Yes exactly 
Nice, just curious. Is there a noticeable or measurable difference?
Most here, including myself, have found a few tens of milliseconds difference. Noticeable? To some, yes but not to me. I still use websocket for ML just because I can.
2 Likes
that's what I measured too (~50 ms)
You can use 2 device nodes (on the same device ID) linked to a debug node and use each kind of configuration (webhook and websocket) for each device node. Then look the difference between timestamp when you receive an event. I think the debug node doesn't print seconds in the timestamp, but you can add an extra node between device-debug to add seconds
2 Likes
@jpoeppelman1, it's been several months and both the HE platform and @fblackburn's nodes have been updated but here is a test I ran
Hubitat vs Node-RED response time
It's not comparing wensocket to webhook but it may interest you anyway.
2 Likes
FYI that on v1.4.0 initialization works differently. I notice that after a node-red reboot my device nodes don't show current value until after they update one more time.
e.g. - notice the device nodes show no status underneath. This is a few minutes after a node-red reboot:
After the device updates, things look normal again (1 device updated):
Oh yeah it's weird, I'll look at it
1 Like
Yep, I'm noticing the problem too. It's causing unexpected results when using the Inject node immediately after restarting NR. Please help! 
I haven't had any issues with inject nodes into the device nodes. They have all been working as expected for me.
EDIT: Never mind, the issue I was referencing was unrelated to the hubitat nodes.
I hope he fixes this soon.
You might need to explain further on what your issue is though. Saying you are getting "unexpected results" probably doesn't help francois understand what is broken.
The only thing I see is purely cosmetic - the status text on device nodes. Everything else seems to be working as expected for me (a couple things I thought were broken was something else I messed up that didn't relate to the 1.4.0 update).
1 Like
As you can see, the value is not showing under my Life360 device after NR restart. This inject node runs .1 seconds after reboot, yet it fails on the James path, literally the msg stops after going through the device node. My wife's showing because she was actually driving/ changing the value.
.1s after reboot? I'm surprised they don't all fail.
I wait typically no less than 10s reboot before triggering flow updates. It takes a bit of time for the Hubitat nodes to initialize after a node-red (or Hubitat) restart.
1 Like
Just want to call out that this was working before the NR HE release. Also, if I'm simply querying for status shouldn't it pull from the NR HE device cache? I guess the cache is building at startup, maybe that is the problem...
UPDATE: changed inject node to run after 60 seconds, it now works. I think HE NR node was building the cache as I was querying and it was failing.
According your setup/configuration/devices/hub/etc... the cache can be longer to build than before. You can look at the NR log to see the time needed for initialization and set your init offset according the result.
But it should not take more than a few seconds
Fixed in 1.4.1, thank you for reporting 
The diff from 1.4.0 
- node.devicesFetcher();
+ await node.devicesFetcher();
4 Likes
Thank you fblackburn!
2 Likes
Does anyone know of a good NR thermostat controller / scheduler?
I'm not sure I understand what you're looking for. You can build the equivalent of the Hubitat thermostat scheduler app, or even the new thermostat controller app, using the built-in nodes in conjunction with the hubitat nodes.
