On Fri, 05 Jul 2013 13:18:49 -0700
Adam Williamson <awilliam(a)redhat.com> wrote:
...snip...
We already have test cases for all of these. No need for new ones.
ok. How can we denote the ones we desire tested for these?
Basic Desktop?
> At the Alpha milestone, all approved spins will be created by
> rel-eng.
>
> Each approved spin MUST have at least 2 people fill in the Basic
> spin test matrix for at least 1 TC or RC. If the image fails,
> maintainers can try again at the next milestone, but the image is
> NOT shipped for that milestone. if an image doesn't get 2 people
> filling in Basic spin tests, the image is dropped for that cycle
> and IS NOT SHIPPED FOR THAT RELEASE! Such images can be reapproved
> by the spins sig for the next cycle.
I think this is going way too far.
ok.
Specifically: I think *one* test for each spin is more realistic. I
don't think we need to block spins for being oversize, or for SELinux
denials. They're spins. I don't think we can overreach on quality
requirements.
> - Is 2 people too many? Should it just be 1? or More?
See above.
Fair enough. If thats the consensus, thats ok with me.
> - Are there other Basic tests we should require?
No. That's actually all we require from the *release blocking* live
images at Alpha.
> - Is there a better way to track things than that wiki page?
Not at present, no. WikiTCMS isn't so bad once you get used to it.
Right.
kevin