João Matos

@brunnerant Wish I had a better answer to tell you but the fact is our time to work on these kind of fixes is extremely limited, given all the other...

Tagging our release-master @ddobrev, he might know what's up.

Hey, it's been a while since I have touched this part of the code, so I am not exactly sure what needs to be implemented here. That said, I would...

Not sure what the errors could be... we do test the C++/CLI backend on CI, and all tests are passing right now, so the most likely issue is on the...

@ddobrev I think the problem is that we generate broken code by default. It's fine to not support it, but it should be ignored, not fail like this.

We've had a pending PR that should help with this: https://github.com/mono/CppSharp/pull/983 It just needs a bit of a push to get it over the line.

1. The name should be `libStd-symbols.so`, but you should compile your own, since the one provided by CppSharp for its own usage may not have all the symbols your binding...

> It seems CppShart try to inject a new method `string.__CreateInstance(new __IntPtr(__ptr));` into System.string type, but I have not got it loaded/discovered. We do not try to inject any methods...

Not sure what it could be either.. You can check if the archives for LLVM are properly extracted in the LLVM build path: `/mnt/c/Users/user/workspace/CppSharp/build/llvm/llvm-0c8f9b-linux-x64-gcc-9-*` Also check the verbose build output...

Ideally when we bind those kind of callbacks, the C# code would be dealing strictly with C# `string`, and not with the bound version of `std::string`. We have an issue...