Prepare qml-power-app for testing and polishing

[ArrayBolt3]

OK, this ended up being slightly trickier than it looked like on the surface, however I believe this is mostly ready.

Full list of changes:

• Replaced all calls to test.py with calls to kfocus-fan-set and kfocus-power-set. (No changes were needed to either tool as both already had all needed features.)
• Removed a spurious fanSlider.value = 1 from line 250 that was causing the app to revert the user's fan settings.
• Fixed a bug uncovered by the removal of the fanSlider.value = 1 line which resulted in the fan profile description to appear as "undefined" if the app was started with the fan profile set to Soft.
• Fixed a race condition in the fan curve code between kfocus-fan-set and kfocus-fan-set -r that was causing the app to sometimes revert the user's fan settings.
• Fixed a nearly identical race condition in the SPLINE tuning code between kfocus-power-set and kfocus-power-set -r that was causing the app to sometimes revert the user's SPLINE tuning settings.
• Added some auxiliary Bash code to get the output of kfocus-fan-set and kfocus-power-set into the proper format for the app's internal use.
• Made the app window a bit wider by default so that the name of each power profile under SPLINE tuning wouldn't overflow the colored box it was in.
• Moved all testing data and code into a testComponents directory. (I also apparently changed the contents of one of those test files at some point - I can revert that if it matters.)
• Made SPLINE tuning profile colors nicer.
• Fixed a bug where the user's SPLINE tuning profile choice would be reverted if kfocus-power-set -r specified that a "custom" profile was selected, which it seems to do randomly every so often.
• Enforced larger minimum window dimensions so as to avoid text spillage or out-of-window UI elements.
• And apparently accidentally added some random whitespace to one line that wasn't needed :P Cleaned that up.

If needed, I can split these changes up into individual commits.

Things that may still need to be done:

• Resizing the app window vertically leaves all the controls at the top of the window, unlike the existing KFocus Power Profile Utility, which keeps them centered. Not sure if that's acceptable or bad.

[ArrayBolt3]

For the problem with text spillage if the window is too small horizontally, I'm pretty sure that can be fixed by just bumping up the minimum window width to be the same as the default width. I'm happy to implement that if that's an acceptable way to resolve that.

[mmikowski]

That sounds like a very reasonable solution, and I trust your judgement on that.

On Fri, Mar 17, 2023, 15:37 Aaron Rainbolt @.***> wrote:

For the problem with text spillage if the window is too small horizontally, I'm pretty sure that can be fixed by just bumping up the minimum window width to be the same as the default width. I'm happy to implement that if that's an acceptable way to resolve that.

— Reply to this email directly, view it on GitHub https://github.com/kfocus/kfocus-source/pull/3#issuecomment-1474476232, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACV5LKYCY2QO7AT3R5GCRLW4TRSVANCNFSM6AAAAAAV63NLK4 . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[ArrayBolt3]

K, I also noticed that the original minimum vertical size let some elements go off the bottom of the screen, so I enforced a higher minimum there too.

[mmikowski]

The work you've done is very good and brings this one major step closer to completion. I've reviewed the changes and the app, and here is my assessment of what remains to be done. Most of this boils down to fixing kfocus-power-set -p to work better with the app, and then updating the app to take advantage of these changes. Also, now that we have a POC, it is a good idea to remove the most hacked features. All of these ideas are expressed below:

1. Header information should also come from kfocus-power-set -p.
2. When keyboard LEDs are not supported, they should not be returned from kfocus-power-set -p nor in the header.
3. Model sensing should likely be done in kfocus-power-set to decide whether certain capabilities are available and return information accordingly (again, see LEDs)
4. When present, the keyboard LEDs should present as 'On' or 'Off', not 0 or 1. This might be handled by kfocus-power-set -p which would return those strings instead of the numbers.
5. See kfocus-power for guidance. Once this project is complete, the wrapper will BECOME kfocus-power.

[ArrayBolt3]

Glad you're happy with where things are going! I'll see if I can look into the points you've mentioned soon. Also, on IRC you said some things about scaling and capability sensing - is the capability sensing stuff already in the above five points, or is that another point that needs work? (EDIT: Nevermind, my brain finally finished registering that yes, you did indeed mention capability sensing in the above.) I'll also hopefully fiddle with display scaling and see if I can find what it's doing wrong and how to fix it there.