STEPS TO REPRODUCE 1. systemsettings kcm_kaccounts 2. Click on Add New Account > Google OBSERVED RESULT Nothing happens. In the terminal window, I see: QDBusError("org.freedesktop.DBus.Error.ServiceUnknown", "The name is not activatable") EXPECTED RESULT An account wizard or something opens. ADDITIONAL INFORMATION This is in a built-from-source Plasma 5 session on top of Fedora 37 packages. Have not tested with the distro-provided session yet. Will do so shortly.
Reproduces in the distro-provided session.
Do you have signond installed? Do you have /usr/share/dbus-1/services/com.google.code.AccountsSSO.SingleSignOn.service?
(In reply to Natalie Clarius from comment #2) > Do you have signond installed? Yep: $ which signond /usr/bin/signond > Do you have > /usr/share/dbus-1/services/com.google.code.AccountsSSO.SingleSignOn.service? Yep: $ ls -la /usr/share/dbus-1/services/com.google.code.AccountsSSO.SingleSignOn.service -rw-r--r--. 1 root root 84 Jul 23 2022 /usr/share/dbus-1/services/com.google.code.AccountsSSO.SingleSignOn.service
Created attachment 158181 [details] working live session I'm inclined to say this is a Nate vs. DBus problem more than a bug. It works just fine in the fedora live session
I reproduced the issue on a distro-provided session with no source-built packages being used, though. Regardless, lowering priority since no one else seems to be able to reproduce the issue, so it can't be *that* common.
Well, our install-session script did/does mess with dbus configurations. I wouldn't entirely discount the possibility of that having something to do with this report. Getting ServiceUnknown while the service file clearly exists is suspicious at the very least
As usual, my response is that all I'm doing is following the instructions on the wiki to get built-from-source DBus stuff working, so if that documentation produces incorrect results, someone who understands how DBus works needs to revise it.
Fair point. We may need a place to track mystery dbus problems though, having them on individual products when they require a particular set of skills to debug makes them difficult to connect the people with the skills to the bugs :) Meanwhile I have a grand idea, I'll PM you.
Right, it's definitely not ideal to have that scattered in random bug reports that are really about setup issues. Perhaps it's a good topic for in-person debugging during the upcoming Plasma sprint. Either way, I'll happily close this bug report if we can trace it to a local setup issue.
Harald has helpfully been debugging this with me and it doesn't look like there is any issue in KDE code here. So far it's looking like it might be caused by a bug in dbus-broker itself. Either way, no KDE issue; closing.
Actually it was our bug after all. It's fixed with https://invent.kde.org/plasma/plasma-workspace/-/merge_requests/2835.