Fibaro universal binary sensor - very confused

I got it to work (sort-of) fiddling with the Associations in HABmin.
1: Input 1 -> openhabController
2: Input 2-> openhabController

Had to set them multiple times, one at a time and restarted binding/openhab several times.
Strange thing is that after a restart, coming back to openHAB; both entries are empty, but inputs still work:

2018-02-04 23:15:51.109 [INFO ] [.model.script.UniversalBinarySensor2] - Update: UniversalBinarySensor2=ON
2018-02-04 23:15:51.383 [INFO ] [.model.script.UniversalBinarySensor2] - Update: UniversalBinarySensor2=OFF
2018-02-04 23:15:54.735 [INFO ] [.model.script.UniversalBinarySensor1] - Update: UniversalBinarySensor1=ON
2018-02-04 23:15:54.795 [INFO ] [.model.script.UniversalBinarySensor1] - Update: UniversalBinarySensor1=OFF

3: Controller updates; which is the one I really want for reporting scenes, is not possible to set, ever. HABmin only mark it as Pending… after I hit save trying to associate it to openhabController.

Oh, almost forgot: I’m on snapshot #1198 and Z-Wave development binding: 2.3.0.201801212218

@chris, any chance you could have a look at getting the scene channel working.
Could be quite useful:

thanks… perfect. i am using your shared workaround with the Associations in Habmin. Without that, my Fibaro sender does not provide the alarms.
it would be great if that can get fixed. … i am happy to help with testing/logs when needed.

how weird - I set the associations in the same way, but never get anything from my items.

They’re defined as:

Switch Lift_Test1 { channel="zwave:device:zstick:node56:sensor_binary1" }
Switch Lift_Test2 { channel="zwave:device:zstick:node56:sensor_binary2" }

and when openHAB starts I see entries in the event log:

2018-03-12 22:23:38.145 [vent.ItemStateChangedEvent] - Lift_Test1 changed from NULL to OFF
2018-03-12 22:23:38.186 [vent.ItemStateChangedEvent] - Lift_Test2 changed from NULL to OFF

but no event ever fires when the inputs are triggered. I know they’re working, because I can capture the results with an awk script monitoring the zwave.log file (extremely hacky, even by my low standards)

Using the latest experimental binding but a slightly old snapshot - #1149

Dan