How to solve Openhab and it's UI confusion?

Add-ons UIs
41

I’ll try to address everything as well as I can, not point by point, but addressing the main points each of you is making. I think anything else would be counter-productive.

@mstormi
I think it’s maybe you that misunderstood me. I don’t for a minute doubt your intentions, I know that you do what you do because you want to help and contribute (and you do). But that is not the point I was making.

What we have here is a user who is providing feedback (this is the way I’m seeing it). Instead of trying to understand why he misunderstood what he misunderstood (assuming it’s just a matter of misunderstanding and not an actual problem on OH’s design/architecture), you’re just telling him it’s his fault for not reading the docs. When he said he did, you told him he didn’t read them “well” (maybe he doesn’t have the “capacity” to understand them well enough? It could be read that way). And then you insisted he didn’t even click the link you posted (let alone read it), although he had but was STILL having trouble/doubts. Again, his fault, I guess.

Regardless, as I said in my original post, all of the above are absolutely FINE. As long as we agree that OH is just what it is today and not trying to be something more/better. Only then and only in that light are your comments “unacceptable” as I called them. Otherwise you’re just a forum member that “comes across as terse and in some respects rude”. You have every right to be.

@Bruce_Osborne
I think your reply illustrates the problem very well. What you’re basically saying is that not one of my points/“complaints” is valid and everything around OH is as good as it possibly could. Maybe that’s true, but I have my doubts. So do others.

@rlkoshak
Thanks Rich.

Regarding the first part, maybe you are right (and also @Andrew_Rowe). I still have some thoughts/concerns but I’ll not analyze them here, I might come back with some thoughts at a later stage.

Regarding the second part, there’s nothing in what you’re saying that I don’t understand or didn’t already know. But to me, what you’re describing is a perfect illustration of the lack of leadership problem that I mentioned. Having individual users contributing and voicing their own thoughts/ideas is perfectly fine, but doesn’t help towards the goal I described. Besides individual users, there should be some with more authority (call them “team members”, “staff”, “moderators”, doesn’t matter) who set the tone, listen to user feedback, keep things within limits and provide direction.

Again (sorry for repeating myself but I want it to become 100% clear), I’m not saying it has to be that way. As long as we (and the project’s leadership, above all) agree that we want to maintain OH’s status quo and are not trying to take it to the next level.

1 Like
42

I fundamentally disagree this is a binary choice. We have to try to act like a commercial offering or else we are sticking with the status quo?

We are not maintaining the status quo nor do we want to maintain the status quo. I would think that all of the replies and links above that basically say “we’re working on it” would be enough evidence of that. But at the same time “the next level” does not have to be a situation where openHAB looks and acts like a commercial product.

43

If I can sum-it up then:

  1. Despite all its ‘power’ and unique features, openHAB is not yet a mature product (in a commercial sense) i.e., it is not suitable for the masses.
  2. As such, do not expect it to be more than what it already is. Learn what it can do, before you start ‘whining’.
  3. Have respect for the many people (and I really mean this) not only keeping it alive, but also continuously developing it (and those answering our many/repeat non-sensical questions).
  4. Read as much as you can about it. If in the end, you convince yourself it is right for you, then all is good (You can come back and contribute). Else, move on and see what else is out there (Alexa/Google/…) and that is more in-line with your needs/skills and budget (time/$/€)
  5. Keep one thing in mind: This is ‘free’ software. You have to invest your own time to get it to work.

Benefits:

  1. You’ll learn quite a bit
  2. You can customize the interface and your rules as you wish
  3. Your info/data is stored locally … not in the cloud, where there is potential for someone to ‘spy’ on you. That alone, is a big plus!

Fair assessment ?

1 Like
44

It depends on what you define as “being commercial”. If being commercial means listening to users more and taking into account their feedback then yes, OH should aspire to become more like commercial offerings. If it means setting priorities based on maximising profit, then obviously this doesn’t (or shouldn’t) apply to OH.

To me, “taking OH to the next level” is making it more mainstream by making it simpler and easier to use for a lot more users. What I said is that in order to achieve this, we have to look at and take lessons from what bigger (not necessarily commercial) projects are doing.

To me, it’s not about being commercial and I never said OH should be. The word I used was “professional”, as opposed to amateur. That doesn’t have to do with being commercial or not. A lot of open source projects work in a very professional fashion and they’re not at all commercial.

45

I don’t think most of use don’t want to maintain a status quo. But the reality is. Home automation is complex, and thus products that try to make.something of it require a lot of effort. With only volunteers, with limited time, it’s just nearly impossible to realize a ‘next level’. Even to just guide people in contributing would be a full time job. The whole idea of the move with eclipse smart home was to get a more solid base. However that didn’t work out and we’re in the after math of that situation. So do we want a next level? I think yes that’s what motivates people. To make things better. In the mean time we just stumble along to get things work just for ourselves and others. But in a context where there are no financial incentives to just pay one’s bills there will not be a next level that puts others first and that might mean it takes longer to get something that might be easier for less technical users.
To indicate my point. The fact I’m replying here means some other developer won’t get his/her pull request reviewed, sorry.

Now regarding openHAB 3. The answers proof people want it better as is mentioned several times it will solve all kind of issues. The reality is, it just depends on what will be contributed.

I can’t think of a large open source source projects that operates professionally (as in very user centered) that is not backed in some way by a corporate sponsor or has financial backing (or at least I can’t remember one). So unless something changes in that area we’ll be guided by the individual contributions of each other. To end somewhat jokingly: Ask not what openHAB can do for you, but what you can do for openHAB (which actually would mean what would be required to get openHAB to a next level, but that really is an individual choice).

4 Likes
46

I’ve debated for a while whether to weigh-in on this topic, but a number of worth while points have been raised in last half dozen posts, that I thought worth building upon. I consider myself still to be a relatively newcomer although I have been using OpenHAB now for about 9 months. As I had noted in a few other posts, I took a look at most/all of the open source HA solutions before deciding to move forward with OpenHAB. In my opinion, they all have relatively steep learning curves and are not meant for the casual user. Flexibility and power comes at a price of complexity. Also as I have said in other posts, I do not see this changing anytime soon, if ever. As a result, open source HA is not the right solution for many and they are better served by using a custom installer if they have a complex installation or being satisfied with Google or Alexa.

When I first decided to implement OpenHAB I had no clue about HA, but I read the docs, and re-read them and looked at examples and began to see the potential. Being retired, I had the luxury of time and so I dove deep, took large bites, and spent long hours setting up my installation. When I hit a wall, as a last resort I would post a question on these forums. Very few to be honest. After investing several hundred hours learning and implementing hardware and software solutions using OpenHAB I know I have made the right choice. I now have a fully automated house using OpenaHAB. I mostly view OpenHAB as an appliance, residing in a docker on my server, waiting for an event to respond to. In between events, it is forgotten. I would say for me OpenHAB is 80% appliance, 20% hobby. Still this not for everyone, which brings me to my next point.

Having spent most of my life developing, advancing and commercializing technology, one of the dangers I see with a project such as OpenHAB is that it tries to be all things to all people in an attempt to attract a larger following. I say again solutions such as OpenHAB are powerful, complex and require a strong technical foundation to be successfully implemented in a meaningful fashion and tap its full potential. So while I agree with rlkoshak that the choice may not be binary, the efforts to ad=vance OpenHAB should be well focused. While there is no true authority to mandate direction of OpenHAB, we all of have the power to influence direction. Perhaps some more than others, but nevertheless we all are influencers by how we choose to interact on these forums.

Lastly, to the point of forum interactions, I think one of the major benefits OpenhHAB is the strong user community as exemplified by these forums. It is global in nature and as such provides unique and varied perspective from many cultures. In my work life, I had the good fortune to deal with people from every continent, and culture does play a role in how each of us views things and responds to what we see. Some of us on these forums are more direct in our language and others more restrained and polite, but I believe we all have a common goal in making OpenHAB the go to platform for HA. In my view it’s not about political correctness, or catering to sensitivities, but getting things right to advance that goal for our mutual benefit.

So my advice to newcomers to any open source HA platform is to be prepared for a steep learning curve. Do not embark on this adventure if you have a weak technical foundation, are not willing to spend time, effort and energy as an active learner (learning is hard work), be prepared for frustration but power it through it as the end result is worth the effort.

Finally, for me at least the UI is not all that confusing. Could it be improved? Sure, but it is definitely not an impediment to implementation or use of OpenHAB.

2 Likes
47

It saddens me and feel somewhat insulted that you think we don’t listen to users it takes their feedback into account. So I guess I’ll just stop here. Nothing i can say will comply with the forum’s terms of use.

Did you even look at the GitHub issues and and PRs and posts above?

  • Replacing PaperUI with something that works and is feature complete
  • replacing Sitemaps with something you can build in browser and is unified with the PaperUI replacement
  • moving to NGRE as the default which means
    • no more unique programming language only OH uses (as the default)
    • ability to create rules through the UI, no “coding” required
    • ability to create complicated rules using a main stream programming language
    • Libraries of reusable rules so users don’t even have to write complicated rules in the first place
  • new features to the core like an honest to God scheduler
  • a concerted effort to migrate as many of the OH 1.x bindings to 2.x as possible
  • development of rest API and points to handle the suffer that currently can’t be configured through a UI like persistence
  • development of features in the replacement UI so users can paste old formatted text configs from examples and the forum and import them into the UI based config

But no, I guess we don’t care at all about making OH easier to use by non technical users.

All of the above is stuffy actually being worked on right now. Much of it is in response to user feedback.

Don’t think that just because it’s not being actively discussed in the form every day nothing is happening. Don’t think your ignorance of what the developers are doing to mean nothing is happening.

1 Like
48

Well, just because I did not react doesn’t mean I wasn’t hurt… The part that hurt me was the claim I did not read the documentation and have not understood the architecture. However, I learned to ignore that in the past as I had to fuzz with german “engineers” of type Dr. Ing. working for BMW, Audi, … Not funny.

49

@rlkoshak This sound great! And yes, I had a quick look at the Github issues. Wish I could contribute with something… What I miss is for example security, something like WSO2 IoT started. Found an interesting blog post om medium recently. Security is quiet the weakest point in ALL home automation solutions.

Interesting thing is: I started this thread with a simple question about the different UIs and the different ways of storing the configuration of our “things” and “items”. However, this developed some discussion about more fundamental parts around OH and even the “management” of the project. Which was not my intend, even though some of my reactions contributed to this development.

My original question is already answered by @rlkoshak , thanks you very much for that: PaperUI is considered abondonware. That relieves my worst fear.

The other part about the different ways to stores the configuration as far as I can see that is heavily dependend on if I make use of auto detection and ZWave (thus being dependend of @chris and his ZWave binding).

That I can solve quit easily: in either ditching my ZWave sensors and replace them with Zigbee or make use of “zwave2mqtt”.

So could any of the board admins be so kind close this thread? I would like to start another discussion about the mentioned security questions somewhere else…

50

@hilbrand
Thank you for this very pragmatic post, I think it puts things perfectly in perspective. At least for me.

@rlkoshak
Sorry if I insulted you, that was definitely not my intention. I’ve already mentioned how much I value your contribution.

To use your previous point, what I said (how OH handles user feedback) is not binary either. Saying it can be improved doesn’t mean it’s non-existent or even bad.

In any case, I don’t think there’s anything to be gained by continuing this discussion here. I, for one, apologize for not getting my point across to everyone as well as I intended, I’ll try to do better next time.

Regarding contributing, I will look at GitHub in more detail and come up with ways I can contribute. I have some ideas, especially on the UI front for the new version. I might need some help (guidance) with that, I’ll come back to you for that.

@mstormi
Please don’t make this about race or culture, it’s not and it shouldn’t be. There are at least three people on this thread (including me) from very different backgrounds that agreed that you are more than “direct”. Everyone can be “direct”, it’s the other thing that’s difficult. And no, being sympathetic and polite doesn’t mean you should be any less direct, that’s just a rationalization. Of course, it’s your choice if you want to act on this feedback or not, no one can force you.

P.S. Maybe this example will help you better understand what I’m trying to say: You said you’re not a native speaker (I’m not either, for that matter) and maybe that’s one reason why it’s more difficult to properly convey tone. So, imagine if someone’s reaction to your posts was “Well, I think you haven’t really understood how the English language works, have you?”, “This is a link to an English course, go and read it and then come back here to write”, “I think you still haven’t learned English, have you?”. Would that be OK?

1 Like
51

You misunderstood me. I mean that YOU need to take less offense due to culture differences.
You ended the above thread on a very offensive note in any language.You need to be more tolerant of others and perhaps take some communication courses yourself. You are very close to violating the forum policies on being respectful to others and their viewpoints.

52

How could I have misunderstood you when I wasn’t even replying/talking to you? :thinking:

As to my personal development, I am open to any feedback and suggestion, thank you for providing some.

53

Ok, then let me explain what was going on when I read and responded, maybe this helps clearing things up (@symos: I didn’t want to indicate it was about culture or tone. I think it was a misunderstanding.)
Some background first:
I answer about a post a day where people really are just too lazy to read the docs and post right away about things that are in the docs. Second, that very specific part of the docs I had written myself after I have been answering for months felt gazillions of posts and critics of users to raise this topic again and again. Only to get that question again because again someone was too lazy to read the docs. Many of those to blame us (forum volunteers) for the shortcomings of OH (which - being neither lead nor developer - we can do very little about except to enhance docs and answer posts).
So I was and still am pretty annoyed by this continuing to happen because it means people waste my time instead of honoring the work I have put into.

Now on to your post: the link I gave wasn’t clicked by the time I wrote that although I had already asked for doing so before twice. Note the forum SW counts clicks and displays that so I was sure you didn’t. So while you could have read and remembered all of the full documentation, that’s pretty unlikely. And without clicking you cannot have known what specifically I was referring to so hence my assumption you still haven’t read it. Add to this the fact that even in your last post you were referring to files input and JSONDB as somehow having the same function of being an UI (that’s at least how I read it). Which is wrong. That’s not the point though, I didn’t want to blame you for misunderstanding that (maybe I did).
But at least it was another strong indication to fit to the picture that you didn’t read the docs. That sort of was my “WTF?” moment in the conversation.
Logical or at least understandable, I hope ?
So it was a misunderstanding. And remember when I realized that, I apologized. But maybe with this post you now understand why your posts gave strong indication I was on the right track in believing .
No I’m not an Audi engineer. I, too, know - and hate - that attitude you mean.
Now calling me one isn’t fair and hurts me in fact.

1 Like
54

To everybody asking for “commercial state, suitable for masses, next level etc.”:

Just keep in mind, almost all of these “commercial state, suitable for masses, etc.” systems achieve their state by restricting functions and/or use cases. It’s quite easy to just support a few selected hardware things or even just stuff from your own ecosystem.

As soon as your system is wide open to everything, it is a huge effort to streamline stuff. That’s one of the reasons the commercial systems are always restricted. Off course in addition to keeping the user locked in to maximize profits with selling only own stuff.

2 Likes
55

You were talking ABOUT me.

I was just suggesting that you might have thought Markus was too direct.
This forum is designed for polite, respectful discussion. Not for ranting and making offensive comments.

56

No, wasn’t talking ABOUT you either. You’re not one of the three people I meant.
I think you got all riled up with no reason.

(finally found out how quoting works, yey! :laughing:)

57

Well, it’s all about the architects and developers. Not the users being stupid or lazy.

Seriously?

Then you read wrong. I differ between UI and configuration.

Nope. Presumably, because you believe you already know what I’ve read and/or understood.

I cannot remember calling you that. Maybe you should read my statement again?

58

Security too is being looked at but it’s not as high a priority as it should. I don’t know if there are issues open yet but adding role based authentication and creating a security issue reporting pathway are both discussed. The latter has seen a little bit of work. I’m sure there is more.

OH needs a lot of work security wise. But we are not alone.

This is where you have a misunderstanding, perhaps. Pretty much ALL OH 2.x bindings support automatic discovery, not just Zwave. And it’s very possible to use .things files for the Zwave binding as well if that’s what you want. Search the forum and you will find lots of posts, and I think the binding docs have a subpage that talks about what the Things need to look like.

I at least, recommend taking advantage of automatic discovery where ever it is available. Consequently, to keep consistency, I recommend creating all your things through the REST API (notice I didn’t say PaperUI). Not only do you get consistency, you will never have syntax errors in a .things file, never have to look up what options are available on a given Thing type, and it will be obvious not only what options are available but what they mean and how to configure them. I can’t tell you how much time is spent on this forum for the MQTT binding alone helping people with problems they never would have had if they had just used the REST API to create their Things in the first place.

Now there is one valid complaint about doing so that I’ve encountered so far (most of the complaints I’ve encountered boil down to “this is how I want to do it and I don’t want to change” or based on a misunderstanding of what is possible with JSONDB) which is it is really clunky to create lots of Things (e.g. MQTT Things) that are all basically the same with just a few differences (e.g. which topics are involved). You can’t copy/paste/edit in PaperUI. This is indeed true. But…

  1. Create the first Thing in PaperUI
  2. Query for the Thing just created in the REST API Docs
  3. Copy/paste/edit that Thing to make the changes for additional Things
  4. Create the new Things by pasting the JSON into the REST API Docs

We are not usually in the habit of closing threads except under certain circumstances. In this case I think the discussion is and continues to be productive for the most part. I think you can unsubscribe from alerts on the thread at the bottom of the thread; change “Tracking” to “Muted”. If you insist I will close it, but that is not something we do very often on this forum.

In Markus’s defense, your posts came across the same to me as well. A plain reading of your post clearly was using “JSONDB” as synonymous with UI, and not even just UI but synonymous with PaperUI in particular. As stated that is incorrect and a misunderstanding of how it works. JSONDB stores configuration created through the REST API or through the Karaf Console. Whether you do so through PaperUI, the REST API docs, curl on the command line, the new replacement UI for PaperUI, or by logging in to the Karaf Console and issue commands there, the changes all get saved to JSONDB.

Other misconceptions about JSONDB:

  • I can’t back it up: of course you can and it even backs itself up automatically
  • I can’t use it with git: It’s a JSON formatted text file. It works perfectly fine with git. I do so myself. While the order of entries is not guaranteed, in practice I can’t think of a case where the order of entries has changed for me. This could make the history a little awkward if it does happen to you though. YMMV.
  • I can’t edit it by hand: It’s best not to but it is just a text file. Just be sure to stop openHAB first before editing it.
  • Clicking through PaperUI to create a lot of similar Things is slow and awkward: Indeed it is, so don’t. Use the procedure described above.

@symos, @Bruce_Osborne, @wwebers, @mstormi, I think everyone is clear, misunderstandings occurred. Maybe it’s best to just accept it and move on, at least on this thread.

3 Likes
59

I quote the official document

Thing Configuration

There are a large number of things supported by the Z-Wave binding, so configuration can not be covered here and you should refer to the device manual. A summary of supported devices can be found here and this links to the list of configuration parameters the binding provides.

#Textual Thing Configuration

Things configured manually through text files require the following minimum configuration to be set. -:

Configuration Description
zwave_nodeid Sets the node id of the node within the network.
zwave_manufacturer Sets the manufacturer ID for this device (as decimal). This is used to get the thing type from the database.
zwave_deviceid Specifies the device ID for this device (as decimal). This is used to get the thing type from the database.
zwave_devicetype Specifies the device type for this device (as decimal). This is used to get the thing type from the database.
zwave_version Specifies the application version for this device. This is used to get the thing type from the database.
60

I referred to the documentation Markus mentioned: Configuration | openHAB

The very first row in the table says: automatically discovered things and items not recommended (are not automatically added in .things files, I guess).

and further on:

Use Paper UI or habmin to manage ZWave things, but use configuration files to manage ZWave items.

This is what I wrote. My misunderstanding was probably that I thought the JSONDB was introduced in the wake of PaperUI. However, I got it now … and still not like it :slight_smile:

I did not mean security for user authentication and authorization but security for device management. Look at this example here:

https://medium.com/@ramesha.karu/integrating-openhab2-with-wso2-iot-server-5fda14fd382e

If you like its all about secure device enrolement and control. Maybe not the best example, reading the documentation of WSO2 IoT is probably a better idea.

https://entgra.atlassian.net/wiki/spaces/IoTS360/pages/352781035/Architecture

Problem is: This does only work for IoT devices being under your own control as far as I got it. That rules out stuff made by Philips, Ikea, Xiaomi and so on if not encapsulated by a facade (like the binding Ramesha coded).

Don’t get me wrong, this is just one example. However, it casts a different light on security in home automation: Can I rely on a device just because it’s communicating in a secure network (like ZWave or Zigbee)? Probably not. The example builds upon authentication and authorization of devices (not users). It’s pretty much in line with the zero-trust paradigm I dare say.