Advice reqeusted to migrate MSR from Bare Metal to Container

[toggledbits]

Sorry, but the image looks like you're showing position, not watts. I would like to understand which of the two power values provided by the device you are trying to reconcile.

That said, a position of 0.91 would be correct, as the scale of the position.value attribute in Reactor is 0.0 to 1.0, computed from the device range of 0 to 99 (90/99 = 0.91 rounded to two digits).

[therealdb]

@toggledbits said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

Sorry, but the image looks like you're showing position, not watts. I would like to understand which of the two power values provided by the device you are trying to reconcile.

no screenshot. As I've said, I'm finding that the device is not reporting that well. I'm 100% sure it was OK under Vera, and maybe it was part of the custom code they wrote for the device, Idk.

@toggledbits said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

That said, a position of 0.91 would be correct, as the scale of the position.value attribute in Reactor is 0.0 to 1.0, computed from the device range of 0 to 99 (90/99 = 0.91 rounded to two digits).

ok, but I've asked for 0.90 via Reactor and I got 0.91 inside Reactor and 0.90 inside ZWaveJS. Imagine something that tries to position the device until it is OK...

[toggledbits]

The rounding is a function of the scaling.

I believe I mentioned in another thread that comparing equality of floating point numbers is a well-known Bad Idea in pretty much all programming languages.

The raw value is always available in the x_ extended attributes.

[therealdb]

so, the correct way is to look for x_zwave_values.Multilevel_Switch_currentValue and divide by 100?

EDIT: anyway, I still think that this behavior is not OK, because value are not consistent, but the call is yours.

[toggledbits]

I can change it, but it may be a breaking change for all users at this point. Not something to be taken lightly. I have considered it.

I still don't have enough information from you to do anything about the power issue, assuming there still is one.

[therealdb]

@toggledbits said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

I can change it, but it may be a breaking change for all users at this point. Not something to be taken lightly. I have considered it.

No problem, I found it strange and I was reporting. It's also strange 'cause after a refresh the value from the controller is overwritten.

I still don't have enough information from you to do anything about the power issue, assuming there still is one.

Yes, but it's zwavejs that's not updating. I suspect Vera was doing some magic behind the scene. I'll keep watching it, now that the new system is almost stable.

[toggledbits]

@therealdb said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

Yes, but it's zwavejs that's not updating.

I see. I guess I didn't understand that before. OK. Let me know if you figure anything out that helps. There may be something I can bolt on to help with that. There do seem to be delta-sensitivity configuration settings on the device. I wonder if those are related?

So for the moment, have we covered all the hot-button items? Anything I haven't addressed yet?

[therealdb]

@toggledbits said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

There do seem to be delta-sensitivity configuration settings on the device. I wonder if those are related?

Maybe. I'm experimenting with parameter and I'll report back. I still have a couple of strange behaviors to fix before I could dedicate to fine tuning.

@toggledbits said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

So for the moment, have we covered all the hot-button items? Anything I haven't addressed yet?

Yes, the only thing pending is tilt position. I managed to use it via raw ZWave commands, but I think it could useful to have a separate endpoint for that. RS 3 and 4 are already exposing multiple endpoints, where the first is controlling the blind and the second the lamellas (and root does nothing).

Unfortunately, those from Fibaro are the only ZWave devices in Europe that are supporting lamellas. I'm sure eventually Shelly will catchup with its new ZWave series from Qubino and I'll eventually replace my RS2s, since they're not ZWave+ and are 8+ years old (I bought them for the old house, where I started with just roller shutters, so they served their purpose well).

[toggledbits]

@therealdb said in [ZwaveJS] Position and cover not working for Roller Shutter 1/2/3:

Yes, the only thing pending is tilt position.

OK. I've lost the thread there. What's the problem with tilt at this point? In my bench-top use, changes in the Z-Wave reports feed to the position attribute correctly, so I'm not clear where the issue remains.

[therealdb]

Sorry, I was referring to this thread: https://smarthome.community/topic/1576/zwavejs-venetian-blind-tilt-for-fibaro-roller-shutters-2-fgr222/4

[therealdb]

ok, meter fixed. I don't know how it worked all these years, but the associations of device 4 were completely wrong (assigned to a non-existence device), thus it was somewhat working on Vera. Go figure. I reset the values and now it's reporting ok.

The very last thing about the device is that, after configuring a parameter, they support scene activation v0 and it would be nice to have a separate scene device, as in the switch. Otherwise, I'll simply go with x_zwave_values. Thanks.

[toggledbits]

Try activating the central scene on the device. Since these values are events, they aren't created until the first event is received. You may also need to refresh the browser, or at least "stimulate" the redraw of the entity list (by modifying the filters, etc.).