openHAB 4.0 wishlist

In the OP there is a section for bindings. Aren’t bindings a part of OH? With every version there are new ones.

Your request is sorta X-Y problem request type of: if you have a specific device you want to get supported in OH then tell us here.
If there isn’t any then there’s actually no problem and you don’t need what you’re currently asking for.
A generic interface to another home automation system isn’t a value in itself (particularly as that’s already available in HTTP and MQTT) so you’re unlikely to find any dev to be convinced of that idea, capable of and willing to implement it for you.

No, don’t get me wrong. It is a way of taking advantage of HA faster development and wider device support.
I believe that as it is of use for me it would be for more people, that’s why I posted here. But if nobody believes the same it’s ok, there’s always another way to do it.

Have you looked at the many examples posted to the Marketplace? That’s a great source for examples and you might find what you need is already done and you can just install and use it.

Beyond that, CSS is a huge topic. I don’t think we have the man power to duplicate documentation for all of CSS. But with specific suggestions what we have could be improved. Though for the most part, if you are modifying CSS or messing around with raw F7 components, you’re an advanced user almost by definition.

To make things easier for beginners we have the Locations, Equipment, and Properties tabs. What’s easier than not needing to do anything at all? Next in complexity is finding and installing widgets from the Marketplace. Beyond that we also have examples for building custom widgets and the layout for the Overview tab in the Getting Started Tutorial.

Many use cases are represented in the Marketplace with more being added all the time.

There are significant technical challenges here that make that hard to achieve.

You should file a feature request issue on the Google Assistant Integration repo: Issues · openhab/openhab-google-assistant · GitHub

There used to be an Xbox binding IIRC but no one volunteered to migrate it from OH 1 to OH 2.

OH has no priorities. Each individual contributor has their own set of priorities. Unless and until someone volunteers to write a binding for this :shrug. No one can tell anyone else what to work on. Unlike Home Assistant, we have no employees.

Even so, this has been a priority for a number of contributors with many thousands of words having been exchanged discussing approach, feasibility, etc both here on the forum and on GitHub and as I understand it on Discourd as well…

There are several threads with tutorials showing how to do exactly that on this forum. MQTT is generally the technology used to achieve that integration instead of REST API (note OH has a REST API too). OH provides an MQTT Event Bus rule template to make it easier, at least from the OH side of things. But again, if someone were to volunteer to create such an integration it would not be refused.

I find through the extensive support I provide on the forum (which is my primary contribution to the project) that it’s those OH 2.5 users who neglect to read the Getting Started Tutorial are the ones who have the most trouble over all. Next come those users (experienced or not) for whom text based documentation isn’t the best for their learning style (we do our best to help those users). Finally it’s a tie between those who simply don’t want to learn anything and just want a big red easy button and those who are OK with some learning but they want infinite control (usually over the workings and appearance of the UI) with supreme simplicity. Unfortunately, TANTAFL. The more control you want the more complicated and the more work it’s going to be. MainUI tries to find a happy medium. You can get to a pretty usable and good looking UI with minimal work. And there are lots of room for configuration and customization where needed. The Marketplace fills in a lot of gaps as well and gets better with every post. But ultimately, if you want something different or want to tweak everything, it’s going to require a lot of learning and work.

Note, the Marketplace is part of openHAB in every meaningful way.

It’s also still under development. I’m pretty certain that it will become published when it’s ready.

No, there is a flag in Settings → Community Marketplace to show unpublished entries. Toggle that and they will show up and you can install it just like the published ones. But given that it’s still under development and not tagged as published indicates beginners shouldn’t be using this widget yet anyway. So these small hurdles are like speed bumps, deliberate safety features. Eventually though that “published” tag will be added and it will show up in the list automatically.

1 Like

Indeed Rich, there is massive work going on under the hood, especially as we are seeing performance issues on bigger installations.
Our aim is to provide a stable and really usable solution by the time openHAB 4 will be released.

2 Likes
  1. Support for homekit devices connected directly to openhab.
  2. an easy “remap” for icons animations in sitemap. For example my heater drain 1200 watt so when watt is 0 should shown icon 0 and when is 1200 should be shown icon 100.
  3. A binding for BTicino Living Now. (also if the point 1 should be enough to use living now.)
2 Likes

Of course, and also here. But I rarely found just what I was looking for.

:smiley: Thanks, but detail oriented, I would say.
I completely understand the complexity of CSS and wasn’t asking for that. Most of the times it’s little things I’d like to tweak and I find it very difficult. For example, here I learned that oh-list is a special case:

I totally agree. The semantic model is great, but the next thing is to get a nicer UI and there it gets more complicated.

I didn’t know that. Let’s hope the Brick schema gets extended…

I’ll look into that, but I believe they are different.

I know. Maybe I didn’t choose the right word.

I’ve read a lot, before and after the migration. Much of it was yours :wink:

Yes, I know than option, but I didn’t know it then. And I understand why it’s there.

Thank you all for your help, and especially you, @rlkoshak!

11 posts were split to a new topic: Timestamps and persistence

I’ve submitted a PR that should enable expanding the ontology at runtime, according to our own needs. If that PR is accepted, further work can be done to propagate full support to MainUI as well.

I will split this, but primarily because this thread here isn’t meant to discuss use cases and potential solutions in detail.

But it’s useful to point at any such solution that exist already in OH, today.
Quite some wishes on this list in fact need not be put on here as solutions for what’s the real intent behind such a wish are there but the requestor isn’t aware of.

@liaskt are you aware of profiles ? The timestamp-change one for example ?

Thanks for splitting, let’s continue the discussion at Timestamps and persistence

The timestamp-change (and timestamp-update) profile is another solution for the 1st “Task: Access the item lastUpdated timestamp”, thanks for pointing this out. I will post in the separate thread why it does not work well for my setup, but in any case it is a useful solution for others.

  • Energy management with corresponding attractive dashboards (without having to do a master’s degree in software development and hours of studying the documentation and community contributions). With a focus on the operation of a photovoltaic system.
  • Data aggregation with persistence of results.
  • Full implementation of persistence services (use case with the REST API). Examples:
    • Allow the REST API to write data to persistence.
    • RRD4jPersistenceService, getItemInfo returns an empty list.
  • Better and more attractive charts
  • Better and more attractive widgets as part of OpenHAB.
  • myUplink Binding
  • Backup and Restore from the UI
  • Create new item, thing, channel with other item as a template
  • Certificate Management from the UI (support to create, sign and install self signed certificates)
  • Intuitive and complete schedule concept with configuration of special days (holidays, public holidays, etc.). With a graphical schedule editor worthy of the name.
  • Alarm management (alarm lists, alarm classes, alarm history, …)
  • Possibility to integrate own HTML/JavaScript/CSS pages. With a templating system such as Velocity (or whatever). The icing on the cake, with a JavaScript library to communicate with OpenHAB (with WebSocket, subscribing changes, etc.).

I consider my wish list as a kind of brain storming. They are ideas, from a developer, but who is an openHAB user here. It doesn’t have to be immediately shot down with “it will never happen, you don’t understand openHAB, it’s too difficult, it won’t work”, or whatever.
And if “Today openHAB isn’t providing applications”, then maybe it should! Without significant improvements, you can stay on version 3.
Version 4 should just get users to switch from HA to openHAB, not the other way around :wink:
I wish you much success and am excited to see what version 4 will bring.

4 Likes

I believe @mstormi has a turn key system built on OH he sells for this. There is also an issue and PR with a lot of activity related to this in core.

This is already possible. If using a Group to aggregate you’ll need a custom .persist file that lists the Group name separately. If aggregated in a rule, put the result in an Item.

This is already supported.

This needs more context. The REST API and persistence actions all work for me.

Grafana has about the best charts you can find. Beyond that, someone will have to find an open source web page library other than the Apache one we use now. Or file some actionable issues on that upstream project.

Actionable and specific issues are welcome. “better looking” isn’t actionable and subjective.

There was an issue and PR opened for this. I’m not sure the status.

I believe there is an issue open on this too. Though I think it morphed into a more general improvements to bulk editing. I’m not sure if any PRs have resulted from this yet. But in the mean time copying the text from the Code tab and pasting that into the code tab for a new one. The Code tab for Items was just merged into main in the last week or so.

Needs more detail. Everything is an Item and the Semantic Model had tags for Alarm and these can be grouped, listed, queried (using HABot), etc.

Should already be possible, unless Velocity requires server side stuff. Anything that Jetty can support serving up, should be supportable. Just place your html files and such in $OH_CONF/html and it’ll be accessible from http://ip:8080/static/filename.html.

It’s all REST API calls. It would be worth checking out what the various UIs do and see if they centralize OH interactions. If so it might just be a quick lift from there to get a working library.

And there are several widgets to integrate Grafana into the openHAB UI available on the community marketplace, e.g. Grafana chart with time ranges (with time range picker) or my fork of it openhab-conf/UI/widgets/community.openhab.org/Grafana_Integration.yaml at main · florian-h05/openhab-conf · GitHub (fullscreen).

I also have a simple full-screen webview widget, see Fullscreen Webview Widget.

There is a GitHub discussion on the WebUI repo, I am super busy right now and will be at least for the next three weeks, however the “Duplicate Item” (and other stuff like Things, Rules) functionality is on my to-do list.

It hasn’t been merged yet, because some REST API improvements are required to make it fully operational, but the code view will come.

3 Likes

Yes, I know. I guess it’s one of the reasons why he thinks it’s a bad idea to have something like that in openHab.

EDIT: Oh dear. I think I used the wrong term. I didn’t mean aggregation. But what is it called? Rollup?

Now, let’s say I have a historicised meter reading. Can I (retroactively) calculate a 1/4 hour interval consumption from it? And from that again a daily consumption? Without gaps if data is missing (interpolation)? And suppose I want to do the same for a level from an oil tank? Without programming a rule? And without Grafana or whatever? Out of the box? That sounds great. Are there any instructions?

Ok. I’m sorry. Either I did something wrong or it wasn’t possible before. I will try this again. Thank you.

API Call for example with rrd4j

/persistence/items

Some persitence service (e.g. influxdb) implement this method. Many return an empty list.

Right. But Grafana is Grafana. Home Assistant has nicer charts than openHAB. Is that why I should switch? I hope not. I thought the point was to improve openHAB. I know that some of these are difficult proposals. They are also just ideas.

You are right here too. Unfortunately, I’m very bad at graphic design. But I can tell what I like and what I don’t like. What is beautiful is always in the eye of the beholder. And what looks beautiful and modern today will look outdated tomorrow.
But somehow there is very little that you can simply use. Not even for the simplest things, which probably everyone (or almost) has.

What I have read so far (from some people) was quite disappointing. According to the motto “a backup does not belong to openHAB”. But there is already a backup via the command line. I just created a script called with Crontab. But to configure something like that via the UI would be nice. Not everyone knows how to do that.

Yes, I realise that. I have to do with BMS professionally. There, it is common to have an alarm management system. For example, I measure a room temperature. This data point (item) has a configurable “alarm extension”. For example, if the value is less than 19°C, an alarm is generated. This alarm can then be forwarded via “Recipients”. I know, until then it is possible in openHAB. Although (I assume) everything with Rules, right?
Then I would like to see the alarms in an alarm list. Is an alarm still pending? Or gone? Has it been acknowledged (i.e. taken note of)?
Then I want to see in an alarm history which alarms (let’s say yesterday) have come and gone (with timestamp) and when and by whom they were acknowledged.
Ok… before I write my fingers to the bone, I’m waiting for the sentence “This doesn’t belong to openHAB” :wink: I honestly don’t know if that should be part of it or not. I just miss it.

Ok. That sounds good. I did not know that. I will try that out.

Thank you for your valuable feedback.

Super thing. I hadn’t seen that at all. We must have something like that. Very nice work! Thank you.

1 Like

Please elaborate a bit more.
There are loads of ready to use widgets on the marketplace.

From what I understood, it is quite difficult to determine the underlying installation method and therefore select the correct way to do the backup. I did not see a solution for that.

Before this can be achieved, we would need a complete user management, which does not exist by now.
And yes, this seems far above openHABs scope :wink:

Probably but without a rule. But by the time you cover all the different variations and edge cases in but sure it won’t be so complicated you might need a rule anyway.

But just because it requires a rule doesn’t mean you have to program it. That’s the short of thing rule templates and the marketplace are for.

I don’t think there is a template for these right now but they are great ideas to be added.

That’s a bug, please file an issue… It should work the same.

And I explained what you can do now to get better charts and what it will take to get better charts in OH. I’m not sure what the problem is with my response.

And if:

  1. charts are that big of a concern
  2. Grafana isn’t an option you want to consider
  3. You are not volunteering to make OH charts better by finding an alternative library or working to fix the upstream library MainUi uses

you might be happier with HA. Unlike the HA community, we have no animosity and the competition between the two doesn’t drive OH development. We want you to use the tool that you are hallway using and that meets your needs best, even if it isn’t OH.

I suppose this needs specifics too because I have a pretty nice UI that I like the looks of and I didn’t need to do anything but set up the semantic model. So what specific “simplest things” are you looking to do UI wise?

yes

Sounds like a job for a repeater card which can use some logic to show it but show stuff based on Item’s states among other things. There are several on the marketplace.

That one is a little harder but I don’t think impossible. it would probably take a bit more work than it should. There are some new stuff coming which might make it easier too but for now you’d need to model everything using Items with rules and such. It’s short of possible but it’s not going to be flexible.

You can open a new thread and we can brainstorm some options for how to achieve that.

That’s quite the opposite of what’s true and not fair. I’m actually pretty actively supporting the OH development in this area, backporting concepts and even some code.
You should read my earlier post on this and speak for yourself only, please.

2 Likes

OK. I apologize. It really wasn’t fair and I retract my statement. So of course it’s important that there are maintainers like you.
I’ve just noticed that a lot of the answers you’ve given are very negative. But I don’t know the reasons for that. It’s a bit frustrating sometimes when an idea is just immediately crushed without giving it any thought (or so it seems).