It's unlikely but possible that this could be a breaking change. If someone implemented CommandExt in a separate crate, then adding this new associated function will break that implementation. I know, it's really unlikely, but still.

This could be prevented by either 1) moving force_quotes to a new trait, or 2) providing a default implementation of force_quotes which did nothing, but returned self. I'm not sure that's possible, though, given how the function signature is defined.

Does trait are possible to implement defaut implementation?

@lygstate Traits can have default implementations for methods, if that's what you mean.

@lygstate For example, you can do this:

pub trait Foo {
  fn foo(&self);   // <-- has no default

  fn bar(&self) {
    println!("I am the default behavior");
  }
}

It's impossible to implement default implementation

    #[unstable(feature = "windows_process_extensions_force_quotes", issue = "77728")]
    fn force_quotes(&mut self, enabled: bool) -> &mut process::Command {
    }

Well, if there's no reasonable default implementation that satisfies the expected behavior, then it doesn't make sense to have a default implementation.

However in this case you can probably just return self as @sivadeilra suggested. Having an empty body is definitely not going to work though.

@camelid I am sorry for that, return self are also not working and I've tried.
It's trait and not inherit from process::Command, so return self not working