Lock/logout interactions

[Michael Rudolph]

On Friday 07 March 2008 15:48:24 Will Stephenson wrote:
> You'll have seen the recent crop of blogs and dot comments about
> logging out of KDE.
>
> It seems we have 2 problems:
>
> 1) Forced double choice of leave options when leaving KDE via
> Kickoff, because the user opens the Leave menu, then selects how they
> want to leave KDE, then these all open the ksmserver dialog* with the
> same choice of logout/reboot/shutdown, and a 60 second countdown
> until the previous choice happens automatically.
>
> * the one that dims the desktop, KSMServer::logout() in
> kdebase/workspace/ksmserver/shutdown.cpp
>
> 2) 'logout' and 'leave' terminology is confused
> Examples:
> * In Kickoff, 'leave' returns to the general process of leaving the
> KDE desktop.
> * In Kickoff, 'logout' logs the current user out and returns to the
> display manager.
> * In the desktop context menu, 'logout' shows the ksmserver dialog,
> including 'logout' again - end session, 'shutdown' - shutdown and
> suspend options, and 'reboot' options.
> * In the lock/logout applet, 'logout' shows the ksmserver dialog, as
> above * Ctrl-alt-del shows the ksmserver dialog, in the keyboard
> shortcuts list for krunner, this is described as 'Logout'
>
> 2 suggestions:
> 1) Make the Kickoff leave action_s_ all lead to a ksmserver countdown
> dialog which only offers the previously chosen action, do it now, or
> cancel.
>
> 2) Disambiguate logout so
> "Log out": end session, return to display manager
> "Leave": offer choice of ways to leave
> Possibly change "Log out" to "End session" or "Sign out"
>
> So the context menu gets a "Leave..." entry, the Lock/Logout applet
> becomes the "Lock/Leave" applet, and the action defaulting to
> ctrl-alt-del (or acpi power button) becomes "Leave".
>
> Thoughts?
>
> Will
> _______________________________________________
> Panel-devel mailing list
> Panel-devel at kde.org
> https://mail.kde.org/mailman/listinfo/panel-devel

Hello everyone,

I think that's an interesting thing to tackle.

And from a developer's point of view it's of course quite obvious to 
start from existing infrastructure, but I'm not so sure this will lead 
us to better logout interaction from a user's point of view.

I guess, I won't manage to come up with a comprehensive list of actions 
a user might want to take when he's "done", but ACPI states, even in 
their "humanized" form: sleep and hibernate, or logout and reboot won't 
be found on such a list.

Users might rather say: "I'm done for now" or "I'll take a break and 
come back later". It's our task to translate this into appropriate 
power states, I'd say.

It's funny, while thinking about it, it occurred to me that rebooting is 
actually quite a stupid thing to have in the logout menu. Of course 
these formidable electronic devices are able to pull of this quite 
amazing feat to power themselves off and boot strap themselves again, 
all on their own, without human intervention. But do we need this 
option to clutter up our user interfaces, when a user interface without 
it would only require one action more on the user's side, which, by the 
way, might be the most natural and organic interaction there is on a 
computer: pressing the power button to power up the machine?

I think we should see if we can come up with a sensible list of actions 
a user might want to take when using his computer's "leave system". 
Then we should see how many of these actions the computer can reliably 
guess on its own (like S3 after idle time or lid switch) and just the 
rest should make it into the "leave system".

Just my two cents.

michael