James Carlson wrote:
Sebastien Roy wrote:
I'm merely providing my input. In the end, if you're leading this
project, then it's your decision. FWIW, I don't believe that the term
"consolidation" applies at all here, as consolidation applies that it's
one of the pieces of the WOS, the collection of consolidations that
comprises the Solaris product that Sun provides. This, to me, is simply
a proposal to create a software repository managed by an OpenSolaris
project.
Mostly +1 to that. But I don't think you're right that a consolidation
must deliver through the WOS. A consolidation (such as, say, the HCTS
consolidation or Sun Ray or Sun Cluster) can also deliver independently.
The important features of a consolidation are that it's managed as a
single software unit, and it has carefully controlled and understood
boundaries with other consolidations. In this case, that's not true.
Garrett's charged that I've only been negative on the proposal, and I
don't think that's right. I've positively suggested running it as a
project rather than as a separate consolidation (and you've added,
rightly I think, that running it as a child of ON would be the easiest
solution). I've also suggested running a group of projects that deliver
independently (in as much as we really do need a *simple* independent
delivery mechanism). And I've suggested that, at least in some cases,
solving some of the barriers that this proposal is trying to work around
(private GLDv3) would be a more useful contribution.
The main barrier that I'm interested in solving is not interface
compatibility or stability. The main barrier I'm trying to solve is to
find a vehicle for maintenance and delivery of software that Sun will
not let into ON.
There are other ways this could be done. Running a child of ON *would*
work, but I think it will ultimately mean more work for the gatekeeper
(who at least initially is probably going to be me).
The dozen or more separate projects idea also can work, but has (IMO)
significant drawbacks. There is a bunch of co-related software
(platmods and drivers and such that really all support the same
workstations), and separating them out just feels like makework. It
also elides the benefit of having a single repo -- leveraging a common
build infrastructure, common gatekeeper, and _some_ level of
oversight. Part of what I want to do is make these bits more
acceptable to 3rd party distributions, and I think having some of the
attributes of a normal consolidation will be beneficial in achieving
that end.
-- Garrett
_______________________________________________
opensolaris-code mailing list
opensolaris-code@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
