waitForCompletion is disabled?

[trmdi]

Is this a typo?

https://github.com/Zren/plasma-applet-commandoutput/blob/9003024463e9e7cceaa07200da7ca0f10af27602/package/contents/ui/config/ConfigGeneral.qml#L45

I mean the option to make the timer runs only one time, in case the command does not need to update.

[Zren]

Kinda sorta. The executable data engine can't run the same command twice at the same time, so I disabled it. If the interval is shorter than the command takes to run, then I think it'll skip the 2nd call. But tbh that's nothing to worry about. I guess we can enable it and make the config default waitForCompletion=false.

[trmdi]

Hmm, when Interval = 0, it still runs only one time, even when Timer.repeat == true. So, waitForCompletion is redundant for this purpose.

Though I don't know if it is still needed for other usages. Can we remove waitForCompletion? Instead, we stop the timer right before executing the command, and restart it after the command exits?

[Zren]

Might as well keep waitForCompletion. I may make limit the interval to 1ms if waitForCompletion=false though since it seems QML Timer's will stop. Even timer.restart() like below doesn't seem to be working for some reason.

if (config.waitForCompletion || config.interval === 0) {
    timer.restart()
}

[trmdi]

Maybe this works?

test.patch.txt

[Zren]

Your code is if interval=0 then the timer doesn't repeat, nor does it "restart" when the command is finished. So the command will never be run more than once...

[trmdi]

Yes, we use interval=0 for making it run only one time intendedly.

[Zren]

Oh, you want a runOnce option?

[trmdi]

Yes, an example use case is just to write some texts on the desktop.

[Zren]

Do you really need it to be a "command", or do you basically want a "Note" widget?

[trmdi]

Do you really need it to be a "command", or do you basically want a "Note" widget?

Hmm... Sometimes I want to write some texts on the dektop just to decorate it. :P Like this:

The "Note" widget can't change the font family and size like this one. Anyways, I can use a very large interval... Though the timer is still a bit wasted.

[Zren]

I can use a very large interval.

Yeah, that's what I was about to suggest.

Edit: Also that is adorable. xD

[trmdi]

Because when removing this, we could make the config window a bit simpler, but I don't know if there is any cons. Could you give an example in that waitForCompletion == false is useful?

[Zren]

A "clock" needs to update every 1000ms exactly. A clock running every 1010ms will skip a second every once and awhile.

I'm more likely to remove waitForCompletion == true than == false.

[trmdi]

I think you can't make a clock update at 0ms 1000ms 2000ms... (the real millisecond of the clock e.g. 10:45:02.000 AM -> 0ms...) exactly, it will update at e.g. 5ms 1235ms. 2001ms... randomly depending on when it starts, how long the command takes, the event loop... there will be always a delay between this clock and another. So I don't really understand your example. :/

[Zren]

There will always be a delay from X000ms, however I assume it should still start the logic around at X000ms. If there's a 100ms delay caused by the onTriggered logic before restarting the timer, then you'll skip from 9s to 11s.

• 0ms
• 1100ms = 1000ms interval + 100ms logic delay before calling restart
• 2200ms = 1100 + 1000ms interval + 100ms logic delay
• 3300ms = 2200 + 1000ms interval + 100ms logic delay
• ...
• 9900ms
• 11000ms

Notice how it'll skip 10sec.

With a repeating timer, I assume the math when the timer is triggered would be the following to attempt to trigger with the same interval between calls:

sleepFor(interval - ((now - timerStarted) % interval))

[trmdi]

Do you know why the output sometimes (1 time per minute) disappears then reappears when I use this command? The latest git version. [[ $((date +%S% 2)) -eq 0 ]] && echo "openSUSE KDE 😗" || echo "openSUSE KDE 😘"

Ah, actually that is an error command: https://stackoverflow.com/questions/36238961/script-error-value-too-great-for-base-error-token-is-09 So, there should be some warning when the command is not successfully exited.