Bug#397973: Processed: upping severity, as discussed with Steve ...
On Fri, Jan 12, 2007 at 03:08:21PM +0100, Frans Pop wrote:
> On Friday 12 January 2007 10:33, Debian Bug Tracking System wrote:
> > Bug#397973: [powerpci/mac] partman-md appears to not write back the
> > raid flag to partitions. Severity set to `serious' from `important'
> I don't agree that this issue is RC. It does not match any of the
> 1) This issue does not actually break anything. It just makes it
> impossible to make use of an optional feature (software RAID) during
> 2) The issue affects only a limited group of users as it is architecture
> specific: software RAID support in partman works just fine on at least
> i386, amd64, sparc and hppa. I'm not sure about other architectures, but
> at least we have no reports of breakage there.
AIUI, this bug does make partman-md unusable for most powerpc users, who can
only use powermac partition tables for compatibility with their firmware.
That alone, IMHO, qualifies it as an RC bug; if a package is built and
unusable on a given arch, it has a grave bug on that arch. Being
arch-specific is not a reason to discount a bug as non-RC.
Separately, I feel that md support is important enough to enough users that
I would not be comfortable releasing an installer with etch that couldn't
handle software raid at install time; I imagine you would agree that if this
happened to the installer as a whole, it would be an unacceptable regression
against sarge and therefore unreleasable. I think for the same reason that
we shouldn't be too hasty to release with md support broken on one of our
> 3) There is no regression, or at least, I do not see how there can be as
> software RAID support and general partman code has not been touched at
> that level during Etch development. This rather looks like an incomplete
> implementation of software RAID support for this particular architecture.
> As such, and since there currently (unfortunately) is no lead partman
> maintainer, it is primarily the responsibility of the PowerPC community
> to provide the missing bits and pieces needed to implement the support.
That it's the responsibility of the powerpc porters to fix it doesn't make
it less of an RC bug.
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.