Attempt to fix force_zero_scaling for tablet and touch

[Ant01n3]

Describe your PR, what does it fix/add?

Tries to fix tablet and touch events not scaled back (mouse is OK) when using force_zero_scaling on XWayland windows and output scale is not 1. (Fixing #3573 )

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

My knowledge of the code base and event propagation mechanisms of Hyprland is really limited. I hope I haven't forgotten anything.

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

It needs to be thoroughly tested. I'm currently using it with Krita (which doesn't support Wayland) and I've done some tests with xev to check that mouse, tablet and touch positions match.

[vaxerski]

I think that's it

[vaxerski]

also, needs a clang-format

[Ant01n3]

also, needs a clang-format

Done.

[vaxerski]

you resolved conversations without addressing them

[Ant01n3]

you resolved conversations without addressing them

Sorry, I do not understand what you want me to do. Be kind.

[vaxerski]

in both cases your calculations are wrong and I commented the correct formulas

[Ant01n3]

in both cases your calculations are wrong and I commented the correct formulas

(g_pInputManager->getMouseCoordsInternal() - m_sTouchData.touchFocusWindow->m_vRealPosition.goalv()) * m_sTouchData.touchFocusWindow->m_fX11SurfaceScaledBy

corresponds, for me, to (coords - windowPos) * scale ? But maybe I do not understand well the meaning of these functions/fields ?

The other one is more convoluted due to the use of touchSurfaceOrigin but I fail to see how the formula you give can be implemented (I tried to change as few things as possible in the original code, but as I need to scale the coords, I needed to scale touchSurfaceOrigin also, that is the only things I changed).

I tested the code (and I am actually using the modified version) and it seems to work (I admit this is not a proof). I did tests with xev to see if the computed pixel positions are the same as the mouse.

[vaxerski]

that is correct

[Ant01n3]

Ok, so should I consider the code is now ok and resolve the conversations ?

[vaxerski]

well, no, you are still calculating it differently and I am pretty sure it is wrong

[vaxerski]

I'll test it on my laptop tomorrow and edit myself. But it definitely is wrong

[vaxerski]

patch.txt

this fixes both my comments, and works properly.

[Ant01n3]

Done. The last commit contains your patch.

[vaxerski]

lgtm, thanks!