+1
On September 17, 2019 10:34:25 PM GMT+02:00, Ben Cotton <bcotton(a)redhat.com> wrote:
On Tue, Sep 17, 2019 at 9:55 AM Kamil Paral <kparal(a)redhat.com>
wrote:
>
> Considering this, I believe it makes sense to move the current
criterion from Beta to Final.
>
I'd be okay with letting the responsible team specify a buffer they're
willing to accept for Beta (e.g. so Workstation might say "we can
allow 10% over our specified limit" and Server can say "if we go over
the DVD won't burn, so what's the point?") In other words, any overage
would be an automatic blocker, but teams can optionally pre-approve
exceptions. This way, teams can still make determinations that are in
the best interests of the deliverable they're targeting, but we don't
end up in a position we're slipping a week because a deliverable is 5
bytes larger than desired.
I like this approach better than just moving the criterion to Final
because in cases of large overages, the package gymnastics that might
be required to make it work could prove to be pretty impactful. I'd
have a lot more confidence in our ability to deliver a good experience
to our users if we're trying to trim half a gig from an image before
Beta than if we're doing it on the Monday before Go/No-Go.
Julen Landa Alustiza <jlanda(a)fedoraproject.org>