maybe we can also mention that semver is more of a rough guidance than anything else. you dont really know if an X+Y that runs fine will really continue to run

maybe we can also mention that semver is more of a rough guidance than anything else. you dont really know if an X+Y that runs fine will really continue to run with X+Y' unless you tests that cover that workload and you've actually run them.