Andreas: what I meant to say:
karaf-features-plugin has some basic agreement about maven scopes and transitive dependencies and how to translate them into osgi run-time consumable feature.xml; to me - that sounds close enough to what pax exam would also need; it would be great if Harald and Jean-Baptiste could agree on the same approach for both pax-exam and karaf; cheers Andrei. -------- Original Message -------- Subject: Re: pax exam config duplication From: Andreas Pieber <anpie...@gmail.com> To: General OPS4J <general@lists.ops4j.org> Date: Fri 28 Sep 2012 04:04:23 PM CDT > > Not really. Karaf uses my paxexam-Karaf which is basically configured > with an entire distribution where your test bundle and exam is > deployed to. If your devs are based on Karaf this might be the easiest > solution anyhow. > > Kind regards, > Andreas > > On Sep 28, 2012 6:37 PM, "Andrei Pozolotin" > <andrei.pozolo...@gmail.com <mailto:andrei.pozolo...@gmail.com>> wrote: > > also: karaf sort of solves this problem via karaf-features-plugin, > right? > > -------- Original Message -------- > Subject: Re: pax exam config duplication > From: Harald Wellmann <hwellmann...@gmail.com> > <mailto:hwellmann...@gmail.com> > To: General OPS4J <general@lists.ops4j.org> > <mailto:general@lists.ops4j.org> > Date: Fri 28 Sep 2012 11:10:03 AM CDT >> This is an FAQ, and if it were easy, I guess we'd have implemented a >> solution already ;-) >> >> You have to worry about transitive dependencies, API bundles vs. >> implementation bundles (like org.osgi.core vs. >> org.apache.felix.framework) and lots of other things. For every >> scenario or proposal I've seen so far, it is easy to construct >> counterexamples where the suggested approach would not work. >> >> I'm not saying it can't be done at all, but my feeling is it would be >> a lot of effort to catch all sorts of weird cases, and I don't think >> it would pay off. >> >> Another benefit of listing provisioned bundles explicitly is working >> with different configurations in different tests of the same Maven >> project, which would not be possible if the configuration were derived >> from the POM. >> >> Best regards, >> Harald >> >> _______________________________________________ >> general mailing list >> general@lists.ops4j.org <mailto:general@lists.ops4j.org> >> http://lists.ops4j.org/mailman/listinfo/general >> > > > _______________________________________________ > general mailing list > general@lists.ops4j.org <mailto:general@lists.ops4j.org> > http://lists.ops4j.org/mailman/listinfo/general > > > > _______________________________________________ > general mailing list > general@lists.ops4j.org > http://lists.ops4j.org/mailman/listinfo/general
_______________________________________________ general mailing list general@lists.ops4j.org http://lists.ops4j.org/mailman/listinfo/general
