My point is your suggestion shows an extraordinary misbalance in terms of the efforts required vs. the benefit this is giving and the number of people this is beneficial to at all.
The only benefit is to save maybe 100€ or less on HW one-time which is marginal.
At the same time, the effort to implement this is HUGE, ressources are short and way better invested into other areas of OH development.
Another major reason you are probably unaware of is that any of the few OH core developers to be capable of writing this binding in theory is a Deutsche Telekom employee anyway to work on the Magenta Smart Home product in his daily job. Thankfully, Telekom’s already supporting ESH and thus OH development to benefit from it a lot despite OH can already be seen as some sort of competition.
I assume you see the potential conflict, and hope you understand why noone will be willing to take that route.
I agree that the resources might be invested better into other things, although there is not really a constant of resources, so there might be someone doing this who would not do something else instead. Also the number of people who benefit from it might rise after implementation, since that would be currently the cheapest available option.
To me that is far away from marginal, but as a student I am probably not the average user
I was half aware of that, also of the pretty big (indirect) support of the Telekom (but originally I thought about some other developer who might pick that up like most of the not highly demanded bindings).
Also I can understand that the Telekom might not want to see this stealing them users from their product, although I am not sure whether that would be the case. Unfortunately that leads to closing down the ecosystem although it was originally thought and is advertised as an open solution to connect as many different brands as possible.
An example is that there apparently used to be a REST API available at http://home_base_ip/rest that could have been used by a binding, but unfortunately that is gone now (https://community.qivicon.de/questions/api-offen-fur-entwickler).
openHAB often went the way of breaking that lock-in, even if that means using the devices in other ways than originally intended by the manufacturer (like hue-emulation and the Amazon Echo Control binding). However I see that Telekom is some kind of taboo and I don’t want to cause trouble. Nonetheless I hope it is ok to further pursue this possibility (binding or direct install)?
Also maybe the Telekom can overthink the strategy of completely closing the Home Base and maybe can offer things like an API again.
I did some experiments today with the network tab of the Chrome developer tools and it seems the REST API is still available on that address, but needs authorization included in the header.
So with that it will probably not be too complicated to create a binding for openHAB and I might do that if I have some time.
Before I want to clarify whether it is ok or would harm the relationship with Telekom.
The goal would be to get the binding into the official distro of course, otherwise it is too much effort to develop a binding just for private use.
I will for now probably use the exec binding to do curl requests to the Qivicon API. For my two devices connected to the Qivicon Home Base that is enough. But I will still think about the creation of a binding.
Feel free. I don’t feel you’re willing to truly listen, one last advice, though, from an experienced Home Automation operator:
I would drop that idea because you will have to rely on two boxes to run and fully interwork with each other. There’s a lot of implications to this on software architecture and detail device communication level why this will lead to problems in one way or another.
One should only ever take this as a last resort approach when you cannot move devices.
Marginal short-time savings will be causing a lot of trouble in the long run.
Hans-Jörg has already given the best possible answer to you, the OP and everyone to still join this thread asking for a binding:
So my advice is: if you already own homematic or zwave devices, get the proper stick(s) - they’re in the 50€ range - and move your devices. And if you don’t own Homematic, there’s no point in using the Homebase at all anyway.
Alright. It is not that I don’t listen, it is just that I don’t really understand the arguments. To me connecting to a Qivicon Home Base is similar as connecting to a Homematic CCU. The REST API should do the work, so that there should be no big problems caused, but I am not that experienced, probably you have good reasons to say that.
As for Zigbee: Apparently the Qivicon Zigbee Stick is also working with openHAB now, so I might give that a try.
It isn’t. And even if it was, you obviously have no idea of the efforts required for this development as well as of the implications this will have for your Home Automation setup and operations.
So if you can’t understand, you should believe when experienced people tell you. Or go burn your fingers.
Again:
You’re only staring at the money (and just yours, btw). Always a bad advisor. If you’re serious about HA, you would not ever even just consider such an architecture-wise bad move just because of 50€. 50€ is also the average price of a single actuator which is really nothing in HA.
Just to give you an idea, 50€ is also even less than the avg. hourly rate for this sort of SW development. I’d estimate the efforts of a comprehensive, fully tested binding (and you will want to have no less) to sum up to end in the three-digit hours range, so do your math of the value you’re asking for.
If you don’t want to take advice, ok - go burn your fingers to save those 50€.
But please don’t expect or ask developers to put work in to help you with that.
I think I have. At least if you are correct with the three-digit hours range assumption. That is also what I assume it to take.
Not only. Browsing the internet a bit showed that there is indeed a demand for that. I assume there are at least one hundred people theoretically interested in it.
Since I will probably develop it alone, I am not asking for anything, instead I am giving something to the community
Overall I think I will continue with writing the binding and risk burning my fingers. I think it might be a good opportunity to get into binding development. I don’t like having effectively dead hardware with big capabilities laying around.
I don’t expect anything, but since I might need some help, I will ask, while trying to keep the requests minimal.
Yes, that is right, but I think there is no one interested in doing that probably and also the benefit is not that good compared to the effort as you stated already. So I will try to do it myself.
great that you share the idea, really like this. If you need a beta tester for something, give me a call I dropped OH after the experience of this strange denial of open integration idea in this discussion, but the next winter will come with long evenings & time for trying…
sorry for my very late reply. Unfortunately I had some other stuff to do in the meantime and did not read all my mails, so I just noticed about your reply recently.
I did some early tests and development of a binding seems possible. However I ran into some problems getting Java to authenticate correctly with the Home Base. So maybe as a first start I should try getting a binding to work with the API without authentication. Unfortunately I don’t have that much time, so this project might also not make it. But I have it in my head.
Sad to see you leaving openHAB because of this. That should not really be a reason to leave.
I’ll try to make sure to notify you when there is progress, but don’t expect anything
good news for you. I’ve been working a long time on a magenta smart home / qivicon binding.
At this time, i have successfully implemented some divices like thermostat, power plug and door window sensors. But it isn’t completely ready yet. I hope to get the rest of the stuff done in the next weeks.
Maybe i have to talk to qivicom / deutsche telekom before creating a pull reqeust because of legal issues related to their authentication system.
Just want you to know, that something is on the way
I dropped OH after the experience of this strange denial of open integration idea in this discussion, but the next winter will come with long evenings & time for trying…
