My understanding is that the tag naming approach was derived from the initial Apple HomeKit integration.
There have been a few discussions on how to continue having a unified configuration across virtual assistant integrations while introducing customized metadata configuration. The main consensus seems to be to use the recently added semantics metadata support concept. Another idea was to expand the current list of supported tags across integrations but we never got to an agreement in the end.
I think that is a terrific idea. It is not unprecedented to have two binding using the same technology. Matt1 has developed a great ip camera binding when there was already an ip camera binding and has developed many new features
How would a second binding get registered with Google? I cant see Google allowing Openhab to have multiple implementations allowed…
I’ve stated this above, that I’m more targeting self hosted solution rather than hopping to become official addon, I mean even now I have my own hosted myopenhab service with test action wired on google.
Regarding Google allowing multiple implementations, probably not, but then again, have a look at their suggested bindigs, e.g. “Daikin AC Manager India”, “Daikin Smart Home”, “Daikin D’SmartHome Lite” “Essence Australia Smart Home” “Essence Smart Home Beta”. They do look like they should be single items, yet somehow they are split. Also I have no idea how hard or vigorous their approval process is. But afaik this is way too early to think about something official.
Cough home assistant cough
PS before I get flamed I come here from home assistant because I have an interest in any open source projects. In looking around I saw this comment and just had to respond. However I am not here as a HA shill. Cheers.
Guys, can you explain for me that please:
is integration with Google Assistant should work now while syncing of new items is disabled to myopenhab.org (according to Myopenhab not reachable) ?
I’ve recently tried to integrate with Google Assistant but was unsuccessful.
It’s very long.
Marzima spent several weeks to push and wait for Google official approval. (Thanks to him and his perseverance!).
Main obstacle was that Google employees had no product to buy and test with, as it would happen with regular smart home device manufacturers, so this was kind of first open source project to be integrated with Google smart home part. Maybe now is better.
BUT there is beta , where you can invite specific google accounts to test your Actions on Google beta, and only those users will see the openHAB-test in Google Home app.
I like also the option for self hosted solution. This is available also with marzima version, but you idea of maybe using Docker or ready-to-use VM, and maybe with a very clear and possible simple setup manual, would make possible to many users to go for the self hosted path.
Another important part to consider is the direct local communication instead of going through myopenhab.
This was suggested as possible improvement (not available initially from google) in one of the issues
Integrate with Local Home SDK #102.
Strangely Google Assistant integration with OH is 1-2 seconds slower than Alexa integration (I can test both on same OH instance and same room).
So maybe by using Google Local Home SDK may improve a lot the response time.
Finally I also think opening a new thread for following you implementation, asking for help on testing , progress etc. is a good idea.
The title here is not appropriate anymore for your new implementation effort.
Yes If it can work avec the local SDK , this one will be perfect
Yes, this is exactly how I use the current integration right now
Hm this is strange. I also have Alexa, but haven’t really noticed this. Can you given more examples? Like is it for simple things like “Turn the light on” or complex ones “Dim kitchen light?”
Afaik it would require changes in OH instance, and you still need approved/self hosted solution to sync devices
You’re more than welcome here on the openHAB forum. We share the same passion, just use a different solution 
Just let it go. Go to your account preferences, select the Users menu and add him to the Ignored list. That’s what I have done. I’m no longer wasting my time reading his assays about how everyone except for himself must do something for the community. If it’s coding, then he can’t code, if it’s going though a list of repositories looking at the dates of the last commit then he doesn’t know how, if it’s organizing then he wants others to do it. The list goes on and on and on.
It’s a total waste of time and energy drain.
You keep on saying “they´re not willing to”. Did it ever occure to you, that maybe they CANT!.
You should get your facts right! I actually said I would.
But it´s pointless unless it gets organised and structured. Otherweise it will end up the same, cause thats whats missing in the first place. You may disagree upon that part, but then you´ll have a problem explaining how come this thread ever got started in the first place.
I find it highly destructive blaming anyone for not knowing “how to”, when they offering their help.
You had about twenty posts time to think about some orga yourself, but you have chosen to post here instead.
If you need a fixed structure, develop it yourself and present it to the community.
No one will do that for you, since it would be the wrong one anyways.
You want someone, who gives dedicated tasks to you and this instance is not existing.
That has been said enough times now.
You have to start yourself and get an overview.
That’s what I have done too as I startet contributing to the docs and it was the same for other parts and events I am helping out now.
I got an impression of the base and found tasks that I am able to solve.
No one poked me or have me a description/blueprint.
So maybe it is time to stop posting here and dig into the GitHub repositories.
And as a second statement like days ago:
This is the wrong thread for this discussion and should have been continued in a separate place.
Yes even the simple “turn light on” take a variable 1-2s more on GH than Alexa, same light, same binding to control the light, same OH instance. This happens always to me.
Maybe the difference could be your connection goes through different myopenhab backends? (They depend on location, I am in Italy)
I Just read briefly in the Local Home SDK docs and they state the home device should not be changed. What they say you should do is create javascript code to be run by Google on the GH device which invokes your home device using local connectivity. Sort of transfer the cloud part of the google Assistant integration to be run on the google home device.
In our case the local home device is of course OH Rest APIs.
From the docs
the Local Home platform scans the user’s local area network using mDNS, UDP broadcast, or UPnP to discover smart devices
So afaik local openhab instance would have to expose item via mDNS/UDP/UPnP . I am not tha familiar with openhab but could not find anything with a quick google search?
@All,
I am out of this Thread.
It went in many strange directions from community members, I still don’t understand.
It is also frustrating for me to read how minimal this integration is valued. I will add the last feature hubs and release it, afterwards maybe gonna fade me completely out. No matter if there are volunteers or not.
@rlkoshakis right:
People invest free time, develop a base for all of us since we are a big family. This doesnt mean that the maintainers or initiators are your free developers and will always implement new features. I think they should at least look that the released stuff is working and check that it is secure etc.
As from this whole discussion I just learned that it is not worth to spend so many time actively in this OH GA project. I mean why should I when there will be new ones created beside this project and no willing to join or improve the existing repo? I always could have created a version running for my self hosted system, but rather tried to make it an official integration. The whole current GA integration is fine and maybe need some small features added/adjusted. So I see your points as a luxury problem. Of course it can always be improved, but to call it not maintained is a strong missvaluation. Taking in consideration how happy all been with it 3-4 months ago.
It is the overall moral /mindset here and how the community thinks. Maybe I should invest my energy to other OH projects where they need me. Or I will start a new project for an openHAB feature missing for long.
I am not pissed, I am just tired of the ego tripping. I ve no time to write 1000 posts here for the same fact: No one wants to make hands dirty. There was one contributer helping out. You guys started to open threads with „…no longer maintained…“ even there been PRs merged. But the feature hubs you wanted to get in as quick as possible where missing. You guys dont pay my bills. If you want a project where the „maintainer“ will work non stop for your requests all time, please first pay him. It is hard hearing this, but otherwise you wont recognize.
You can’t expect that volunteering people (even thos started the efforts) will forever develop on that project. We are a community NOT a marzima-tity.
Really need to re-think the future of this integration and my involvement.
BR Mehmet
I can reiterate what I already said: I don’t find Tag solution flexible enough and like the metadata idea better. You did not like this approach as I understood from this thread, quote ‘Good look with your Approach’. I also like typescript better, and creating node app instead of deploying it via google functions. And well, it’s always easier to just say screw it, no backwards compatibility. So sometimes it’s easier to start from scratch and rewrite everything. Plus, as I always stated I’m more targeting self hosting stuff than trying to take away someones Ego and create an official integration. And the most important thing - I find it fun what I’m doing and how I’m doing it - fun coding, fun learning. I don’t care if it goes to trash bin the next day, or no one uses it. “Time you enjoy wasting is not wasted time”
Yes, and no. The whole point of voice control is to be able to talk conveniently like ‘Arm’ instead of ‘Turn the alarm on’ or ‘Lock the door’ instead of ‘Turn the door on’ or ‘Open gate’ instead of ‘Turn gate on’ and have those nice related device icons. Luxury? At least not to me.
If I offended you by opening this thread, or if you are offended by effors, please accept my appologies.
So isn’t it time to close this thread? Even the thread title is negative.
Will someone open a new thread, positive and productive GA integration (I would but not qualified to do so YET).
Thank you All,
Looking forward to advancing toward an enthusiastic thread!
Steve-
I hope the two of you are real pleased with yourselves. You’ve burned off another talented developer