pointer: map devices across all outputs by default

[JManch]

Describe your PR, what does it fix/add?

This fixes a problem with wlx-overlay-s where multi-monitor overlay pointer movement is broken. Somewhat related issues: #6023 #6889

Is there anything you want to mention? (unchecked code, possible bugs, found problems, breaking compatibility, etc.)

I'm not sure if it's correct to map across the entire space by default.

Is it ready for merging, or does it need work?

Ready

[Agent00Ming]

Mapping to all outputs by default only for pointer devices makes OpenTabletDriver's absolute mode (pointer device) and artist mode (tablet device) map differently despite otherwise identical internal mapping configurations.

[vaxerski]

isnt that just #7586

[Agent00Ming]

Not exactly, this PR changes the default mapping for pointer devices only while #7586 only gives the option for tablet devices to map to all outputs.

[JManch]

#7586 doesn't fix the problem since boundOutput is empty. Would it make sense to change the default mapped area to all outputs instead of just the current monitor?

Hyprland/src/managers/PointerManager.cpp

Line 682 in 88e9e03

CBox mappedArea = currentMonitor->logicalBox();

[JManch]

I've changed it so that tablet, touch and pointer devices map across all outputs unless they are explicitly bound to an output. This changes how tablet regions are applied. When bound to an output, region_position is relative to the bound output, otherwise, it's an absolute position.

This fixes wlx-overlay-s whilst hopefully maintaining consistent behaviour between absolute mode and artist mode for tablets. I've done some brief testing with my tablet and it seems ok but I don't have a way to test touch devices.

[JManch]

I've added an absolute_region_position option that, when true, treats the region_position as absolute. Otherwise it's treated as relative like before. It doesn't apply when the tablet is bound to an output.

[JManch]

Just tested wlx-overlay-s with one of my monitors in negative space and the overlay pointer problem is back so need to figure out what's going on there...

[JManch]

Turns out it was an upstream issue galister/wlx-capture#3 so this is ready for review

[JManch]

I've added the ability to set input:tablet:output to current for replicating the old behaviour.

I could make current the default to keep default behaviour the same as it is now but, as @Agent00Ming said, this would make the OTD absolute mode map differently to artist mode.

[Agent00Ming]

I think that's a reasonable compromise for now as pointer devices lack the configuration granularity that tablet devices have. Tablet tested.

[vaxerski]

wiki mr needed

[vaxerski]

oops conflict

[vaxerski]

clang-format missing xdddddd

[JManch]

I don't think it's from my diff

[izmyname]

I don't think it's from my diff

0.45 broke the stuff, prolly

[fxzzi]

seems to work fine like this:

but not like this (tablet rotated 180deg):

[JManch]

Rotated tablet works for me. In what way does it not work?

[fxzzi]

how i have rotated it on that screenshot causes the mouse cursor to still be able to reach my monitor on the left, which should not be possible considering how I have mapped the tablet area to the display. Moving the area over to the bottom right of the tablet (top left physically, as it is rotated), seems to work fine:

EDIT: my apologies, I think this is intended behaviour by OTD with "Clamp input outside area" turned off. Turning it on and tablet is now working as intended again. Thanks!

[fxzzi]

further testing done with artist mode shows it is now working, however still being funky on osu!lazer. It seems to be stuck on the right side of the window. Looks to be a bug on hyprland considering the same config seems to work fine for me and others on sway.