[Discussion] Replace UTF-16 with UTF-8, where possible

[davidebeatrici]

The project originally always used UTF-16 for Unicode strings, possibly because it was initially developed for Windows.

UTF-8 is considered the proper standard (https://utf8everywhere.org), in fact the new JSON API uses it.

The problem lies in the rest of the codebase, where we have UTF-16 variants for a lot of functions.

Ideally we should only use UTF-16 when calling WinAPI functions, which can be done with MultiByteToWideChar().
Any UTF-16 strings we receive from WinAPI should be converted to UTF-8 as soon as possible, using WideCharToMultiByte().

Compatibility with the various management tools can certainly be maintained, however porting new code from/to the stable version (https://github.com/SoftEtherVPN/SoftEtherVPN_Stable) could be more difficult when Unicode strings are involved.