Just to get clarification - if I am going to try to control my Insteon devices from Hubitat, I need to either bridge to the Insteon hub or to a PLM, both of which are no longer available off the shelf. Bridging to either device also requires an intermediary computer running server software under Linux. Please correct me if I am wrong.
Why canāt the server software be incorporated into Hubitat? The radio and power line signaling is done in either the PLM or the Insteon hub, Eliminating the need for a bridging computer would make all of this do-able for a lot of folks, I suspect.
I think it would take a good amount of work to get a nodejs application to run on a JVM (the Hubitat platform runs in a JVM). I also think that doing so requires low enough access that it could not be done by a community developer.
I'm not sure what the commercial incentive for Hubitat Inc. would be to devote resources for an Insteon integration, because, as you state:
I canāt help but think that, with tens of thousands of Insteon hubs sitting out there with no way to access them, it might be worth a day of a Hubitat developerās time to build the code into the system. Hubitat would instantly attract many of those stranded Insteon users who want to preserve their substantial investments in home automation.
When Hubitat added support for Iris V1 devices, they could do so because Iris (Lowes) published specifications for the V1 zigbee protocol with sample code.
When Hubitat added support for other LAN integrations (like YeeLights or LIFX lights), or cloud integrations (like ecobee), it was because an official API was available; the latter being important so Hubitat users don't end up relying on something that is possible ephemeral.
None of that is true in this instance. There are differences in the two Insteon bridges (2242 vs 2245), and the PLM. And for some of this the only guide available is reverse engineered code. So it would involve more than a day of a developer's time.
A possible indication of the magnitude of the effort involved arises from the Hubitat team's decision not to prioritize developing an Insteon PLM integration even when Insteon was in rosier times, as indicated in the linked post:
Interesting background on this decision - I appreciate your insights. What might have changed is Insteon has gone from a large base of supported users within the Insteon environment when that decision was made to that same base now being abandoned and desperately looking for a new home to host their investment in devices.
Apparently Home Assistant has developed very portable code that runs on a variety of platforms to connect with any of the Insteon hubs or the PLM (link below). That seems like a viable alternative for those looking for a new environment but my preference would still be for Hubitat to provide the same functionality.
If Hubitat did anything integration wise, and to be clear we have no plans to do so, by the time we have something we are comfortable and proud of most of the Insteon base will have moved to other tech or already implemented our community integration, or some other solution...
When Lowes Iris tanked, we had notice and time to work a solution before D day...
I bring other devices that don't work well with Hubitat into Hubitat as virtual devices using NodeRED. Everything stays on my LAN and works well and is fast.
I certainly donāt have the market perspective that you have, @mike.maxwell, but from my perspective, throwing a life preserver to tens of thousands of Insteon users inviting them to buy a Hubitat and continue controlling their investment in Insteon devices seems like a good business decision. Those users will go somewhere to continue protecting their investment. My hope is that it will be Hubitat because I believe it is a superior environment.
It would be a horrible business decision for Hubitat to pursue additional Insteon support. There are already several companies/services pursuing this finite and falling market (Home assistant, Homeseer, others) and they have a head start because they already supported Insteon. Some (e.g., HA) can be made to work with Hubitat. Come on, people, put emotion aside and think about this. Not. Gonna. Happen. Nor should it, if you want Hubitat to thrive.
There IS a Lutron Caseta Dimmer (PD-5NE-WH) that has a round favorite button in the middle of the dimmer that can be set to any level you choose. When you want less than full on, instead of pressing the top on button, you just press the round favorite button. The dimmer is pricey, but it is rock solid in terms of network connection and functionality.
I have simply automated all of the lighting in the house with motion sensors, contact sensors, and/or time-based scheduling. We rarely need to touch a light switch, and thus all of my Caseta dimmers always come on at the correct brightness level. Also, Caseta dimmers can be configured manually to limit the upper and lower ends of the dimming range. That may be helpful for some users, instead of paying the premium for the PD-5NE.
Same here -- the lighting in our main rooms & areas is all automated one way or another - there's very rare need for anyone to ever touch those Caseta dimmers. The only switches we actually ever touch are non-dimmer on/off switches (mostly Caseta and a couple Zooz) in sporadically-used areas like closets, random nooks etc.
Everything is 100% automatic.
My wife does not even know where all switches are (they do exist and well accessible).
The only one manual thing - is a little zigbee button for turning off whatever is still on
when she is about to fall in sleep. Of course, I can automate this if I come up with
nice and reliable algorithm how to detect "we are both sleeping" condition.
Home Assistant isn't a commercial product. They have some revenue streams, but it's just not the same business model. Honestly, while I get your passion for the product (and to be clear that's why I worked with @cwwilson08 and Scott Kuester to test as much as we could to bring the Insteon Websocket server to fruition), the business case for a commercial product to support a currently dead product line (I'm not saying useless because it's certainly not, but unless the line is picked up, for which they claim to have already tried and failed, then it's not going to come back to life. End of story). Chris put in the lion's share of effort, developing three different versions of support on HE over time, for a system he didn't even own or use!
With the passing of time and little users responding to the efforts, we just could not keep testing and iterating. For myself and most, it does exactly what we set out to do, and with more responsive results than even the HA integration that they spent years working on. Contrast that with the efforts of Chris, building three different integrations over a single years time! I'm not dissing the HA Insteon integration. It may be slower, but FFS it works. What's to complain about there? And they have agreed to now integrate Tom Harris' excellent Insteon Panel too, giving back device configuration so you don't have to stand on one leg and go through complicated device button presses. They also claim they will be putting more development effort into it. So if that actually happens, what's not to like there? Setting up HA to run on an old computer with VirtualBox takes 10 minutes or less. Unless you're really savvy, setting up the node.js instance is going to take you a lot longer than that. But if you've got some time and patients. it's worth the effort.
Probably you can use Home Assistant Device Bridge to bring HA discovered Insteon devices back into HE. I did a quick test a week ago, and it didn't work. Might just need some updates or new device types to make that work properly. You can always. help in the community support efforts by setting up HA and testing that, but please post in that thread about Home Assistant Device Bridge issues, not here. I can test myself at some point, but since I already run the HE node.js setup and have for years, it's not a priority. Don't get me wrong, I want to help, but I can only afford to volunteer so much of my time for this. Bottom line is, there are already several viable solutions, so pick one.
Iām not familiar with how double tap works with the Lutron Caseta dimmer. Could you please explain? Does it temporarily override the trim adjustment, or something else?
