Hello Chris, Yes sure. I've just done it: https://issues.apache.org/jira/browse/KARAF-3267
Thanks for the help. I might have some time to provide unit test for this bug. I'll attach it to the ticket. WBR, Igor On 9 October 2014 10:13, Christian Schneider <[email protected]> wrote: > Looks like a bug. I think the current model for displaying state does > not handle the case well where there are more than one > *BundleStateService* that applies to a given bundle. > > Can you open an issue? > > Christian > > > On 08.10.2014 10:13, Igor Raizin wrote: > > Hello! > > I gave a try to extend the bundle state in bundle:list command. > > I tried to implement the > *org.apache.karaf.bundle.core.BundleStateService* in the same way as > *org.apache.karaf.bundle.core.internal.BlueprintListener* provides state > for the blueprint. > > At the end I faced the issue... > A bundle is in *Active *state, blueprint started successfully and my > *BundleStateService* returns *Failure* state -> the result state in > bundle:list is *Active, *but I expect to see *Failure*. > > In debug mode I saw in > *org.apache.karaf.bundle.core.internal.BundleServiceImpl#getInfo > *the *BlueprintListener *is the last in list to check the state and it > overrides the *Failure *state from my *BundleStateService*.... > > Is this a bug or it was done on purpose? > If the second than if there any other way to show the bundle state *Failure > *basing on some condition other than blueprint errors or OSGI bundle > state? > > > Thanks for your help. > > WBR, > Igor > > > > -- > Christian Schneiderhttp://www.liquid-reality.de > > Open Source Architecthttp://www.talend.com > >
