Christoph Läubrich

For example `org.osgi.test.cases.jpa.junit.JPAService_1_1_TestCase.testUseStandardDataSourceConfig()` seem to require the h2 jdbc bundle because it uses an url of `jdbc:h2:mem:` but the test bundle has no requirement on this, therefore the TCK happily...

Currently the JPA test do not properly cleanup as they install a bundle but only uninstall it if the EMF was found. This now moves the wait for service inside...

I currently try to upgrade [pax-jpa](https://github.com/ops4j/org.ops4j.pax.jpa) to use the "new" `org.osgi.service.jpa` instead of [osgi.cmpn](https://mvnrepository.com/artifact/org.osgi/osgi.cmpn) but noticed a major problem. If one [looks into the jar](https://repo1.maven.org/maven2/org/osgi/org.osgi.service.jpa/1.1.1/org.osgi.service.jpa-1.1.1.jar) it declares the following in...

The test JPAService_1_1_TestCase#testReportsCorrectPersistenceProviderDetails can run into a NPE as it accesses the providerBundle without a null assertion. Can be seen in this tck run: https://github.com/ops4j/org.ops4j.pax.jpa/runs/27083276801

[Preferences] specify a way to listen for preference changes

2

The [PreferencesService Specification](https://docs.osgi.org/specification/osgi.cmpn/8.0.0/service.prefs.html) currently does not specify any way to be notified about preference changes. This currently requires to use vendor specific extensions like [IEclipsePreferences](https://github.com/eclipse-equinox/equinox/blob/master/bundles/org.eclipse.equinox.preferences/src/org/eclipse/core/runtime/preferences/IEclipsePreferences.java) or bake own solutions. It...

With: - https://github.com/osgi/osgi/issues/509 there was a new attribute added to deprecate a bundle, but this has to be hand-crafted, it would be good to have a `@BundleDeprecated("The reason ...")` bundle...

Assuming a component at least provides one service, one currently has two options: 1. `immediate = true` what results in the component being activated and never deactivated 2. `immediate =...

More than one event should be allowed in case of coordinated delete

2

Fix https://github.com/osgi/osgi/issues/718

Currently Equinox fails this Coordinator TCK test-case: https://github.com/osgi/osgi/blob/11d440e336ac2c22bade54105ed4ef2c9aab79d8/org.osgi.test.cases.cm/src/org/osgi/test/cases/cm/junit/CMCoordinationTestCase.java#L384-L441 the reason is that here is is asserted that *exactly one* event is triggered: https://github.com/osgi/osgi/blob/11d440e336ac2c22bade54105ed4ef2c9aab79d8/org.osgi.test.cases.cm/src/org/osgi/test/cases/cm/junit/CMCoordinationTestCase.java#L439 because the test-case assumes that the block...

Fix https://github.com/eclipse-ee4j/jaxb-ri/issues/1807