Simon Richter <s...@debian.org <mailto:sjr%40debian.org>> wrote:
> On 01.12.19 02:54, Russ Allbery wrote: > > >> Russ> Could you provide some more information about what your > >> Russ> concern is here? libsystemd-dev depends only on libsystemd0, > >> Russ> which has a pretty tiny list of dependencies and shouldn't > >> Russ> require that systemd be running so far as I know. Are you > >> Russ> thinking of test suites that assume systemd is available? > > >> pkg-config for systemd is in systemd not libsystemd-dev > > > Oh, okay. Also, I remembered some previous discussion that some libraries > > end up with dependencies on systemd components for functionality. Thanks, > > I think I understand the issue now. > > That was a thread I started on debian-devel: > > https://lists.debian.org/debian-devel/2019/08/msg00278.html > > The resolution of that thread seemed to be that people were mostly fine > with the dependency chain because every decision in the path makes > sense, even if the end result does not. > > Without the package pin of systemd-sysv at -100, installing the > "libwxgtk3.0-gtk3-dev" package causes an init system switch for me. With > the package pin, apt finds no resolution even though one exists (#940965). It's bin:libpam-systemd that pulls bin:systemd-sysv (the package that makes systemd the init on the system), not bin:systemd. Here it's dbus-user-session that pulls it because it needs a logind (dunno if it works with elogind) session opened to have the session bus started when the user logs-in. So I'm not exactly sure why the systemd.pc should be moved to an other package.
