On Fri, Jun 15, 2012 at 4:24 AM, Ciprian Dorin Craciun <ciprian.crac...@gmail.com> wrote: > On Fri, Jun 15, 2012 at 12:36 AM, Kok, Auke-jan H > <auke-jan.h....@intel.com> wrote: >>> As such I don't see what I'm doing wrong... >> >> lots, not helped by the fact that you're not using systemd --system as pid=1. > > Please don't get me wrong, but I think that if systemd "makes" me > either use it to replace my existing init system or not at all, then I > think it is **very** broken...
How is that logical? The way it is intended is that systemd --system sets some stuff (PAM session, XDG_RUNTIME_DIR, and various other stuff, setuid...) for the --user session. >>> XDG_RUNTIME_DIR=/tmp/systemd-tests-run >> >> needs to be /run/usr/<user>/ instead > > Why does it need to be that particular value? If it **must** be > that one, then why have this variable? well, you could redirect everything. Technically ~/.local or ~/.config will work too. However, pam_systemd.so will create some stuff for you that will make this work, so, I wouldn't start out trying the hard way. >>> DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-xcSICndjkx,guid=e28e135122b7479666f6bdda000032b0 >> >> should point to XDG_RUNTIME_DIR/dbus/user_bus_socket. > > I guessed it, but I don't know how to convince it to do so... > > >> I think you really don't want to try and run a systemd --user (highly >> experimental!) without a systemd --system (stable), that would be the >> world upside down, and you're going to get every little part of it >> wrong. > > I think I'll give it a rest for a while until the `--user` gets more > stable. it's stable, but, you're trying to make use of it while it's currently building heavily on assumptions that your system cannot provide, and, you're not providing an appropriate replacement to meet those expectations. Auke _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel
