Trying to install rawhide: madness and where do we hope to go ?

[Dave Jones]

On Wed, Sep 07, 2005 at 07:26:01AM -0400, Daniel Veillard wrote:

 >   Can't basic bootup of kernels on i386 uniproc machine be tested before
 > pushed to rawhide, channels, mirrors and users ? 
 
I won't hold up release of a rawhide kernel because it fails
to boot one on architecture, even if that is the most prominent
architecture we support.

Whilst i386 was broken, x86-64 for example worked fine.
With daily rebases to huge amounts of upstream code-change,
we can't afford not to have a kernel rpm of each build to
test when a regression got introduced.

The last week brought a huge amount of change as we moved
from 2.6.13 to 2.6.13-git. Breakage is to be expected
every time we do a rebase. Yes, this means rawhide can
be expected to be uninstallable at that point.

		Dave