On Wed, Nov 14, 2018 at 06:12:11AM +0100, Michal Schorm wrote:
We, as a distro, just take a different approach.
To be bleeding edge requires to have releases often.
But being "bleeding edge" has never been what Fedora's about, despite
getting that epithet slapped at us. Yes, we want innovation and new
software, but we also want software which *works*. We want to provide a
consistent, pleasant, and functional user experience, without getting
metaphorical blood all over.
[...]
What I wanted to express by this message is the fact, prolonging
software support time in Fedora means *a lot* of low-level work TBD.
I can maintain it either "bleeding edge style", geting users new
features literally ASAP, or "LTS style" defend the database from any
update to not break anything, bugfix and security fix only. (I'm doing
it for RHEL after all).
But not both.
Lots of good feedback, but I want to focus on this last comment.
It sounds like you are actually already doing both approaches -- the fast
branch for Fedora and a slower, RH-internal branch for RHEL. What if you
made that slower branch *also* in Fedora as a direct upstream for your
internal RHEL work? Fedora is supposed to be the upstream for RHEL, after
all. Why not make that true all of the time, not just every N years when
RHEL branches?
--
Matthew Miller
<mattdm(a)fedoraproject.org>
Fedora Project Leader