JDK 9 Rampdown Phase 2: Process proposal

[mark.reinhold at oracle.com]

2017/3/20 10:28:40 -0700, joe.darcy at oracle.com:
> On 3/16/2017 3:51 PM, mark.reinhold at oracle.com wrote:
> [snip]

Thanks for snipping extraneous text!  I wish more people would do that.

>> JDK 9 Committers are invited to comment on this process proposal.  If
>> no serious objections are raised in one week's time, by 23:00 UTC next
>> Thursday, 23 March, then this qnexiis the process that we'll use.
> 
> A few other situations which might merit explicit consideration:
> 
> * Backing out a change: Despite reasonable care being taken, there are 
> times that after a change is pushed the change is found to be 
> problematic. If the problems are severe enough, the change needs to be 
> backed out quickly while the fix is reworked or reconsidered. Perhaps 
> backing a change could just be subject to normal code review procedures.

With sixteen potential reviewers, spread across both European and North
American time zones, I don't think it's unreasonable to have to ask just
one of those people to review an emergency backout.  I'd prefer to keep
the process simple, at least until we see actual data that would justify
complicating it.

> * Problem list maintenance: to deal with unexpected test failures, we've 
> often wanted to update the problem list within a few hours of a 
> changeset going in. For example, a fix is generally fine, but causing an 
> unexpected test failure in another area. It would be helpful for problem 
> list updates to be handled under the "noreg-self" rules.

That makes sense; I'll mention it in the web-page text.

- Mark