Qt 5.14.0 introduces a new `qfloat16.h` file that spams this error, with no way of ignoring it on MSVC.