It's doable in theory, but the issue is that in reality nobody looks at it until it's out the door. It's a similar issue with feature branches, they get merged, and it's not until after the merge that people say "hey, this breaks all sorts of stuff", because the devs don't necessarily know all of the use cases, and the users are not devs.
Even if it is doable, I think it's probably a bridge too far. The reality is that devs don't really like to work on bugs as much as they do new features, so I'd like to get there one step at a time, rather than present something massive sounding that will just turn everyone off completely. On Mon, Sep 9, 2013 at 2:47 PM, Daan Hoogland <daan.hoogl...@gmail.com> wrote: > Why can't we cover every use case, Marcus. We will need help from > users, but if they do help it will be easy to do so. > > On Mon, Sep 9, 2013 at 10:43 PM, Marcus Sorensen <shadow...@gmail.com> wrote: >> I was actually talking about separate things in relation to this >> thread and the other where I mentioned that I'd like to see a release >> focused on bugfixing and testing. With that, I'm advocating a test for >> every api call and focusing on broadening use case test coverage. >> >> Here, I'm simply talking about taking the support matrix and doing >> some vary basic testing. This can be a dozen or so tests, each >> platform we say we support needs to successfully deploy a zone and a >> vm on every storage type that is in the support matrix. I don't think >> this would include plugins (or maybe those are left to the developer >> of the third party plugin). For KVM, this is literally a marvin script >> away from being there, I don't think there's a ton of work. I have no >> idea what we have or can do with vmware, and I'm guessing Xen is >> largely covered already. >> >> We'll never be able to cover every use case, I may be able to deploy a >> zone with my KVM setup, but not someone else's special network layout. >> I'd just like to see sanity checks to say it works, at all, on the >> handful of 'supported' systems. >> >> On Mon, Sep 9, 2013 at 2:24 PM, Mike Tutkowski >> <mike.tutkow...@solidfire.com> wrote: >>> Do we have any statistics that say how many of our customers are using >>> feature x, feature y, etc.? >>> >>> If not, I would say if we know about a feature that has regressed to the >>> point of breakage in 4.2 that it should be fixed before releasing (or at >>> the very least well documented, so - if it is impactful to someone - they >>> do not upgrade until it has been fixed). >>> >>> >>> On Mon, Sep 9, 2013 at 2:12 PM, Chiradeep Vittal < >>> chiradeep.vit...@citrix.com> wrote: >>> >>>> I think that Animesh is trying to stress what is "key". If it hits 1% of >>>> cloud operators is it key? >>>> >>>> >>>> On 9/9/13 7:42 AM, "Simon Weller" <swel...@ena.com> wrote: >>>> >>>> >-1 from me as well. >>>> > >>>> > >>>> >I know we're trying to hit timed releases, but I think it's very >>>> >important to preserve key underlying functionality across releases. If a >>>> >supported and documented feature is known to be broken, we need to >>>> >address it...if we don't, it's going to cause lots of pain, and reflect >>>> >badly on ACS as a project. >>>> > >>>> >----- Original Message ----- >>>> > >>>> >From: "Chip Childers" <chip.child...@sungard.com> >>>> >To: dev@cloudstack.apache.org >>>> >Sent: Monday, September 9, 2013 9:24:23 AM >>>> >Subject: Re: [VOTE] Apache CloudStack 4.2.0 (fourth round) >>>> > >>>> >On Sun, Sep 08, 2013 at 12:40:30AM -0600, Marcus Sorensen wrote: >>>> >> -1 ... sorry guys, especially with Simon chiming in. >>>> >> >>>> >> I'd request f2c5b5fbfe45196dfad2821fca513ddd6efa25c9 be cherry-picked. >>>> > >>>> >Agreed. >>>> > >>>> >I'm -1, given simon's perspective as well. Since we have the fix, let's >>>> >get it into the release. >>>> > >>>> >>>> >>> >>> >>> -- >>> *Mike Tutkowski* >>> *Senior CloudStack Developer, SolidFire Inc.* >>> e: mike.tutkow...@solidfire.com >>> o: 303.746.7302 >>> Advancing the way the world uses the >>> cloud<http://solidfire.com/solution/overview/?video=play> >>> *™*
