Sync Mode (AutoSync vs Batch)

Remote tables support two synchronization modes:

Single Sync Mode
Batch Sync Mode

Synchronization mode determines:

When changes are sent to the server
How requests are structured
Whether changes are transactional

The mode is configured when creating the table.

Single Sync Mode (autoSync)

In single sync mode, each data change triggers an immediate remote request.

Enable it by setting:

autoSync: true

Example:

const dataManager = spread.dataManager();

const orders = dataManager.addTable("orders", {
    remote: {
        read:   { url: "http://localhost:3000/api/orders", method: "GET" },
        create: { url: "http://localhost:3000/api/orders", method: "POST" },
        update: { url: "http://localhost:3000/api/orders", method: "PUT" },
        delete: { url: "http://localhost:3000/api/orders", method: "DELETE" }
    },
    autoSync: true
});

orders.fetch();

Behavior:

Insert → sends create
Update → sends update
Delete → sends delete
Each operation sends a separate request
No manual submission required

Use single sync mode when:

Immediate persistence is required
Each operation is independent
Simplicity is preferred over batching

Batch Sync Mode

In batch mode, changes are stored locally until explicitly submitted.

Enable it by setting:

batch: true

Example:

const orders = dataManager.addTable("orders", {
    remote: {
        read:  { url: "http://localhost:3000/api/orders", method: "GET" },
        batch: { url: "http://localhost:3000/api/orders/batch", method: "POST" }
    },
    batch: true
});

orders.fetch();

Behavior:

Insert/update/delete operations are cached locally
No immediate network request is sent
All changes are submitted together

To send changes:

orders.submitChanges();

To discard changes:

orders.cancelChanges();

Behavior Comparison

Feature	Single Sync	Batch Sync
Network calls	Per operation	One request
Manual submission	Not required	Required
Transaction-like behavior	No	Yes
Failure handling	Per operation	Entire batch

Batch mode is appropriate when:

Multiple changes must succeed or fail together
Network traffic should be minimized
Server-side transaction handling is implemented

Change Lifecycle

In single sync mode:

User modifies data
Remote request is sent immediately
Server response updates table state

In batch mode:

User modifies data
Change is stored locally
submitChanges() sends all changes
Server returns aggregated result
Table state is finalized

Important Notes

read must always be defined.
Single sync mode requires create, update, and delete.
Batch mode requires batch.
Only one mode should be enabled.
All tables must call fetch() before editing.