You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Marking a task done never stuck — the real cause behind 1.10.2's
symptom. A subtask id is UNIQUE across the whole eav_entities table, but
two copies of the same task arriving from sync carry the same subtask ids.
Saving the second copy hit the unique constraint, saveTasks reported
failure, and the transaction rolled back — so every task's save was
discarded on every write, forever, while the app's in-memory list drifted
further from the database. Writing a subtask now claims its id from any
stale row first, so one duplicated task can no longer wedge the whole store.
Every SQLite failure is logged (log show --predicate 'subsystem == "com.sharingan.app"') instead of being swallowed, so a save that stops
persisting says why.