On Fri, Jan 22, 2021 at 10:04:44AM -0600, Michael Catanzaro wrote: > On Fri, Jan 22, 2021 at 3:58 pm, Zbigniew Jędrzejewski-Szmek > <zbyszek@xxxxxxxxx> wrote: > >AFAIK, it's just implementing an obsolete concept. The session bus is > >started by systemd --user, which is triggered by the pam stack > >when the user > >session is created. So it is generally fair to assume that the > >session bus > >always present. I don't know why gdm still uses the helper. > > I think container images typically do not run systemd at all, right? > That's where dbus-run-session is generally used, e.g. for CI images. > I'm not sure how else you would expect CI would ever work. At least, > many GNOME tests are going to fail if a D-Bus session is not > available. > > Yes, dbus-daemon can be installed manually in CI images. But if you > want dbus-broker to *really* replace dbus-daemon, such that the > dbus-daemon package can be retired, it would need to learn to cover > this case. %check dependencies are fine. Even systemd pulls in dbus-daemon (or at least did in the past) to test dbus interoperatibility. But this doesn't mean that it should be required in normal installations. (And I don't the goal is to retire dbus-daemon. It remains the reference implementation, and I hope it remains available in Fedora, as long as it is supported upstream.) Zbyszek _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
