+1 for code cleanup, obviously pending merge conflict resolution and failed test examination. I'm not confident in how `AnalyzeSerializablesJUnitTest` desire to not break backwards compatibility works in practice.
It also looks like we've let this PR get stale enough that the test failures are a bit strange. It looks like `StressNewTest` failures were `UnitTest` failures, but didn't fail in `Build`. But `Build` has been replaced by a dedicated `UnitTest` job in these past three weeks. Still, somewhat concerning that the stress went red where the `Build` didn't. [ Full content available at: https://github.com/apache/geode/pull/2546 ] This message was relayed via gitbox.apache.org for [email protected]
