In the past the analogy between the Linux kernel and Linux distributions has been used (even though, like any analogy, this one's not perfect either).
Felix could be seen as a project providing a kernel (the OSGi framework) and lots of drivers (the OSGi bundles). Karaf, using some of the components (framework + bundles) provided by Felix, creates a distribution out of it. Like in the Linux world, there is no single distribution that "fits all", even though some can be designated as being more "server" or "enterprise" oriented. In that same sense, Karaf seems more oriented in that direction. Greetings, Marcel On Jun 8, 2010, at 15:09 , Guillaume Nodet wrote: > I don't have much problems dropping the enterprise work if that suits you > better: > "an OSGI-based runtime for creating servers" > or even > "an OSGI-based runtime" > I could even go with your proposal if dropping the servlet reference: > "an OSGI-based runtime container that allows various components and > applications to be dynamically deployed" > though the last part is pretty much what OSGi is about so it does not really > refine the scope at all, it's just redundant. > > As far as Sling is concerned, I don't think Sling's scope is about the > runtime container itself. Sling is defined as "a scriptable web framework > that uses a Java Content Repository, such as Apache Jackrabbit, to store and > manage content" so it's unrelated to both OSGi and providing a container, > which *is* the scope of Karaf. > > As far as Felix is concerned, its scope is "OSGi Service Platform and other > software that is associated with or related to the OSGi Service Platform" so > pretty much anything related to OSGi could fall into its scope. FWIW (in > case it was missed), Karaf is currently a Felix subproject. > > Thoughts ? > > On Tue, Jun 8, 2010 at 00:12, Roy T. Fielding <field...@gbiv.com> wrote: > >> On Jun 7, 2010, at 1:22 PM, David Jencks wrote: >>> On Jun 7, 2010, at 12:52 PM, Roy T. Fielding wrote: >>> >>>> Hi Richard, >>>> >>>> This isn't too important, but I have to admit that >>>> >>>>> charged with the creation and maintenance of >>>>> open-source software related to an OSGI based runtime for >>>>> creating enterprise servers >>>> >>>> doesn't actually say anything meaningful to me because >>>> it has too many happy market terms. >>>> >>>> Would it be fair to summarize Karaf as >>>> >>>> "an OSGI-based runtime container that allows various components >>>> and applications to be dynamically deployed within a Java servlet >>>> environment" >>> >>> No. There is no necessary connection between karaf and servlets or web >> servers or web applications, and even with a more general definition of >> "server" I think this misses the point. I'm sorry you don't like our >> wording, but we tried a lot of variations and this is really the best we >> could come up with, and to me it's quite specific and accurate. >> >> There is no way that "creating enterprise servers" is either specific >> or accurate, unless you have some special definition of BOTH enterprise >> and server that I am not aware of. It is incorrect use of terminology. >> >> You can describe your technology any way you want, but I need to >> know what the scope of the project will be in order to vote for the >> resolution, and the above is what I could guess from reading the >> website content. I don't know what Karaf is capable of doing. >> I just want to know how to distinguish it from what Apache Felix >> and Apache Sling are already responsible for overseeing, and I know >> for a fact that Sling is already "creating enterprise servers" >> by at least one definition of that phrase. Please be more specific. >> >> ....Roy >> >> > > > -- > Cheers, > Guillaume Nodet > ------------------------ > Blog: http://gnodet.blogspot.com/ > ------------------------ > Open Source SOA > http://fusesource.com
