On Sat, 30 Jul 2005, Roy T. Fielding wrote:
> At this point, it is clear you guys were not paying attention
> to the contents of the thread.  The disagreement is over the
> community having the ability to work on a branch that is not
> stable.  All of the Solaris releases would be on a stable branch
> that has the exact same interface stability requirements as it
> does right now, for the same reasons that you described.  The
> difference is that the community *outside* Sun can do visible
> work within OpenSolaris rather than having to create private
> forks of the community, and Sun could back-port the compatible
> portions of that work to the stable tree according to its own
> review processes.

Having been away from my mail for 5 days, and carefully digesting (with
great interest) this thread in one sitting, it seems to distill
down to two key premises:

- The CAB is the arbiter of what (the entirety of) OpenSolaris will become.

- Sun's architecture and product committees are the arbiters of what
  Nevada (SunOS 5.next) and its successors will become.

And from these premises, I get these conclusions:

1. Currently Nevada is the sole "OpenSolaris branch" not by decree, but
   by default. After the CAB governance is finalized, it will be by
   decree. In other words, Nevada will be a CAB-blessed, stable branch of
   OpenSolaris.

2. In addition to Nevada, at least one other CAB-blessed, stable branch
   should be created (ASAP I hope) to accomodate community developers
   for whom backwards compatibility is not a #1 priority.

3. Developers will be highly motivated to work on their unstable
   branches and other projects on opensolaris.org because they will
   want their work to become part of one or more of the CAB-blessed,
   stable branches.

Eric
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to