On Tue, Sep 24, 2013 at 12:06 PM, Chip Childers <chip.child...@sungard.com> wrote: > On Tue, Sep 24, 2013 at 10:00:32AM -0600, Marcus Sorensen wrote: >> I guess I'm in the minority though, since we're the >> only ones discussing it. > > No you are not. Personally, I'm thinking about this problem a bit > before adding my 2 cents. Don't take that silence as thinking we are > "ok". > > Short version of my thoughts: > > 1 - a schedule is really important to a project like this > 2 - not releasing crap quality is really really important to a project > like this > > Note the "really" vs. "really really". > > -chip
I agree with the sentiment in the above two points. I, personally, do not think that the problem is with a time-based release. I don't think that hurts or helps us in the quality department. I do think that we have to get to a point where we can unequivocally verify the status of 'master' or a 'release' branch. I'd prefer that we be able to do this before code gets pushed to it. The network effect of even small changes in CloudStack can be profound. I am almost to the point that I'd prefer us to focus the 4.3 release cycle on nothing but building out QA infrastructure so that we can 'prove' that proposed changes don't cause harm, and that the software is actually releasable. --David
