In the course of public interest, I am taking the risk of publishing the e-mail without Haier's consent.
### Is Haier's claim true?
I think the points are very questionable, but I'm a software developer and not a lawyer or judge. So I can only try to explain here what the plugin does, but the legal assessment must be made by others.
In the first moment of getting the mail I was absolutely shocked, I didn't think that someone cares about me and my little plugin and I know Haier is a billion dollar company, so I answered
after that i announced to take it down and then you guys happened.
**2024-01-19**
I'm getting so much support, and you started a huge wave, I mean we are now an example for the [Streisand effect on wikipedia](https://en.wikipedia.org/wiki/List_of_Streisand_effect_examples#By_businesses) xD
I wrote another mail on and tried to get some clarification and reach some agreement:
Haier US [answered on X](https://www.reddit.com/r/homeassistant/comments/19a615l/haier_us_supports_home_assistant_and_open_iot/) that they have nothing to do with it and support open IOT platforms.
Haier Europe [created a blog post](https://corporate.haier-europe.com/press-release/hon-app-a-message-about-our-iot-and-ecosystem-vision/) and said they are _committed to enhancing the smart home scenarios in line with authorized usages and intellectual property rights of Haier Europe._
To create an account for Haier hOn you have to accept the terms of service. Without it, you can't connect your appliances to hOn and so you can't use Andre0512/hon.
Haier sells home appliances with internet connection and offers the free hOn app. As far as I can see, there is no ads, no subscription and nothing else obvious to generate money with it.
The connection only works with the Haier servers, so your appliance sends data to the cloud and the hOn app communicates with it, there is no direct connection.
I used [HTTP Toolkit](https://httptoolkit.com/) to monitor the HTTP requests between hOn and the Haier servers and then rebuilt the requests in Python (with aiohttp). I have tried to make the requests in the same way as the app does, except for the ones we don't need.
The pretty complex login can be found in [auth.py](https://github.com/Andre0512/pyhOn/blob/main/pyhon/connection/auth.py) and the API requests that I have adopted as relevant for the integration are these [api.py](https://github.com/Andre0512/pyhOn/blob/main/pyhon/connection/api.py).
Beyond that, there is no communication with the hOn servers in the code.
### Why is the plugin divided into two repositories?
**Andre0512/pyhOn**: Is a python library that I publish in the [python package index](https://pypi.org/project/pyhOn/) (pip). The library is used for communication with the Haier's hOn api.
**Andre0512/hon**: is the integration for home assistant. This is the part that for official integrations is located in homeassistant/core. Here I have defined how the data (which is read out by pyhOn) is displayed in home assistant.
This division is common for home assistant and hacs repositories and is helpful to include Andre0512/hon in homeassistant/core at some point.
In my opinion, it would be much more difficult for Haier to enforce claims to Andre0512/hon. So [Mazda also only claimed the library](https://www.home-assistant.io/blog/2023/10/13/removal-of-mazda-connected-services-integration/), but without Andre0512/pyhOn, Andre0512/hon becomes useless.
### How does the plugin uses the api?
This are all requests the plugin sends to Haiers servers
**Restart of Home Assistant or manual reload of the plugin**
- Authentication to the Haier api with the stored username and password
- Loading of all appliance functions (In [hon-test-data](https://github.com/Andre0512/hon-test-data/tree/main/test_data) you can have an overview of which data this is for each appliance)
- 1 request every 10 seconds to fetch the current state for each appliance ([something like this](https://github.com/Andre0512/hon-test-data/blob/main/test_data/ac_312/appliance_data.json))
- If any action is triggerd, e.g. start some appliance or set a new a/c mode, some data have to be posted
**Creating a new releases**
- If I create a new release, program names and translations in all languages are fetched from the api and loaded to the [translation folder](https://github.com/Andre0512/hon/tree/main/custom_components/hon/translations)
Polling every 10 seconds is a bit much. The default interval for most integrations is 30 seconds. Even if the hOn app makes more requests more frequent, but it does it only in use and not 24/7.
As Haier explained in their answer, this generates a lot of traffic on the not so cheap aws hosting. I understand if Haier wishes a higher value here and will hopefully find a good solution with them.
_I had initially claimed 5 seconds, but it is actually "only" 10 seconds, see [this constant](https://github.com/Andre0512/hon/blob/main/custom_components/hon/const.py#L10)._
Since the latest versions are downloaded [about 3000 times](https://tooomm.github.io/github-release-stats/?username=Andre0512&repository=hon) each on GitHub, I assume 2000-5000 active installations.
### Are there some secret keys stored in the repository?
There is a constant for a [client ID](https://github.com/Andre0512/pyhOn/blob/main/pyhon/const.py) and an [api key](https://github.com/Andre0512/pyhOn/blob/main/pyhon/const.py). They seems to be static because they are the same for requests from every account I saw.
The client id is necessary for doing the OAuth of the login process. The api key is to get some static data (the readable names of the programs etc) and would not necessarily be included in the release.
