Steve Langasek wrote: > I don't think there's any justification for a package ever calling u-a --set > from a maintainer script. For what it's worth, independently of how that requirement gets enforced, I agree completely. Thanks, Jonathan