[Release] Xiaomi / Aqara device drivers

community_driver

#403

There are virtually no other ported ST drivers that I'm aware of that will be effected by this change.

There were no changes required to any of the Hubitat built in drivers in support of this.


#404

Cool thanks for the info, are you going to release the 2.0.5? like today? :stuck_out_tongue_winking_eye:


#405

It is it urgent this firmware release/change? Or better are there any ramifications by not upgrading?

As this can create a lot of issues to a lot of users due to not only the drivers but rules that depend on those drivers.


#406

@veeceeoh has already given his time line, the decision as to whether to update to 2.0.5 or not is up to you.


#407

OK, but I still maintain it would be better to advise users of exactly what is changing.

I've knocked up several of my own drivers for various no-brand UK ZigBee devices, plus am using some other Aqara drivers (not from @veeceeoh), a custom Hue motion driver from somewhere I can't remember, and an ST ported driver for an Osram 4 button switch.

How do I know if the next firmware update is going to break any of that or not?


#408

Not what I asked.
What I asked was if there would be any reason for me to update straight away. For Example if there were changes into the GH integration that requires you forcibly to upgrade to continue to use.


#409

I have no idea what changes you're looking for, you'll be able to review the release notes before updating though.


#410

If you're parsing the description string directly (as these drivers are) you may have an issue.
If you're using the parse helper methods as one should be using, parseDescriptionAsMap you won't...

If you are parsing the description string directly, or need further clarification feel free to PM me.


#411

I think these guys are looking to start an Update War!


#412

I don't know why people do that... lol. The answer is the same, no matter the software: when it is ready/done.


#413

I have several Aqara sensors in my house - leak sensors, door sensors and motion sensors.
Yesterday I have updated to the new firmware (2.0.5) and so far looks like the basic functionality of the sensors I checked is working correctly.
Motion is detected and door open/close is detected as well (haven't tested the leak sensors yet).
I believe the firmware update is not as bad as feared (at least for me).


#414

Did you check if you are getting any errors on the logs? Thanks for the info.


#415

There are problems on the vibration and temperature sensors. Leak, door and motion sensors didn't give me any issues off the bat but I didn't dig too far into them as the basic functionality seemed to work.

Specifically on the vibration there was an issue with battery reporting and tilt reporting. I didn't even bother to test more than that as that's all I use them for right now.

On the temp sensors, the temperatures will report wrong and the battery reporting will break.

Best bet for everybody with Xiaomi devices (to save yourself some headache) you might just want to wait on the drivers to get updated and tested properly. It shouldn't be long. Just unfortunate timing.


#416

I suppose you click configure in those devices after the update?


#417

I'm guessing thats for me?

The issue is in the code itself and how it parses the information received from the device. Clicking configure will not fix it. Won't take long for it to be fixed... probably testing will take longer on the vibration sensor because it does soooooo many things.


#418

Yeah.

I was looking at the temp sensor code this morning. I think I can see what needs to be changed, but I'm not home to test it - and since I use my Xiaomi temps directly for thermostat control, not sure I want to break it right now. :smile:

I'm sure veeceeoh will get it fixed up sometime next week.


#419

I'm in a similar boat with the temp sensors...they are controlling the Keen vents in the bedrooms so I think I may hold off as well.


#420

yeah, if you want to the full on manual route, attribute string data > 1 byte needs to be swapped end to end
IE D01D -> 1DD0 then convert to its int representation...
both temp and humidity are 2 bytes


#421

Yup. :slight_smile: That is what I thought based on your other comments.

Isn't hard to do, really. I just have to think if I have time before Sunday, since I'm out of town next week. If I leave the thermostat broken, my wife and kids would not be happy. lol. Well, not BROKEN really, it just have to control from the thermostat temp instead of the Xiaomi.

EDIT: Ok, based on more information, I will say it isn't a hard change for SOME of the sensors... Luckily, or unluckily. I only own the Aqara temp sensors, so that is the only one I am comfortable potentially modifying (which should be easy).

I'll see what my schedule looks like tomorrow.


#422

I just gave a quick look in the logs of the door and motion sensors, didn't see any errors there.