It's just a name, and it comes from the days before user-mode/kernel-mode separation.
There's stuff going on that's hiding behind the innocuous-looking name.
Using what you know about how functions are imported to construct a theory that matches the evidence.
Mapping the generic concept to specific file types.
You forgot to include the namespace header file, didn't you.
Trying to look like somebody else.
They go all over the place.
You can shadow it with something useless.
INVALID_HANDLE_VALUE , and it worked: What just happened?
Coincidentally valid, but not what you think.
It sort of depends on why you're comparing them.