A Humble Plea: Announcing the Home Assistant - Amber Edition

Social
1

Dear Nabu Casa, the Core Dev Team, and the entire universe of contributors,

First, let me say thank you. Thank you for creating this magnificent, sprawling, endlessly powerful, and occasionally baffling platform that has turned my home into a semi-sentient being that mostly listens to me. My lights turn on, my thermostat is smart-ish, and my coffee pot knows my soul. It’s a miracle of open-source software.

And it is with this deep and abiding love that I ask you, with all the sincerity I can muster:

Please. Stop. Changing. Things.

I propose we take the current release, let’s call it version 2025.7.2 (a truly vintage year), and declare it “feature complete.” We will encase it in digital amber, preserving it for all time. This will henceforth be known as the Home Assistant - Amber Edition. The one true HA.

Is it perfect? No. The new dashboard control I wanted isn’t there. My Bluetooth proxy sometimes decides to take an unscheduled siesta. The settings menu is a labyrinth that would make Daedalus weep.

But you know what? It works. My automations have achieved a state of delicate, precarious equilibrium. I have spent countless hours, gallons of coffee, and a significant portion of my spousal goodwill to get to this point. My configuration.yaml is a teetering Jenga tower of genius and desperation. I have finally memorized which obscure submenu hides the one setting I need to toggle once a year.

And then the .8 release drops.

The release notes read like a thrilling horror novel. “Breaking Change” is a phrase that now causes a Pavlovian stress response. I see a UI element has been “streamlined,” and I know my muscle memory is about to be betrayed. An integration has been “refactored to a new config flow,” and I feel a phantom pain in the YAML file I so lovingly crafted.

So, let’s just stop. Let’s agree that this is it. This is Peak Home Assistant.

The tenets of the Home Assistant - Amber Edition:

  • No New Features. Do we really need an integration for that new smart toaster with blockchain-enabled crumb analysis? No. We have enough integrations.
  • No UI Changes. That button you hate? You will learn to love it. It is now a permanent feature of your landscape, like a mountain or a troublesome mole.
  • Bug Fixes Only. These must be applied with the stealth of a ninja, leaving no visible trace. If a bug fix changes the alignment of a single pixel, it will be considered a breaking change.
  • A Moratorium on “Improvements.” What one person calls an improvement, another calls “the reason my wife is asking me why the bathroom lights are now a strobe disco.”

Think of the peace. The tranquility. No more frantic pre-update backups. No more reading 3,000-word blog posts to understand why your Zigbee network has suddenly developed a personality.

Let’s put the “Home” back in Home Assistant—a place of comfort, stability, and predictability.

Yours in stable, unchanging, and beautifully stagnant automation,

A Concerned (and Tired) Automator

(P.S. On a completely unrelated note, I just bought the new “AquaGlo Smart Fish Tank with AI-Powered Algae Monitor,” and it’s not supported yet. The API looks super simple. If a developer could just squeeze that one little integration in before the Great Freeze, that would be amazing. My fish would really appreciate it.

Also, my wife has declared the current Areas dashboard “an unusable catastrophe,” and I happened to see a preview of the 2025.8 release where it’s completely redesigned and looks… well, perfect.

So, my final proposal: we add the AquaGlo integration, implement the new Areas dashboard from 2025.8, and then we freeze it. For good this time. That’s all I ask. Promise.)

4 Likes
2

Stop upgrading.

Well, not yet. But when 2025.8 hits which you seem to think you will really like. Then stop upgrading. Well, unless you see a future update that you really like. Then, really, stop upgrading.

Lather, rinse, repeat!

3 Likes
3

This is Apple and Google fault.
At some point people began to think that upgrading is good. NO. It is not.

Upgrades are purposeful.
New feature you like. Upgrade
Security flaw fix. Upgrade
Like life on the edge. Upgrade

But if you have something that works and honestly even through some non critical security flaw. Why upgrade? You don’t need to. You don’t have to. HA was created to allow you freedom. Freedom to do what you want. Take that freedom and choice you’ve been given and DO NOT UPGRADE.

Change is not always good and you have the choice not to change. And that is OK.

But LTS version is kinda meaningless and no more than a name. I guess if there were critical flaws maybe but your asking a lot as the addon devs would need to support this which would immediately cause bigger problems.

4

[chuckle] Name a corporation, or an application, or an app. :slight_smile:

5

I like to think there’s a happy medium here.

Yes, we need to update. Things change. New things come out. Old things fade away.

What we don’t need is Change for Change’s Sake.

The UI is a great example. The moment you hire a UI designer, they just have to start re-arranging the furniture. It’s what they do. It’s what gets you noticed in the designer community. It’s what gets you job offers. It’s survival.

Here’s a little secret: Most users hate that. Oh, sure, a new splashy interface gets people to say “Wow!” and maybe look at the product a bit closer. Journalists, YouTubers and other “influencers” love to have something shiny and new to talk about.

But the rank-and-file user knows that just moving things around and changing the “look” isn’t helpful. It’s just another frustration which keeps us from enjoying what we already have and know how to use.

This is not a dig at HA, or anyone in particular. It’s this way everywhere in the IT field.

I ask you this: When was the last time you looked forward to a new version or a new UI from any of the big-name tech companies?

2 Likes
6

People could have said that a long time ago… then here we are with a better version. How can you claim that today or next month is ‘peak HA’?

If you scared of breaking changes, don’t update. If there is nothing else you need HA to do which it can’t currently but may in the future, don’t update. If you don’t want to ‘play’ with HA and it’s working fine for you, don’t update.

You are in control of your HA install. Don’t try to change what the rest of us appreciate because of your own feelings.

3 Likes
7

Hello @michaelwacey ,

OK… Done.

You can stop looking at the HA repo, it will not update again.
(unless you look at it)

1 Like
8

I’m with the upgrades

I just don’t understand why people complain about HA upgrade cycle when they can simply not do it.

Without monthly upgrades I’d be bored

1 Like
9

Picks up my LTS sandwich board and a popcorn bucket…

This again? :rofl:

5 Likes
10

I feel your pain. My rant below. But of course, I will still use and try to add capability to home assistant in the future!

It is the little things that get annoying. Developers (amazing as they are) feel they need to rename a unit constant for some esoteric reason and every custom integration breaks. Just leave the old constant. It was fine for years, does it really need to be renamed? I had some custom integrations crash due to a change that was not even documented and when you post an github issue, you get zero response e.g. https://github.com/home-assistant/core/issues/107004 Requiring you to be freaking Sherlock Holmes to find a fix deep in the bowels of the github source code from some other person that happened to fix a similar issue.

I’d love to get some of these custom integrations into core so other developers can maintain with skills beyond mine, but it is so much work. Code needs to be perfect to be accepted. Some code needs to be in packaged into external python pip packages for an unknown reason, instead of just in the integration source code.

Another example, badges were made totally ugly a year ago. At least make it not require a Ph.D in python to get the old ones back.

I guess devs want tight control to keep it code base from spiraling out of control, but…

1 Like
11

Nailed it, I almost lost my mind when I couldn’t find the integration “options” select from the integration 3 dot menu. Then I started clicking buttons and found out that now, a cog wheel is the options button.

You know the cog wheel config icon that’s been for settings/configuration forever? Yeah. Apparently they decided to make it for options and leave configure in the dropdown. Wtf?

12

The big problem with not updating every single month is that eventually there will be a reason you have to update - a security vulnerability that needs fixed, a update to an external API that renders a necessary device/service inoperable in HA, etc, etc.

then you’re screwed.

because then you need to go all the way back when you decided that updating HA wasn’t worth the hassle and struggle to deal with breaking changes and start updating one major version at a time to make sure that if something is a breaking change you can figure it out then.

it’s just easier in the long run to be forced to update every month a bit at a time than to do a years worth of updates all at once.

I’ll bet good money there isn’t anyone that can honestly say that they will never need to update HA ever again.

3 Likes
13

Every two months is also quite fine.
It’s still recent enough that you will fairly easily find relevant information if there was something breaking.
If you just look through the braking changes, if they are good then two month update is usually ok.
I have done it several times.

14

You hold the power for your systems! You can stop updating, backing up,… and live a long and happy life :raised_hands:

But may also think about that others are not that simple satisfied and might prefer to get continuously monthly improvments. :star:

15

At this point I feel like the OP should just delete this thread before getting burnt further…

2 Likes
16

Personally, I think it’s a good discussion to have.

This is a real pain point for users, and one which could have negative impact on the whole project. It needs to be talked about. And, especially, it needs to be a consideration when contemplating changes.

As a developer, I was hyper aware that the things I did off-hand could cause serious frustration and reflect badly not only on the product, but on me personally.

That’s a good thing. Living in a bubble with other developers, never being exposed to how the product is being talked about in the “real world” isn’t healthy. It can foster an “us-vs-them” mentality which would be a sure sign of a failing project.

So, yeah, complaints should be aired. And developers should fear them. Likewise, users need to take the time to read, listen and understand why things are being done a certain way, too.

Me? I just update monthly. It’s a lot of work to read all the update notes and follow all the problems others have reported here. But HA is worth it.

2 Likes
17

I think this describes almost everyone who has been using Home Assistant for more than a year or two.

And, that’s where we part ways.

Home Assistant is a DIY hobby and will never (I hope) become an off-the-shelf appliance with a closed ecology. The number of available integrations grows exponentially as the user base grows. But manufacturers disappear, cloud services vaporize and your formerly working device becomes junk. (The main reason that NO CLOUD is a primary consideration for most of us).

And as new devices are introduced to the IOT world, someone, somewhere will find it compelling enough to create an integration for it.

Breaking changes are endemic to our world. Not just Home Assistant, but Node Red, Arduino, Apple stuff; all have them. (I really can’t speak for Apple because except for my wife’s iPhone, you won’t find an Apple logo anywhere here).

I would like for Home Assistant to break the 1-million installations milestone. I would like for device manufacturers to see value in the “Works with Home Assistant” program so introducing new products with the “works with” badge will be a priority. (Tip- Analytics should be OPT-OUT, not OPT-IN).

Dashboards- I have been using Home Assistant for more than 5-years. My wife loves it. But she has never seen a dashboard. Her primary involvement with Home Assistant is through Amazon Alexa. She peaks and Alexa (mostly) obeys. A few days ago there was a few hours where all Alexa Skillls were offline and my wife was thoroughly frustrated that Alexa wasn’t doing her bidding.

So, no. don’t freeze development. You can freeze your system at any point in time that you wish. But if any of your devices are cloud-connected, they won’t be frozen. They will eventually be incompatible with your frozen installation.

1 Like
18

You expect such default settings from Google, Apple, Microsoft, and other companies that don’t respect your privacy, choice but rather be as intransparent as possible. :put_litter_in_its_place:

Why would somebody want HA/Open Home Foundation to steer to such consumer unfriendly vias like opt out (essentially leech data the second the program launches and just being able to stop data transfers after) :question:

1 Like
19

I agree with most of the posts here and certainly do not want development to be frozen. I believe the OP doesn’t want development frozen as he notes he is looking forward to the ‘Areas’ and wants a new integration.

I believe the biggest problem with updates is the fact that some of us have systems using lots of addons and hence it is impossible for the developers to ensure nothing breaks. Therefore, I think we need to separate the user base into hobbyists and regular users.

I believe the ‘Works with Home Assistant’ badge should only be allowed on devices that will work with an out of the box installation - any add-ons, integrations or other bits that are required should be part of the official repositories.

I would then suggest they have a separate branch for the ‘Works with Home Assistant’ release which is basically the last release of the previous month. Other branches still continue to release the way they do, so for many nothing will change and in effect we become testers for the next ‘Works with HA’ release.

I would hope this allows the user base to increase a bit more rapidly and I would also hope this encourages more third-parties to look at the ‘Works with Home Assistant’ program as well as pushing to get more included in the official repositories.

20

Google and the other online services sell search data, usually in aggregate, because they see a profit opportunity. But you agree to it in yout terms of service.

All that Home Assistant is doing is to count their user base. It makes it easier to go to a device manufacturer and say: “we have a half-million verified installations. Don’t you want a piece of that market?” They also track how many of which integrations are used. This would aid in deciding when or if to depreciate an integration for lack of use. How can this be an “invasion of privacy”?

1 Like