Currently, we publish only the jars for the connectors we build, but the runtime requires bundling all dependencies as well. Currently, this is done in the Dockerfile, as that is the main way to run the connectors. In the near future, we plan to run these in distributed mode, meaning we should be able to pull a single file containing all the dependencies (e.g., a .zip file). For example, take a look at the iceberg-connector releases -- https://github.com/databricks/iceberg-kafka-connect/releases. This would be useful, for example, if we want to run the connectors on the Amazon Managed Kafka service or on a distributed kafka-connect cluster, where we can just upload/pull the zip and run.
Currently, we publish only the jars for the connectors we build, but the runtime requires bundling all dependencies as well. Currently, this is done in the Dockerfile, as that is the main way to run the connectors. In the near future, we plan to run these in distributed mode, meaning we should be able to pull a single file containing all the dependencies (e.g., a .zip file). For example, take a look at the iceberg-connector releases -- https://github.com/databricks/iceberg-kafka-connect/releases. This would be useful, for example, if we want to run the connectors on the Amazon Managed Kafka service or on a distributed kafka-connect cluster, where we can just upload/pull the zip and run.