> On May 10, 2016, at 8:32 AM, Akira AJISAKA <ajisa...@oss.nttdata.co.jp> wrote: > > Hi developers, > > Before cutting 3.0.0-alpha RC, I'd like to drop JDK7 support in trunk.
+1 yes please. > One concern I think is, when the minimum version is set to JDK8, we need to > configure Jenkins to disable multi JDK test only in trunk. Two points: * The feature to do this in test-patch is already on the roadmap, since HBase asked already: YETUS-369 . We’re really close to releasing Apache Yetus 0.3.0, so it won’t show up until 0.4.0 though. It’s a non-trivial change on it’s own, and based on some conversations I’ve been having this week with other Yetus users, I’m thinking we may want to rewrite how multijdkmode works anyway. * It’s feasible to just disable multijdkmode in precommit and then rely upon the regularly scheduled builds to tell us if branch-2 gets broken by a JDK8-only patch. (FWIW: I’ll be talking about it during ApacheCon, but I’ve been working on a Yetus-ified replacement for those “build the world” jobs such that they are actually useful, less frequent, easier to read, less likely to break due to running in a docker container, etc, etc.) --------------------------------------------------------------------- To unsubscribe, e-mail: hdfs-dev-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-dev-h...@hadoop.apache.org