On Wed, Jul 12, 2017 at 02:10:13PM -0700, Adam Williamson wrote:
So, I guess I should set some more detailed expectations here. At
from my perspective on it.
Thanks Adam. I'm going to snip most of this but I read it all and will
think more about it too. None of this is easy stuff.
I do hope that as we have more compose failures, people do feel
interested in (and enabled to) fix the problem quickly, rather than
just feeling antsy. I'm not sure how that will go, though, so it's
definitely a big concern.
Overall with the shorter schedule, I think we should focus *primarily*
* What you and Dennis and others need to do this
* The arbitrary branching thing, as that's a big F28 enabler
* the other automated testing initiative (CI)
and if that takes your time away from the normal (I know, gigantic)
amount of work you do around QA, plan to find other ways to fill in.
Honestly - I was kinda banking on us having a reasonable amount of
usual 'slack time' at the start of a release cycle to try and do a bit
of a 'soft launch' of compose gating, with at least a few weeks for us
to shake down all the details and get a feel for how significant the
impact on the development process is. What you're talking about feels
somewhat different. I'm not necessarily comfortable with us banking on
the idea that compose gating is going to be something we can kick in
*immediately* with complete success, to the degree of basing F28/F29
plans on it.
Well, we can definitely decide as a project that this "strict six
month" thing isn't working. I don't really like any of the alternatives
we have available, though, outside of magical capabilities that
modularity might theoretically give us but definitely doesn't yet.
Soooooo, what happens if we don't get "hard" compose gating working?
You mentioned that OpenQA coverage of the current alpha criteria is
pretty good (even given all the caveats I cut out above) already. I
know you do analysis of the failures regularly; what if we start pretty
aggressively after branch at working to get the right people looking at
all of those, and skip the formal Alpha even without a compose gating
Fedora Project Leader