All,
This will be my last post. I will no longer be updating Hubigraphs and I encourage the community to pick it up, modify it, and own it. Feel free to take credit and ownership. I have no plans to delete my Github account so things should continue working.
It has been a journey, but I as of today, I am no longer using Hubitat for my home automation needs. It has been fun, and I have enjoyed learning and implementing HubiGraphs. I have thoroughly enjoyed the Hubitat developers' support and seeing the platform grow. Every comment below is my opinion and not meant to disparage the hub or its team. TO BE CLEAR: HUBITAT ON THE WHOLE IS A GREAT PLATFORM AND I HAVE ENJOYED MY EXPERIENCE. But, the Hubitat platform has come a long, but as in any relationship, it is time to go. Why am I going? How about a list?
-
Hubitat hardware. People have begged for years for better, more powerful hardware. In the Hubitat team's defense, it is sufficient for thier stated goal: Home Automation. However, for my needs which is more focused on Home Visualization AND Automation, Hubitat is ill-suited. I did not come to this conclusion lightly. I had 1 C7 hub running my entire zwave and zigbee networks. Total devices? 150+. I had a separate C5 hub running all the network connected devices as well as HubiGraphs and HubiWidgets. In that configuration, simply POLLING 5 of my non-reporting devices brought my zwave network to its knees.
-
Architecture. The following statements will be controversial (and are only based on observation) but here goes: Hubitat architecture is built up to the point it appears to be collapsing in on itself. The database implementation (H2) is slow, buggy and scales very badly. As a result, in order to be more efficient, saved events went from 2000 to 100 (states dropped to 30). Yes, the developers gave you the ability to change that, but at the cost of hub performance. I do not know if Hubitat takes advantage of multi-threading, but based on what I have observed, I doubt it. To recall 1000 events, it takes Hubitat 7 seconds. During that time, heaven help you if a zwave event arrives or an automation is supposed to kick off. Yes, I tested on both hubs.
-
Inconsistency in Message. "Hubitat is NOT a development platform". "We'll expose file storing for developers". "Making file-store available to apps is coming". "Making file-store available to apps is not a priority". "Subscriptions are bad". "Introducing Hub Protect". "Introducing Remote Admin".
-
Commercialization. "Developer-plus" and the "opening of a store where people can share thier apps and drivers without sharing their code". Ugh. I don't disagree that folks should get paid for their work, but to encourage it? I am not in this for the $. I understand that Hubitat is a company but this feels like a money grab. Why? They are NOT offering insight into their architecture. They are not giving an opportunity to use their in-house API... just a place where you (and they) can make $.
-
Lack of Hearing the Customer I have encountered this in the past. The dashboard - garbage. The IOS app - garbage. My son built a better looking and more functional beta (see HubiPanel) in 3 weeks! Anytime someone complains, it has become a running joke that a certain staff member will disparage those comments. Other deficiencies - Alexa not "hearing back from a device", lack of a roadmap. We don't know what to expect, but based on reading the forum and projecting, it won't be long before we are paying for a new device AND the driver to use it.
-
Groovy - I hate Groovy. I get that it is the heritage language that Hubitat is based on, but in Computer Science you should always use the best tools for the problem you are trying to solve. Can someone really tell me that Groovy is the best solution?
-
Security - I will just say, I wrote a driver this past spring as an experiment. This also highlights the poor architecture but...I was able to successfully intercept every switch event on the hub and redirect it to do anything I wanted to... from a custom driver. It took 5 lines of Groovy. Out of respect to the Hubitat team, I am not publishing details.
Look, I get it. Hubitat is not a developer platform, but THE MAIN REASON I AM LEAVING HUBITAT IS THAT I AM TIRED OF HACKING AROUND ITS ARCHITECTURAL, HARDWARE, AND PHILOSOPHICAL DEFICIENCIES. If that was too harsh, I apologize. The Hubitat team has done nothing "wrong" but for all the hours I spent working on the platform, I do not see it having a future any better than SmartThings.
Be well. Enjoy you hub. It is a good platform, certainly one of the best. Its not you, its me.
