David Jencks wrote:
On Nov 7, 2006, at 1:50 PM, Jacek Laskowski wrote:
On 11/7/06, Rodent of Unusual Size <[EMAIL PROTECTED]> wrote:
The implicit assumption there is that the 'developers' all have
the same interests. That's not necessarily how it works. If
there are some people who are interested in EE 5 but not in the
work on trunk, then they probably aren't contributing anything
there now. So creating a branch won't be sucking those people
away from trunk, whereas *not* creating it won't advance the
trunk work and *does* present an obstacle for those people.
The uncertainty element here is people who are currently
contributing to trunk but would switch over to an EE 5
branch if one were created. Those individuals would be happy,
but the trunk work would suffer.
On the other hand, part of what contributes to community is
individual empowerment. If there's insufficient interest in
the trunk work to advance it, then that's pretty much all
she wrote. It becomes a dead horse. Trying to mandate that
people put aside their own interests and work on stuff
that *doesn't* interest them isn't an Apache paradigm. It
has a lot more in common with commercial meet-the-deadline
models.
Whow! Very well written! I like it. It should be part of our
documentation somewhere. I was leaning towards Dain's and Dave's views
and thought to withdraw my +1, but now it won't happen at all. Thanks
Ken!
My objection to putting the jee5 work into a branch is that it's not a
branch!
I think that we would need to make the Java EE 5 work a true branch
rather than a sparse tree.
It is currently and IMO will always remain additional code
that can happily run side by side with all our current code.
Do I understand correctly that you are proposing to maintain both 1.4
assemblies and Java EE 5 assemblies from the same branch? I'm not
convinced this is possible. If it is then it will probably come at a
significant cost for little value. Items like the console will most
likely have to change to support Java EE 5. If we are attempting to
maintain compatibility with the 1.4 build then we would need to have two
console images (4 configs when you factor the tomcat/jetty split).
Couldn't this result in a number of confusing, mutually exclusive
plugins? I think we should make a clean break from 1.4 to 1.5 and Java
EE 5.
Creating
the jee5 workspace is not going to involve copying any existing
geronimo server branch: its going to involve adding new modules for new
functionality. After reading ken's comments I agree that keeping it in
the sandbox carries an unwanted implication that it's not ready for
prime time.
The sandbox is also treated as ... well, a "sandbox" ... a play area
where it doesn't really matter if things work or not at the moment.
Making it a branch brings more focus and hopefully greater collaboration
as it is now something that all of the developers are responsible to
maintain rather than just a few interested parties.
The only solution that comes close to seeming appropriate to me is to
reorganize the jee5 work into plugins and move each one into an
appropriate plugin area. I'm worried that this will cause build system
disruption but I think we have to deal with that soon anyway.
It would be great to make plugins from much of the Java EE 5 functions
in any case (and we'll need them as we continue to push the modularity
of Geronimo and the RYO server capabilities we want to make easier in
the future). However, I'm not sure this will be possible of all of the
Java EE 5 capabilities.
thanks
david jencks
Jacek
--
Jacek Laskowski
http://www.jaceklaskowski.pl
