Thanks. If required I will move to zigbee2mqtt but at the moment it is online including
zigbee:tuya_ts0044:SLZB-xxxx, 2 of the 3 items I can read but the buttons are not working. For test I use this code
rule "Button 1 short press"
when
Channel "zigbee:device:SLZB-06M:xxxx:button1" triggered SHORT_PRESSED
then
logInfo("default.rules", "Button 1 short press detected")
end
but is not working. Any idea how to work with triggers?
The SLZB OS can be the latest available for the stick. Mine is 2.9.7 and I just saw that they released a 2.9.8.
The second firmware is the one for the Zigbee chip. The protocol implementation up to version 13 inclusive was using a Software Development Kit (SDK) with versions up to 7.something. The chip manufacturer did a complete replacement of that SDK from version 8.something, moving to protocol version 14. This latest SDK causes an incompatibility with the function library used by the Zigbee binding in openHAB which doesn’t yet implement these new function calls. This means you need to pick a Zigbee firmware using the 7.something SDK.
If you don’t use the openHAB Zigbee binding, and instead you use a MQTT broker + zigbee2mqtt, then you do not have this limitation and you can use the chip firmware with the latest SDK 8.something. This will also give you a bit more flexibility in using Zigbee devices that are not fully compliant with the standard, but does mean you have two additional software components to install, configure and maintain.
@chris is the maintainer for the binding, I believe, and I totally love his principle-based approach to adhering to standards. From what I’ve read in various discussion threads, the effort to overhaul the entire library and then the binding is completely unjustified. There aren’t any obvious advantages or new features as incentive for the tens if not hundreds of hours of work, test etc. But to attempt it is a massive pain in the neck, because the chip manufacturer seems very keen on its tradition of never documenting anything and just letting developers figure it out.
It would be more exciting to see effort put towards integrating other hardware interfaces based on Texas Instruments chips instead of Silicone Labs, which would open up the door for more Zigbee coordinators. But that still doesn’t resolve the compatibility issues with the gazillion products straying away from a certifiable standard, so in the end it might just make more sense to put the effort towards improving the adoption and integrability of MQTT, with multiple benefits even beyond Zigbee.
Personally I’d be very interested in the ability to use the SLZB-06 (without M) with the TI chip, because that one has a Hub mode which seems to make it work standalone. That would allow me to define rules for water leak sensors triggering an alarm and the closure of the mains water valve with only the Zigbee coordinator powered on. This would restrain the openHAB computer to a non-critical monitoring role which can fail without breaking key functionalities in the house. Software can hang, computer can overheat, software rules could fail to trigger because of some untested update… None of that should end up with a flooded basement.
Sadly for newer devices or newer revisions of existing ones, there’s no longer firmware available with SDK V7.* - only SDK V8.* - I even see dev firmware about SDK V9.*
So really hoping the Zigbee binding will be updated soon, else the only solution is Zigbee2MQTT - but I prefer a more native approach - and seeing V9 is also on the horizon