Hi.
I’m having a deja vu moment here, need someone (@chris?) to help me remember. I know I’ve had this problem before and discussed it somewhere (might be in this thread thousands of posts up or somewhere completely different).
I’m trying to include some new nodes. When I start the inclusion I just get “Z-Wave network inclusion failed” within seconds and nothing gets included at all. I’ve put a short debug log here but I don’t think it says much useful.
I’m on Debian 8, running a current snapshot (Build #1330) of OH and latest version (2.4.0.201808142206) of this binding.
Here’s a link to my issue that documents it. The channel has always been available, but this firmware did not announce the class so the binding did not use it in mapping items. chris ‘forced’ it to know that class.
Hi, I just noticed a strange thing while using the development zwave binding. I have a bunch of Fibaro dimmers. If I send the command 99 to a dimmer. The light then changes to 100, without me giving any command. The log is as follows:
2018-08-16 20:28:08.821 [ome.event.ItemCommandEvent] - Item 'Hal_Dimmer' received command 99
2018-08-16 20:28:08.840 [vent.ItemStateChangedEvent] - Hal_Dimmer changed from 0 to 99
2018-08-16 20:28:09.884 [vent.ItemStateChangedEvent] - Hal_Dimmer changed from 99 to 100
It seems that this is releated to z-wave. If I send the command 99 to a Hue dimmer, it just changes to 99 and doesn’t change to 100.
ZWave protocol doesn’t have the concept of 100, so to avoid even more confusion, the ZWave binding always sets the value to 100 if you set to 99 - it’s normal.
Hi again. I gave up and just included by just stopping the binding, unplugging the z-stick and using the physical button on it. Included just fine, but now after waiting +12 hours and doing numerous wakeup attempts it still shows as “Unknown Device”. When I push the wake up button I get the following in my log:
It likely means that the device doesn’t report CRC_16 in the list of command classes in the NIF. What version of the binding are you using? A recent one, or something older? I’d suggest to try with the latest at the top of this thread so we understand what you are using (if you’re not already of course).
Ok, thanks. I’ll have a look at this later. I thought that I’d reverted a change that we made a while ago that would have explained this, but maybe not.
1 Like
jwiseman
(Mr. Wiseman (OH 4.2 Snapshot on Pi4))
3869
Hi Chris,
Thank you for your work on this! Here’s my question . . .
Running OH2.3 on Synology with the included zWave binding your built.
Can I run the new zWave 2.4 binding on a OH2.3 install?
I understand the installation procedure you have posted but not sure about the version difference in case there are includes needed from the core of the OH2.4 core or not?
So, I did remember correctly, and the general code to add unknown classes was re-added into the binding already… This is a special case that wasn’t handled since it’s an encapsulation class - I’ve just modified the code to make this more consistent and I’ll update the binding at some point later today with this fix.
Just to note that the 14 August binding is working perfectly for me - all the previous versions with hold-off had been buggy and locked up after 24 hours or so.
Thanks Dan. This is now merged into the master development binding.
For me, I think this is the last outstanding issue before merging this into master - if there are significant outstanding issues that people have, please shout.
Yes, but I’m just wondering if I compiled the main development branch which doesn’t have this fix in it. Just looking at the times of the JAR and the commit, I suspect this is possible…
I’ll do an update now… If it still doesn’t work, can you provide me the log entries around this again as they should change.
No need for feeling sorry, you’re doing a huge job with this binding!
Now it seems it got past that error. I’m still not sure if everything’s working (currently it’s changing back and forth between “Node is not communicating with controller”) but I guess that could be a matter of time so I’ll keep debug logging on and see what happens…
edit: Just for the record, my new nodes are now fully identified by the controller and working fine
What is node 10? My guess is it’s got a problem as the frame doesn’t look correct. It’s possible though that it’s real and it’s doing something super new?