That was me, and yes, that’s the correct interpretation - the feedback as several other contributors*.
We don’t like to take the implied responsibility for maintaining jars we repackage ourselves with OSGi MANIFEST.MF headers for a variety of reasons. Luckily the situation is getting better, based on my experience. Generation tools are better, more convenient and therefore more prevalent (thanks bnd-maven-plugin). OSGi headers are cropping up from source more and more regularly (I’d love to see another scan of central to see how many jars supply them, broken down by popularity). ServiceMix provides distributions, though they sometimes seemingly haven’t been tested. Karaf wrap means things frequently just work at deployment time Best, Dan. * Balazs, I’d meekly suggest grouping the ‘other’ responses and including into the graphs where significant. For example, Karaf shell clearly deserves a place if the extra responses were grouped, as does Vaadin for UIs, etc. > On 8 Jun 2016, at 20:53, Balázs Zsoldos <balazs.zsol...@everit.biz> wrote: > > My guess is that OSGi metadata here mean the OSGi MANIFEST headers and the > sentence mean something like the following: Lot's of technologies do not have > the OSGi MANIFEST headers in their jars, but thanks to ServiceMix the > situation is getting better (as ServiceMix re-packages many popular jars)
_______________________________________________ OSGi Developer Mail List osgi-dev@mail.osgi.org https://mail.osgi.org/mailman/listinfo/osgi-dev
