+1
My feeling is that the highest priority has to be to grow the community.
If we succeed at that, we will have the resources to do the rest. After
my first days of pretending to be a new potential user, I'm convinced
that a better "Getting Started" experience is essential. We may or may
not attract new participants with that. We are unlikely to do so without.
One or more good demos would be a major contribution, and I think should
be prioritized ahead of reorganizing the source code.
On 1/9/2015 1:35 PM, Greg Trasuk wrote:
Hi all:
We had talked about pulling the starter and class server out of the
JTSK and renaming them to the ‘org.apache.river.tools’ package
namespace.
Although I still think it’s a good idea, I’ve become convinced that
there are too many follow-on requirements. Essentially, all of the
infrastructure service implementations have a dependency on the
com.sun.jini.start package. Also, the QA framework uses it.
The logical extension of that restructuring work would be to pull out
the QA framework into a different project, and also pull the
infrastructure services out. So in order to even create a demo, we’d
have to restructure and re-release the whole project, which is not
something I’m planning to take on at this time.
So, I’ve changed to examples project to go back to using the
com.sun.jini.start package. That’s still possible, since start.jar
is published to Maven Central.
I’m not sure at this point whether it’s still a good idea to pull out
the classdep implementation from the JTSK into a separate project, or
just apply the patches to move up to using ASM5. I’ll see about it
after the examples project is ready.
Cheers,
Greg Trasuk
