ARC role [was Re: Include GNU sed 4.1.5]

[James Carlson]

Nicolas Williams writes:
> I repeat: the ARC does not usually care what meta-cluster anything it
> reviews ends up in.  Has the ARC been ignoring its responsibility to do
> so?  Or did it never have that responsibility?

At the risk of belaboring a thread that's already stuffing everyone's
inbox with the obligatory three or four duplicates per message sent,
I'd like to comment on the above.

As with all other matters, the ARC reviews these things *when they are
architecturally significant*.  Please don't think of the ARC's role in
terms of black-and-white lines on a grid, which is implicit in your
question of whether the "ARC cares" (or perhaps whether ARC members
care) about meta-clusters.

Instead, the root issue is system architecture.  If the choice of
meta-cluster forces design or architectural decisions, then it's in
scope for architectural review.  For example, if we had a perfectly
good library for what a core project wanted to do, but the project
team waas being forced to reinvent that wheel because the library was
available only in SUNWCall, then that's clearly an architectural
matter.  The choice of meta-cluster in that case would be an in-scope
issue, and perhaps even something the ARC could take action on.

For what it's worth, the same thing applies to all other reviews.
People who are good at reviewing design, code, and other materials
*often* end up with a wide scope to accomplish that goal.  Telling a
code reviewer that he can't look at the design, or a design reviewer
that she cannot see the test results, is just nonsense.  The same is
true here.  What's in scope is what's necessary to get the job done,
and what isn't exclusively someone else's prerogative.

-- 
James Carlson, Solaris Networking              <james.d.carlson at sun.com>
Sun Microsystems / 35 Network Drive        71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677