sopwith at redhat.com
Thu Jun 2 15:37:50 UTC 2005
On Thu, 2 Jun 2005, seth vidal wrote:
> On Wed, 2005-06-01 at 16:44 -0700, Roland McGrath wrote:
> > To reiterate what Elliot said, the basic principle as I see it is that we
> > endeavor for the current Fedora release always to have the newest stuff
> > that is reasonably stable. After some fairly short interval in the 3-6
> > month ballpark, *something* certainly has a newer and better version that
> > is reasonably stable. So more or less any time "rawhide has newer stuff",
> > then it's a good reason to have a Fedora Core release before too long. If
> > that sensation comes along, and there's something with a bleeding edge
> > that's still too bloody, then that something can roll back to the stable
> > version until the next FC release. It won't be too long.
> > We don't want Fedora Core ever to spend significant periods of time where
> > there is great new fabulosity out there, in some area or other, that we
> > aren't representing in the latest release.
> That's easy for you to say. You get paid to work on linux and/or
> fedora/rhel for a living. You have all day to work on it. If fedora is
> intended to have community involvement you have to remember that there
> are many of us who are doing this outside of our real job. Remember,
> fedora is for hobbyists.
> But if hobbyists cannot participate in development b/c the cycle is just
> too fast, well then it's not too terribly useful for us, is it.
So my answer would be "if FC5 deadlines don't give you enough time to
complete a given piece of work, target FC6 instead".
On the other hand, someone was also pointing out that there are some
pieces that really need to be worked on every single release, so it's more
complicated than that. Maybe we need to try to decouple development work
from the release cycle a bit more. What if we could make it so that
hardcore development work could go on throughout all six months of the
More information about the devel