Solomon Peachy wrote:
So far the only tangible result is that the release date for F21 is
delayed (which is probably a good thing)
It's not. As you say yourself:
A longer release cadence means we lose the 'First' goal (both
in the
First-to-market and Upstream-First sense), and the main beneficiary
seems to be those who think the 'Freedom' goal only applies to
themselves, not their downstream users.
so we shouldn't delay our releases for
no good reason.
The main "feature" I've seen requested is an
intermediate-cadence
support cycle between RHEL/clones' 5-year and Fedora's 1ish-year, but
nobody (especially not those asking for it) seems willing/able to do the
work to provide that support on the (nontrivial!) distro-level scale.
Longer support must not happen at the expense of release frequency.
(I remember all too well the Fedora Legacy folks' pleading for
help..)
The reason Fedora Legacy failed was its unrealistic QA requirements on
package updates, a disease that has since spread to Fedora as a whole. We
could support our releases for much longer (WITHOUT reducing their
frequency) if we just let maintainers push the security fixes without any
bureaucracy. Right now, we even have trouble getting karma for the Fedora
n-1 release that is still supported.
Kevin Kofler