Note that it is because we ran our CI with Java 15 that we were able to report a bug that was actually fixed for Java 16, so that's cool.
Longer term, I'd like us to run builds against LTS versions and the latest version, Java 14 today, with newer versions allowed to fail. Gary On Mon, Aug 24, 2020 at 9:47 AM Matt Sicker <boa...@gmail.com> wrote: > I like the test strategy they’re using at Mina: 8, latest LTS, and latest > release. Those are the main supported lines of OpenJDK (from their > distributors at least; I don’t think the project has official support), and > since other versions won’t receive updates, there’s not much point in > continued use. > > On Mon, Aug 24, 2020 at 04:51 Xeno Amess <xenoam...@gmail.com> wrote: > > > Hi. > > > > There exist a bug on jdk15, and which results a test failed on jdk15 for > > > > commons-lang. > > > > JDK guys said they will fix it on jdk16 but no plan for migrating the fix > > > > to jdk15. > > > > So maybe we should shut that test on jdk15, or we filter out some buggy > > > > parts in that test and do not do it on jdk15? > > > > Any ideas? > > > > -- > Matt Sicker <boa...@gmail.com> >
