kwin wayland turn off screen (xset dpms force off)

[inkbottle]

Hi,

I'm looking for a way to "turn off screen" from command line, with wayland.
Obviously `xset dpms force off` won't work.

Ultimately or optimally, it would IMO be nice the "turn off screen" timer to be
shortened when the screen is locked.

I've found "PowerDevilDPMSAction.cpp" is related to that (even the wayland
case), but I haven't found how to make use of those actions through `qdbus`.

The following line does work, but it seems awkward, and after it is executed
the terminal scrolls as if you were continuously pressing enter key (not
always though):

qdbus org.kde.kglobalaccel /component/org_kde_powerdevil invokeShortcut "Turn
Off Screen"

And yes, I could define a shortcut, but I don't know which to choose which
would be convenient, wouldn't interfere with other shortcuts, and anyway, the
correct shortcut should be the one I already use to lock the screen `Ctrl-Alt-
l`.

Knowing how to cause the screen to turn off from command line would be nice.

Also, the expected behavior would IMO the change the duration of the screen off
timer when the screen is locked. Like setting the timer to 40 sec.

Thanks,
Chris

[Grzegorz Szymaszek]

(Reposting to the list, sorry inkbottle!)

On Thu 06 May 2021 01:34:15 CEST, inkbottle wrote:
> Also, the expected behavior would IMO the change the duration of the
> screen off timer when the screen is locked. Like setting the timer to 40
> sec.

FWIW, <https://bugs.kde.org/show_bug.cgi?id=348529>.

[inkbottle]

On Thursday, May 6, 2021 9:30:54 AM CEST Grzegorz Szymaszek wrote:
> On Thu 06 May 2021 01:34:15 CEST, inkbottle wrote:
> > Also, the expected behavior would IMO the change the duration of the
> > screen off timer when the screen is locked. Like setting the timer to 40
> > sec.
>
> FWIW, <https://bugs.kde.org/show_bug.cgi?id=348529>.

Thanks.
Someone says <https://bugs.kde.org/show_bug.cgi?id=405344> must be addressed
first.
Things might be different, with wayland.
<https://blog.martin-graesslin.com/blog/2015/09/comparing-dpms-on-x11-and-wayland/>

[inkbottle]

The following line does succeed in conveniently setting `dpms` timer.

`kwriteconfig5 --file powermanagementprofilesrc --group AC --group DPMSControl --
key idleTime 60; qdbus org.kde.Solid.PowerManagement /org/kde/Solid/
PowerManagement org.kde.Solid.PowerManagement.refreshStatus`

And again, kwin-wayland seems more consistent than its x11 counterpart in the
matter of turning the screen off, so, so far, the whole operation is far less
painful than it was with x11 version.

See: https://bugs.kde.org/show_bug.cgi?id=340342
for the command line.

[inkbottle]

On Sunday, May 9, 2021 2:27:32 AM CEST inkbottle wrote:
> On Friday, May 7, 2021 12:06:52 AM CEST inkbottle wrote:
> > On Thursday, May 6, 2021 9:30:54 AM CEST Grzegorz Szymaszek wrote:
> > > On Thu 06 May 2021 01:34:15 CEST, inkbottle wrote:
> > > > Also, the expected behavior would IMO the change the duration of the
> > > > screen off timer when the screen is locked. Like setting the timer to
> > > > 40
> > > > sec.
> > >
> > > FWIW, <https://bugs.kde.org/show_bug.cgi?id=348529>.
> >
> > Thanks.
> > Someone says <https://bugs.kde.org/show_bug.cgi?id=405344> must be
> > addressed first.
> > Things might be different, with wayland.

> > <https://blog.martin-graesslin.com/blog/2015/09/comparing-dpms-on-x11-and-> > wa yland/>

>
> The following line does succeed in conveniently setting `dpms` timer.
>
> `kwriteconfig5 --file powermanagementprofilesrc --group AC --group
> DPMSControl -- key idleTime 60; qdbus org.kde.Solid.PowerManagement
> /org/kde/Solid/ PowerManagement
> org.kde.Solid.PowerManagement.refreshStatus`
>
> And again, kwin-wayland seems more consistent than its x11 counterpart in
> the matter of turning the screen off, so, so far, the whole operation is
> far less painful than it was with x11 version.
>
> See: https://bugs.kde.org/show_bug.cgi?id=340342
> for the command line.

I'm trying running the following in a terminal, and then locking the screen
with Ctrl+Alt+L:

#! /bin/bash

# https://unix.stackexchange.com/questions/353998/run-script-on-screen-lock-in-kde

dbus-monitor --session "type='signal',interface='org.freedesktop.ScreenSaver'"
|
while read x; do
case "$x" in
*"boolean true"*)
echo SCREEN_LOCKED;

kwriteconfig5 --file powermanagementprofilesrc --group AC --group
DPMSControl --key idleTime 60;
qdbus org.kde.Solid.PowerManagement /org/kde/Solid/PowerManagement
org.kde.Solid.PowerManagement.refreshStatus

;;
*"boolean false"*)
echo SCREEN_UNLOCKED;

kwriteconfig5 --file powermanagementprofilesrc --group AC --group

DPMSControl --key idleTime 600;
qdbus org.kde.Solid.PowerManagement /org/kde/Solid/PowerManagement
org.kde.Solid.PowerManagement.refreshStatus
;;
esac
done

[inkbottle]

Using systemsetting GUI, I removed a `key`:

[AC][DimDisplay]
idleTime=540000

Also, my kde-plasma-framework... might be more up to date.

Who knows. In any case the above script, which leaves much to be desired, is
now working.

So, with it running in a terminal, you can do: Ctrl-Alt-L, and ta-da, the
screen turns of a few seconds later.
And when you unlock, the default behavior is restored.

[Norbert Preining]

On Thu, 20 May 2021, inkbottle wrote:
> So, with it running in a terminal, you can do: Ctrl-Alt-L, and ta-da, the
> screen turns of a few seconds later.
> And when you unlock, the default behavior is restored.

I submitted a patch to the kernel that fixes this on the kernel side ;-)
No extra scripts needed.
It is not perfect and seems to not work in all circumstances, but this
is what I am running (but I run self-compiled kernels).

https://gitlab.freedesktop.org/drm/amd/-/issues/662#note_909333

Best

Norbert

--
PREINING Norbert https://www.preining.info
Fujitsu Research + IFMGA Guide + TU Wien + TeX Live + Debian Dev
GPG: 0x860CDC13 fp: F7D8 A928 26E3 16A1 9FA0 ACF0 6CAC A448 860C DC13

[inkbottle]

On Thursday, May 20, 2021 10:10:03 PM CEST Norbert Preining wrote:
> On Thu, 20 May 2021, inkbottle wrote:
> > So, with it running in a terminal, you can do: Ctrl-Alt-L, and ta-da, the
> > screen turns of a few seconds later.
> > And when you unlock, the default behavior is restored.
>
> I submitted a patch to the kernel that fixes this on the kernel side ;-)
> No extra scripts needed.
> It is not perfect and seems to not work in all circumstances, but this
> is what I am running (but I run self-compiled kernels).
>
> https://gitlab.freedesktop.org/drm/amd/-/issues/662#note_909333

Yes indeed.

But I understand my scripting is addressing a different purpose in the "turn off
screen" arena. It is not addressing a hardware related issue preventing
altogether things from working. It is more like, assuming lower level things
are well behaved, "can we shorten the dpms OffTime timer when the screen is
locked?" And in that "user experience" concern, actually, and contrary to my
first feeling about it, it is doing pretty well, or so it seems, after several
days of using it. It even seems reliable.

For the present, with my "systemsettings", my hardware, wayland, "unstable",
Plasma 5.21.5, framework 5.82, the script found in the previous post, and
verifying that nothing is obviously interfering with `dbus-monitor --session
"type='signal',interface='org.freedesktop.PowerManagement.Inhibit'"`, things
have been fool proof for days.

So far, very satisfactory.

Thanks,
Chris

[inkbottle]

In some circumstances, that I haven't been able to identify at all, but which
could be related to opened Firefox, but I have really no evidence at all of
that, the script is defeated, and it's a huge PITA!

The screen turns back on at some point, and it's easy to see that it is not
turning off again, because of the timer being set to 30 sec.

$ dbus-monitor --session
"type='signal',interface='org.freedesktop.PowerManagement.Inhibit'"

dbus-monitor is giving no clues of something happening it'd be aware of.

Thanks,
Chris