() function returned 1 without persisting
when no state cache existed. This caused the compaction threshold to
never be reached, leading to unbounded operation log growth.
Also updated review docs:
- Marked compaction counter bug as fixed
- Identified memory leak in TaskElectronEffects as false positive
(take(1) ensures automatic cleanup)
- Identified TODAY_TAG "board-style" claim as false positive
(code correctly REMOVES TODAY_TAG from task.tagIds)
- Identified cache invalidation claim as false positive
(sequence-based invalidation works correctly)
- Identified repair mutex race condition as false positive
(JavaScript single-threaded, no await between assignment)
Tasks created without explicit dueDay don't appear in TODAY view
due to recent selector changes. Add sd:today short syntax to ensure
tasks have dueDay set and appear correctly in tests.
Add Karma-based integration tests for the operation log system using real
IndexedDB. Tests cover multi-client synchronization scenarios and state
consistency verification.
New test files:
- multi-client-sync.integration.spec.ts (16 tests)
- state-consistency.integration.spec.ts (17 tests)
Test utilities:
- TestClient helper for simulating multiple clients with independent vector clocks
- Operation factory helpers for creating test operations
- Add unit tests for isPersistentAction type guard.
- Fix compilation errors in task scheduling components caused by removed reminderId/removeReminderFromTask.
- Fix type error in create-sorted-blocker-blocks.spec.ts.
- Fix lint errors in various files.
- Add syncTimeSpent persistent action for batched time tracking sync
- Accumulates duration per task, syncs every 5 minutes or on task stop
- Local dispatch: no-op (state already updated by second-by-second ticks)
- Remote dispatch: applies batched duration to task.timeSpentOnDay
- Fix remote archive handling to include time tracking data
- writeTasksToArchiveForRemoteSync now moves historical time tracking
data to archiveYoung alongside tasks
- Add flushYoungToOld persistent action for deterministic archive state
- Replaces inline flush logic with action-based approach
- Remote clients replay the flush operation to maintain consistent
archiveOld state without syncing large files
- Vector clock ordering ensures flush happens after preceding operations
- Implement processRemoteOps migration logic in OperationLogSyncService using SchemaMigrationService.
- Add MAX_VERSION_SKIP check and user warning via SnackService.
- Fix Log class to support dynamic level switching (fixes PFAPI tests).
- Fix operation-log-stress imports.
- Skip disabled MultiTabCoordinatorService tests.
- Add unit tests for migration logic.
- Update ValidateStateService documentation to reflect better error handling for isRelatedModelDataValid.
- Update Part C status to Complete as Server Sync is implemented.
- List implemented components for Part C.
- Fix lint error in startup.service.ts
- Store compaction counter in state_cache to share across tabs/restarts
- Add bySyncedAt index for faster getUnsynced() queries
- Remove in-memory opsSinceCompaction counter from effects
- Reset counter in compaction service after successful compaction
Implements A.7.12, A.7.13, and A.7.15 from the operation log architecture:
- A.7.12 Migration Safety: Backup state_cache before migration, restore
on failure, detect interrupted migrations on startup
- A.7.13 Tail Ops Migration: Migrate operations to current schema before
replay during hydration
- A.7.15 Unified Interface: Extend SchemaMigration with migrateOperation
and requiresOperationMigration for explicit migration decisions
Changes:
- operation-log-store: Add backup/restore methods for state_cache
- schema-migration: Add operation migration support and validation
- operation-log-hydrator: Integrate backup safety and tail ops migration
- Update architecture docs to reflect implementation status
