db-core/collection-setup

TanStack DB collection setup involves selecting an adapter like Electric or PowerSync and calling createCollection with a defined schema. This centralizes configuration, reducing boilerplate and preventing type-adapter mismatches across REST, RxDB, and local backends.

Creating a TanStack DB collection supports configuration options including getKey, schema, sync, gcTime, autoIndex, defaultIndexType, and syncMode. It also provides lifecycle hooks like onLoad and onLoadSubset to keep data synchronized.

Yes, you can use RxDB or TrailBase as backend adapters for TanStack DB collections. The adapter pattern supports multiple backends including REST query, Electric, PowerSync, RxDB, TrailBase, and local adapters for flexible data synchronization.

Adapter patterns handle data synchronization in TanStack DB by utilizing syncMode, txid tracking, and direct writes. Lifecycle states and hooks like onLoad and onLoadSubset ensure data remains consistently synced across the chosen backend adapter.

The best way to prevent type-adapter mismatches is to centralize collection creation using createCollection. By enforcing a consistent structure with adapter-specific patterns and schemas, you ensure type safety across REST, Electric, PowerSync, and RxDB backends.