Adam Karim

Looking at it, it only looks like this change would affect the nixosModule. I've been using this in my flake outputs as a workaround but not ideal. ```nix apps =...

``` CGO_ENABLED=0 GOOS=linux GOOARCH=arm go build -a -installsuffix cgo -o kube-plex_linux_arm github.com/munnerz/kube-plex sed -e 's@alpine:3.6@arm32v6/alpine@' -e 's@kube-plex_linux_amd64@kube-plex_linux_arm@' Dockerfile | docker build -t munnerz/kube-plex-arm -f - . ```

I ran into this as well and just settled on: ```python groups = {k: hass.states.get(k) for k, v in hass.data['entity_info'].items() if v['domain'] == 'group'} groups['binary_sensor.stairway_motion'].attributes['entity_id'] ``` Not ideal but works...

And don't get me started on gathering ZHA groups ;-) If you run into that, take a look at `hass.data["entity_registry"].async_get('my.entity').translation_key` which should right something like `light_group` or other similar items...

Same here. Frustrating. Reverting the change [here](https://github.com/mdeweerd/zha-toolkit/commit/2fe9507820e82127ac3540d046466beae7ea6880#diff-583ba145d04c99a9ef3eaee73bc0dbfe8122109d7b72871a18de7527e710516dL660) fixed it for me. Looks like too many levels of `async` ;-). > Just the single change, nothing else needed.