No I’m using one I designed.
This is good news. Especially since the many references to we only officially support the primary-color and accent-color properties and your own text about that in Entity state color changes
How will you merge those 2 conflicting stances?
HA Frontend team might not ‘want to remove that’ but you are making it terribly difficult for even the most savvy tweakers amongst us. Let alone the beginners with their out of the box experience.
Time and time again you state there is no need for documenting changes, or even the used variables, because they are not supported in the first place.
Same goes for breaking changes on these variables and their use: No need to inform the users, because its not supported.
(you= not personal, but ‘the HA dev team’)
5 Likes
What I really like about HomeAssistant - compared to other products on the “market” is its flexibility.
I’ve played around with several other software - and, they are:
- Hard to setup
- Have a very old / limited UI compared to what HA does provide
While they also have components for at least most of the stuff I do use - the choice was HomeAssistant, because it does provide a great flexibility for the UI - and yes, a good UI is a major topic when it comes to controll your devices, and also monitoring their status.
And I fully agree:
HA should not stop with innovative technologies - and either HA should not stop with adding cool and smart stuff.
But the major focus should be its functionallity - there are a lot of Ideas and Feature Requests, where I think: “Yes, that would really fit my needs”… And I also understand that there are a lot of people who are using third party stuff like “mushroom cards”, and those would benefit from a “UI redesign” - if HA will be able to provide the same “feeling” out of the box.
BUT - maybe I need to change the wording here a bit:
Nabu Casa should not become a company, that looses its focus.
To be honest: Without custom components and alternative integrations compared to the core integrations, HA would nearly be useless for me.
Why?
Because I do integrate as much as possible.
HA is more than a “smarthome” management.
It does manage my Energy Consumption, it does provide me Information from my Solar Production, It does provide me information from my Car… It is more a “management System” - which I cannot rebuild in products like Apple HomeKit, GoogleHome - or others.
HA does also provide a “Gateway” for Smart Home Controll - like automations, integration to Alexa for Voice controll … but that can all be done without any UI… that’s the technical aspect.
So… What I wanted to say:
The only usecase for me - for having a UI with HomeAssistant: is getting Information.
Which means - a lot of Entities that just provide information.
Examples?
EV-Charging information:
Changing the Charging type, Starting or stopping charging is something, I can manage completely without a UI… but getting all the information - does not work with a UI…
Solar Production:
My Solar Installation does not really require ANY HomeAutomation (might be different if you only have a small installation and want to switch on some power plugs when your production reaches a specific value)…
EV Car Information:
Again - most value is information, even if the Cloud does not provide all information atm (known issue, unfortunately)…
Room based:
Kitchen:
Oil Consumtion:
So, sure - my dashboards could see a bit more love … but at the moment, my focus is more on “function before design” - and I don’t even see how I could get the same informative value with the new UI approaches.
My major concern at this point:
That NabuCasa / HomeAssistant will loose Monitoring Options in the long term.
Paulus already pointed out, that - with the new design approach, Cards like the entity card or entities list won’t work - because they don’t meet the design guidelines etc.
THAT’s MY MAJOR CONCERN - for me personally.
In addition to that:
Right now, HA does provide great features compared to other competitors.
But if the focus will be the UI - it will become replaceable.
New users will probably stick to the Software they already have on their Smartphones, because with Matter, there is no need to use something like HomeAssistant anymore.
HomeAssitant was the choice when it came to “linking different plattforms together”… despite of the other benefits of being: “local first”…
That’s something, younger generations haven’t in focus as much as we might do - and it can be observed everywhere.
The last one to this sentence:
“Feel free to share your concepts or PR’s.”
While I am working in IT - HomeAssistant is still a “hobby”.
I already spend hours and probably to much time on the Forum, in submitting tickets (which then sometimes aging without any further comment) - and I always try to contribute with improvements to custom components and other stuff - where I can.
BUT: While I am not a developer, I cannot just “create PRs”… (I will - if it’s something I can provide)… Nor am I a Designer.
And I have plenty other things - as well as a regular job - where I need to spend my time on…rather than providing design concepts.
“It’s open source - wait and see what happens, or provide a PR” is something, that can be heard nearly in EVERY open Source project.
On the other hand - Open Source projects want to address newcommers… that doesn’t fit together.
I’ve provided my concerns about the design aproach - that’s already a contribution 
5 Likes
a future HA? hopefully it’s simply so called “headless” with the ability to use addons for visualisation.
Personally my vision of a Home Assistant would be that once set up and validated it becomes as much invisible as possible. So that others do believe in pure magic. No fancy touchdisplays, no pestering with information on my mobile and such like. Simplicity/Purity … nothing geeky/wiggeling/animated.
It should do what it’s name suggests … assist me. What lots others ask for seems to be something I’d name a Home Visualizer. 
3 Likes
But you can do that with HA right now. Nothing forces you to use a UI, outside of configuration and debugging. Once everything is set up, you don’t have to use to UI anymore if you don’t want to.
As far as I see it, there are two different (and sometimes competing) ways to approach a smart home. For some it is the way you described. Fully automated, as few user interactions as possible, no UI is fine, the house does all kind of stuff by itself like magic. For others, like me, it’s less about automation and more about monitoring and keeping situational awareness of what my home does, what happens on my property and alerting me about anomalies or anything that needs human attention. For this scenario a user interface is vital, from smart light switches to tablets and anything inbetween.
Oh and then there is the third group who likes to talk to their house like Tony Stark… But let’s not talk about those, they’re just weird 
3 Likes
Now that I have discovered how to change the tile colour depending on numeric state with card mod (see the climate card below) this is really starting to grow on me.
I still need to investigate conditional cards (or custom state switch) for some dynamically visible info and to set up the other sensor cards with state driven colour but I can see this being put into production - for my mobile dashboard at least.
Are features being considered for media players and the alarm tile?
Also select/input_select features?
1 Like
Please don’t destroy HA with concepts and designs stolen from Homekit/Home apps. If you are going this way, please state an original design identity for HA.
Who wants automatic dashboards has already Dwains for that.
Who wants Homekit, has already tons of custom cards and layouts.
Who wants Google Home, has already UI minimalist and Mushroom.
What lovelace is missing is an officially supported versatile card like button-card, with templating and javascript customization. And a better layout engine than masonry which looks nice on mobile only.
Rooms are nice, but - as previous commenters said - Ilike to organize my dashboards by functionality. An overview of the heating system is better than wasting screen space with room names.
5 Likes
I started like this even before I switched to HA. Influx DB, Grafana, as much information visualised until I asked myself about the difference with brokers at the stock exchange sitting in front of 6-8 displays trying to follower next to anything changing right now.
I then dropped that and now doing it the other way round. Everything which does work could do this without ever asking for my attention, since as long as it works.fine. My car doesn’t inform me each 5mins about the tire pressuredrop, and I don’t want get informed while it’s ok. In case of leakage it should signal it, since then it needs my attention.
But yes, granted 10 users, 10 differnt use cases, for sure.
As long as HA doesn’t end up similar to a previously well known software title from Redmond, which nowadays if ever gets a differnt look and user interface once each 6 months … everything fine.
And for the third group you mentioned … I recently also heard about humans mad enough talking to/with machines. Some are not above anything …
1 Like
It’s not about watching things 24/7 yourself, that’s what HA is for
It’s about notifying us if something is ‘wrong’ (ie. outside of a certain range or a set of parameters) so we can take action before things get worse. The freezer temperature is above a critical point, yet there is no power outage ? Notification that someone probably didn’t close the lid tight. Someone triggered the motion detection in our yard while we’re not home ? Notification with camera footage. Solar production for the panel on one of our outbuildings dropped while the illuminance sensor says it’s a sunny day ? Notification that leaves probably fell on the solar panel again and someone needs to get on the roof with a broom before the battery goes dead. One of the solar well pumps out on one of horse pastures is getting power but the flow rate is zero ? Notification that the filter is probably clogged and… you guessed it… someone needs to get out and clean it. Basically HA orders us around to do chores we’d rather forget about
Add to that things like remotely checking on things when we’re out, on vacation, at work, etc. That’s the way a smart home should be in my opinion. Alerting you of things that are out of the ordinary and giving you peace of mind when you’re not home. Not some fancy motion light thing or your TV turning on some RGB strips while you watch Netflix. But yeah, to each their own
2 Likes
@justone and @HeyImAlex, I’ve got nothing to add. You both said it better than I could.
I’ll never say anything bad about someone who wants to “play” with HA in all kinds of interesting ways. But for me, the real justification is monitoring and alerting on critical systems. The rest is just for fun, or at best, a convenience.
yeah, but that fancy motion light thing makes it so I don’t have to use my arms when traversing my house. ultimate laziness.
1 Like
lol, hard to argue against that
That’s REALLY dis-heartening to read.
Almost my whole UI is based on entities cards. I REALLY hope those don’t go away. That would be horrible.
you can (and should) have both. It provides the best of both worlds.
When worst comes to worse, they’re really easy to reprogram as a custom frontend card. I’m sure someone would do it.
When seasoned users start being forced to move to custom stuff to maintain functionality then it seems obvious that HA is going down the wrong path.
At least to me.
2 Likes
IF they just introduce a new “default” Dashboard for newcommers, that’s totally fine.
IF they replace the existing elements - and require you to use custom components (whats not officially supported) - I agree with the above statement.
1 Like
So true … but i think you see it as a deliberate marketing decision; whereas I see it as being narrow-minded instead of targeting anyone.
All FOSS projects start as one developer making his fancy code available to anyone who may be interested. It is written by developers for other developers. They are driven by the technical challenge of adding the next cool feature. Not interested in money or marketing. Coders have valuable skills - but they do not include writing documentation and user support. It is simply the nature of the beast.
However some flagship FOSS projects such as GIMP and Firefox have become successful in the commercial environment. They are still open - but they have user-friendly documentation, and good support.
The transition comes when the original developers relax their control of the project enough to encourage non-coders to contribute their skills - and the project management takes a broader view which is invariably more user-oriented.
As for Firefox, how did they fail their users ? Or was it that Micro$oft and Google put resource into marketing or creating new features ?
1 Like
Except that it is not true at all.
If it were the case then there would not be such a massive push to develop the GUI editors to open the project up to a wider non-technical audience.
Are we there yet?
No, but it is the direction the project is heading.
2 Likes
Deliberate ? Not really, it’s a side effect of any complex open source software. What I argue for is embracing this, and yes, making it deliberate. Defining your target audience and specifically develop a tailor made product for them. Right now, there’s a heavy push to ‘bring it to the masses’, so to speak. In my opinion this is a huge mistake and I am concerned that it will run HA into a wall. A bit like like Firefox, in fact. I don’t want this to happen.
It’s spelt Microsoft.
@matthiasdebaat - Before HA I spent 8 years with SmartThings and they too adopted this design style (or very similar). It drove me literally up the wall because it became my only option and limited the number of devices I could see (at a glance / per screen) dramatically. When I moved to HA about 1.5yrs ago I loved the ability to fully customize my dashboards and the ability to show a large number of devices/entities in a small space.
While my “design” is definitely not the most attractive, it is highly functional for me. This is an example room:
(on mobile, the 3 columns are stacked so yeah, there is a bit of scrolling - please add HTML style anchors to enable jumping to specific parts of the page!)
Having this amount of information and controls using the new UX would be incredibly aggravating. I am therefore happy to hear that the new UX will NOT replace or remove our ability to build custom UXs as it is fundamental to a lot of users.
IMO, Smartthings made the big mistake to standardize on their average user with 15 or so devices… please do not ever do the same. The number of devices that homes have will only grow so the need to be able to show information and controls with increased density will also grow. I currently have about 50 views spread across 4 dashboards…
The new UX is likely great to add new users coming from other ecosystems as everyone seems to have the same design, but I don’t see myself ever adopting it as long as I have an alternative. I quit ST due to their limitations on device count (~200 at the time) and their unusable “Material Design” app. It was so bad I did anything I could to avoid ever needing to open it, including having more than a dozen Echos around the house.
1 Like