Re: [D-I] Supporting 2.6.14 kernels in base-installer
On Mon, Nov 14, 2005 at 10:06:55PM +0100, Frans Pop wrote:
> On Monday 14 November 2005 21:47, Joey Hess wrote:
> > This seems more in-scope for the kernel team to decide than the d-i
> > team, but if it would help I can flip a coin..
>
> Note that one consideration is consistency between Debian and Ubuntu in
> d-i. Ubuntu have implemented initramfs-tools and in fact Colin has
> committed the code to support initramfs-tools today. Colin has left the
> option to also/alternatively include yaird open though.
Has he already commited the code for that ?
> will include yaird and dependencies by default on CD images to satisfy
> kernel-image dependencies as yaird is listed first in the dependency
> list. So we'd need some debian-cd magic to exclude yaird and include
> initramfs-tools.
Well, that choice was made when initramfs-tools was broken, and klibc not yet
in debian, it was always the intention to reevaluate this choice later on,
once we have more oversight of it, so this is not an issue, whatever default
is chosen, it will go for both d-i and the kernel.
I still think choice is good, and also what users expect of debian. A sane
default, plus the ability to override that in expert mode.
> If an arch has a strong preference for "the other" tool if we should
> select one, I guess this could be supported. This will in practice put
> the burden of maintaining the code and testing installations on the
> porter(s) involved though.
Well, you all speak about burden, but if the tool fails, this is an RC bug in
the first place, and we can't release etch in these conditions, what extra
testing is needed here ?
Friendly,
Sven Luther
Reply to: