Examining the org.apache.flink.table.descriptors.Kafka class in Flink v1.8,
it has the following startUpModes for consumers:
.startFromEarliest()
.startFromLatest()
.startFromSpecificOffsets(...)However, it does not have a method to support starting from a Timestamp. The FlinkKafkaConsumer supports this feature though. Was it a conscience decision to leave that start up mode out of the table descriptor? If so, what was the reasoning?
