QA 'vote' at the go/no-go meetings

[Adam Williamson]

On Wed, 2011-05-18 at 02:04 +0000, "Jóhann B. Guðmundsson" wrote:
> On 05/18/2011 01:16 AM, Adam Williamson wrote:
> > to specify the basis on which QA's 'vote' at this meeting is cast. It's
> > really entirely deterministic; there's no discretion involved. If there
> > are open unaddressed blockers, we do not approve the candidate for
> > release. If there are no open unaddressed blockers, we do approve the
> > candidate for release. There's really no wiggle room in this: any reason
> > we have to not approve the release should be phrased as a release
> > blocking bug in any case.
> 
> Then I simply propose that the go-no meeting will be dropped and instead 
> after we have confirm that there are no unaddressed blockers and we have 
> confirmed that all release criteria have been met on a blocker review 
> meeting then the chair holder of that meeting sends an ack from the QA 
> to fpl leader and cc fesco and release engineering team which then will 
> send their ack when ready and ones fpl has recived ack from all three 
> parties he declares a release ( Alpha/Beta/GA ).

I'm open to the possibility for sure, the meeting usually winds up being
either a straightforward rubber stamp or a dissection of why we can't
ship, neither of which really really needs to happen. It's possibly a
good thing to have just as a failsafe in case something really odd
happens, I guess. But then, we also have the release readiness meeting,
I'm not really sure we need to have a go/no-go that's separate from that
now we have a clear release blocker process.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net