Welcome to the OpenWebNet Encyclopedia #22
Replies: 4 comments 8 replies
|
@GreenGrassBlueOcean, @xtimmy86x , @Interstellar0verdrive, @fedem95 you might want to come have a look around! I'll probably miss many people I've met over the years... @caiosweet, @TheDarkWizard, @mantovanellimatteo, @Depechie, @lyubomirtraykov, and everyone interested in OpenWebNet, this place might be of interest to you! |
|
Hi @anotherjulien, Thank you so much for starting this initiative and opening up the OpenWebNet Encyclopedia! Having a central, authoritative, and evidence-backed reference for the protocol is something the entire community has needed for a very long time. I noticed you mentioned:
You'll be glad to hear that we already have a fully functioning, production-ready Model Context Protocol (MCP) server dedicated to this exact purpose! 📦 Where It LivesThe repository is located right here in the organization: (We also have 🧠 What OpenWebNet Data & Capabilities It HasThe server is built with Python 3.11+ using
⚡ How to Add It to Your AI Coding AssistantsBecause it supports 1. Claude CodeRun this one-liner in your terminal: claude mcp add openwebnet-mcp -- uvx --from git+https://github.com/OpenWebNet-HA/openwebnet-mcp.git openwebnet-mcpOr add it directly to your {
"mcpServers": {
"openwebnet-mcp": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/OpenWebNet-HA/openwebnet-mcp.git",
"openwebnet-mcp"
]
}
}
}2. Antigravity / Gemini CLIAdd it to your {
"mcpServers": {
"openwebnet-mcp": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/OpenWebNet-HA/openwebnet-mcp.git",
"openwebnet-mcp"
]
}
}
}3. Claude Desktop / Cursor / VS CodeAdd it to {
"mcpServers": {
"openwebnet-mcp": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/OpenWebNet-HA/openwebnet-mcp.git",
"openwebnet-mcp"
]
}
}
}We can easily feed any new findings, edge cases, and reverse-engineered frames from the Encyclopedia straight into Looking forward to collaborating on this! 🚀 |
|
I agree with this approach. In my opinion, the Encyclopedia and the MCP have two distinct responsibilities and should probably remain separate repositories:
The important part is avoiding two independently maintained protocol catalogues, because they will inevitably diverge. Ideally, the machine-oriented knowledge base should be generated from, or automatically validated against, the human-facing documentation. It should preserve the same distinctions between confirmed facts, inferred behaviour, limitations, and unknowns, together with sources and version information. The MCP could then consume a versioned structured export from the Encyclopedia rather than copying or manually recreating its contents. CI could verify that the structured data remains in sync and test the MCP parser and validator against known frames and examples from the Encyclopedia. From the MyHOME panel perspective, this could also be extremely useful. The Encyclopedia contains enough structured knowledge to support advanced panel functionality such as:
The panel should not interpret Markdown pages or duplicate the protocol catalogue in its frontend. The MyHOME backend could consume the same versioned machine-readable export, perform all protocol resolution and communication, and expose already validated structured data to the panel through WebSocket APIs. I would initially keep this functionality read-only. Device programming should come later and behind strong safeguards, because the documented I would also avoid merging the repositories for now. Keeping the canonical knowledge source separate from the tools and applications that consume it gives each project clearer ownership and allows them to evolve independently. Regarding licensing, I think that deserves a separate discussion before any content is transferred. Documentation, structured protocol data, libraries, integrations, and tooling do not necessarily need the same licence, but their intended reuse and compatibility must be made explicit. So, in short: one canonical knowledge source, one machine-readable representation derived from it, and multiple consumers — including the MCP, the MyHOME integration, and the dedicated panel. |
|
First of all, thank you so much for remembering and mentioning me after all these years! 😊 A very special thank you goes to @anotherjulien. I still remember very well when we worked together on the thermoregulation support. What impressed me back then, and what I still remember today, is that Julien didn't even have a MyHOME thermoregulation system at home. Despite that, with incredible patience, kindness and a great sense of humor, he implemented and improved the support remotely, relying on our tests and feedback. Thanks to that work, I've been able to control my home heating with Home Assistant for all these years. So Julien, thank you again. I haven't forgotten! 😊 And of course, a huge thank you to everyone who picked up the project again, fixed the problems, reorganized things and, most importantly, brought it back to life. I still need to understand exactly who did what, so I'd rather not start naming people and accidentally leave someone out, but what you have achieved in such a short time is truly impressive. And since this discussion is about the OpenWebNet Encyclopedia, let me say that I think the idea is simply brilliant. It isn't just a collection of information. For those of us who spent years trying to understand OpenWebNet by piecing together documentation, experiments, frames and scattered bits of information, finally having an organized reference like this is incredibly valuable. I've been using MyHOME since 2005, and OpenWebNet has been part of a good portion of my life. Seeing the OpenWebNet Encyclopedia today makes me smile: if we'd had something like this twenty years ago, we probably could have saved ourselves quite a few sleepless nights trying to understand frames, commands, strange behaviors and documentation that was almost impossible to find! 😄 In the meantime, I've reached 53. I still work in IT, and these days I mostly enjoy experimenting with AI, including in my day-to-day work. So perhaps I'll leave some of those sleepless OpenWebNet nights to the younger guys now. 😄 I also noticed Getting back up to speed with OpenWebNet and MyHOME now, and especially understanding how much has been done in the meantime, won't be easy for me and will take some time. But I really hope I'll be able to take a look and at least understand what you're creating... ...or rather, looking at all this, I should say what you've already CREATED. 😄 And please forgive me for using this Encyclopedia discussion for a few thank-yous and some personal memories. Julien mentioning me, and seeing all this work again after so many years, brought back a lot of memories of experiments, tests and discussions around MyHOME and OpenWebNet. Congratulations to everyone and, above all, thank you for bringing this project and this community back to life. P.S. My English hasn't suddenly improved after all these years! 😄 I don't speak English, so I had some help writing this message. After all, I did just tell you that I've been playing with AI lately... 😉 |
Uh oh!
There was an error while loading. Please reload this page.
Welcome to the OpenWebNet Encyclopedia
The OpenWebNet Encyclopedia is intended to become a central, durable reference for the OpenWebNet protocol.
It brings together established public documentation, practical knowledge, software behaviour, and original reverse-engineering work into one structured and evidence-conscious body of knowledge.
A significant part of this project has been dedicated to documenting aspects of OpenWebNet that, to the best of our knowledge, had never previously been publicly documented. This includes protocol behaviour reconstructed from device communication, configuration exchanges, diagnostic mechanisms, and relationships between the protocol and the underlying device model.
Where something is known, the Encyclopedia aims to document it precisely. Where it has been inferred, that distinction is preserved. Where the evidence is insufficient, it remains explicitly unknown.
The project is now being opened so that this knowledge can be preserved, challenged, extended, and made useful to installers, developers, researchers, and anyone working with OpenWebNet.
A separate machine-oriented knowledge base, intended for ingestion by LLMs and other automated systems, is also part of the project. That work is still in progress at this stage and should be considered under active development.
Corrections, additional evidence, captures, documentation, and new discoveries are very welcome.
The objective is simple: to build the most complete and reliable public reference for OpenWebNet available.
All reactions