​

4.10.1

​

4.10.0 New Capabilities

​

Now Supporting Flutter 3.19+

Ditto’s Flutter SDK now supports Flutter 3.19 and later, broadening compatibility and making it easier for more developers to integrate Ditto into their apps. This update ensures greater flexibility, allowing teams to adopt Ditto without needing to upgrade immediately to the latest Flutter version.

​

Experimental Support for Hot Restart in Flutter

The Ditto team is excited to preview support for Hot Restarts in Flutter.

To allow users experiment and provide feedback on hot restarts we’ve shipped a specific SDK version with hot restarts enabled by default.

To use hot restarts update your version to 4.10.0-experimental-hot-restart.0. This version is identical to 4.10.0 but has Hot Restart enabled by default. In a upcoming release this capability will be integrated into Ditto’s standard package as a developer mode config that can be set by the user on Ditto class initialization.

Flutter Dart has limited support for Hot Restarts with native binaries, such as Ditto. The Ditto team is working to mitigate these issues. If you run into any issues while experimenting with Flutter Hot Restarts reach out to Ditto Customer Support.

Known issues include:

• Logging must be disabled. To do this set DittoLogger.isEnabled = false immediately after calling await Ditto.init().
• Attempting to use attachments will cause a crash.
• Hot restarting will leak some resources, notably memory. In our testing, approximately 20-30MB is leaked per hot-restart.
• Data integrity cannot be guaranteed. You may experience data corruption when hot-restarting, which may be propagated to any peers that are being synced with.
• Sporadic crashes can occur.

​

Small Peer User Collection Sync Scopes

User Collection Sync Scopes let you control how data from each user collection is shared with other connected Peers. This is useful when you want to fine-tune which collections do not sync, which collections sync with only the cloud (Big Peer) and which collections sync only with connected Small Peer devices.

Sync scopes can be used to control how documents sync, reduce unnecessary data transfer, and improve performance.

​

Details

DQL Syntax for setting a Sync Scope

ALTER SYSTEM SET USER_COLLECTION_SYNC_SCOPES = {
    collection_1: "BigPeerOnly",
    collection_2: "SmallPeersOnly",
    collection_3: "LocalPeerOnly"
}

Example Setting A Sync Scope in the SDKs

In the example below we’ll set the collection local_mesh_orders to only sync with other small peers in the mesh and not with the Big Peer using the SmallPeersOnly sync scope.

var syncScopes = [
"local_mesh_orders": "SmallPeersOnly"
];
await ditto.store.execute(
    query: "ALTER SYSTEM SET USER_COLLECTION_SYNC_SCOPES = :syncScopes",
    arguments: ["syncScopes": syncScopes]);

​

Store Observers Run Multi-Threaded By Default

Ditto store observers (e.g ditto.store.registerObserver(...)) will now run multi-threaded on all platforms (except for in Web Browsers) by default. Prior to 4.10.0 all store observers would run on the same thread which could lead to performance issues for applications with multiple store observers. Users may see performance improvements on applications using more than one store observer, especially during high-volume usage.

Ditto guarantees order sequencing for a single observer. Because observers now run in parallel there is not a guarantee for order sequencing between observers. This behavior change should not have an impact on your application. If you have concerns or questions about this behavior change reach out to Ditto’s Customer Support.

To reset Ditto to only use a single thread for all Ditto observers use the following system parameter. For instruction on how to set a system parameter in your specific SDK see SDK>Customizing System Settings

ALTER SYSTEM SET STORE_OBSERVERS_NUM_THREADS = 1

​

DQL Support for LIKE operator

The LIKE operator can now be used for pattern matching of string values.

Example

SELECT * FROM my_collection
WHERE some_field LIKE 'abc%'

​

DQL New Objects operators

DQL now supports two new methods for querying over the keys and values in an object.

• object_keys(object): Returns an array with all the keys in the given object.
• object_values(object): Returns an array with all the values in object.

​

DQL LIMIT/ OFFSET / ORDER BY support for mutation operators

Users can now use LIMIT, OFFSET and ORDER BY in EVICT and UPDATE statements.

DQL Example

EVICT FROM users
WHERE status = 'inactive'
LIMIT 10;

​

4.10.0 Flutter Specific Changelog

​

4.10.0 Common Changelog