Moving from OH(this is no rant)

Hello so moved to a new house , and had to make a choice
OH2 - i am long time user , and i got nothing bad to say other then OH3 is in the way…

OH3 - still in beta when i started at least… and moving to a new home , i did not want to be a beta tester and electrician at the same time… too much work around the house to even considering diving in to OH3

so my next choice was HA…
i am not here to tell you that its much better …

but there are some things that i was suprised… and i really hope OH3 will go in that diraction

  1. out of the box UI that can almost do anything in terms of design
  2. the history items/entitys
  3. graphs out of the box
  4. generally a better User experience

this is 5 mins UI work… in HABPANEL it will take me ages

bad things:

  1. not as stable as OH2, or maybe its becuse now i am runing on RPI, but OH was rock solid i can tell you that

  2. need to restart when adding new things, this one is killing me

  3. i just miss OH :frowning:

1 Like

Read this comming from a total NOOB.

I was very recently also looking at HA and was very impressed with how fast everything was up and running.

but to make a long story short ,i am now back to OH.( well didn’t actually leave)


OH3 is a huge step forward for usability
OH in general just seems more “mature” and higher quality for me.
Far better and more dokumentation
an awesome and very helpfull forum.


Fortunately OH3 has a stable release now.

With the new MainUI you can do all this. I have tried all options yet, but I know you can see a history of items and a graph, just by clicking on analyse. You can even add other items to the same graph on the spot.


Personally I am considering at least partially moving back to an HA Supervised installation.
What brought me here was

  • the much better Z-Wave binding,
  • the apparently more robust development cycle at that time
  • Both system were in flux but OH appeared to have a roadmap


  • both are apparently using a 1 month development cycle.
  • senior HA developers appear to be more responsive to the user experience ( They even made a recent choice and then partially reversed it due to unexpected user impact.)
  • HA appears to have a much simpler, better user experience

Apparently the Z-Wave here is still much better than the old integrated one or the newer Z-Eave2MQTt over at HA. I am considering keeping my Z-Wave on an OH version & using MQTT to connect.

This probably does not take longer with the OH3 Main UI.
Needless to tell OH 3 is released now, isn’t it.

1 Like

Main reason for HA for me, are all integrations it’s such a big community and tons of people contributing.
Main drawback is stability yaml and the overall architecture which I don’t really like (personal opinion).

Fancy guis is a nice thing, but not a major thing. I try to automate as much as possible and don’t actually use the gui much. I think oh3 is a step in the right direction.

1 Like

Well, in my experience, when the development cycle changed here, stability took a nosedive. For instance, it makes no sense to introduce changes in a Milestone build that have not already been tested in a snapshot build, but it was done here. One time it even broke things in a major way.

I think you should give oh3 some time to settle, and stability will be back.


I’m not familiar with the HA release cycle, so I’m interested to know if you’re comparing apples and apples here.

The 1-month OH release cycle is for Milestone releases. Stable releases are still on a 6-month cycle. How does that compare to the HA 1-month release cycle?


I’ve never tried HA but I’m with you on GUIs. I’ve spent hours and hours on habpanel but my real goal is to not use it. The smart in smarthome to me means it works by itself without me having to press buttons. I’m the only one that ever looks at the gui or uses the app. The misses just talks to Google or otherwise expects things to work.

Sounds to me like home assistant is more about putting a nice interface on everything being remote control that requires interaction. Either that or a lot of people that end up on HA are more interested in interacting with the software than just letting it work.

The HA GUI is expandable where custom menus can be added. Community involvement is easier since community addons can access the system through an API rather than being locked in to a particular programming language.

yep this was my point also… i have no problem going back and fowerd
must of my smart home is MQTT and rules in node-red…

great to know… as i said i miss OH and i am sure i will give OH3 a try when i will have the time

1 Like

I just did the opposite and moved from HA to OH3.
Stability, software design and zwave support where the main drivers.

I have to admit HA is way simpler. It autodiscovers most things and devices are just that. No things/channels/items, no model.

HA is easier but I think OH3 is more sophisticated and feels like a more serious piece of software.

My wife only uses HAs interface to turn or off some lights, but most often than not she prefers wall switches and Alexa anyway.

?? Context and details, please

ESH is dead.

Sorry, not yet public or anything usable.
I read here that Kai felt forks were OK though. I did not bookmark that post.

Breaking changes in OH3 aside (which there were pretty few given a full year has passed since the last major version), what’s your point ? You think it’s unstable ?

Normally in my experience with software development,

  1. snapshots are used for limited testing of all units of a system.
  2. Milestones are used for wider testing of features after they have proven stable in snapshots
  3. Stable release contains features proven stable in the Milestones.
  4. Stable release does not happen until the features are proven stable, not necessarily on a predefined scheduled date.

That is not the current OH process, in my view.

Something else, which I practice as a developer () and so does HA), is since they are the most knowledgeable on the features they develop, they write at least the first version of the documentation on how they expect their creations to be used. That was part of my initial estimate that 3.0 would not be ready until the end of 2021.

If it’s not public, why did you disclose its existence?

1 Like

Sorry deleted that post. Bad judgement at times :frowning:

1 Like