Thanks :) The TCK can be a little fragile, so a small thing, like a dependency update can trigger a number of tests to fail. In this case, it was something like 300 ejb30 tests and 100 or so ejb32 tests. The TCK is massive so the numbers always look worse than the problem really is when this happens. Hopefully my change won't break anything else!
Jon On Mon, Oct 14, 2019 at 12:16 PM Richard Monson-Haefel < [email protected]> wrote: > Nice work, Jon! > > On Mon, Oct 14, 2019 at 5:50 AM Jonathan Gallimore < > [email protected]> wrote: > > > I noticed we had some EJB TCK regressions following this update. > Debugging > > in, it looks like OWB was tripping up with a NullPointerException because > > no InjectionPointService implementation was configure for it (looks like > > this is a new interface in OWB). > > > > I've added these properties to TomEE's openwebbeans.properties: > > > > # New properties in OWB 2.0.11 > > org.apache.webbeans.container.InjectionResolver.fastMatching = false > > org.apache.webbeans.service.DefaultInjectionPointService.implicitSupport > = > > false > > org.apache.webbeans.spi.InjectionPointService = > > org.apache.webbeans.service.DefaultInjectionPointService > > > > in this commit: > > > > > https://github.com/apache/tomee/commit/b5ebdecbf5e7aff1bb61936cfc93177d71e59ee1 > > > > Jon > > > > > -- > Richard Monson-Haefel > https://twitter.com/rmonson > https://www.linkedin.com/in/monsonhaefel/ >
