0.4.0

• Support exactly-once consistency in BigQuery sink. This is integrated with Datastream and Table/SQL API.
• Add Flink metrics for monitoring BigQuery sink.
• Package unshaded guava dependency for enforcing the correct version used by BigQuery client.

0.3.0

• Support BigQuery sink in Flink's Table API.
• BigQuery sink's maximum parallelism is increased from 100 to 128, beyond which the application will fail.
• Modifies the following config keys for connector source in Table API:

Before	After
read.discoveryinterval	read.discovery-interval
credentials.accesstoken	credentials.access-token
read.streams.maxcount	read.streams.max-count

0.2.0

• Release BigQuery sink with at-least-once support.
• Avro's GenericRecord to BigQuery proto is the only out-of-the-box serializer offered for now.
• BigQuery sink's maximum parallelism is capped at 100, beyond which the application with fail.

0.2.0-preview

Preview of the Apache Flink BigQuery connector sink, offering at-least-once write consistency via Flink's datastream API.

0.1.0-preview

First preview of the Apache Flink BigQuery connector, offering source (BigQuery storage read) functionality for bounded and unbounded tables.