On Fri, Nov 1, 2013 at 3:12 PM, Stephen Gallagher <sgallagh(a)redhat.com> wrote:
-----BEGIN PGP SIGNED MESSAGE-----
On 11/01/2013 03:04 PM, Simo Sorce wrote:
> On Fri, 2013-11-01 at 15:01 -0400, Simo Sorce wrote:
>> On Fri, 2013-11-01 at 14:57 -0400, Simo Sorce wrote:
>>> I agree on this, I think Fedora Server could see the 'stable'
>>> branch of other Fedora products as a sort of 'testing2' branch,
>>> and then a fedora server updates would go "Fedora Server
>>> stable" at a different pace, even skipping updates (except
>>> security ones).
>> Just one thing, I think we need to be careful not to triple the
>> load on package maintainers without good reason.
> Another question, when we get to Fedora Server 1.0 where are we
> going to source packages from ?
> The Fedora 21 git branch, the 23 one ? A mix of the three ? What
> about build dependencies ? It can get hard quickly to try to mix
> and match packages from different releases, OTOH using always the
> latest one (F23 at release) kind of defeats stabilization unless
> you keep some set of package and the base on a leash for 18 mo.
Only Fedora 23. Sorry if that was unclear in the other thread. My
intention was always to imply that each of the previews was based
wholely on the Fedora Base release at the same time. So the idea is
that every third release of Fedora Base would become the *stable*
Fedora Server N.0
That implies that Fedora Base follows this "build to stable"
methodology over the course of 3 releases as well, correct? I'm not
saying that is a bad thing, but it needs to be worked out with the
Base WG as one of their primary goals.
Also, when Fedora Server N.0 forks, is that maintained in addition to
building towards Fedora Server N+1.0 on the now unstable Base? This
is essentially the idea that Spot pitched at FUDCon Blacksburg. If
the resources to pull it off are attainable, I think it will go a long
way but the requirements to do this shouldn't be underestimated.