well there is an automation editor that is currently being put together. It already works for simple actions and conditions will be inserted soon I think. So I’d say we are getting there, in a few monthes or a year there will probably be a more complete UI that allows you to do most of the stuff from frontend.
Rome wasn’t built in one day. 
Especially not if it was open sourced…
Agreed, but I think the thing that trips so many folks up is that this is still a long way from being “production ready”. The trouble is, that’s shown by the version number, not by some big banner labelling it as being pre-production (given the rate of change, IMO it’s not yet a beta).
Currently HA works if you’re willing to invest the time and effort into understanding it. The documentation will remain spotty while it’s evolving so fast, and you can reasonably expect some breaking change at every release (though, whether it’ll impact your install or not is the real question). However, that rate of change means that it’s constantly getting “friendlier” - in the 6 months I’ve been tinkering with it, it’s come on leaps and bounds.
The problem, as I see it, is not maturity.
It’s perception and product design.
Being an early-stage software doesn’t necessarily mean being unfriendly.
When XboxMediaCenter (XBMC, today Kodi) was released, the first versions were buggy and feature-limited, but were usable by the avg layman. Their focus was on creating a super-friendly interface (that did required some tweaking and manual configuration, for some cases) and they delivered. This is what made it popular. Instead of navigating with Windows Explorer, you could navigate with a nice graphical menu on your TV.
I think HA should be made the same. Some simply, generic, core functionality (like “turn the light on at sunset”) can and should be done in a super friendly way.
Putting a UI together won’t solve the problem, I think. You can spend hours trying to teach my father the difference between am automation trigger and a condition - I doubt he’ll understand.
I believe the product layout, the whole paradigm, should be different. Not creating UX to help the average Joe build configuration files, but simply creating building blocks that a user can simply configure. For those who required unique setups - they can ask for more flexible components or even resort to writing code or YAML.
If things were built that way - I would have been in a whole different situation now.
AD is not a real solution. It’s great for me (actually, I’ve already converted most of my YAML to Python code and it’s wonderfully easy for me) - but I don’t believe HA defined its target market as developers or even tech savvy users. It would probably want to cater to the needs of average users, those who don’t want to invest lots of money in high-end sophisticated (and easy to use!) home automation systems.
I’m actually playing with that idea - of building easy to use building blocks that include everything you need, including the UI. It can be based on this wonderful HA engine but will have a completely different approach. It is a huge project, too big for me probably, but I truly believe this can be a cornerstone in making HA a mass-production league of a product.
And thus the cycle of open source projects continues. New guy complains about it, says it should be done a different way, has grand plan, realizes how much work it takes and never contributes.
Meanwhile the few people who actually contribute continue working in their spare time for free to deliver this product and fix the bugs found.
silverr - so your world is only developers and users?
Either someone writes code or they should shut up and be thankful for what they get?
Sounds like a Microsoft Vista approach…
In open source, where no-one is paid, yes it is - if someone was investing in this and running it like a business, of course there would be marketing, focus groups, plans, etc., but that’s not how we got here. We got here because Paulus wanted to automate his house and wrote a cool script to do it. Later on others added functions and it grew organically, to the place it is at now.
When someone is working on this in their spare time we should all be grateful for what we get. As a contributor, I am enthused when people find my stuff useful - if it ever gets to the point where people think they can dictate what I do rather than ask nicely, I’ll respectfully tell them to fork my repository and pick up my code and go home - it already does more than I need it to, everything I am doing at this point is for people who have respectfully asked for stuff. This is not my day job, I get told what to do enough in my day job, I am doing this for fun and satisfaction.
By contrast, when I was using Smart Things, 50% of forum posts were people bitching about how Samsung should fix this and do that - that was at least somewhat legitimized by the fact that people had spent money with them, and rightfully expected a polished product and overall vision, but the forum and community had a much more negative cast as a result.
My point is, that projects like this are driven by people’s enthusiasm, and scratching a particular itch, not necessarily what is “best” for the project for a particular perceived goal. This is both a strength and a weakness, but it is what it is. The cool thing is that anyone can come along and contribute.
You sound like you have some good ideas and the skills to implement them - why not contribute and make it better for everyone?
On the subject of AppDaemon - I’m glad you like it, and have the skills to appreciate it’s approach. You say it isn;t hitting the target market, but go check out Smart Things - they also have event driven Apps, I think it’s a useful part of the ecosystem for power users but I agree there should also be a much easier entry level.
I think aimc wrote what I was thinking, I would add one more group though.
There is a large number of contributors who aren’t programmers. Those who provide testing of new components, write/update the documentation, provide examples, answer users questions on this forum/chat are a valuable resource in this community.
moti_r, it seems like your skill set is something that this project could use. I strongly suggest you give HA another try, when you run into a problem, use this forum or the chat for help. I think you will find that after you pass a certain hurdle it becomes much easier. Then, when you have the opportunity, contribute to make the project better, update the docs if they are confusing, provide a new feature, fix bugs or just help the new users who are getting started.
Just to add to what has been said…each of the comment that you have about automation that you quoted highlights the flexibility and ability to customize HA. You can have a simple automation without all these complications (look at IFTTT), but HA allows you to have significant control over what and how things get done.
Can things be improved? Absolutely. And that is where contributions from the community help. Volunteers spend their valuable time working on writing and maintaining codes and documentation. Not to mention, supporting users, troubleshooting. Given that you are experienced in Python, you can actually help improve HA by contributing and addressing some of the issues you outlined.
There is a reason, HA is not reached version 1.0.
Comparing a media player and home automation is like comparing apples and oranges. A media player is for everyone. It is easy to grasp and should be easy to use. You press a button and it will play what ever you selected. Home automation is so much more complex, especially when you provide the plethora of components. And when you are about to take Kodi as an example, it keeps pretty simple as long as you don’t do anything special that is not part of the standard functionality and will quickly exceed the common users capabilities. So, yes, Home Assistant might not be as easy as pie but if you are not willing to invest time into a working home automation solution neither of the other open source alternatives will be to your liking. You have to invest. And it is either time or money. And I doubt there is a commercial solution out there that provides the flexibility and compatibility of open source systems like home assistant.
So, no, home assistant should not be simpler or more generic. But it will get friendlier and more accessable over time.
I also really have a problem with the statement that programmers or any other kind of IT specialists are having a harder time with home assistant than anyone else. Because I am exactly that. I studied computer science, I am working in the industry and I know how to write code. And I never felt that I’m having a hard time with hass. It has a certain learning curve but implementing relatively simple automation is easy enough even with YAML (and no, I didn’t use YAML before more that I absolutely had to). It might be because this isn’t the first open source project I am actively taking part of, the rules were always similar. You invest time, ask question and you get answers and sometimes even friends along your way.
And I have no knowledge or experience but am grateful there are those who do and are willing to share for no more reward than a thank you.
In fact, I like it not being some dumbed down building block system because I can set it up exactly how I want it and get a great sense of achievement when something works. Equally, I can’t see myself actually creating anything for the community but if someone asks a question I know the answer to I try to give something back by sharing my experience.
If I wanted a Hue like experience that’s what I would’ve gone for but I didn’t.
More than happy to slag anything off you produce and share at no cost though so count me in for testing 
Completely disagree.
It seems we come from extremely opposite directions.
You could say the same about MPC (Media Player Classic) which was superior to XBMC in any way - except it was meant for the high-end users and was too complicated to use. It is still maintained but has a fraction of the user base of Kodi.
A software is a software is a software.
A product is a product is a product.
Step 1: define your target audience
Step 2: define the problem you are solving
Maybe you (the HA team?) defined the audience as “tech savvy people” and the problem as “sophisticated configuration and customization of home automation scenarios”.
In that case - you’ve nailed it.
Just please write that on the tin. Make sure it’s clear this is your goal, on the front page of the website.
My definition would be completely different.
The audience is “Anyone who wants to do home automation” with an emphasis on anyone.
The problem is “Making home automation affordable and accessible”
Yes, there will always be ppl who want highly customized scenarios, I’m not saying HA should not support that.
But if you end up realizing 80% of the ppl need 20% of the (same) functionality 100% of the time - then you’re doing something wrong.
An additional comment: through my struggles with YAML I went through numerous examples (from the cookbook page, there are links to real configurations on github, published by sophisticated users). A brief analysis turned out most of the functionality is really the same. If you look at the (tremendously long) config files you see everyone made it differently… only to come up with the same solution. Does that make sense to you?
On a more general note, if you choose to look at me as the enemy (or the devil’s advocate) and try to convince me I’m wrong or say “well, if you have a better idea why won’t you write some code” then we’re missing the point here. Please don’t expect me to be as invested in this as you guys are (maybe I will, maybe I won’t, I’m just a new user that got into this nice world of home automation, but I’m light years behind you guys). All I’m saying is I’m sure there are plenty of others who went the same path and gave up (and my PM box kind-of verifies that feeling).
moti_r
Completely agree with you. The most successful products aren’t technically the best but they become better than anything else because they become usable, have support and have a huge user base. The niche stuff dies, the effort put in is wasted and lots of frustration that a technically superior product didn’t mature.
I see no point in all this community spirit if the result is going to be too late and too complicated for the average user. It’s time the ‘community’ realised this and looked for some funding to employ some professionals. And I don’t just mean software gurus. This needs the steps moti_r suggests and marketing and drive to make this go.
I think a nice analogy is VCRs.
When VCRs first came out, it involved manually ‘programming’ it to set the time and day for the machine to come on. It was complicated and the average user couldn’t (or couldn’t be bothered) to go through the process.
20 years on, DVRs came out with on screen displays and user interfaces that the average user can manage sufficiently easily to be useful.
Currently, HA (and the rest of the IOT world as far as I can see), is still at the VCR stage. It should be looking towards the DVR stage, but to be honest, at the moment, I can’t see how to make programming sufficiently easy.
Something like NodeRed seems to be a step in the right direction, and I did see an HA interface to that announced in the forums, but NodeRed scripts rapidly become the ultimate spaghetti code with anything other than simple on off commands.
VCRs were never complicated unless you got in all the wrong queues at birth though perhaps that explains why some people have trouble understanding this, just like Ikea instructions which are written for four year olds
What some people are failing to grasp is all the hand holding systems are out there already, Hue, Hive, Xiaomi etc etc, HA is for the people who want to bring all these competing, solutions together and then, if they desire, get clever with it.
The majority are happy with these (not) smart overpriced systems, the people who come here are those that aren’t.
Perhaps a better comparison is Linux -
It starts out as a hobbyist OS, you have to get down and dirty to install it, patch and upgrade the kernel etc. etc. And you can still do that to this day if you want to, but a couple fo things have happened over the years:
Maybe that is the future of HA - as a base system that others will use and enhance while still retaining the ability for the bare metal approach that some of us love. Maybe in 20 years time, HASS will be the Linux of Home Automation, used in hundreds of devices and hubs, and wrapped in all sorts of cool UIs
The reason I use that metaphor is that like HASS, Linux just was, because someone wanted to build something cool one day. Over time, people saw the value and it started to be enhanced, added to and used everywhere.
I have been involved in M2M and then IoT for around 20 years… I am by no means a coder… and knew sweet FA about Python before I stumbled across Home Assistant…
I tweaked, built, learnt, and asked a very responsive, supportive community, when I got stuck. I then helped where I could passing on what I had learnt either by trial or error or from other users…
That to me is what this community is all about… not simply coming in, ripping it apart and then telling everyone how it should be fixed…
Pick an area you feel you can contribute and help plug the gaps… or piss off… choice is yours… no-one is forcing you to use HA…
But in my 20 years of seeing various OEM’s try build a business model in this space, isolating users in order to get hold of their cash, I have yet to come across a more flexible platform with as many supported components…
As a programmer I find it hard to believe that you have an issue with this… When coding software there are many many ways to achieve the same end goal with some being arguably better than others. I might develop something similar to my co-worker, we sit down talk about it, figure out which is best and move on. Doesn’t mean the software should prevent the “worse” solution…
Although I agree that there are many quirks with the yaml config options, they are far from deal breakers for me (especially with the support here). I also appreciate the flexibility and the community’s input when trying to solve similar tasks. To me it’s fun seeing different solutions to the same problem. 
There are side projects such as the mentioned configuration tool that when flushed out might give the point and click experience some may be looking for.
HA would probably still be considered pre-beta in my opinion but the ecosystem surrounding it is amazing. It’ll be interesting to see where it is in 5 years.
I’ve sat back and read these for a while now. As a group we need to stop looking at things from our own personal perspective and look at things from the other persons point of view. As moti has said, he is new to this. His statements are from his perspective as a new user. I have to admit, there are times when I was starting out that I cursed YAML up and down.
I believe that HA is for people who like to tinker with something and find a way to do things that the other retail solutions don’t give you out of the box. Some people want the retail box. I have a good friend in Houston that just wants the box. He doesn’t want to be bothered with all the extra stuff. He has very simple needs and the retail boxes work just fine for him. And that’s ok. I see HA vs Wink for example as Linux vs Windows. My wife can use windows. I would never dream of sitting her down at a Linux command prompt. It just wouldn’t be for her. And that’s ok.
But, with that said, if we ever want to see a broader audience like the people who just want a box, then we need to do more along the lines of making configuring HA more user friendly and intuitive for those people who just want a box. If that’s not the path the dev’s want to go, I’m ok with that. I don’t like boxes anyway.
End of Rant
Boxes are made to be broken, shredded and turned into rubble.
This is actually a very interesting point. I think you actually nailed a lot of things that continually frustrate me bout HASS. Personally I feel it is easier than OpenHAB and everything else cost more money than I am trying to spend on hoe automation. So far the best part about HASS for me is that I can pretty much purchase any mainstream or very cheap HA product and somehow get it to work.
You do have a point when talking about how fragmented some of this coding can be. It can be frustrating when you are looking at the components section for an answer (which most people tell you to look at first) and it has outdated examples that dont actually work anymore. I also had that same exact issue you had when dealing with the kodi component working sometimes and not working other times.
At the beginning it took me loads of hours to find answers to some of these questions but thats when I learned that the answers are actually in the forums and rarely in the components section.
You also have a pretty good point about some of these very common components. I can almost guarantee if you did a poll of the top 40 components used you could easily capture 70 to 80 percent of what people are actually using.
But honestly even though I complain I know I can always come here and find some kind of answer to whatever problem I am having and honestly I will say it has been continually improving and I think some of the devs see some of the same issues you are having and I see them trying to make things more common and streamlined. The newer configuration section, the new automation section, the much needed zwave section, the broadlink learn command. These are all sections that make me feel there is honestly hope for HASS.
I think your approach probably couldve been a little better but I completely get it. Ive had some stressful nights dealing with HASS only to find that they made a change and didnt really document it and I had to find that one single post in the forum to fix the issue (eg. Add two equal signs at the end of a broadlink code if it isnt working correctly…).
Long LIVE HASS though. keep up the great works guys.