You and I agree on that. But those using it, they believe its smarthome
Remote control is a subset of smarthome, no? Therefore they are not entirely wrong.
To be honest, I started with sort of remote control, too.
But smart home is much, much more.
The ideal smart home would be, you don’t need no interface - neither graphical nor lingual.
The home would be so smart, it would do all things, you would normally do manually, but it would do it automatically.
We may never reach this goal but this would be really smart.
I agree, and OH is a fantastic platform to achieve this. In our home OH controls the energy use of our home so that they run when cost of electricity is cheapest during the day. At the moment the following applicances are included
Ground source heat pump
EV car charging
In addition my wake up light is controlled by OH.
The product features of OH are great. Our marketing challenge is to “productify” OH in such a way, that a non-engineer can install and take it into use. Productification is the process of making a product easy to buy, take into use and use.
They are not wrong. But they are limited to their knowlegde. And they will never archive what openHAB (or something simular) can do.
That is smart, if thats what you need. But if you dont need the electricity during the day, maybe because you´re not at home. Its not smart.
This is often the paradox of trying to deal with the electricity costs in a smart way, at least here in Denmark where the electricity costs is lowest during the day (mostly), where Im not at home.
Note, when I speak of electricy costs and power consuming, I dont mean the dishwasher/washingmachine and things like that. First of all, they´re easy to start at a different time of the day. Second, they´re limited in how much power they actually consume in the over-all picture.
I have an electrical heater (heatpump) and I have a EV driving at least 130km each day to/from work every weekday.
Those two devices brings my overall consumption to almost 13.000kWh/year. In all I use about 15.000kWh/year. So all the “rest” is only aprox 2.000kWh/year.
Why would I concentrate on electricty costs for the 2000kWh/year? Makes no sense to me
My EV needs to charge at night, as I work from 08.00 - 18:00. In Denmark we have an extreme high electrcity costs (tarifs) between 17.00 - 21.00, and again from 06.00 - 09.00. These hours noone should be using power at all. Or pay a very high price when doing so.
There isnt really much leftover to do any charging beside at nigths. Which means these few hours left, is what I could deal with. Is it worth the hassle? Maybe! Maybe I could save a coupple of thousand kroners (danish valuta) in a year.
My heatpump is consuming aprox 6-7000kWh/year. This is controled by the simple use of thermostats in each room of my house.
The issue is, my house has underfloor heating with concrete. So any kind of controle/change to the thermostat setpoint will take several hours, before the temperature is at the setpoint. (Several hours means at least 12 hours). Does it make any sense to do something “smart” controle from the electricity costs. In my opinion, it only makes real sense if the buffer-tank is big enough to hold at least 24 hours of needed hot water. My buffer-tank is not that. If it was, it would be easy to controle the heatpump only to run when the costs is at its lowest, within a periode of 24 hours/day.
The reason I mention all this is, that I dont consider myself as beeing unique. Lots of people in Denmark lives like this, and I even think lots of people in Europe lives like this as well. Denmark may be different due to the high electricty costs periodes. But the living is still the same.
I do differ in one situation though, as Ive got a solar system running (no battery though, yet). But then again. When the solar system is producing power, I/my car will not be home to consume it. And then its useless, unless I can do something about either the heatpump consumption or my EV
At least I´ll try, but at this moment its mainly because I can, and openHAB/smarthome became a hobby to me years ago. Not because I really need to do anything. I could just buy a huge but highly expensive battery for my solarsystem, and all my concerns would wanish. But thats not fun
OpenHAB is great, because its highly flexible. But beeing highly fleixble also means beeing rather difficult, and can include hard work to archive something rather simple.
This is why openHAB needs to change or at least try to. Cause its not a simple task to change openHAB. Maybe its even impossible.
In our market, Finland, industry use seems to dictate energy cost which leads to high costs during the day and low cost during the night. The lowest cost timeslot is not at the same time every night, and here OpenHAB comes in handy, it finds the cheapest consecutive cost every day, automatically = smart home.
Because I am an engineer and love tinkering with technology. Like adding WIFI to a washing machine that had none. The relays cost 5€. The washing machine uses about 230 kWh a year. If I can save 5c/kWh by running the machine by night I will have paid back the relays in half a year. I guess total saved money equals a Pizza and a beer per year. So I am not doing it for the money but for the love of technology.
I love OpenHAB because it suits my needs so well. But I am an outlier, the mass market will never do what I did. Productification, productification, productification!
Been there done that, too, just because I could
I know the feeling
The easiest thing to do is look at the other pages that are there. Find a page that has a TOC and look at the markup. There is some Vue wizardry to add to create the TOC. At it’s basic it’s just
[[toc]] on it’s own line, but if you want to configure it (e.g. control how many levels to include) you’ll have to find an more involved example or look at the Vue docs.
For the left hand menu bar, those are handled in https://github.com/openhab/openhab-docs/blob/main/.vuepress/docs-sidebar.js, but I think there is another step needed when you edit that file. Look at the tutorial I linked to above for contributing to the docs. I think it’s covered there. I might be completely remembering incorrectly on that extra step.
As far as I see it, you can have it one of two ways.
- Easy to do the easy things, hard/impossible to do the hard things
- A little bit harder to do the easy things, a lot easier to do the hard things.
It’s clear based on what’s been said here HA is shooting for 1. Alexa and Google Assistant are definitely shooting for 1.
Besides OH, is anyone shooting for 2? Is that not our niche? Shouldn’t we lean into that?
Thanks Rich, I already did, but [[toc]] is not working. That’s why I am asking. I am using in place editor and did also not find any wizard. Never mind, toc is not important.
It won’t actually do anything until the markdown is compiled by Vue and the HTML documents are rendered.
When you create the PR, your PR will be rendered and you’ll get a link where you can see what the docs will look like with your changes applied. That version will have the TOCs in place.
I would prefere 2, ofcouse.
It all depends on how much “a little bit harder” is. But I would suggest openHAB try to go for 2, mainly because this is how to archive most.
My experience thus far is that the code generated by LLMs (I’ve tried ChatGPT and Bard) is more often than not incorrect and sometimes complete nonsense. You basically have to be an expert in writing rules in what ever language you choose to read and fix what ever the AI produces.
In general they are pretty good with basic language stuff but they get the OH specific part wrong almost all of the time.
Here’s a Bard example I just ran.
Please create a rule in openHAB in JS Scripting that turns off Light1 at 10 pm
The following openHAB rule in JS Scripting will turn off Light1 at 10 pm:
rule "Turn off Light1 at 10 pm" when System.time >= 220000 then Light1.sendCommand(OFF) end
This rule will trigger every time the system time is greater than or equal to 22:00:00, which is 10 pm. When the rule triggers, it will send the OFF command to the Light1 item.
To create this rule in openHAB, you can use the following steps:
- Go to the Rules tab in the openHAB web interface.
- Click the New Rule button.
- Select the JS Scripting rule type.
- Paste the above rule code into the rule editor.
- Click the Save button.
- Click the Enable button.
Once you have enabled the rule, it will be active and will turn off Light1 at 10 pm every day.
Please note that you will need to replace the Light1 item name with the actual name of the Light1 item in your openHAB system.
I’m not going to go through everything wrong above, but hopefully everyone here can see that this simple request produced nearly complete nonsense, though it’s very convincing nonsense. That does look a lot like rules you might see here in the forum.
If AI can get to the point where it can produce working OH rules more often than not yes. For now, it’s got a long way to go. I think trying to use AI for new users would be worse than no AI at all.
The whole point of ChatGPT is, it’s not A.I.
Instead, it’s a bunch of (really clever) algorithms which steal (err… borrow…) code and/or text from internet. (did you ever see some sort of copyright or at least a given source in ChatGPT answers?)
But as the database of openHAB is really small (only some ten thousands of rules, many of them are not entirely correct or complete nonsense itself) there is no chance at all that ChatGPT can get the correct answers to these tricky questions
In my career I’ve noticed a pattern when it comes to AI. It’s called AI until it works. Then once it works everyone is disappointed and complains “that’s not AI!”.
I think a cahtbot could work to create OH rules at some point. We are just a long way off right now. It might be an interesting problem to train an LLM just on “good” OH rules and see what it can do. But I have neither the time nor the hardware to do it myself right now. I bet someone could get a good paper out of it though if any grad students are reading this.
I live on a farm in country area of Australia where there is a lot of cattle and sheep and A.I means something different here. (Artificial Insemination)
training a chatbot like chatbase on this documentation would likely be a very useful tool. in my medical practice we have loaded all our internal training documents to chatbase. it has made onboarding new employees easy and provides a quick way for existing employees to find answers. with the right effort, there is no question it would make using OH far easier.
Hi All. Hope this provides some perspective. I have been a “user” since OH2 using config files. The draw card at the time was habpanel as HA had nothing like it. This was my main requirement I wanted a simple interface. Habpanel allow me to have a small display at the front door ( an old android phone) panels in a wall if I wanted of simple use an iPad on a stand etc.
I have looked at HA every time OH upgraded but each time it was obvious I was going to spend more time in yaml than just rebuilding/fixing. OH has more and more in the UI and this is now my only interface. ( Maybe time to put more UI in the Doco and remove the File config examples as that is a put off for new users)
I have never used groups or more recently the Model as Things and Items do the job for me. I guess I am not a power user with my slow growing environment. I have very informative panels into my home devices (Sonos, IKEA,MQTT for lights ipCamera Weather Station, Alexa echo etc). For me this is the key and to be able to do some simple rules to turn lights on at sundown and close binds OH does job.
So we should just cut off the file based users and not provide any reference documentation at all?
I think that’s a pretty drastic and controversial proposal. The reference docs are for everyone, not just new users. Even I use them almost daily.