Are these two ways the same for overriding a function?

Help
1

Nix Pills | Nix & NixOS provides a way of overriding a function f:

rec {
  makeOverridable = f: origArgs:
    let
      origRes = f origArgs;
    in
      origRes // { override = newArgs: makeOverridable f (origArgs // newArgs); };
}

Is it the same as:

rec {
  makeOverridable = f: origArgs:
    let
      origRes = f origArgs;
    in
      origRes // { override = makeOverridable (newArgs: f (origArgs // newArgs)); };
}

?

In particular, are

newArgs: makeOverridable f (origArgs // newArgs)

and

makeOverridable (newArgs: f (origArgs // newArgs))

the same?

Are they the same as

makeOverridable (makeOverridable f)

?

2

I don’t have a computer at hand to test it out, but from reading the two functions I think they will behave identically, but the second one is conceptually quite confusing, because it passes the newArgs to the origArgs argument of makeOverridable (remember that as you’ve written it, the function is eta-reduced, it takes an extra argument that’s not explicit in the definition) and then modifies the function f in order to do the right thing in the end. Especially the meaning of origArgs is confusing, since once you call override, this argument is actually receiving the newArgs and your modified value for the function f is merging this into the actual origArgs (which f closed over when it was created).

So they are functionally equivalent (afaict), but the second version is conceptually pretty confusing and the variable names do not reflect the semantics.

The first version however is very clear, the function f is always the same and we merge the newArgs into the origArgs and pass the result to mkOverridable as the new origArgs for the next level of overriding.