I just wanted to comment on the "2.4 is deprecated" thing. Just because the kernel team is muttering[1] about not supporting the 2.4 kernel does not mean that Debian as a project has decided not to support users using their own versions of this kernel. As Steve notes in #361024, we have to support 2.4 anyway to support users upgrading from sarge. Some other good reasons for the project to continue to support 2.4 include:
- There is still hardware that is only supported by various 2.4
kernels. For example, I have various arm boards and mips machines
that are running Debian with, 2.4, non-debian kernels, which still
work fine (until this bug). Dropping support for 2.4 will simply make
this hardware useless, since Debian is the only reasonable
distribution that runs on it, and since doing the work to make 2.6
run on it varies from far too much effort to nearly impossible (think
binary 2.4 only kernel modules).
- We can't all upgrade to 2.6 trivially. I have production machines that
are colocated thousands of miles from me, and upgrading them to 2.6,
while scheduled, involves a plane trip, and considerable expense.
- Making debian unstable not work in a chroot on a stable machine that
happens to be running 2.4 is not a good idea. Consider that Debian
has a lot of machines running stable with 2.4 + chroots. Also, it would
make remote cross-distribution debtakeovers of machines running some
horrible ancient version of redhat difficult.
- Debian's userland has *always* supported at least the previous major
kernel version, and most often the previous two, or sometimes I
think, three major kernel versions.
PS, Petr Salinger's glibc test package fixes #361024 for me on my 2.4
machine. Unfortunatly, since that machine is responsible for the d-i
i386 daily builds, which involve copying glibc into the d-i images, and
since I do not want to ship d-i images containing an unofficial glibc,
I've had to take those builds down until this is resolved in a glibc in
unstable. Hope it's resolved soon..
--
see shy jo
[1] Or at least some of them are, it's not clear to me if the d-d-a
mail captured the consensus of the team.
signature.asc
Description: Digital signature
