reviving Fedora Legacy

[Benny Amorsen]

David Woodhouse <dwmw2 at infradead.org> writes:

> What I _don't_ understand is why these requirements are not met by
> CentOS. Isn't that _precisely_ the 'market' that RHEL and CentOS exist
> to serve?

RHEL and CentOS are simply too slow for some things, unless you get
very lucky and they happen to have a release at the perfect time.

Imagine starting development on something to be deployed in half a
year or a year. RHEL right now is on kernel 2.6.18. If you e.g. need
some of the many new features of netfilter since 2.6.18, you're
"stuck" with Fedora. Not that I mind in the least being stuck with
Fedora, of course.

Also, let's take Asterisk, which is included in Fedora. You get to
have 1.4.x if you use Fedora 8, and 1.6.x with Fedora 9. With EPEL,
you currently get neither, but that is relatively easy to fix. What
isn't so easy to fix is that you get either 1.4.x or 1.6.x with EPEL,
and it's a long wait for the next release of RHEL just to get a new
Asterisk. You can solve that by adding updated packages yourself, of
course, but it can easily devolve into your very own distro where RHEL
updates don't apply safely.

So I want a RHEL with a new release every 6 months and 5 years of
stability. And a pony. If I can't have that, I'd like to have whatever
updates people can be bothered to make for old Fedoras. It's ok if
they break things, it's ok if they don't manage to cover all the
updates that RHEL would have done, it's ok if everyone gives up on it
after 3 months. I'll take what I can get.


/Benny