cloud-sync

Offline-first sync requires a local data store that operates independently, paired with a cloud sync engine that reconciles changes when connectivity returns. Implement using SwiftData or similar local persistence, CloudKit's CKSyncEngine, or iCloud Drive for file-based workflows, handling conflicts through merge strategies or last-write-wins resolution.

CloudKit suits structured data with relational queries and fine-grained sync; iCloud Drive works better for file-based workflows and large assets. Choose CloudKit for database-like schemas, multi-user sharing, and incremental updates; choose iCloud Drive for document storage and when file-level versioning suffices.

Conflict resolution strategies include last-write-wins, merge-based reconciliation, or user-prompted selection. Track modification timestamps and device identifiers; apply CloudKit's CKSyncEngine or custom logic to detect conflicts and resolve them consistently, optionally logging conflicts for audit or recovery.

NSUbiquitousKeyValueStore provides basic cloud sync for small key-value data but lacks fine-grained conflict control and incremental sync. It suits simple preference syncing; for structured data or complex conflict resolution, CloudKit or SwiftData with CKSyncEngine offer greater control and reliability.

Offline-first architectures persist all changes locally during network failures, queuing them for upload when connectivity resumes. Your local store remains readable and writable; sync engines replay queued changes, detect conflicts against server state, and reconcile using your chosen strategy without data loss.

Full sync works for small datasets but scales poorly with large assets or frequent changes. Incremental sync tracks modifications via timestamps or version tokens, reducing bandwidth and latency. CloudKit's CKSyncEngine and iCloud Drive both support incremental patterns; choose based on data volume and update frequency.