Hi Antonio, I just saw something similar today, where someone needed to instruct the bundle plugin to process the annotations early in order to get the tests run:
<groupId>org.apache.felix</groupId> <artifactId>maven-bundle-plugin</artifactId> <extensions>true</extensions> <executions> <!-- Configure extra execution of 'manifest' in process-classes phase to make sure SCR metadata is generated before unit test runs --> <execution> <id>scr-metadata</id> <goals> <goal>manifest</goal> </goals> <configuration> <supportIncrementalBuild>true</supportIncrementalBuild> </configuration> </execution> </executions> regards, toby On 22 June 2017 at 23:22:17, Robert Munteanu (romb...@apache.org<mailto:romb...@apache.org>) wrote: Hi Antonio, On Thu, 2017-06-22 at 07:12 +0000, Antonio Sanso wrote: > anyone :S ? > > On Jun 21, 2017, at 9:12 AM, Antonio Sanso <asa...@adobe.com.INVALID> > wrote: > > > hi, > > > > I was investigating the test failure related to [0] and I thing I > > found an issue/limitation with OsgiServiceUtil. > > Indeed the activateDeactivate method doesn’t need to be able cope > > with the new R6 annotations. > > Fairly enough for case like ActivateDeactivate (aka [0]) being the > > Config interface an inner one I would not know how OsgiServiceUtil > > can be able to manage this scenario… By my understanding the OsgiServiceUtil works on the XML metadata. That metadata should be the same no matter what annotations are used. I suggest to get an isolated test case and file a bug against the sling osgi mocks. Robert > > > > any idea? > > > > regards > > > > antonio > > > > > > [0] https://issues.apache.org/jira/browse/SLING-6957 > >