* James Carlson <james.d.carlson at sun.com> [2009-06-09 13:38]: > Christopher Kampmeier writes: > > I have to believe that there are many functional areas with similar > > rallying point interests that don't immediately or ever require spinning > > up their own project with a source repo and all. > > > > How does OpenSolaris support these interests? > > I think that's an excellent question. > > The key issue, it seems to me, is whether an OpenSolaris community > should endorse projects that may have functional overlaps and/or > conflicts, or whether the community must act as though it were a > consolidation or distributor that needs to avoid high-level conflicts. > > I think that having overlaps at least at this level is goodness. It > allows projects -- even those that may be in some competition -- to > coordinate on common issues, which is something that's certainly to > the benefit of OpenSolaris. That's true even if these competing > projects end up delivering through different distributions, or if some > just never deliver at all. > > The alternative seems bleak to me. It would mean that communities > would effectively be able to endorse only a single distributor's > (indeed a single vendor's) vision of what projects are valuable, and > would consign non-compliant projects and simple experiments to an > unnecessary purgatory elsewhere on the net.
In general, I agree with your above points. In specific, we have a proposal from relative newcomers to the Install CG that contains substantial overlap with existing work. The point of my objection is that we haven't discussed JeOS enough to determine whether or not it needs to be an independent effort, rather than enlarging one or more of the existing efforts. I am asking that we attempt to have that conversation in existing fora, rather than splitting prematurely. I do disagree with your last point, however: projects that don't want to work within a particular CG and whose leaders have no interest in persuading others of their merits via reasoned argument should be sent into the wilderness. (That's not this project or these leaders.) > We have and use the idea of "release vehicle" elsewhere to allow > projects to deliver in ways that are meaningful to the users, why not > allow it here? > > Thus, I'll reiterate my +1 for this project. I think it's a great > experiment, and deserves some real estate on the opensolaris.org web > site in order to do its work. That's all we're really discussing here > -- not whether this project will ever integrate anywhere, not whether > we're giving it Sun's imprimatur -- but just whether this community > should allow it to exist as an OpenSolaris Project, and I think it > should. I had thought that projects were supposed to produce something tangible; perhaps the current instantiation policy dropped that requirement. If the goal is merely a separate, focused discussion, the Install CG can request the creation of install-jeos at opensolaris.org. The pkg(5) project is happy to see package dependency fixes, refactored package boundaries, and new metapackage recipes; anyone producing three non-trivial patches is granted commit rights. Maybe it would be more clear if the JeOS proposers identified which distributions they hope to influence? - Stephen -- sch at sun.com http://blogs.sun.com/sch/
