Unnanounced migration phase at its end, non-transparency, "open" community project ... (was: No more kernel-source(code) ???)

[Axel Thimm]

On Fri, Jun 25, 2004 at 12:53:02AM -0300, Alexandre Oliva wrote:
> On Jun 24, 2004, Axel Thimm <Axel.Thimm at ATrpms.net> wrote:
> 
> >     kernel-source(code) is needed for a migration phase
> 
> Err...  You shouldn't need kernel-source to build kernel modules.  The
> right procedure to build kernel modules is to use /lib/modules/`uname
> -r`/build.

No, it's not, that is the procedure for building kernels modules
against an installed kernel (and even running kernel in your
syntax).

And as it has been pointed out at too many parts of this thread that
kernels for different arches and their headers collide, so you will
never be able to build in the same box a kernel module for i586
running the same kernel in i686 (yes, you can in a chroot, don't
bother mentioning, that is not the point).

And if kernel-source(code) were to be phased out it _has_ to be
preannounced (and discussed not dictated!). You don't throw something
into rawhide and hope people to understand the deeper meaning of this.

> This works with 2.6, so you had the entire FC2 development cycle to
> migrate.  It's time to move on, and leave behind those who didn't
> migrate in a timely manner, wouldn't you think?

So, leave behind everybody? Who _did_ migrate to the Unknown Plan of
God (TM) in a timely manner? The 50.000 Red Hat kernel-source related
pages? The dozen of professional kernel module writers? Any ISV? Red
Hat itself? Oh, yes, Red Hat never builds kernel modules, so it is not
affected.

Right, noone did. And in fact just days before the kernel-source(code)
was dumped, there was a discussion on its name change reveiling the
wide use of kernel-source(code). Back then the only comment was "don't
rely on implementation", only to blown the "implementation" a few days
away.

What ignorance is this, when people already point out that the name
change will make problems to continue to remove it altogether
promising replacment in form of future documentation?

That is neither transparent, open, an announced migration, or sensible
to do in any way. If this is a community project it has to involve the
community, not pester it. The current attitude is only driving people
away. Stop the social experiments!

Please revert this, until we find a real solution, can you?

P.S. Note that _I_ can still cope with changes like these (until now,
let's see what will come next).
-- 
Axel.Thimm at ATrpms.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20040625/e2b5a2b6/attachment.sig>