On Sun, 12 Feb 2023 12:40:19 +0100 Marco d'Itri <m...@linux.it> wrote: > On Feb 12, Michael Biebl <bi...@debian.org> wrote: > > > I think this only makes sense if systemd-boot is actually the active > > bootloader, which is why the existing hook scripts also have a "bootctl > > is-installed --quiet || exit 0" check. > I think that you are right, but then we currently have a significant > usability problem because "bootctl install" does not generate the > configuration by itself. > > Are there still any reasons at this point why systemd-boot should not be > enabled if the package is installed? > What grub does is ask a debconf question, but is there any NEED to > actually ask a debconf question, considering that we do not have all the > complexity of BIOS systems anymore and bootctl can automatically figure > where sd-boot needs to be installed?
I agree, we should just make it enable itself on install, like we did for systemd-resolved. It's a separate package, not pulled in by anything - if you install it, it means you want to use it. It's been available in testing for a while, and apart from a few fixes for the scripts at the beginning, there's been no complaints so far. > Also: at least we need to run "bootctl update" on package updates. Ah good point, completely forgot about this, will send a MR. -- Kind regards, Luca Boccassi
signature.asc
Description: This is a digitally signed message part
