Patrick Ziegler

That would be great, thanks! Let me know if you run into issues. Getting older versions to compile can be a pain

For older commits, you need to cherry-pick these two commits to fix some build issues: https://github.com/polybar/polybar/commit/16102c258ae5eb619e6b494a93be58897281118d and https://github.com/polybar/polybar/commit/432e19df013d0227b30e987119d93aba32b24aee

In `battery.cpp` replace `on_event(nullptr);` with `on_event({});`. Or just skip with `git bisect skip`

Thank you so much for tracking down the exact commit, this already helps a lot. In that commit I rewrote a lot of our handling around pseudo-transparency. To confirm the...

Thanks a lot, your experiments have the expected results, it is indeed the pseudo-transparency (and only the pseudo-transparency) that causes this. Finding the issue in the code will likely take...

Got some time to dig into this again. One thing that looks suspicious in your logs is this: ``` - background_manager: root pixmap (0x600001:24) 1920x1080+0+0 - background_manager: Copying from root...

Alright, now we're getting somewhere, I can finally reproduce on my own! Steps: 1. Turn off second monitor 2. Set wallpaper `feh --no-fehbg --bg-fill /path/to/image`. 3. Turn on second monitor...

I have a fix in #3096 Please try it if it also fixes the issue on your end (it should simply show a black background and print an error message)

Well that's not great 😕 Can you run with `-l trace`? Maybe I didn't account for some edge case

The xworkspaces module does not look at [`_NET_WM_STATE`](https://specifications.freedesktop.org/wm-spec/wm-spec-latest.html#idm46044244303104) at all. It only uses the urgency bit (in [`WM_HINTS`](https://x.org/releases/X11R7.6/doc/xorg-docs/specs/ICCCM/icccm.html#wm_hints_property)). My WM (herbstluftwm) seems to set the urgency bit when `_NET_WM_STATE_DEMANDS_ATTENTION` is...