[Rant] Getting tired of Hubitat's issues

💬 Lounge Debate Chamber
1

Tirade for Hubitat devs... TL/DNR: Either provide more support in both process and feature-set to app makers, or make the system actually user friendly, but having an arcane undocumented development environment with zero customizability AND having a system UI from the 1960's for cumbersome Rule Machine routines is pushing me away, and never has appealed to my spouse.

I've been working on building some apps, and have had my fair share of difficulties. Examples: No support for rooms (as far as I can tell - seems to just tag used for the devices page?). state/atomicState variables don't support dictionaries properly (eg state.varName.remove(key)) (not to mention, state/atomicState being weird to begin with). Input buttons (input buttonName, 'button') don't work consistently. There are weird tichy rules about where and how in an app UI for parent->child calls, accessing settings variables, etc. (ie before an app.label is entered, prior to between "preferences" and "page", etc). HTML sometimes sorta works, and sometimes doesn't (ie in section titles), and isn't "supported" (read: Hubitat developers don't care if there's a system bug related to HTML). There's zero customization for UIs, and the existing options are... shall we say "limited". The documentation is just pathetic. There's no support for JavaScript [in app UIs], but maybe kinda sort of use it, if you can figure out how. The #include feature is janky at best - calling it "library support" would be generous. There's no debugging. Often runtime errors don't log correctly, and when they do, sometimes don't provide line numbers. The list marches on.

Sometimes I spend hours trying to debug some issue, only to realize.... I made a dumb mistake. But other times, I realize it's an undocumented system issue. And when there's a fair chance of it being either, it makes it so much harder. With sketchy documentation and zero debugging tools, how the hell is anyone supposed to know? Should I be looking for an error, or trying to find a work-around? To be blunt, to do anything remotely complex, there's a LOT of workarounds needed.

When I ask questions or mention a problem (which TBO is rare) the problems are just ignored. Don't get me wrong, I usually get an answer and quick! But the response for anything beyond "wtf is the problem" is always "that's not a priority" and/or "only a few [of the few crazy enough to try to program for Hubitat] are likely to ever have a problem". For instance, when state variables were behaving bizarrely, after hours testing and debugging, I go searching the forums and come up empty. Look at the documentation, nothing. Ask in the forums (taking the time to mock-up and test an example), and (quickly) given the answer. Thing is, I missed a forum post... from 2021. In at least three years after being identified as an issue by a user, not only has it not been fixed but it hasn't even been documented either! To be clear, this is a BUG - no programmer should just magically know that state variables that look and act like variables aren't somehow actually variables! edit: And how often do people ask about state vs atomicState? Getting answers is great, but fixing the problems would be greater. /edit

I'm not going to pretend like none of my issues aren't related to my programming skills.... I won't argue that some by themselves aren't minor issues (including the example of state variable). But, I've seen no interest in actually addressing issues for developers either. Hubitat, you need to figure out which way you want to go. If you're aiming for average end users (like my spouse) the complete lack of user-friendly interface won't allow it. She won't touch it. You have a LONGGGG way to go. If you're aiming for real power users, you don't have the development environment for it. I'm damn frustrated and considering bailing out. Why waste so much time developing an app in groovy that has pretty much zero portability for... maybe some day, having my apps used by a few dozen or maybe hundred users? Since it's not user friendly enough for my wife, and I'm not given the tools to build something user-friendly enough for her, seems to be the target audience is ... who exactly? A few people who are already Hubitat super-fans?

edit: I know people will say "Well, don't make apps. Just use Rule Machine." Well, first off... This isn't for you! :stuck_out_tongue: Second, I've tried. For anything other than simple If X AND/OR Y, THEN Z sort of stuff, it's too difficult to use, and still has the same issues - finding workarounds and making compromises to suit it's limitations. /edit

Hubitat, either rebuild the platform with a nice system UI, making Rule Machine/Basic Rules [edit: correction, Simple Automation] easy enough for normal people to use, or fix the development environment so that we have the ability to do it for you.

3 Likes
2

Good luck on your next home automation platform of choice. I hope it fulfills all those gaps and needs for you.

15 Likes
3

Admittedly, that's part of my frustration. There's precious little options. But, I hope very much that Hubitat's business model isn't based on "Haha. Good luck finding something else!"

But, I will give Home Assistant a shot, despite a list of problems I have with it even before seeing it in person.

4

Is that your hope?

It seems like since your complaints are directly targeted at Hubitat staff and developers (and not the community, nor really a community "discussion" topic) that if you were really trying to provide feedback and improve the product you would have sent them that list directly instead of blasting it on their public forum.

But whatever?

Anyway, as I said - Good luck! There are many different platforms out there for these exact reasons. I genuinely hope that HA or something else can meet your needs.

1 Like
5

This sounds like a line from a list of Festivus grievances.

8 Likes
6

Have you tried Node-Red? You seem to have an affinity for Javascript so you might be right at home with it. The learning process is relatively quick, but moving all your existing rules might be a little tedious. I ran it with Hubitat for years relegating Hubitat to basically a hardware coordinator.

7

That was my answer. I mainly use Hubitat as a Z-device hub, and do all logic and programming in node-red. It also makes it really easy to mix and match devices from Hubitat, home assistant, some cloud providers, etc, in my logic.

It isn't for everyone, but was a solution that worked for me.

1 Like
8

It seems like since your complaints are directly targeted at Hubitat staff and developers (and not the community, nor really a community "discussion" topic) that if you were really trying to provide feedback and improve the product

That is correct. I love my Hubitats (plural). I just don't love it enough to put my love for it behind my actual system needs.

edit: And to reiterate, where it is lacking is in either 1) Being user friendly enough for my spouse, OR 2) Making it at a minimum easier for me to do that for myself. I know I sound like an angry troll, but 1) I'm frustrated, and 2) I have two volume settings - I either deal with it, or I vent. /edit

you would have sent them that list directly instead of blasting it on their public forum.

Not sure what you mean. The forum is literally called "Feedback". Was I supposed to call Mike Maxwell's personal phone and invite him to lunch?

3 Likes
9

I would urge the dev staff reviewing the @Roguetech message to consider that while Hubitat cannot meet many of @Roguetech needs he does make several good points. It’s easy for the whole message to be dismissed in totality rather than consider all the elements of the message.

10 Likes
10

affinity for Javascript

Oh, gawd no! :joy:

What I'm referring to is allowing JS in app UIs, like "onclick" events and such. That is, JS being processed by the browser client when viewing an app [edit: interacting with an app UI]. By extension, I'd like to see being able to swap variable values between UI JS and Groovy, but I have no clue what kind of security issues that might introduce.

I would like an alternative to Groovy, but... whatever. Groovy gets the job done.

11

No matter what anyone says, I don't think there's a single system out there that will meet everyone's needs. Hubitat, IMO, has done a good job targeting the 80/20 rule. For the 20, a combination of tools (node-red, HA, etc) can be utilized to go that last mile.

5 Likes
12

Yes, most if not all of the message will be lost on most of us due to the title of the thread and the approach they took to submitting feedback. :shrug:

There are effective, and ineffective ways of providing feedback. This thread is the ineffective style.

Anyway, putting this on mute now so that I don't derail it further.

8 Likes
13

Yes, most if not all of the message will be lost on most of us due to the title of the thread and the approach they took to submitting feedback. :shrug:

I changed it to "tired of hubitat*'s issues*", but open to suggestions on how to not rile everyone up for no good reason. Not my goal.

3 Likes
14

Agreed, this approach has a high likelihood of causing a flame war in an online forum, but less likely to result in any meaningful change along the lines OP is presumably hoping for.

Anyway, flame on.

image

3 Likes
15

Will you please expand a little on this particular point? What is not clear to me is how your spouse interacts with the system. Are they trying to use the Hub's Admin interface directly? Or using a Hubitat Dashboard? Or Voice control using Siri, Alexa, or Google? Or using the Apple Home interface on their phone?

My wife pretty much never touches any part of the my home automation platform. She simply enjoys the automated portions of it. Once in a while she'll use Apple Home on her iPhone to remotely open/close a garage door. Otherwise, the house is pretty well automated, and thus she has no reason to ever see the Hubitat Hub's web admin page. She does use Amazon Alexa voice control from time to time, but even that is fairly rare.

It sounds like you have some programming background, correct? Does your spouse also have a similar background in technology? Are they trying to write their own automations? I am just curious as this would be interesting... Most home automation systems are not really designed to have multiple developers on the same platform, as there are so many dependencies between automations and devices. I could see it becoming very cumbersome to try to have multiple Devs on one hub. :thinking:

Thoughts?

13 Likes
16

^^THIS^^

My wife is a Software Engineer with decades of experience, and she has absolutely no desire to know or understand how to set up any home automation. She could learn and master it, but has no need. She just wants to enjoy automated things that just work intuitively ... and Hubitat has not only made that easy to set up, but it lets them operate reliably.

13 Likes
17

Take a look at Homey Pro. But be prepared $$$$

1 Like
18

edit: Your general point is a good one. It's not fair for me to blame Hubitat for my wife :joy: My point is, even if this one specific spouse of mine has no interest in "automating" as distinct from "automation", I feel like the alternative of writing custom apps is also needlessly not user friendly, even for "developers". I could instead complain they've made it too hard for her to use. That too is a legitimate complaint, but my personal frustration happens to be framed as the other way. /edit

What she has done: 1) Create a (simple) dashboard, 2) Use the dashboard, and 3) Create a Simple Automation Rule (?) to turn something on/off at certain times. With Hubitat, she even balks at pulling up a device to manually change settings - which, TBF, I have like 90ish devices, and my naming scheme could use maintenance.

But, to go through the UI shortcomings would be another long post. I've literally just now pulled it [Rule Machine] up, here's a short (non-exhaustive, 8 page) list of issues I have (which I feel confident the wife would largely agree with :stuck_out_tongue:). this is essentially a functional spec for my apps (so I know a lot of it CAN be solved). It [RM] is a great app, super powerful, but anyone who calls it "user friendly" deserves a smacked bottom for lies.

  1. Well, it's an ugly form from the 1980's :joy: That's not good, but I'm not saying RM should be prettied up, rather developers be given some latitude to pretty up custom apps. Built-in defaults aren't always pretty.

  2. (I rather quickly got a Java error, but... whatever. I'd post a bug report, and I know Hubitat would most likely patch it, but I don't use RM so don't care.)

  3. No instructions.

  4. Obtuse options.

4a) Not plain-English, average end-user options, like "Contact sensor" as opposed to "door sensor".

4b) Contradictory options. The "[Type] sensors *" option has a page title of "Select Trigger Events", but "Event" is not a "Device", and therefore is confusing.

4c) Options disconnected from context. The "And stays?" field is not next to the "[Device type] reports", so makes no sense - "And stays" what?! Even next to would be less than obvious, since the user does not need to select a "reports" option.

  1. New pages open without breadcrumbs, so the user has to select options with no context, no reminder of wtf they're doing,

That's where I stop - I selected a "contact sensor", and I need to select "open" or "closed" (or "changed"), but I want to do something with both. I don't know if I need to setup two different RMs, so I have a gazillion different RMs? If staying with "open", will an action for "closed" show up later? This is when I say "I can make this easier to use". (And that's when writing an app takes me months longer than it should.)

1 Like
19

This is exactly my approach for the Home Automation. Near everything is 100% automated (I am running out of ideas what to automate next and which way). My wife requested an ability for manual control of everything (done) but right now she is not sure where all switches are and what exactly they are controlling. She is using only few buttons on her night stand remote.

7 Likes
20

Honestly reading through the few items you listed it seems you want a much more limiting device then what HE provides.

I don't think many people will argue that the UI could be prettier. It does come across a bit dated but has gotten better overtime. Personally I would much perfer functionality over pretty.

As far as your #2 it is hard to understand where that is coming from since you don't list it. I never see java errors on my hub, so without contexts it is impossible to know what the heck this is in reference to.

For item 3 Instructions are fairly extensive actually, When you setup the hub initially I believe there is a walkthrough to help out setup a few quick things and point you in the direction of some built in apps. Could it be better, possibly, but it is there. You can also review the documentation that provides basic directions on how to use the platform.

As far as your items around item 4 I think you are being dismissive of the amount of stuff and the type of items this platform can control. Complaining that a device is called a contact sensor and not a door sensor is short sighted. Of the 18 contact sensors I have only 4 of them are on Doors. Should I demand they be called window sensors rather then the type of sensor they really are? Perhaps the verbiage could do better at explaining itself, but that isn't exactly easy in a system that is design to support such a robust set of abilities.

Lastly item 5, Breadcrumbs are nice, but I would expect you should know what you are trying to do already. I have never gotten halfway through a rule in RM and then though what was I trying to do. Your example of a "Contact Sensor" and then not being sure of the difference between "open", "closed", or "changed" actually doesn't make much sense to me.

I am not trying to be condescending and apologize if it came across that way, but though I think there is room for improvement as nothing is ever perfect. I expect the problem is a balance between trying to make things super simple and directed vs flexible and robust. Sometimes those two things don't go well together. What other Home automations platforms have you tried.

10 Likes