new daylight savings time

[Bruno Wolff III]

On Sat, Feb 24, 2007 at 11:09:16 -0800,
  Benjamin Franz <snowhare at nihongo.org> wrote:
> that Fedora (and even maybe Redhat itself) is dying. Not quickly, but the 
> trend is clear and reaches back years now. I *hope* F7 will reverse that, 
> but without a willingness to support Fedora for more than 12 to 18 months 
> it simply isn't worth the hassle for most people.

Based on the results of the Legacy project, getting longer support isn't
compatible with getting Fedora for free. It seems to only make sense for
Redhat to pay for it for about a year. The people who actively volunteer
to work on Fedora, don't seem to be that interested in supporting old
versions for free.

So it looks like if you want a distribution similar to Fedora but with a
longer support lifetime, you should use something RHEL based.

> Despite Redhat's protestations that Fedora isn't just RHEL beta/technology 
> testbed, in practice that is how it is perceived. Each Fedora is supported 
> just long enough to 'get stable' and transfer technology to RHEL, and then 
> the end user is forced to do an OS upgrade. There just aren't enough 
> people who love reving the OS every 12 to 18 months to a newly unstable 
> release make that viable.

It seems pretty viable to me. It doesn't have to be the most popular desktop
to be a viable one.
 
> If Redhat really wants their technology testbed for RHEL to reach an 
> expanding rather than a shrinking audience, they are going to have to bite 
> the bullet and provide some method for inexpensive support for at least 
> security patches for an extended time ala the old RH boxed sets. I realize 
> they have a problem with it cannibalizing RHEL (which is their cash cow), 
> but CentOS and Ubuntu are already doing a great job of that right now.

You seem to be assuming that all users are worth the same to Redhat. That
isn't the case. People who are willing to participate in getting bugs fixed
are going to be worth more to them than someone who doesn't. Assuming that
there is a net drop in desktop users of Fedora (which googletrends doesn't
prove), if they are gaining in users that like what Fedora is good at and
are actively participating in the Project, they may be better off.

That said, it would be nice of more effort was put into making upgrades
between Fedora versions smoother. Having core and extras combined should be
a big step in that direction. Being able to use other (user specified)
repositories (in particular Livna) at install time is another big step.
This won't handle config file incompatibilities, so there will still be issues.
However handling config file changes during upgrades is a lot trickier.

I think that spending effort making upgrades better rather than extending
maintenance is a much better way to provide value for the people that use
Fedora for what it is good for (having up to date versions of software)
rather than people that just show up in netcraft stats or google trends.