Just curious, isn't Unicode support a C11 thing or am I getting it wrong ehat is meant by this

@emilengler on Windows, the macro UNICODE is used to specify which API should be used, mainly when dealing with path names. But the name of the macro is misleading. It actually sets the character encoding used for path names : ANSI or UTF-16. These API exists at least since Windows XP. For example, when you call Createfile() API :

• if UNICODE is defined, CreateFileW() is called
• otherwise, it's CreateFileA()

Ok my mistake, thansk for the clarification 👌

Windows docs sometimes say Unicode but actually they mean a certain type of Unicode encoding, UTF-16.

Do we need -municode, what does it actually do? Use wmain entry? I thought that was detected automatically by the compiler (or at least that's how we have it set now):

@jay for -municode : https://gcc.gnu.org/onlinedocs/gcc/x86-Windows-Options.html

wmain() isn't the only solution; it's also possible to stay with main(), but in that case the command-line needs to be parsed manually after retrieving it via GetCommandLine() in UNICODE. Normally, Windows doesn't do a whole lot with the passed command-line, so parsing it is relatively straightforward. I'm not sure however what extra magic (globbing?) may be done by MinGW's *main() entry code – this is something to look at.