On Mon, May 12, 2025 at 11:34 AM Ludovic Courtès <l...@gnu.org> wrote: > > Hello, > > Greg Hogan <c...@greghogan.com> writes: > > > I would like to second Leo's idea from the kernel-team branch discussion: > > > > Changing the subject, it would be nice if grafts were committed > > along with an ungrafting > > commit to the team branch that would be responsible for the code, > > assuming such a team > > exists. That way they wouldn't be missed like I did here. > > > > This sounds like a solid improvement. > > I agree. > > Also, once ‘core-packages-team’ is merged (or with a tweak to exclude > the glibc graft), I’d like to resume my experiment in auto-ungrafting: > > https://issues.guix.gnu.org/74654 > > That would potentially allow us to ungraft “anytime”. > > Ludo’.
Is this not generating a large number of package rebuilds? What would the priority be? How different is this than skipping the graft and simply performing the security update on a special feature branch at priority 1? But either way we lose the build bunching.
