[Dev] Tizen 3.0 proposal for fixing OSP/WRT/Core hard-coded UID issue

Jussi Laako jussi.laako at linux.intel.com
Wed Oct 16 09:59:12 GMT 2013


On 16.10.2013 12:41, Patrick Ohly wrote:
> This might not be the case for this particular example: on the IVI image
> using the systemd user session management, the session D-Bus socket
> address is hard-coded per user. Initially, systemd listens on that
> address and then starts the dbus-daemon on demand when a client tries to
> connect to D-Bus via that socket.

OK, on openSUSE (also using systemd) that is the case only for system 
bus. Session bus still runs with "dbus-daemon --fork --print-address 
--session" and has random dbus session bus address.

Fixed per user address may become tricky in case there are more than one 
user session per userid.

With our current plan, there would be a separate guest user per seat so 
it shouldn't become an issue. But are we sure we don't need to support 
multiple sessions per user?



More information about the Dev mailing list