[Fedora-packaging] Re: Attention kernel module project packagers!

Tim Jackson lists at timj.co.uk
Tue Aug 15 21:42:36 UTC 2006


Jesse Keating wrote:

> 3. Kicking kernel module packages out of Core / Extras and moving the debate 
> away from us so that we can focus on the more important things.  I vote for 
> #3.

This one has come out of left-field to me. Nobody AFAIK has previously
proposed throwing out kernel modules altogether. I also don't think it
makes sense because, notwithstanding the RPM problems, there are plenty
of useful kernel modules that exist and are maintained (for various
reasons, good or bad, right or wrong) outside the kernel tree. I don't
see why kernel modules are any less useful than other random FE
packages, just because they're in kernel space. FE is supposed to be a
maximal set of software, right? That's at odds with a rule disallowing
any kernel modules. Why shouldn't kernel modules be included, exactly?
Excluding them just because of the problems with packaging seems rather
like throwing the baby out with the bath water.

FWIW, although the kmod stuff seems "nicer" in a way, I've been
relatively convinced by what Axel has had to say about kmdl. They're
both trying to make the best of a bad job; kmdl seems to have slightly
fewer problems.

And for those that say RHEL is nothing to do with Fedora: true
theoretically, but if you look at:

a) the number of Fedora devs who are RH employees

b) the number of Fedora users who also use/build packages for RHEL and
derivatives either for work, fun or both

then I think it's worth at least *considering* the potential
annoyance/confusion of RHEL/Fedora having two different schemes.


Regardless of the outcome I'd like to thank Axel in particular and all
the other contributors to the debate so far for their time and energy. I
don't think anyone finds this stuff terribly exciting as I think we'd
all rather be *using* packages than debating the finer details of which
packaging scheme is better than another, but this stuff is important in
the big picture both of getting the best technical solution and making
Fedora the best distro with a wide choice of well-packaged software.
Whilst we shouldn't pander to any special interest, I do believe it's in
all our interests to at least pay serious attention to the input from
well-established third party contributors like Axel who have done so
much in the past few years to bring a large base of add-on software to
RHL/RHEL/Fedora.  They have very useful practical experience of various
techniques applied to a relatively large audience and this knowledge
(and established packages) can only benefit Fedora.


Tim




More information about the Fedora-packaging mailing list