### System details

Omarchy 3.4.2, Hyprland 0.54.2, hypridle 0.1.7, Chromium 146.0.7680.153

### What's wrong?

After upgrading to Chromium 146 (landed in Arch repos around March 11), the display never turns off via hypridle. Chromium is asserting the Wayland `idle-inhibit-unstable-v1` protocol on tabs that have no video, audio, or Wake Lock API usage — just normal web pages.

You can verify with:

```
hyprctl clients -j | jq '.[] | select(.inhibitingIdle == true) | {class, title}'
```

Which returns something like:

```json
{
  "class": "chromium",
  "title": "Some Regular Website - Chromium"
}
```

This did not happen on Chromium 145. The `systemd-inhibit --list` output shows only `delay`-type sleep inhibitors (rtkit, upowerd, electron, hypridle, signal-desktop) — the idle inhibition is happening purely through the Wayland protocol.

### Upstream bug

This appears to be a Chromium regression. There's a related upstream issue about Chromium firing idle inhibit via both D-Bus and the Wayland protocol simultaneously:

- https://issues.chromium.org/issues/386887875
- https://issues.chromium.org/issues/40264445

### Suggested workaround

Add a default Hyprland window rule to ignore idle inhibit from Chromium:

```
windowrulev2 = idleinhibit none, class:^(chromium)$
```

Or, to preserve idle inhibit for fullscreen video:

```
windowrulev2 = idleinhibit none, class:^(chromium)$, not:fullscreen
```

This could be added to the default Omarchy Hyprland config until Chromium fixes the regression upstream.