On Tue, Jul 11, 2017 at 01:30:32PM -0400, Colin Walters wrote:
> Hopefully, by the time we are at F28, Modularity will provide a
way for
> us to offer faster streams for people who want them -- but let's also
> focus on stable releases.
But with Modularity, how much does it even make sense to talk about
"Fedora" releases in a generic fashion with a 6mo cadence? Aren't we
likely for many modules to only have a single stream (or multiple)
that may not match that cadence?
I think we're still going to have a large number of packages that
aren't on that model in a year. And, I expect we will still have a base
runtime (or something) that comes out on a six-month cycle -- and
we'll probably want to have some release artifacts (like Workstation)
which also follow that cycle.
It seems to me offhand that some things like the Change process will
be around
modules, and then changes in those modules get reflected into any editions they
affect? A lot of the Changes listed here:
https://fedoraproject.org/wiki/Releases/26/ChangeSet
would seem to be for the base module, but there are several desktop/Workstation
specific ones. And we get into a lot of interesting questions around the intersection
of the languages and Workstation, depending on what gets installed by default.
(My take would be to reduce the amount of things installed by default, and really
encourage doing most development in containers, decoupled from the base host
lifecycle, like Atomic Host)
Sounds good to me!
--
Matthew Miller
<mattdm(a)fedoraproject.org>
Fedora Project Leader