Vinod Johnson - Sun Microsystems wrote:
Part of what is in play here is how much more time and effort (I would
imagine the effort part would me more process oriented rather than
technical) it would take to do a 0.1 vs a 0.2. If the answer is 'not
that much more' then there seems to be little to be gained from
splitting the two. If it is 'reasonably more' then I think there is
value in having a binary distribution as an initial artifact for people
to be able to download and touch and feel. This value judgment, of
course, comes with the assumption (perhaps misguided) that the slant
initially is towards showing people that we have a binary that works
rather than the magnitude of active ongoing work.
My 2c.
Perhaps this sounds odd, and maybe it is, but if River were to get an
early release out, even one that I don't use. At least I'd have
something which is River and is open to change. Right now it feels like
much of River is opaque and within Sun. The sooner there is a release,
the sooner River will feel like a project.
I think there is a lot of value in seeing a 0.1 release (unless of
course 0.2 won't be significantly longer), even if everyone knows
0.2/0.5/... is going to change everything. The value here is in having
something which people can call River. Then River will become something
concrete and open to suggestions and critique.
This said, perhaps the code drop is enough, but I'd rather not have to
track changes and discussions at that level of granularity right now.
In many ways I think my feelings follow Mike Herrick's. I have a lot of
hope for River, and I hope my comments don't come across as being too
critical. I'm convinced that Jini's move to Apache is the right thing,
it's just frustrating that it's taking so long.
--
Rick Moynihan
Software Engineer
Calico Jack LTD
http://www.calicojack.co.uk/
