Bug#726763: Bug#729576: duplicate of bug #726763

[Vincent Lefevre]

On 2013-12-22 17:47:42 +0100, Andreas Cadhalpun wrote:
> On 22.12.2013 17:38, Vincent Lefevre wrote:
> > If #726763 occurs with GNOME and without lightdm, and is the same
> > problem as #729576, then this could confirm that it is a systemd
> > bug.
> This is no bug in systemd,

Sorry, I mixed up two bugs related to systemd and with similar bug
number (the other one is 732623, where the installation of systemd
breaks some lightdm features).

> but rather a missing dependency of gnome-shell and probably also
> gdm3 on systemd-sysv.

Perhaps. This would mean that gnome-shell / gdm3 would be completely
tied to systemd, and concerning gdm3, admins who don't want to use
systemd for init would have to use a different DM. A better solution
for gdm3 could be to use the traditional shutdown/restart/... method
as a fallback, which doesn't need systemd at all; then systemd-sysv
could just be recommended. But that's up to the gdm3 developers to
decide.

> But this dependency cannot be added, because systemd-sysv conflicts
> with the essential sysvinit package.

If GNOME developers want to require systemd as the init system,
I don't see this as a reason not to add the dependency. Users
are not forced to install GNOME packages. And if they want GNOME,
they would have to accept the consequences about the init system.

> Have you tried using systemd as PID 1?

No, I don't want to do such kind of risky test. Since I was not
using the GNOME desktop environment, but just gdm3 and some GNOME
apps, I decided to drop those that depend on systemd. For me,
this meant:
  * gdm3 -> lightdm (but systemd must *not* be installed due to
    bug 732623),
  * gnome-control-center (to set up the volume) -> pavucontrol
    (though it is inferior).

-- 
Vincent Lefèvre <vincent at vinc17.net> - Web: <http://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)