On Sat, Jul 12, 2008 at 1:58 AM, John Plocher <John.Plocher at sun.com> wrote: > [Cc'd to ogb-discuss] > > * Come up with an initial list of "top level things". > > For governance, we need a rational set of top-level groups with which > the OGB can communicate. They need to be our "existing" top-level > groups, > there need to be enough of them so that the "soup" is distributed > across > them without large concentrations under a single top-level group, and > few > enough to make regular reporting feasible.
It seems to me that the Groups directly under the OGB needn't be instances of the collectives they oversee. Indeed, they need not be collectives at all. If a reporting structure is what is desired, then an appropriately scoped set of subcommittees seems more appropriate - that way, you don't have to force all parts of the wider community into a structural straitjacket for reporting purposes. > The architectural principle behind this tree is that it matches the > actual > practice we have today. > > The wiki page at > http://www.genunix.org/wiki/index.php?title=OGB_2008/010 > has an initial stab at such an hierarchy. Is the notion of a hierarchy valid? It might work in the context of a command and control infrastructure, but not if we believe that OpenSolaris is an open community that anyone can contribute to. In particular, the notion that SIGs cannot begat projects is (or ought to be) false. Looking at the hierarchy, basing it around consolidations to the exclusion of other entities seems wrong. Are we designing Sun's management structure for the commercial Solaris product or trying to evolve the OpenSolaris community? -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/
