regression: cursor_inactive_timeout is broken

[andresilva]

Hyprland Version

1c9d6b9

Bug or Regression?

Regression

Description

Dispatching movefocus now activates the cursor. Did a bisect, regression was introduced in 7a5234a.

How to reproduce

• enable cursor_inactive_timeout
• move window focus
• cursor reappears

Crash reports, logs, images, videos

No response

[vaxerski]

why didnt you select "regression" then?

[andresilva]

What do you mean? There are two issue templates: bug and feature request. Nevermind, because I was dumb.

[vaxerski]

:)

[andresilva]

friendly ping. I tried to debug this but couldn't figure it out.

[dusanx]

I have started working on this, MR should be ready in a few. Cursor is being rewritten/forced in a weird way so fix requires slight logic change with cursor handling. I didn't go all the way trough code, so far seems very fixable.

[andresilva]

Fixed in #4184. Thanks @dusanx!

[andresilva]

Broken by: b500e56

When you move focus to a place where the cursor should change it gets re-activated.

[vaxerski]

patch.txt

what if

[andresilva]

Solves the issue, although when the cursor is recovered it might not be using the correct cursor for the surface.

[vaxerski]

hm, right.

[andresilva]

FWIW this isn't any different than it was before so it's probably worth it to merge that patch anyway (the changes in setCursorFromName aren't needed though).

patch.txt