Alex Scheel wrote:
In an ideal world, Modularity would've been proposed to Fedora,
it would've been discussed by the community, improved, perfected,
and _then_ brought into the RHEL fold only when stable! Fedora
would've _driven_ the innovation, rather than having to keep up
with whatever RHEL decided to do, being careful not to break
anything.
Well, that would have been the only way to do things that would have made
sense.
We don't live in an ideal world and deadlines, initial rejection
by Fedora, and issues implementing Modularity forced at least _some_
of it to go the other direction, RHEL -> Fedora.
So now we're stuck trying to improve Modularity in Fedora, without
stepping on any RHEL toes, and knowing full well that if the
community rejects Modularity in Fedora (again!), we're stuck supporting
it in RHEL through RHEL 8 (and, likely into RHEL 9). And I think that
worries some people, hence the attempt to influence discussions, name
calling, &c. Others however, want Fedora to remain fully independent
of RHEL, and thus have asked that _failure_ be an option explicitly
on the table.
It was purely a business decision of Red Hat to ship RHEL 8 with an
unfinished alpha version of Modularity instead of giving it the time to
mature in Fedora (either pushing back the RHEL 8 release (it's not like any
release date was actually announced, nor like pushing back RHEL waiting for
Fedora to stabilize has never been done, see how RHEL 6 was based on Fedora
12 rather than 9 due to the major changes in Fedora 9) or pushing back
Modularity to RHEL 9) and it is not fair that Fedora has to deal with the
consequences of that decision.
The only mistake Fedora has made was to accept Modularity in its current
state at all, giving in to the pressure from RHEL developers.
Kevin Kofler