openHAB vs the "The Rest" from a Newbie

I’m very, very impressed by openHAB. With the Covid-19 Lock down I’ve had time to re-look at HA solutions over the last few weeks. In that time, I’ve had a play with the following and here are my impressions (from bottom to top):

SmartThings: 2.5/5 : A cloud Controller :frowning: in a pre-rolled Appliance. I’m not a big fan of having stuff “in the cloud” but for modest price, you do get some hardware, nice looking GUI, and it is all reasonably good… but…not great. They don’t seem to know what they are doing between the new IF and “Classic”. I’m note sure the know where they are going. Also not much you can do locally for config which just quickly becomes a PITA. Verdict: On the Shelf (not worth returning).

Home Assistant: 2.5/5 : A local Controller :slight_smile: in a roll your own environment. I wanted to really love this one. It was pretty, easy to use + there are bindings for everything I’d ever want (and more)… but the Z-Wave experience is appalling and for for me it just was not that stable. It seems it is all flash and no depth. Verdict: Uninstalled.

Domoticz: 3/5 : A local Controller :slight_smile: in a roll your own environment. I’ve actually been using Domoticz for the last couple of years on a basic version for monitoring sensors are sending alerts. It has been solid and reliable but is pretty limited in the bindings available and it’s Z-Wave support is only fair. Verdict: Retired.

Fibaro Home Center 2: 3.5/5 : A local Controller :slight_smile: in a pre-rolled Appliance. I almost kept this one. It is very expensive but comes packaged with what looks like an Intel CPU/Mobo with a custom Z-WAVE daughter board. The reception and range on their Z-WAVE hardware is spectacular and flogs what the likes of the Aeotec Gen 5 USB Controller Stick can do. It is also easy to run, setup and configure. The results are nice and pretty and what it does, functions well and is reliable. The downside is it does not support that many bindings, but the killer for me was their lack of timely Z-Wave Template support. On an unkown Z-Wave devices you are prompted to send in the automatically generated info, but seems that is the end of the engagement. Verdict: Returned

openHAB 2: 4/5 : A local Controller :slight_smile: that is fully transferable between OS and HW. I’ve already had it on my main WinPC, then just copied the files to a Pi4, then to an Intel ComputeStick and it just runs. I normally bag Java but wow… this is a great benefit of the architecture! It has all the bindings I need (and then some) and it’s Z-Wave implementation and support for the templates (thanks to Chris Jackson) is just #1. Seriously the top of the heap. Did I mention, I’m very impressed? Now all I need is the Z-Wave HW & Antenna from the Fibaro :slight_smile: . It’s not all milk and honey. openHAB2 has the highest learning curve of all HA soln. Started with PaperUI in simple mode that found everything, but my frustration trying to work out the syntax to create my own basic UI resulted in some stern spousal rebuking! What got me over the learning curve (or wall) was not so much the Into Info on the website but looking at the demo site code at https://github.com/openhab/openhab-distro/tree/master/features/distro-resources/src/main/resources Verdict: Keeper

I really look forward to openHAB 3 and the promise of the unified UI. You already have all the depth, now a pretty top cover to make it easy for us newbies (if that is ever possible) and it will be a 5 / 5 product.

Thanks
Nathan

18 Likes

For me:

I have tried home assistant many times over the last 2 years, it is getting better and easier to use but, having to pay for google home assistant integration is a no for me. Smart things not tried and don’t like the cost. not tried the others.

I am very much a diy person, started with a few sonoff basics and a few light bulbs. Didn’t really like being connected to a Chinese server so tried home assistant first, struggled a bit with yaml as i hadn’t done any programming since BBC basic days. Then tried Openhab and it found my TPlionk bulbs straight away. Then a few minutes with the guide on the tasmota website and all my sonoff stuff was working. I have since spent many hours struggling with code and syntax. but everything is working and has been reliably for 2 years now I would have trouble living without it now.

3 Likes

Yes, this is not recommended and, in fact, is removed in 3.0. Yay!

5 Likes

To throw my hat into the ring…


HomeAsssitant -

I tried it and just couldn’t get my head around it.
The installation process left me confused as to what I had done, Vs. what I should have done.

Eventually I got it running, but trying to add the Velbus hardware just didn’t pan out well. (This is my benchmark, as I can’t recommend a solution that I can’t support.)

Conclusion - Abandoned


HomeCentre -

While this is a ready to go system for Velbus (by a third party), which automatically scans the network and creates a very basic UI (that can be modified with a Windows configurator app), it’s layout is limited.,

However, it is aimed at installers that want a "Plug in, power up, confirm it works, walk away) solution.
(leaving the customer to edit the layout, if they want to)

There is a 2 year buy in if you want the core software updating (which I don’t disapprove of, as the single individual that maintains this has to pay his bills one way or another)

The biggest issue I had with setup is the big price tag on day one.

Conclusion - Reserved for a select few installers


OpenRemote V2.x (not V3.0 that is in development)

I spent a lot of time and money getting Velbus hardware added to this platform.
It is by far the most advanced platform for any kind of integration.

What you can do with Velbus in OpenRemote, from a JAVA 6 setup is pretty amazing.

However…

The UI design is restricted to the EXACT screen dimension that are used in the designer.

You can create as many bespoke “screen” and groups as you wish, but there is NO way to copy elements from one screen to another, so each Phone , PC or Tablet layout has to be created from scratch.

I do really like their colour picker solution, where you just load up an image with all the colours you want and when a user clicks on an area of the image, that’s the RGB value that is sent to the Item.

For example -


or
colourfan

HOWEVER…

It takes HOURS and HOURS to get every fine detail of control built in the background, before you can start to think about your first UI.

For example.

A simple Dimmer requires 4 seperate (but linked) elements.

ANYTHING other than Java 6 and it fails

Conclusion - Reserved for a select few corporate customers with BIG budgets and very narrow needs regarding the UI

Note - V3.0 promises to be a lot easier, with incredible functionality— but the system design costs will be big. (Time or Money, depending on your skill set)


openHAB2

It just works

It’s easy to deploy on whatever hardware and OS you wish to use.
I have my favorites, but who’s to say I’m right or wrong.

Adding bindings is easy and after a bit of hard work and careful reading… most things work beyond what you would normally need in a setup.

(Read / watch some tutorial/s… if it doesn’t work as expected… read it again… slower…

If that doesn’t help you, just ask a polite and clear question on the forum)

I’ve had a few hiccups along the way, as I’m sure everyone has, but there hasn’t been a single thing that I haven’t been able to overcome (with more than a little help, for which I am always grateful).

Conclusion - I’m happy to use it, recommend it, support it, shout about it… and fund it

Remember… .
"Free Software" is only “Free” when someone else is paying for it.

and

It ain’t a cloud… It’s someone else’s computer, that still needs to be paid for.
You either pay for it in cold hard cash or in the case of Alexa / Google etc, your information has a $$$ value to them.

myopenhab.org “cloud” service is the only “Free” service I know of that doesn’t harvest your information.


If you think openHAB is amazing, please join the foundation for as much or as little as you can afford.

7 Likes

Hehe, this happens quite often to me, and still I get the things to work (most of the time) :rofl:

1 Like

Thanks for posting! It’s rare to find someone who has had the time and expertise to perform such a broad survey of the home automation market place right now. This is a very valuable post.

@MDAR, your responses too are very valuable. I’m going to bookmark this thread so I can link to it when the topic comes up.

I would be very interested to hear if anyone has experience wtih Hubatat. Watching the Wink subreddit (Reddit decided I must be very interested in Wink after they went to a subscription model) most of those users are moving over to it.

2 Likes

I could not get a Hubatat to try as the Australian model is out of stock. This is also true for the Australian Version of the Fibaro Home Center 3. So for me they both get a 0/5 Verdict: Unavailable

Anyway, my next project is to see if I can improve the Z-Wave hardware to that I was getting with the Fibaro Home Center 2 box. I’ve ordered a 902-928MHz Whip (2.4dBi) antenna and I’m going to see how modding an Australian (919.8-921.4MHz) Aeotec Gen 5 USB Stick goes. I’ll post the progress of this in another thread once the bits arrive.

1 Like

Good Point - application sent to become a sustaining member

3 Likes

Another package that is opensource and created by the Mozilla org is this one, still very new but I hear good things about it. Would love to hear someone thoughts on how that project is going compared to Openhab.

https://iot.mozilla.org/

Yes Home Assistant have just this week moved to a new Zwave ‘binding aka component’ using Openzwave project as the base. So it would probably be a nightmare to try it out in the middle of that change over and in a few months the situation may improve. Chris Jackson (and many other contributors) from what I hear from users, have done an excellent job and the Openhab binding is written from the ground up and carefully crafted to work well.

HA puts new features above stability and breaking changes. Releases don’t seem to be rated and every user has to beta tests any changes.
Openhab puts stability above new features and clearly marks a tested build as ‘stable’ giving the user the choice if they want newer sooner (daily snapshot or milestone builds) or more stable releases.

This is why I choose Openhab currently as I personally prefer the second situation. I was about to install HA to do some testing and eval but then some major changes occurred over there in the past few months and appears to be getting worse. If Openhab 3 was ready and lives up to my hopes, with good documentation, we would be seeing a mass migration from HA to Openhab right now.

Originally when I started out it was only Openhab V1 and Openremote to choose from, I quickly forgot about Open remote as it required a cloud to design the UI and it was moving away from FOSS and towards a half commercial approach just as HA has done. Openhab was free and opensource and I had the ability to keep a server and all source code fully frozen in time if I wished, so Openhab won. Openremote if the project died, the cloud died and so did all my work and that was not a risk I saw as acceptable. I am glad to see they are still around.

1 Like

What is the recommended way? Could you point me to the related article or so on?

Turn off Simple Mode. Manually create your Items either in PaperUI or in .items files.

1 Like

^^^^^

What he said. :wink:

1 Like

Not true, you don’t have to pay.

1 Like

I came from ioBroker to openHAB and now to Home Assistant.

ioBroker: 4/5
positive:

  • great community
  • lots of adapters
  • very fast development for new hardware
  • even adapters for openHAB and Home Assistant

negative:

  • UI really sucks

openHAB: 3.5/5
positive:

  • I like HABPanel
  • I like the rules-DSL (but it seems I’m the only one left) :wink:

negative:

  • development was a bit slow in general and for new devices
  • I never really got warm with the community (may be my fault too)

Home Assistant: 4.5/5
positive:

  • quick start
  • zigbee2mqtt is so easy
  • fast development (and bugfixes)
  • Lovelace UI is very cool

negative:

  • same discussions as in openHAB about deprecating textual config
1 Like

Thanks for taking the time to post your experience.
In regards to slow development, I can understand that comment, but it is simply not getting seen. Openhab V3 is having a huge amount of work done on it and only once it reaches stable status will it get seen how much work has been done. Removing Eclipse namespace and adding Java 11 support, wont perhaps be seen as new features, but at some point a project has to take the time to prepare for the future and features slow down when that is happening.

A new UI is also coming in Openhab 3 with the goal that it is easier to develop with which will speed up changes after V3 is here. It is not just a quick cosmetic change, but a big re-write to make changes easier and quicker.

Which platform you choose will probably get based around the hardware you use. In the case of Zwave, Openhab wins currently, but for ZigBee perhaps it is HASS? For other brands another platform may be better, purely because of the person doing a good job on developing the support.

I could be wrong but I believe that is not the case anymore for Openhab, it was talked about and then the feedback was listened to. The feedback in Home Assistant was not listened to and that is causing issues over there that a single person dictates the future of the project. Openhab 1, 2 and 3 will all have textual config and 1 person does not dictate the future which does slow things down as they are discussed and agreed upon before the work is carried out.

6 Likes

Hmm I get a bit worried when I see, easyness of development is dependable on a UI. Developing for systems like OH should be the same, no matter what UI beeing used. I see a UI as a interface to the actual core system. If its part of the core system, then it will have an influence on developement.
Or maybe I misunderstand something here?

The reason PaperUI has languished with few enhancements over the years and almost no new features is that it was written by an organization and donated to the Eclipse Smarthome project by a third party who has since abandoned it’s support. Furthermore, it is written using JavaScript frameworks that the maintainers who are left on openHAB do not know and have no interested in learning. The new UI, in addition to replacing a bunch of different UIs with one consolidated UI is being written in a way and using frameworks that are easier to learn and use so we don’t end up in the same situation again.

The easiness of development is for developing the UI itself. The goal is to make the UI itself easier to maintain and therefore improve the ability of the developers to fix bugs and add features to the UI.

5 Likes

I used Homeseer for a couple of years and now I’m trying OpenHab to compare. Homeseer has great community support from a lot of dedicated plugin authors. Plugins are generally easy to use, setup guides are good and great advice is easy to get. The number of plugins is amazing - one of my favs is easy triggers that allowed me to have heaters running at 3 different temps at different schedules, with an on/off/auto switch. I didn’t get very far with it because my next goal was lighting and then I met DMX, MQTT, Arduino and the system ended up powered down for various reasons while I did research.
The down sides of Homeseer for me where that the Arduino plugin was hard to use and unreliable so that made really cheap DIY hardware inaccessible (it may be fine 2 years on), but that’s when I shifted focus to MQTT which does work with Homeseer.
Also, it’s American so AU/NZ (and possibly UK to a lesser degree) products are less available. I see the zwave mouse trap is finally available for AU/NZ so all is forgiven.