On Sun, Feb 21, 2021 at 7:19 PM Tom Seewald <tseewald@xxxxxxxxx> wrote: > > > I understand it's bad UI/UX and I agree. But what I'm asking is *what > > criterion*, i.e. > > I would argue that it fits best in the shutdown category [1], specifically that the system must cleanly shutdown such that "storage volumes (e.g. simple partitions, LVs and PVs, RAID arrays) are taken offline safely". Given that a 2 minute timeout will result in many users holding the power button rather than waiting, filesystem(s) will not be cleanly unmounted. While the ensuing frustration and impatience resulting in pulling the plug is a reasonable possibility, it's an additional step involving user intervention that the criterion doesn't anticipate. The criterion says it must be possible to trigger it. And it is possible, and has been triggered, the logs show that. The clean part is a reference to not doing something like reboot -f or sysrq+b, that would e.g. require a resync, log replay, or fsck on the next boot. >I personally have interacted with a user who was force powering off his rawhide machine due to this bug, so this isn't pure theory. Oh I'm aware, I pull the cord myself. Although it's a bit handier to enable the early debug shell, switch to tty9, and reboot -f >Have there been other recent Fedora Workstation releases that shipped knowing that the system is unable to reboot in less than 2 minutes by default? I'm not sure about recency but yes, it's happened since that criterion was written. It's been an on-going effort for some time. > Could you please respond to the rest of my previous post in order to help me understand why this is not a gnome-session issue as Benjamin Berg seems to have indicated? Ben Berg is certain to be correct, so disregard any comment from me that suggests otherwise. As it turns out I even got confused about what component I originally filed the bug against. :P > I will also mention that after I manually added "Slice=-.slice" under the [Service] section of /usr/lib/systemd/user/gnome-session-restart-dbus.service (which is the exact change that Benjamin committed to gnome-session[2]), I no longer experience 2 minute timeouts. Overall I think there is very strong evidence that this bug is exactly the one Benjamin Berg describes, and hence this is an issue with F34 and rawhide's gnome-session package needing to be updated to include the commit previously referenced. That sounds reasonable. -- Chris Murphy _______________________________________________ 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 Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
