The current openHAB release is still 2.4 and to the best of my knowledge it still contains the MQTT 2.4 binding including “tons of bugs”. Since OH2.4 is still the current release, that is what most people will use when they upgrade.
Just like most developers are not always reading every topic on the forum you can’t expect every user to read every topic on this forum. Fact is, this quote is still on the blog this very day.
I can’t change the blog. I can’t change the fact that MQTT 2.4 was released prematurely. But I can provide answers to questions and explanations for why things are the way they are.
Huh… There you see… Lots of damn kernels with strange names… Of distributions or whatever… They´re all mumble jumble to me, no matter what name they have 
Serious - What exactly is a kernel update then? I thought is was the OS.
Does it work for items and things files as well?
Agree. However, testing new releases should be done in a test environment if your automation is critical.
Only one of those things is the kernel, Linux. The kernel by itself is not sufficient to have an operating system. It is just the interface between the hardware and the rest of the software that runs on a computer.
When you do a kernel update, you are just updating this one piece of the operating system.
The hundreds/thousands of additional libraries and programs that actually make a useful system is what makes up the operating system. All of these scores of libraries and programs are developed by hundreds of thousands of developers.
Each different version of a full Linux operating system is the Linux kernel plus a carefully selected and tested subset of the thousands of available libraries and programs. These versions are all given a name. For RPi, that babe is Raspbian.
Yes, just change the extension of files passed to cat.
I might be a tad late weighing in on this conversation, but would like to share my experience.
I was a long time user of OH, (Since v1.5 if memory serves me correct). I struggled through the learning curve, got the basics setup and things went well for a while.
By the time we got to V2.1 things started to go pear-shaped. I jumped on the Sonoff and automations train, and started running into all sorts of problems. My OH install kept corrupting, resulting in at least 10 complete rebuilds. I could not get simple timers to work for pool pumps and aquaponic units and irrigation system, presence detections etc. Rich helped me with some of the issues, but it was a struggle for me, not being a coder…
A less technical colleague could not get OH working and so moved to HA (Hassio) with the ease, so I stuck my toe in that pond and for the last 2 or so years have perfected my HA deployment with a whole lot of Sonoff (23 devices), NodeMCU (11 devices) and Broadcom IR blaster (2 devices) and automations (admittedly plagiarising others brilliant shared code). My corruption issue vanished (still using the same Raspberry pi, SD card on which OH kept breaking). Yaml was easy to learn, and the layouts of the files made sense, and I could break them up to fit my illogical way of thinking.
Tasmota and ESP home are the absolute jewels of my implementation, and with an Arduino mega, I could replace my rather dated CSC Alarm system with the standard Hassio alarm panel components. Presence detection also worked quite well.
For me it was home automation paradise… So what went wrong?
Well no fairy tales last, The constant updates with their breaking changes kept me on my toes, forever having to tinker after many updates broke things. As the honeymoon phase wore off, my system slowly began to fall apart as my interests moved elsewhere, and I stopped tinkering. The exciting Lovelace UI was released, (pretty sweet) but it was as big a shock to me as the OH V2.0 was. Another steep learning curve. The final nail in the coffin was a massive failure after the latest updates, My deployment had fallen behind by about 4 months, and I rushed the updates in. Everything broke, The addons were a mess, the Yaml files were a broken due to deprecated options. It took me a solid 4 days over Easter to get things to where I could see the UI again. Many hours of Yaml fixing still lie ahead.
And my biggest bugbear was the rather poor phone companion apps for HA. I always loved the OH app.
So I am back testing the water with OH again, having stumbled across some folks who have built some rather sweet irrigation systems. (I can use tweaked irrigation timer code and rules to automate my aquaponics, Pool pumps, garden lights, bedside alarms etc.) and I get to use the sweet phone app again. This time however, I will be running both OH and HA though. Reasons for both:
Having said the above, if ESPhome and TasmoAdmin can be integrated into openHab, I would possibly ditch HA… maybe, maybe not. 
For the DIY’er approach that I have taken, these 2 tools are heaven sent, and made the setting up of the Sonoffs, NodeMCU and Node32’s an absolute breeze.
Finally, just to be clear so as not to offend anyone, this was my experience on the familys home automation journey and not a criticism of either system. Both HA and OH are marvelous systems, each with their own benefits and drawbacks and as such, I am quite happy to run both side by side, leveraging the benefits, and mitigating the drawbacks. A big thanks to all the DEV folks and forum contributors that have made these systems what they are.
It is worth noting that those corruption issues have not been a problem since 2.2 I think. And automatic backups of the jsondb have been around for around this amount of time easing the burden of dealing with corruption issues.
I’m not saying that you made a bad decision, but since OH 2.1 a lot of attention has been paid to these sorts of issues.
YAML is the front runner for the replacement of the current text config formats.
I bet a lot of folks would love a posting describing more about what you are doing. Sounds pretty awesome!
It depends on what your original problems where but there have been whole books written about presence detection and tons of improvement on that front since then. See Generic Presence Detection for a generic approach to presence detection and management.
The MQTT 2.5 M1 binding supports the Home Assistant MQTT standard for automatic discovery so I think this part is done.
What would you want from this integration? Lots of users here use Tasmota but this is the first request/idea I’ve seen with integrating it with OH.
I doubt anyone would see this as offensive in the least. It’s your personal experience and it seems well reasoned and described. Thanks for sharing!
Thank you for the prompt reply and pointer on the improvements , (this is another thing I missed once moving to HA, Those forums were not quite as helpful and friendly as the OH forum).
I will most definitely be sharing info on my deployment.
Great to hear about the ESPHome, apologies for not doing beter rersearch before posting, I literally made the decision to return to openHAB yesterday.
As for the TasmoAdmin. I am a rather long in the tooth Data Centre manager / IT infrastructure architect and as such have a great fondness for tools that allow one to centrally manage technology devices. With the amount of tasmota devices I have, TasmoAdmin made managing these quite easy.
It was convenient to have the TasmoAdmin console part of the HA mangement console… However I must state that I had completely forgotten that TasmoAdmin is a standalone solution, whos web interface was presented in the HA gui. I just realised that I can do the same whith OH.
To a point that someone made much higher in ths discussion, HA does tend to impress with the number of components and easily deployable addons. TasmoAdmin and ESPhome are offered as addons that a non-it person can install and configure very easily. The same with NGINX, LetsEncrypt, Spotify plugin (used for my multiroom audio), Samba and SSH terminal. (These last 2 I needed because I am am a real novice when it comes to Linux.) Whilst these community addons are not part of the core system, they made the complete setup of the Raspberry Pi easy.
SIDE NOTE: The term Addon does not mean the same thing between the two systems. OH Bingings are similar to the Components in HA. Whereas the Addons in Hasio are seperate applications hosted in docker containers that make up the Hassio ecosystem.
REQUEST: and slightly off topic, I would realy appriciate it if anyone on the forum can direct me to similar 1 click deployments for NGINX, and LetsEncrypt for Raspbian.
Aren’t you aware of openHABian ? It’s all in there. You can use the image or install on top of Raspbian (well or any Debian compatible).
Please distinguish the application layer (= openHAB) with its bindings (e.g. Spotify) from the underlying OS layer and helper tools (e.g. NGINX). That bundle is openHABian then. It complements and includes the application.
Same for HA, there’s HA and there’s Hassbian (and Hass.io).
I’d be interested to hear from those who have experience with Hass and OpenHab as to how often they start down a path with a component or an addon to find out that it really doesn’t work well enough to depend on. From a self reflecting standpoint I tend to go far down the rabbit hole and then if I cannot succeed I am unlikely to try again. I’ve only encountered this once when trying to use the OH embedded MQTT broker but after having ran into that barrier and not seeing a sign that clearly said wait until X.X.X I switched to Mosquitto and will be reluctant to switch back. As a hobbyist I had that option, but after hours of struggling emotionally I cannot go back.
My choice to start with OH was based on avoiding the failed attempts. So far I think I’ve made the right choice.
OH is not dying in any way, my system is alive and kicking
looking from a heavy user perspective
i have 3 OH setups talking to each other at all times
i can say at least this from experience stop using RPI!!!
i have been using only desktop PCs, and even on HDD or SSD i never had any boot fails
i think the OS SWAP mem is the key here for a long life system but maybe i am wrong, just talking from what is seen in my own setups
right now I am trying to combine HA > Node-red > OH
i am a heavy user of OH and NR, i think we can take a lot from both worlds…
i will keep my standing Home and can add items from HA
for example:
if HA GUI is better for user i will use it…
if HA can give me some function that is not yet on OH i will be able to add
but i will be also able to use the OH free cloud …
if anyone has done such a thing please let me know … i have a bit of learning curve on HA
and last thing thanks for this great community beside OH, i think we have great people here and only for that its worth sticking around !!!
Have you seen habpanel? There is a whole thread showing off what people have created. If you want it to look a certain way you can, very flexible.
There are a few people running both connected by mqtt, however if you learn one system in depth you will probably end up with a better system, except if you need both to get hardware support that the other does not have.
yes but i am too lazy to create my own GUI…
and i was just giving an example thanks!
I tried searching for posts on the Home Assistant forum that may have been unhelpful or unfriendly to you. However, it does not contains an account named “ScottyDS”.
Did you use another account name? Alternately, can you provide examples of replies that you felt did not help you or, worse, were unfriendly to you?
The Home Assistant community forum is rapidly approaching a membership of 40000 registered members. Like the openHAB community, only a fraction of the members regularly participate in answering questions.
Over the course of six months of participation, and after having read thousands of posts and assisted dozens of users, I would estimate there are, perhaps, twenty highly-active members who provide detailed, helpful, and courteous replies.
Just like the openHAB community, there’s no guarantee a question will be answered. People who volunteer their time are free to choose which ones they answer. Obviously, clearly phrased questions, demonstrating some degree of preliminary research and effort, are the ones that get answered first.
I run three different kinds of home automation software. They work together by communicating via MQTT.
For my purposes, it was easier to achieve my UI goals with Lovelace than BasicUI or HABPanel. I’m not stating its ‘better overall’ just that it was better for my requirements.
Sounds interesting. What’s your experience with the rule engines? I’m not that impressed by the Eclipse Xtend based OH solution. (Maybe because I’m no fan of Eclipse and it’s platform either, …)
Xtend is not the only OH based solution. You can use the JSR223 Rules and code in Jython, JavaScript, or Groovy. There are a number of third party options that run outside of OH including NodeRed and HABApp. Soon there will be a usable way to code Rules using a more GUI based approach for the new users. It’s kind of there already but the PaperUI interface leaves much to be desired so I’m not sure I would recommend it for general use yet until a better UI get’s developed. The UI Rules and JSR223 rules are ultimately executed on the same rule engine.
This one looks interesting, I tried NodeRED once and do not really like this kind of “visual programming”. Will give HABApp a try, thanks Rich.
OpenHAB is not dying. The Home Assistant group makes a lot of noise and prides themselves in having a release every 2 or 3 weeks but every release breaks things. The last release they released a patch the same day.
Home Assistant needs paid users of their cloud service to support them. If you value stability stick with OpenHAB. I just moved from Home Assistant to OpenHAB. .
Bump 