oberkc

Yes, I am very hesitant to use matter beyond experimentaion at this point. Most of the devices I have tried work intermittently, at best.

I forget which of the apps that preceded UDMobile had a Tasker interface. I used to use Tasker for geofencing. I also recall that I could create direct buttons on the Android Home Screen with a prior app. I found these useful at the time and not available with UDMobile, but my interests have evolved over time and I have found other options. Still, these options might be missed by some.

There is. As IndyMike says, link the scene to Alexa, rather than an individual device within the scene. When linking, define the scene as a "light" to Alexa. If you happen to have any programs that turn this light on or off, make sure they command the scene, and not the individual device. One other possibility (but I doubt this is a factor unless you did this without knowing) is the scene definition itself. Yes...you have everything as controllers...which is correct. But, it is possible that devices within a scene can be commanded to turn OFF (or at a dimmed level) when a particular controller (including the ISY itself) commands the scene to turn ON. In your scene, select each of the controllers and observe the table to the right, making sure the "action" is on or default. Also, as near as I can tell, if a controller in a scene is a dimmer switch and one dims it to a relatively low level (say less than 50%) a responder keypad will not light up.

Last I checked, PLMs were available.

When you ask Alexa to turn it on, is it turning on the scene, or an individual device within the scene? If the latter, this could explain one way your devices are getting out of sync.

If the problem appears global and random, I usually suspect comms problems or failing PLM. Perhaps a failing Insteon device has the ability to flood the communication paths, but I am not sure how to troubleshoot such an issue besides trial-and-error. Unfortunately, I am not much help with regards to error messages.

If turned on/off by the keypad, or by the switch/module, the keypad button would reflect this status because of the scene. Is it possible that there is a program that is turning on the switch/module, or that the switch/module is a member of another scene?

In my mind, the variable adds unnecessary complication UNLESS you are trying to track current state (I am currently at home, I am currently away) versus just tracking when one crosses the geofence boundary. Having said this, if one is trying to keep track of a presence (or any other binary-value variable) I suggest considering virtual switches, rather than a variable, to keep track of presence. Using a variable switch adds some convenience to manually changing the presence status should that become necessary (such as when the geofence doesn't fire for some reason). Clearly, though, there is some use in knowing current state of presence. For example, one might have a program to turn on some lights at sunset, but only if home.

The first thing you do is disable the program so that it does not self-initiate. The second thing you do is to add a command after the wait statement to turn the lights off. The third thing you do is change the trigger in your geofence. Tell the geofence to "run if".

Does it call for "Run If", "Run Then", or "Run Else"? Given this, the lights will come on at sunset and turn off as sunrise, regardless of any geofence activity. I do not believe that the presence of one phone inside the geofence would cause the second phone to fail triggering the enter command. Why do you have a "wait 10 minute" action at the end of the THEN clause? Next time you enter the geofence, soon check the admin console (or log) and see when this program last ran. Hopefully, there will be a record of the program running at the time you entered the geofence.

That program appears to be set up correctly to me. (The two yellow programs below it have something wrong, but this should not matter.). I would right-click on you "cleat and hearth" program, then choose "run then" to confirm that this is the program that is not working. If you follow @Geddy advice and find that turning the scene off directly from the admin panel does not work, then you might restore the affected devices. Failing that, you might consider a new communication problem that may have arisen.

Yes, hue emulator has same IP adress as EISY. It just seems likely in my mind that changing the port number will break something. I fear that my experimenting with 20 different node servers and separate hubs for hue, casetta, yolink, harmony, and others, and integration with alexa and google nest, I have this sytem that breaks any time I make a change.

Summary: What are the consequences of changing the IoX HTTP port to something other than 8080? Background: I have a renewed interest in trying to get a brilliant smart switch connected to the EISY via the Hue Emulator. I recall that the brilliant switch looks for a hue hub on port 8080, which creates some sort of conflict if the emulator port is the same as IoX. If I change IoX port to 808X and the hue emulator to 8080, what are the unintended consequences? Will the finder not be able to connect? Will the EISY-UI not work? Will I brick my system?

I must admit that I was quite disappointed to learn that the "failure" of a single z-wave device can cause such broad consequences to the other z-wave devices.

Strange. This is how I do it and it works every time.