>>>>> "John" == John Plocher <John.Plocher at sun.com> writes:

John> If LSARC were to become the JDS ARC, what would/should happen to
John> the rest of the traditional LSARC caseload: Explorer, SunVTS,
John> Woodstock, the open source database stuff, NetBeans, the
John> compilers, SunMC, Cacao, etc?

Unless things have changed since I sat on LSARC (1997-1999), LSARC's
domain is mostly a grab bag of things that weren't a natural fit for
some other ARC.[1]  Restructuring the current set of ARCs (as you
suggest later in your email) makes sense to me.

jek3> What happens when this change which would normally be reviewed by
jek3> LSARC is submitted?

John> Answer #1: If we have only one OpenSolaris ARC: The project goes
John> to "OSARC" for review.  Hopefully the LSARC members who are also
John> JDS community core members are participating in the operation of
John> OSARC and take a leadership role in the review.

One of the concerns about being on an ARC is the time commitment.[2] In
the past, I've suggested that this could be remedied by letting members
only pay attention to some subset of the cases that come before the ARC.
This idea didn't get much traction.  The thinking was that we were
better served by having the same set of people review all the cases over
a certain period of time.  It gives more consistent decisions and
better institutional memory than having one set of people review the
cases one week, and a somewhat different set of people review the cases
the next week.

With that in mind, how would things work in a single OpenSolaris ARC?
Would members still be expected to participate in all cases?

Also, I'm a little concerned about the rationale that if a single
committee doesn't work, we can reorganize into multiple committees
later.  How long would it take to implement such a reorganization?
Would we have sufficient warning, or would it end up being a firedrill?

The reason I ask is that AFAIK, the historical pattern for PSARC is that
case load goes up significantly as various release deadlines approach.
So I think we can assume that the case load for the OpenSolaris ARC(s)
*will* go up; our planning should take that into account.

mike

[1] "LSARC reviews everything *else*!"  --Dean Stanton (I think)

[2] During my tenure, the nominal commitment was 20%, but realistically
it was closer to 40-50%.


Reply via email to