I guess this isn’t a surprise that the result is the same - isn’t the lidl device basically the same? Probably the implementation is broadly the same.
I really have no clue about Home Assistant and how it works. All I know is that I have literally 10s of thousands of installs out there with direct connections (and that’s just the ones I know of - I’ve no idea how many customers German Telekom supports with this) and they also work flawlessly.
I’m not really sure where to take this - if you can try and debug that would be great, but otherwise I’m not sure what I can do.
I don’t really want to start digging through Python code - at the end of the day it’s just two implementations of ASH. My implementation of ASH was produced 4 or 5 years ago and has worked without any issues on lots of projects. Of course it’s always possible there’s a bug, or some sort of tight timing issue that’s causing problems here, but it works well on the serial connection and I’m not really sure how to debug this further at the moment (sadly).
So I guess if the problem can be solved in the OS and it works on openHABian, then the issue lies somewhere in the OS or RFC2217 implementation rather than the binding?
All we can say is that we cannot connect the OH ZigBee binding to a ‘serial’ port using RFC 2217 (as of now - tested on Windows 10 and openHABian). However, the Z-Wave binding seems to work in combination with RFC 2217 (Share Z-wave dongle over IP (USB over IP using ser2net / socat ) guide - #253 by mvbergen). In other words: there must be be a problem in the chain of the components. Unfortunately, we don’t know where - and IMHO it would be difficult to find the root cause. Besides the RFC 2217 problem the OH ZigBee binding seems to work flawlessly. The fact that the Z-Wave binding does support RFC 2217 and the OH ZigBee binding doesn’t could be an indication (not a proof though) that the problem lies within the OH ZigBee binding or is caused by an incompatibilty between the ZigBee binding and the OS/OH. Whether the OS/OH or the OH ZigBee binding is the culprit, I don’t know.
If you are interested in debugging the ZigBee Binding/RFC 2217 problem any further, I could send you a hacked Silvercrest ZigBee gateway (EFR32MG1B-256k @ 6.7.8.0). As an alternative I could send a hacked Sonoff ZigBee gateway (EFR32MG21, IIRC), but the Sonoff device is a WiFi - ZigBee bridge which might make debugging a nightmare due to WiFi latency issues.
Thanks @anon71759204 and sorry for the slow response - I’ve been ill for the past few days. Anyway, if you’re happy to part with one, and on the basis there’s no guarantees that this can be fixed as I suspect it could be difficult to trace, then I’m happy to take a look. I do find it strange that this occurs, and am concerned there might be some underlying issue that is (mostly) only revealed when run like this, so it would be good to take a look.
As the problem with RFC 2217 seems to be independent of the [Ember] ZigBee gateway used, I’d appreciate if you could take a closer look into the problem. I’ll PM you for the shipping address.
