Thanks for pointing this out. Yes, it is necessary to depend on beam-model-pipeline because it contains the proto pipeline definition. If we want to load the max timestamp from the proto we have to depend on it.
I agree, it's not very elegant and causes some issues which we still have to resolve (looking at the Java build failures). [ Full content available at: https://github.com/apache/beam/pull/6381 ] This message was relayed via gitbox.apache.org for [email protected]
