I have no axe to grind regarding Tuya in particular as I donāt use it but these quotes seem somewhat incongruous to me considering the HA ethos.
Such a big fanfare for a third party (developed) integration that appears from what I have read here to not have been particularly well received.
If it werenāt Home Assistant who I (still) trust it would conjure up images of some kind of backroom deal.
Just sayingā¦
Admittedly it does have a bad taste to it although this could be mere wording and not intentionally .
Not helping much that it seems to be rather broken in home assistant.
Thank you very much for this great release! I have 105 integrations and 10 add-ons, and they all work nicely. Iām inpressed. Keep up the good work. For next time, iām seriously considering to join the beta, my wide eco-system would hopefully be a good test case.
Iād like to encourage others to join the beta as well: it will help the development greatly.
The wizard doesnāt tell you anything like this, because itās impossible for it to know. Maybe you are confusing that with the inclusion method selection?
When I added it, all seems okay. But is there any way to see that it is actually using S2 in favor of S0? It just says āSecure: Yesā
It depends on how you added it (which inclusion method), whether the device actually supports S2 and/or S0, and whether the device is a security device (lock or garage door opener) or not.
If you want to check, you can download the network dump and look for the "highestSecurityClass" field of the individual node and check the value. -1 is no security, 7 is S0, and 0-2 are the S2 security classes. The UI will show this in a friendly manner in (probably) the next release, instead of just āYesā.
Using HA for over 5 years now and this is the first time I have ever considered rolling back an update(boy Iām I glad I did a full backup prior to the update). I spent my entire night fixing things after the update. I went from 59 devices the older tuya to 39 in the new one most of which participate in automation routines which have all broken too. The tuya integration is worse than the one it is trying to replace. It is not ready for prime time and the old integration should still be kept in tandem until the new one gets to where the old one has gotten to.
There should be no need for that anymore as the device entries will keep retried even if they are not available during the startup. This fix that will be in the upcoming .1 release will improve the situation further.
After doing countless tests (I also tried with different versions of python), the only way to start HA without the CPU overuse problem was to delete the .storage directory and reconfigure all integrations.
Of course, by doing this, you will lose statistics, additions and more.
I restored the old storage directory so as not to lose everything. I hope this problem will be fixed in a next update, as I haven t found any way to debug.
Has the issue been written up? Probably shouldnāt assume itās going to get fixed unless thereās an issue to back it up. Iāve seen these threads go on for days without fixes because the initial issue was never written up by someone who has it.
I have a problem with Z-Wave. I tried to re-add my Fibaro Roller Shutter 3 via the new security - S2. I was not successful. Node is added, but I can see it as a general node only (without any functionality or sensor data). In Z-Wave JS addon log, I found this:
Error during node interview: Node does #30 not support the command GetV1!
Am I doing something wrong or is there a problem that needs fixing? When I re-add with no security, it works fine.
Using Aeotec Z-Wave Stick Gen5+ on Raspberry Pi 4.
Iām getting the exact same behavior with a Zooz motion detector. Also using a Z-Stick Gen5+ and a Pi 4. I only see one entity now for the motion, with the other 4 disabled. And the device doesnāt have any properties i can use in automations. I can use the entity but the device definitely seems crippled in S2 mode.
An interview error would be a problem with the driver. I would submit an issue to node-zwave-js, and attach debug-level logs that illustrate the error.
Logs can be enabled by going to Configuration ā Integrations ā Z-Wave JS ā Configure ā Logs (tab at top). Set level to Debug. Keep the log window open, and in a second browser window perform the inclusion. The log window must stay open and visible to record the log messages. When the process is finished, click the download button and create the issue, attaching the log file.
Seems this release is causing some errors. Anyone any ideas?
Error doing job: Task exception was never retrieved
18:06:10 ā (ERROR) /usr/src/homeassistant/homeassistant/runner.py - message first occurred at 17:44:11 and shows up 250 times
I can not find it (using open zwave integration)
Thereās no migration for the ozw integration, only zwave.
Yes youāre totally right. I did the exact same thing on about 30 devices late last night. I just did it on another monitor while watching a YouTube video.
All my automations went back to working except for my TP-Link dimmer HS2xx āturn onā. I found as an interim solution I made the HS2xxs turn themselves off, toggle state and then increase brightness. With those steps I could replicate the old āturn onā functionality.
Thanks for the info.
I had an embarrasing number of typos/type conversion issues in templates (both sensors and automations) uncovered by the latest changes. I thought I was good about parsing most of these through the template editor, butā¦ nope. Thanks for keeping the features coming!
What, I am not giving up Tuya v2 for this.
Is it even confirmed that the new Tuya can update devices with dp_refresh / command18 every 5 seconds or so?
And does it display power, voltage and ampere of electricity meters?
At the moment it does none of that.
Mostly because it and home assistant dont actually work together in any reasonable form.
This home assistant update is not worth the update.,
For those with TPLink dimmer issues after upgrade, I noticed both of mine created a second entity (light.XXX_2). I simply deleted the original āunavailableā entity in the integration and renamed the working entity by removing the _2.
Not sure if this impacts everyone, but worked flawlessly for me.
