Bug 460965 - On X11, Panel position is slightly pushed away from screen edge and breaks Fitts' Law
Summary: On X11, Panel position is slightly pushed away from screen edge and breaks Fi...
Status: RESOLVED DUPLICATE of bug 460896
Alias: None
Product: plasmashell
Classification: Plasma
Component: Panel (show other bugs)
Version: 5.26.2
Platform: Arch Linux Linux
: NOR normal
Target Milestone: 1.0
Assignee: Plasma Bugs List
URL:
Keywords: regression
Depends on:
Blocks:
 
Reported: 2022-10-25 04:55 UTC by Slavi
Modified: 2022-11-04 06:21 UTC (History)
2 users (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Slavi 2022-10-25 04:55:39 UTC
This appears to be some regression in 5.26.1 or something. It used to work fine before.

STEPS TO REPRODUCE
1.  Create a panel which fully extends from side to side
2.  Add a Pager widget or an Application Launcher or a "Peek at Desktop" widget as either left-most or right-most
3. Position the panel to be either at the top or the bottom of the screen
4. Try clicking in the corner containing your widget

OBSERVED RESULT

The panel appears to fully extend side-to-side and be at the screen's edge.

However..

Regardless of these configurations (top-panel vs bottom-panel; type of widget; left or right positioning in the panel), corner clicks (or mouse-wheel scrolls) do not work.

It's as if the panel is slightly pushed away (1px or so?) from the actual screen corner, causing usability issues - one can no longer push the mouse to the corner and expect to be aiming at an element.

EXPECTED RESULT

Left-most or right-most widgets in a fully-expanded top-most or bottom-most panel should be possible to activate simply by pushing the mouse to the screen edge and clicking or using the mouse-wheel to scroll (on widgets that support this).

SOFTWARE/OS VERSIONS

KDE Plasma Version: 5.26.1
KDE Frameworks Version:  5.99.0
Qt Version: 5.15.6
Comment 1 Nate Graham 2022-10-25 17:08:29 UTC
Are you using Wayland or X11?

Are you using any scaling?

Are you on X11 with scaling and the PLASMA_USE_QT_SCALING=1 environment variable set?
Comment 2 Slavi 2022-10-26 05:01:35 UTC
I'm on X11 with 150% global scaling applied through Display Configuration. I've tried 100% global scaling as well, but it didn't change anything.

Rebooting with just the laptop screen (no external screens with different resolutions involved) also didn't help.

I'm not using PLASMA_USE_QT_SCALING.
Comment 3 Nate Graham 2022-10-26 18:07:51 UTC
Thanks for the info. A few more questions, if you don't mind:

- Did this work properly in 5.26.0?
- Can you try on Wayland to see if it works properly there?
Comment 4 Slavi 2022-10-28 04:57:49 UTC
> Did this work properly in 5.26.0?

Yes, it works correctly in 5.25.0. I've just tried downgrading and cannot reproduce the problem there.

It's reproducible in 5.26.1 and 5.26.2.


> Can you try on Wayland to see if it works properly there?

Just tried Wayland for the first time. Performance seems worse and colors are washed out, but.. this problem is NOT reproducible there. It seems to only be present on the X11 session.
Comment 5 Slavi 2022-10-28 05:00:16 UTC
Sorry about the flood of messages.

I had a typo in my last comment - I meant to say that it used to work correctly in 5.26.0, not in 5.25.0 (although 5.25.0 worked fine too back in the day, from what I remember).

So, as I said, it appears to be a regression caused by 5.26.1 and to only be reproducible on X11.
Comment 6 Nate Graham 2022-10-28 16:52:44 UTC
Thanks. Cannot reproduce with a bottom panel on X11 with 200% scale.
Comment 7 Nate Graham 2022-11-03 19:56:15 UTC
I can reproduce this 100% when compositing is turned off.

Is compositing turned off for you?
Comment 8 Nate Graham 2022-11-03 19:57:55 UTC
Never mind, I'm 100% sure it's Bug 460896 now.

*** This bug has been marked as a duplicate of bug 460896 ***
Comment 9 Slavi 2022-11-04 06:21:42 UTC
I can confirm that compositing was turned off for me, for whatever reason. Turning it on immediately fixes the problem for me.

It indeed seems like a duplicate of bug 460896.