Marcus Ilgner

Another update: I saw the `opcua_console_logging` crate in some of the samples and activated it. This is the output: ``` 2021-09-24 08:20:30.706 - ERROR - opcua_core::comms::secure_channel - Cannot read our...

@jsnnerve what server are you connecting to? And what operation are you running when it crashes? Because I can happily report that with the latest changes everything is working great...

That's a good observation! I had also removed the pragma from my entry point but didn't remember the specific reason except realising that it wasn't needed. Also note that the...

there's a [patch here](https://github.com/re-gmbh/opcua/commit/1986f0cd1e7705ec8ad554b42900233de2fa902c) - I'll try to find some time to create a MR for it.

> And just to clarify: I completely understand if this is a kind of "not our problem" issue :-) There will always be bad server implementations, I just wanted to...

> > Note that this is not a bad server implementation, only a side effect of OPC UA being mind-boggingly complex. A full-fledged and semantics-aware OPC UA client is supposed...

I wonder whether this is still an issue with the current `master`? I saw that the version of `write_bytes_task` that @ctron referenced has been changed in the meantime, so maybe...

I encountered a similar situation and would be willing to assist with implementation. Let me know if there's anything I can do to help.

Thank you for elaborating! I'll have a closer look at the source then and think about potential solutions.

Nice! That sounds like it would be also be in accordance with the specification if decoding would return `Variant::Empty` in case the array length is 0? :thinking: