Re: CDI?

2017-08-07 Thread Matt Sicker
I'm not so sure about deprecated, but DS is the only dependency injection standard in OSGi that respects the dynamic nature of services. CDI, blueprint, etc., all have to rely on hacky proxies to emulate support while adding nonstandard extensions at times. On 7 August 2017 at 17:02,

Re: PaxWeb: Whiteboard and compliance to the R6 specification

2017-08-07 Thread Marc Schlegel
Hi Jérémie We definately want to be spec-compliant and I didnt realize that there are so many points missing. I know we have some parts missing regarding the whiteboard-dtos (mostly error cases). Do you already have access to our Jira? If not, we can add you there so you can submit a

PaxWeb: WebSocketTracker and Object

2017-08-07 Thread Jérémie Brébec
Hello, In pax web 6, a WebSocketTracker is created. This tracker "track" with a ServiceTracker every registration of services with the class "Object". This tracker resolves every service through bundleContext.getService(). As a consequence, every component registered with "Object" are resolved,

Re: [pax-exam] How to simulate beforeClass / afterClass hook *outside* of the OSGi container

2017-08-07 Thread Nicolas Brasey
Hi Christoph, Yes, that is exactly what I tried already, to bundle neo4j and all the dependencies with the maven shade plugin (like the ones in service mix bundles). I realized that this is not really ideal for the following reasons: 1) The build time exploded, to create this huge jar takes more

Re: PaxWeb: WebSocketTracker and Object

2017-08-07 Thread 'Achim Nierbeck' via OPS4J
Hi, the problem I see with this is, if one registers a WebSocket as Service, it usually doesn't need to be of a special Interface ... I can see that a property might help. Will think about this a bit more. Right now you can't disable it. regards, Achim 2017-08-07 14:00 GMT+02:00 Jérémie

Re: PaxWeb: WebSocketTracker and Object

2017-08-07 Thread 'Achim Nierbeck' via OPS4J
BTW, just created the following improvement: https://ops4j1.jira.com/browse/PAXWEB-1119 2017-08-07 18:22 GMT+02:00 Achim Nierbeck : > Hi, > > the problem I see with this is, if one registers a WebSocket as Service, > it usually doesn't need to be of a special Interface