45 automations, that’s a little. In this case, it’s easy to not have issues with finding automations.
Lots of us here have 100+ automations, even more.
I am switching from an other home automation system to Home Assistant, I have 169 automations on my other system. Because of the way automations are made in my other system and how they are made in Home Assistant, some of them needs to be split into 2 automations. At the end, there will be more automations in Home Assistant, without counting the new that can arrive.
While I just switched some of them, already struggling to find automations in Home Assistant, I never had the problem on the other system. All the automations there are tagged and the interface shows you groups/tags that you have to unfold to see the wanted automations.
So yes, for someone with some automations, this is not a problem. But for those of us with a lot of automations (Home Assistant is made for this after all), it is a big problem. This page needs to have something to group automations, tag them, something.
When you go on web automations websites like Zapier, automations are grouped by folders or tags. This is the way it need to be.
Until this is implemented, I recommend grouping similar automations by combining them into one and using a “Choose” action. For ex, all my automations related to the whole house fan (notifications, auto-off, auto-on, air quality) are all contained in one automation.
The only downside is if you need to mess with the automation mode, since you’ll need to keep it in parallel or queued.
I use the trigger ID and give the trigger and “choose” option a matching alias so it’s easy to see what triggers what.
its almost 2024 and home assistant still doesn’t have simple folders? just for “advanced users”. come on, how much of a essential UI thing is folders.
Come on.
Add automations to areas. That is possible, but kind of hidden.
Filter the automation list by area. Again, possible, but doesn’t work. Even when I have just one automation in “Balcony”, when I filter by Balcony area, it shows 3 of them - 1 correct and other 2 doesn’t have any area assigned.
Because they’re spending their time working on utterly useless [at this point] half baked shopping lists. I’m a long time HA user (7 years?) and subscribe to nabu casa and I have a lot of love for HA but over the years I have noticed a trend. They rush to incorporate more more more without polishing the existing stuff. I can understand why so many users in the release notes comments thread were unhappy when the tasks feature was introduced. It’s yet another feature that is being rolled out before it’s time, unfinished, with a ton of ‘workarounds’ suggested for various shortcomings when more pressing matters should be prioritized. It’s been suggested here that people don’t like working on UIs… well that didn’t stop HA hiring someone specifically to give the settings page pretty little multi color icons. How about the aged lovelace interface? Why hasn’t mushroom been incorporated? Why haven’t automation groups? Why are issues like the HA OS bluetooth debacle happening? The last time I checked I was on stable, not dev. I lost count of how many times Paulus said HA was so stable in the recent birthday broadcast… like he was trying to drive the point home… well, it’s not. That’s why a huge percentage of us virtualize it. As some user on git recently commented, it’s not a great day when the computer running your house has issues and falls over. Or when the feature request with the most votes keeps getting ignored in favor of a task manager. We have that on our phones, for now, thanks. And at some point when it’s ready it would be great to move that away from Apple/Google and have it in HA. But lets fix the stuff that needs fixing first, yeah? I really want to keep supporting Nabu Casa, I do. But IMHO some serious ‘head out of bum action’ needs to happen, and preferably before the next About Page icon is beatified or the shopping list gets even cooler emojis! Oh and I’ll add a bit about the automations showing up as errored in the recent release too: Yes, maybe they should have been entities not devices, but there should have been some sort of a) warning, b) helper script to resolve. But there were people actually saying it wasn’t an issue and the users automatons were already broken. Well, they weren’t, they became broken. Yes ok, it was due to poor automation design, but the attitude of “nah… it was already like that, must be you” sucked.
I expect to be flamed for this, but whatever. It had to be said.
Well, i kinda agree with you… too many unfinished stuff is happening lately… the proof for this is almost regularly a “correction” release a day or two after first monthly update, which suggests that new things aren’t thoroughly tested before publishing. There are people that don’t even install first monthly updates, but they wait for a second or third “correction” because of that…
I can’t get rid of the feeling that they keep ignoring many of user’s requests (as this one, which is being requested a few years now, but still unsolved).
Too much time was wasted on “year of the voice”, if you ask me. If you’re not in english speaking country (as i’m not) it’s pretty much useless, not to mention that many people doesn’t use voice anyway. I really can’t see a good reason to mute my radio, then “shout” into my tablet “close the cover” and unmute radio again, when it’s faster just to tap on the icon (or make an automation anyway)…
I totally agree about “aged lovelace interface”. Say, from the beginning i was wondering why on earth edit lovelace cards window is so narrow and can’t be expanded to full screen? Who benefits with that blurred background while we could have a decent full-screen editor? And, no yaml mode is not the solution… i like UI interface, but i also like to “tinker” a bit sometimes.
Don’t get me wrong - HA is still a great (and free) app, i can’t imagine a life without it anymore, so many thanks and congrats to all who develop and maintain it (and all it’s addons, official or non-official ones). But, listening to users wishes just a bit more would be nice…
I could say my piece about year of the voice too… I too had high hopes for it. But as you say, if you’re a non-english speaker, it’s not really a thing. Tbh, I agree with you about pushing a button. I have more buttons around the place than you can shake a stick at, because I find it easier to hit something than to stop whatever I am listening to in order to give a command. But: assuming the voice feature is built on top of the existing text input engine, I can’t even get that to work half the time.
Yeah it’s hilarious that I can get an ancient RJ45 phone from yesteryear and talk to HA… but why, when all this other stuff isn’t done?
HA is great for the most part - and that’s what makes this so disappointing. I have stuck with it through a lot of rough times, but I do find myself looking at cancelling my subscription (I really only subscribe to support HA, because I don’t use any of it’s other features like remote access or voice, Google/Alexa integration).
A friend of my is a hard core OpenHABer and he thinks I’m an idiot for putting up with HA for so long. Most my stuff uses Zigbee2MQTT, which is now well supported by OpenHAB. As soon as a couple of other things are supported, I may well give it a try.
FYI there are a few common miss conceptions in your post that should be cleared up.
Mushroom cards have been added into HA via the Tile card and chips are on the way. Every single release this year has added features to tile card. It’s not the default card, however it is something you can add yourself. If you have a full UI of Tile cards, it looks nothing like 2018 lovelace. Not to mention, each release has also improved the more info cards to match the look and feel of the tile cards.
The guy they hired to add mushroom cards into HA is doing just that. In order to support mushroom card functionality, they changed how coloring worked so that users could color every aspect of HA with a theme. I’m sorry that this upset you, but it did pave the way for additional tile card features.
Year of the voice was mostly handled by 1 Nabu Casa developer, the developer specifically hired last year to do all this work. Very little voice work was done outside that. The core team focuses on core HA features to support every aspect of home assistant. Just like the frontend team focuses on the frontend. “All hands on deck for year of the voice” is an assumption many people made this past year, and it’s not correct in the least bit.
HA core team typically does not work on new integrations like most recent to-do list. There has been a feature request and multiple WTH’s asking for multiple shopping lists for years. A volunteer decided to add this recently and he’s planning on adding more functionality next release. Please be patient and also understand that most of HA’s development is built by volunteers (roughly 90% of new integrations are built by volunteers). Features you may not care about can be added by anyone at any time. Typically Nabu Casa likes to highlight highly sought after features in release threads, which makes it appear as if they were developed by Nabu Casa.
The frontend team and backend team have been adding a long list of features in order to specifically support this Feature Request. There’s even links to the backend work (that started a year ago) in this thread. It’s not being ignored, but it’s also not an “easy add”. At this point, we are waiting on the frontend team if I remember correctly.
If you are having issues with bluetooth, please open up an issue on github. No clue what you’re even referencing with this “bluetooth debacle”. Seems a bit miss placed? Are you referring to the library switch? If yes, that was necessary to add other long awaited blue tooth features.
If you’re having stability issues, please create issues with details that lead to the instability. Personally, I’m sitting on a system I built in 2018 and have been simply updating since then. The only stability issues I’ve ran into are related to the daily (alpha) builds.
Lasly, please try to understand that I’m just providing information, I’m not here to start an argument.
Thanks for the info. Some of that I was unaware of, so thanks for taking the time to post. I am always grateful for additional information, and I have no intention of getting agro or starting an argument - I guess I’m just voicing this month’s frustration at HA, and this is where it landed.
1 and 2. I’ll look into, thanks - should totally be the default though. Even OpenHAB wins here.
The fact that it’s a common assumption, I can guess is because such emphasis was put on it. We were all hyped up - THIS IS IT! YEAH! And as you’ve just said, it’s really come to very much less than the artificial hype.
4 and 5 I was also unaware- thanks for the info
6 and 7 however, totally surprised you’ve not come across these.
I don’t have the link to hand, but the Bluetooth issue (that is now resolved) was something to do with the upstream stack used on HA OS 10. The unofficial solution was to just use the ESPhome Bluetooth proxy. Whilst not technically the fault of the HA OS team, showed lack of testing - as I understand it, pretty much all USB Bluetooth adapters failed.
As for stability issues, you must have a stash of Felix Felicis and take your daily dose, because I regularly have to roll back to a snap I took before a core update. I daren’t update anything except perhaps the file editor and samba plugin without snapping first.
And it’s not like my HW is even in the slightest bit old. I was running a Rocket lake i5 until recently, now an Alder Lake i7. Prior to the Rocket Lake it was a Coffee Lake, but seriously - this should run on any old dragon within reason. In an ideal world I’d just nuke the lot and fresh install as a last ditch troubleshoot attempt but that is… a lot of work.
It wasn’t lack of testing, it was an active decision that needed to be made. They had to stick with a library that would allow HA to expand bluetooth into a “new frontier”. This unfortunately, killed some old integrations that used bluetooth. Yes it was a tough decision there were technical limitations to having both bluetooth python libraries in use. That and I believe the old library was no longer maintained.
I’ve been using HA for 8 years now, this is hands down the most stable HA has ever been. I have to purposely find things to add/change into HA, otherwise the upgrade process is me just pressing “install”.
Well I think your first problem is that you’re associating your computer hardware specs to stability. Stability typically comes from the devices you connect to HA (and how you connect them), not necessarily the hardware HA is on. HA mashes up thousands of different python libraries to make specific devices work. Any one of those libraries could have bugs that cause stability issues. Keep in mind that these libraries are just used by HA, they are not maintained by HA.
Secondly, are you ignoring current errors in your logs?
Lastly, if you have zigbee or zwave, are you using device by id instead of the canned usb aliases? If you use the usb aliases, those are not guaranteed to find the same device every boot. That’s a linux thing by the way. Your best option to avoid that issue is to use device by id.
Other than that, I’m not sure what can be causing your issues. It might be time for you to debug into what integrations are causing instability and report them to core.
Apologies, let me clarify what I mean when I imply that some faster hardware should improve stability: I started out (like most do, I’d imagine) using a low powered device like a Pi. But the experience was just horrible. More often than not you’d reboot after an update and it would never come back again. So I moved to a more powerful Pi - same. So I moved to a more powerful AMD64 (albeit an old Turion but more powerful than the Pi) - same. It was only once I started moving to serious hardware that the daily issues stopped, but not completely - and that’s where virtualization saves the day. This is a condensed experience of many years.
With regards to integrations, I actually use very little. Mainly Z2M and Hue. I toyed with the Bluetooth for a bit (then I came across that OS issue) but have since abandoned it.
The stability I am referring to is rebooting and it never coming back again, or performing some sort of DB maintenance that causes boot loops. I also frequently come across an issue where I’ll reboot after an update and the UI won’t load, although the CLI is ready and I have to manually reboot again to get it to come ready.
Yes I am using the dev ID, this is basic stuff and the only thing I need this for is Z2M which causes me little to no headaches at all. Occasionally a cheap naff Tuya device will fall off, but that’s nothing to do with HA. I realize a lot of people don’t use the dev ID which causes problems, but that’s not the case in my setup.
The majority of my logs are warnings that the ADB service cannot access our Shield TV’s because I kill power to them when not in use (something which annoys me greatly) but as for the logs containing a load of red that would certainly explain the behavior, no, I wish it was that simple.
It’s not known to cause issues, but if you’re running adguard on HA, the HA container starts up before adguard. Meaning HA doesn’t have a DNS server because adguard isn’t running, which causes all sorts of issues. This is a downside of using adguard as an addon with HA. This would happen on openhab too btw. It’s probably best you move augard off HA or have a backup dns server that your router uses.
FYI this will only occur on reboot. Which would explain your issues.
Oh, no in which case that isn’t the issue. AdGuard is running in an LXC container on a different box. I don’t think it’s wise to tie anything internet access related to HA.
Like I said, I use very little add-ons etc. things like Let’s Encrypt, Vault Warden etc are all hosted on a different machine.
I have to agree with @petro HA has been extremely stable for me the 3+ years I’ve been running it on RPi 4. I initially started with a VM on my Synology. That worked ok, but I definitely had some stability issues there since that system is also my primary data storage for my workstation.
I find it so stable that I actually have an automation in place to pickup every update when one is dropped! Mind you, it only does the update / restart of HA itself between 3 and 5 am so that it doesn’t disrupt any of the primary usage for my family.
I’m going to have to do some digging then. I am literally running this on a HP Elite Mini Gen 12 i7 virtualized in Proxmox and nothing else…! Manual updates only after snapping.
