Final Release Criterion for Xen DomU

[Adam Williamson]

On Wed, 2011-10-05 at 16:02 -0400, Konrad Rzeszutek Wilk wrote:

> > What's necessary to make it a release blocker is not so much 'it'd be
> > nice if it worked' arguments - I think in general everyone agrees it'd
> > be nice if it worked :) - but more 'it would be terrible to release
> > without it working, because' arguments. That's the level of impact we
> > need for release criteria, because something being in the release
> > criteria means that if it's not working, we don't ship. Which is a
> > pretty high bar to get over.
> 
> There are kernel developers who are willing to write and test code.
> There is also a community of folks who are willing to test install/provide
> patches for grub/grub2/grubby/etc.
> 
> However, I am ignorant in the ways of making a release go out the door
> - so I am not seeing the full picture. What are the missing pieces?

It's not so much a question of 'do we have the resources in place to
probably make sure it works at release time' but 'is it a terrible
disaster if we make a release in which it's broken'. That's what the
release validation process is meant to ensure, and _only_ that. It's not
like, if we don't make it a release blocker, it means no-one will care
about Xen and it will always be broken.

I'm still madly catching up with stuff, but for me the only
consideration so far which falls into this category is the EC2 one, but
that is a *big* consideration. It's pretty close to being enough to make
me vote to have it as a criterion. It would certainly be unfortunate to
ship a release you couldn't install as an EC2 guest.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net