I mean every single step takes 100 steps to accomplish. If you are a bit adventurous in what you want to achieve, you are going to fall in the pitfall of tabs and space and errors of indentations and this thing will just keep saying, this key or that indentation is wrong while you just pasted some code from the next LLM into your poor pathetic attempt into become the next professor of Home Assistant.
No really, I have never felt this two sided. It’s awesome to be able to achieve this much with this great program, but it really seems it is mostly aimed at rocket scientist programmers and their ‘logic’. You guys and girls do understand that there are many people out there that just want to get things done? Without following endless tutorials and then just feel stuck because the Advanced SSH and Terminal thinks they should not be allowed to SSH into their own system, and god forbid when you try to change the config, it will just revert back to it’s original state after saving and restarting, without any sort of clue! I have many stories and I feel I’m not alone. It feels like everything is changed, but nothing really is. It seems that HA keeps focusing on the technical part instead of the solution part.
Just a simple example, I tried using my PS3 Eye cam as microphone and Smart/Bluetooth speaker as a speaker, I needed a study to in the end find this Motion-eye stuff (what’s in a name) with a non con-formative design of a pull-down menu to be the place to find my microphone back. I expected a Skype kind of interface with a selection of possible microphones and bars that indicate the sound pickup, but nothing like that ever occurred. Then let’s not even get into getting my speaker setup, since I’m still stuck and I think many with me.
What is next, I ask myself, will it just be a couple of years until I can just tell my “Hey HomeAss, rewrite my whole system!”, because I want it to do this and that and ohh… especially that too, and do that while laughing out loud at indentation errors that are not resolved by the very system that sees what is wrong, but won’t tell me why…because yeah, if only you were a rocker scientist!
Likely the reason you are looking at Home Assistant and not one of the many alternatives (Alexa, Google Home, IFTTT, Homey, etc) is because it offers more compatibility, customization, and features. While the developers work very hard to make it intuitive and simple, doing harder things will always be harder.
In other words, if you want something that is not basic, it is reasonable to assume that you will have to learn more. Do you want to learn how to iron a shirt? Set aside 5 minutes of your time. Do you want to learn how to make a shirt? Set a side a year of your life. If you want to be able to do something harder, you will have to work harder.
That being said, technology will get better, innovations will come about, documentation will improve, and things will generally get easier. So, if you have specific examples of ways to better things, please mention them. A general statement of, “Why is it so hard to do complicated stuff?”, is not going to go very far.
Ohh that’s why I sad it’s a two sided feeling. It’s a bit of a tongue in cheek story, anyway… it will probably be not long before we have an AI Home Assistant wizard, that can do stuff for us… that’s more what I was aiming for…
Even though many people on here disagree strongly, I’m with you: The HA project needs to take a break from pushing fancy new techie-functions and start investing heavily in UX/UI developement for the average non-techie user. As it stands now it’s unsable for 99% of people and thus will never live up to its potential of being the gold standart for home automation.
This has been the push for the past 4 years, I have no idea why you keep saying HA is not aiming for this. Since you joined HA, there has been about 8 new hires specifically added to address ease of use concerns with HA. Frontend developers, product managers, marketing, product designers, and a community manager.
Please understand that when a person or group of people disagree with a singular narrowed WTH that does not mean these people are against ease of use changes in general.
To me, those “techie-functions” are just as much a part of the user experience (UX) as being able to push a button on your dashboard to turn on a light using your smartphone is.
We just had a heated discussion about this in another thread. Lets not start this again here.
I have repeatedly said I appreciate the work being done. Nonetheless I sympathize with anyone feeling frustrated coming into HA and failing to do the simplest things because as OP put it, they are not “a rocket scientist”.
I’ve been there and still get frustrated rather often - even though it has been years of developement on the side of HA and years of learning on my side since I started using it.
I watch a lot of anime so maybe classes based on skill level needed to complete. 4 classes should be enough
S - requires advanced networking or software knowledge. May need Linux administration skills
A - requires basic networking or software knowledge
B -able to access network router and find device IP. Can login to device and change settings or update if needed
C - able to access webpages and scan QR code using phone
These posts look at HA integrations from same view and expect same level of ease from all but this is simply not possible when items are newly added or early support is provided (think matter or voice). Many integrations are extremely easy to add considering the technical complexity behind them. This came from years work. It’s unfair to compare these but I understand the confusion. I can read docs and say “yeah this is gonna be hard” but some don’t see the forest as they look at a few relatively small trees. Having a difficulty ranking could help this, especially as HA tries to move to more mainstream users.
I’m curious, can you name some integrations that would get class S?
Also, how would that be simpler or more helpful than just checking the literally first paragraphs of integration documentation, which are always instructions on setting up (and a lot of integrations have autodiscovery, so it’s basically add the integration and turn on the device and HA will find it).
As far as I know it has never been the goal.
Unusable is probably a stretch but doing more than the basic auto generated dashboard is probably not as common as you believe.
Look at the statistics, 12 automation and 260 entities is the average.
That means there are a lot of installations with very very few automations.
I don’t believe it’s that many that do home automation, so 99% of the planet that is not interested in automating or smart home seems pretty reasonable.
With 8 billion people on this planet that means 8 millions is possible.
And currently there is less than 0.5 million installs according to analytics.
Its not about simplicity. Part of product release is marketing and controlling user expectations. A user may think from the first few paragraphs they can do it but “you dont know what you dont know”. Like I said, I can tell from the first few sentences while a non-experience newb will say “yeah, I can tackle this dungeon alone even though I dont have the right gear and 0 experience” The goals are
-inform user what we recommend they know before try
-reduce the “HA sucks” from users that think they can do something they clearly cannot
-ultimately the will reduce complaints about integrations and it will clearly inform users which integrations are for everyone and which will require some serious skills or patience to make work
some of thise mat be dependent on install method. I spent all of 30 seconds thinking this over but honestly I think this could be useful but need more time to really think through details of how it would work
S - voice/matter
A - zwave
B- amcrest…looks easy but could get complicated since it may require some specific setting
C- Version, Onkyo…either use discover, just work or ip address entry
I open matter integration - it’s marked as beta and has a big statement it may have bugs and issues. Then there’s a huge intro and long explanation of how to set up everything, with a whole list of prerequisites. How is this not clear that this complicated?
Similar for zwave.
Then there’s amcrest. Right away I can see it requires manual setup with yaml in configuration file. Again, it seems clear to me this is not easy to set up if you haven’t used YAML before or can’t find IP address of the device (this is literally shown in the example configuration). I don’t see the benefit of introducing yet another scale that would really be in practice a complete dark magic to people it would be aimed at anyway.
Meanwhile petro is right, there’s the quality scale, and although the focus there is not on skills required for installation, you will notice that any new integration is now required to fulfil at least the bronze level, which includes easy installation through UI. And the gold tier also has requirements specifically about ease of use and autodiscovery of devices.
Thank you. I looked for this before I posted. I remembered there begin something but didnt realize the “special tiers” existed and missed it on the few I checked.
A friend of mine who is technically capable of the more advanced things in HA just isn’t interested in diving into it that deep. They have however recently picked up a HA Green. On initial setup they were greeted with HA finding pretty much everything that they already had on their network. And getting those integrated was actually very painless for them. They have yet to find a reason to dive into YAML anywhere.
And in the case of their Wemo lighting devices, not only were they painless to setup in HA, they have been more reliable than ever when run from within HA instead of the native Wemo app. He was impressed.
They also recent picked up an Ecowittt weather station, it too was rather painless to tie into HA. Though interestingly, they don’t seem to be interested in fully using it from within HA. They do still prefer the app for the Ecowitt. Same for their IoTaWatt. Most of that I think has less to do with HA’s abilities or setup effort than it does them simply not thinking in terms of just making HA the unified centerpiece of their home tech. They are still just so used to everything having it’s own phone app.
For me being able to avoid all those one-off apps was precisely the main draw of HA. And then along with ESPHome, I can create sensor sets that you simply cannot buy anywhere. Plus I avoid all those dubious phone apps that may stop being being supported by their manufacturer at any time. And with HA, they are all equally accessible on my tablet devices and my desktop machines.
Yes there were definitely times I was frustrated getting the unique things I have working. But the payoff was definitely worth it, and in reality, those “unique” things will never be supported by any of the big name automation systems.
At the end of the day, I just don’t think that HA is truly that much more difficult than most of the other bigger players in the home automation space for the popular and common devices. And those big name players are just so limited in what they can actually do. I watch various vids for those different big name automation systems, and always end up thinking “is that all it can do?”. Never mind the privacy aspects and vendor lock-in with those others. No thanks.
I agree with the general feeling for this WTH because much of the discussion seems to talk around the sources of frustration instead of addressing them directly. For example:
The quality scale talks a lot about ease of installation and good support but at no level does it actually require integrations to be able to be used entirely within the Graphical Editor with no need for the user to make any YAML entries.
Newcomers feel there is mixed messaging because the video events talk about wanting to make things more available to non-technical users but on the forums if you ask for a simple GUI based way to do something you are told to go learn templating, YAML, or whatever and try to implement some convoluted work around to accomplish what you need.
New users are often puzzled when simple things that they are used to using on other platforms do not exist in Home Assistant. It is understandable if some special request that few people would use gets low priority but very simple features that other platforms have already implemented clearly have been determined to be a good resource investment for them but are ignored by Home Assistant.
New users are also surprised at which features are not useable from the Graphical Editor and require YAML editing. An example would be TTS which despite an entire “Year of Voice” seems to languish as a low priority with quality voices depending on cloud services and no way to speak entity states or values without using templating which is not supported in the Graphical Editor.
I think the problem is not that it takes a rocket scientist to get things done in HA but it does take an impractical amount of learning and effort to get some very simple things done that don’t seem to be on the radar to get simplified.
Thank you for this very, very well-put and detailed explanation. THIS is exactly what I was trying to say but was lacking the proper wording / knowledge to formulate properly.
It is exactly this overly steep learning curve and that feeling of “I just wanted to switch the light on, why do I have to learn to code now?” that is frustrating to newcomers (and some existing users like me).
