On 28 January 2014 13:38, Matthew Miller <mattdm(a)fedoraproject.org> wrote:
On Tue, Jan 28, 2014 at 08:34:23PM +0100, Robert M. Albrecht wrote:
> >* Although it's certainly not the only reason, Fedora as _solely_ a
> > hobbyist desktop is not ideal for an upstream for RHEL server and
> > cloud products.
> No other system can be reinstalled / upgraded every six months. That
> single fact IMHO kills all other use cases.
Fedora actually has a 13-month lifecycle, which is still fast but less
dramatic than that. And although that makes it *challenging*, it doesn't
preclude real use in non-desktop cases. I know because I've seen people in
real life doing all sorts of things, from high-performance computing to
running hosting services to high-frequency stock trading.
Well on paper it has a 13 month lifecycle, but the perceptions are a lot
shorter due to three factors:
1) If you have a complicated stack for applications then you are going to
need 3-6 months to port, test and verify that it works with the latest
perl, python, ruby, java, and haskell you ended up using to get your job
2) If you are really a part of the community, it feels like a 3 month life
cycle because that is the time between X and X+1-alpha, and the constant
and needed drumbeat that you test, use and promote X+1alpha so that it can
be ready in 3 months time.
3) When you end up being on X-1 release, you quickly find your bugs being
answered that you need to either move to X or X+1-beta to see if it is
fixed there because the package maintainer only has so much attention to
spend on things.
The biggest things I can see from Fedora.Next is working on solving 1,2,3
by making it easier and faster to either port or carry your own versions of
the apps you need and making as much of the OS 'metric' as possible so that
it can be forgotten by the meta-os people.
Stephen J Smoogen.