FESCo: Feature process and release blocker process

[Adam Williamson]

On Mon, 2011-07-18 at 09:42 -0400, Denniston, Todd A CIV
NAVSURFWARCENDIV Crane wrote:

> <I am pretty much out of the fedora process, but the wording above
> leaves me a bit more queasy for using the downstream products.>
> 
> I assume you(QA) are at least doing a cursory review to see if it does
> "constitute an infringement of these quality standards", so that IF the
> feature is still present at the release(a choice of FESCo), then the
> quality(a choice of QA) will still be at the level we expect.
> i.e. *just* because the bug is on a "feature XX", QA is not just
> pitching it back to FESCo, and when final release comes the bug gets
> missed (from a QA perspective) because it was part of a feature that
> FESCo is accepting.
> 
> Just looking for some more clarity on this concept, thanks.

Yes, that's correct. The way I look at it is that the feature process
just isn't really reelvant to the release validation process at all. The
release validation process takes the code in the pre-release to be
tested and makes sure it meets the quality standards - the release
criteria. It doesn't matter if that code happens to be part of an
'Official Feature' or not. If a bug in a Feature infringes the release
criteria, it's a release blocker. The issue here is just that we
shouldn't have 'feature XX is not complete' as a release blocker.
'feature XX causes the system not to boot' would certainly be a blocker.
=)
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw
http://www.happyassassin.net