Personally, I’d love if we can simplify our build like that. I’m not sure if I have the expertise around it though.
On Wed, Aug 26, 2020 at 08:47 Volkan Yazıcı <volkan.yaz...@gmail.com> wrote: > IMHO, we should stop adding any new features until we get "verify" working. > > It is really confusing to commit any change in the presence of test > > failures. Thinking it the other way around, if we are okay with making > > releases given these failed tests, let's delete those tests. > > > > For the records, package+verify works fine on GitHub Actions and my machine > > (TM). (Actually, "package" phase is required to get the right Maven reactor > > context for "verify".) Hence, I am strongly inclined to make "verify" work, > > rather than reaching out for work arounds, e.g., "install". > > > > On Wed, Aug 26, 2020 at 3:31 PM Matt Sicker <boa...@gmail.com> wrote: > > > > > I’ve never been able to get the verify task to work properly. I always > need > > > to use install instead. Some modules referring to other modules seems to > be > > > confusing Maven. > > > > > > On Wed, Aug 26, 2020 at 07:16 Volkan Yazıcı <volkan.yaz...@gmail.com> > > > wrote: > > > > > > > Nevermind, fooled by the -Dmaven.test.failure.ignore=true in the Maven > > > > > > > > options. Nevertheless, now we have plenty of tests that are failing. > > > > > > > > > > > > > > > > On Wed, Aug 26, 2020 at 1:19 PM Volkan Yazıcı <volkan.yaz...@gmail.com > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > Hello, > > > > > > > > > > > > > > > > > > How one shall interpret "./mvnw verify" success where there are > > > failures > > > > > > > > > in the test (surefire/failsafe) reports? Should this be marked as a > > > > > > > > > successful build or not? If not, why did "verify" succeed? > > > > > > > > > > > > > > > > > > Kind regards. > > > > > > > > > > > > > > > > > -- > > > Matt Sicker <boa...@gmail.com> > > > > > -- Matt Sicker <boa...@gmail.com>
