I was thinking about move to Home Assistant, but after I installed, setup a few integration and started to use, I have some doubts. I use OH for a few years, and I’m using 4.1.1 now, everything in text files.
Why move:
Mobile App: Seems like there is a delay when I open the OH ios app, it doesn’t show the updated status, I have to wait ~10s. My android tablet stops my habpanel when I wake it up, and I have to restart the app. I tried openhabviewer, I cannot remember why I removed that app, but something didn’t work. I tried main UI on tablet, but it didn’t like my letsencrypt cert. Maybe it is user error, I need some time to find out what is happening there.
Integrations: I have a lot of items, rules, habpanel, main UI, etc. But I have to use shell scripts, python packages to integrate everything. HA has these integrations, and so much more. This is my main reason.
HA community seems bigger, but to be honest I could solve everything with OH community.
After all these years I have some messed up rule, not needed items. I really like to start from clean, but I can do that on OH as well
Why not:
I thought HA supports text-based config. It is, but not really. I have a config yaml, an I can split it to multiple files, but as I read HA pushes you to use UI and some integration doesn’t have text-based config. Seems like they won’t support this at all. I hope OH will support.
Items history not as straight forward for me in the database, but maybe it is just different.
I cannot find any scripted or step-by-step solution to move my OH history of items to HA. Probably, I can create an sql script to copy these lines but it doesn’t seem easy.
At first I think I can create more flexible and custom dashboards with habpanel or main UI.
What do you think? Let me know if I’m wrong in any item of these pro, con lists, or you can add more.
I’m pretty happy with basicui on android and computer(web). Although the ios app is lagging behind the android, I heard it’s usable for basicui. I find mainui too hard and time consuming for me, but I do have one mainui interface (a photo frame).
I’d agree here, HA does seem to have a lot more integrations. We need more people to contribute to OH, and as discussed before, that means OH needs to attract more new users. The current pool of users have already done what they can so the progress is somewhat plateauing when it comes to “new integrations”.
Have you configured the url with a static ip in the app? Maybe that will speed up the startup?
Tbh I don’t see an issue that way have you own integrations. If it works it works.
If you want something new you can always try HABApp and write some of your integrations and rules in python.
Maybe that’s the fresh breeze you need to fall in love with openHAB again
I use main ui on iOS, It looks great, but it needs some time to show the fresh data.
I use my local fqdn. I can only reach my OH in local network (or vpn).
Not all of them works properly. For example the python package for Midea AC needs 3-5sec to send my command (and sometimes it cannot), or I send http API requests to my reolink cameras, which is not working all the time. (I should try OH4 ip camera binding with reolink).
Do you have specific requirements for using iOS app? If it is just because of a cleaner GUI style, you can use MainUI in a browser as a PWA and style the page so that it appears as an app.
Are you seeking confirmation that openHAB will continue to support text-based config?
I’m just another openhab user here, but I am pretty sure openHAB will continue to support text-based config. It is also still maintained, fixed, and continue to be improved by those who use text-based config.
There is no reason to drop text-based config. Many automated provisioning systems also rely on text-based config, so the chances of it being discontinued is zero.
In one hand, yes and I think they will as well. On the other hand I more interested what I felt about HA. Is it true they “force” users to use UI not text based configs?
That has been reported by others on this forum by others but we are hardly experts on HA’s intentions and we should not spread second hand information as if it were first hand info.
I’ll second that is probably a question better for their forum than ours.
I would suggest not using the word “force” if/when you ask on the HA forum.
Since this is going to keep coming up, I spun up HA on a Dietpi server to get a better sense of it. I was impressed by how it immediately identified devices on my WiFi network, including:
CUPS printer
octoPrint server
Google Nest/Cast devices
Samsung TV
Router/modem provided by my ISP
It strikes me that they’re trying very hard to present HA like a consumer-focused web appliance, and that’s a good thing for attracting new users. It’s much closer to the simple Alexa/Google/HomeKit experience that consumers have become accustomed to over the past decade, so it’s not as much of a leap as OH is. I imagine that some HA users never feel the need to touch the Linux command line or manipulate files directly.
I would expect that power users can still get into text-based configuration if they want to do so, but that’s just a guess. I don’t plan on going that deep, as I’m more interesting in evaluating the “average consumer” experience.
It sees my printer and Google TV/Chromecast devices. The rest I’ve already installed and use. Work is ongoing to improve that first run experience in OH.
Cool. I’m planning to build a fresh system when openHABian shifts to Bookworm. I’m also keen to try the Network UPS Tools option that’s now available in openHABian.
Yes try it as I and some testers just went through and made some improvements which will come out in the next milestone for Reolink. Jar is already available if you want to run it on any v4 core.
I can confirm this. Overnight a private company (NCasa) made the direction change decision to just stop textual config for all new style integrations. This should not be confused with textual definition of rules/automations, its just the setting up of their form of binding, thing and items(?). It annoyed a lot of long time users and it happened probably two years ago when we were releasing V3 and killing our V1 bindings.
All newer integrations are like our v2 bindings and in HA they can only do UI config. Their form of our V1 bindings cannot use the UI for config, and only work with textual config. This was the same with our V1 bindings, they could not use the UI. So in HA it’s not consistent and a huge pain to be forced to learn two ways and know which one needs to be used.
There’s also talk that their V1 bindings/integrations are not getting upgrades unless they are fully rewritten in v2 style. What the future holds over there is not clear what will happen to the older integrations, it makes sense to get rid of them all, or they appear to be leaving them broken to force people over to the UI only newer integrations instead of making an announcement that will anger users. The result will still be the same. This direction is made by a company and not a group of volunteers that discuss the future.
When these changes went down, they then introduced tracking and data that gets sent home so they could work out how to kill off old code with the least amount of backlash based on statistics of how many users are doing what.
Our projects are similar so we both face the same growing pains, like getting rid of V1 bindings, but we are making different choices and run in a different way.
Yes I agree, and it’s why I am so certain that their old V1 bindings are going to get the axe and very soon, and it may not be a direct shutoff date. Cutting off any improvements and fixes so people have to move, until the tracking shows the use is low enough to remove the code. They are working towards targeting users that want to pay for a cloud service, because it is the easier way to get remote access.
Its great we finally have this after HA has had it for 3+ years. But we are still not close to what they do to ease new setups. Without any steps needed, you can go to a ready setup widget for your chromecast, type in some text and have text spoken out your speakers. Auto thing accepting, auto widget adding to main UI and setup of text to speach and tying it all nicely into a ready made solution. They have done a good job of it, just for everything I like, there are two things I do not like and its only a matter of time and setting up to gain the same stuff here.
Thats all I do, I test and compare how cameras and other bindings I have written work like in their platform to ensure it is better over here.
Textual setup of a binding = NO can not be done for all the modern good ones, it was killed just like we killed V1 bindings to keep it fair.
Textual rules = yes
Textual setup of the UI = yes
There is a bluetooth based integration for my Astral Pool Chlorinator, the energy usage stuff is pretty good once you plugin your solar meter, Big Ass Fans integration is working with the new firmware, Tuya Local allows direct connection to tuya devices, there is also a tight integration with Frigate. I also like the HA Phone App, as it can track my location and also works with Android Auto so easy access to open the garage door, etc.
