Challenge 3: Why a Red-Cut Predicate Can't Be Reordered Freely — Possible Solution
====================================================================

// Pure Prolog facts and rules, without cut, genuinely mean the same
// thing regardless of the ORDER they're written in -- this is what
// "declarative" is supposed to guarantee. Reordering:
//
//     parent(tom, bob).
//     parent(tom, liz).
//
// to:
//
//     parent(tom, liz).
//     parent(tom, bob).
//
// changes nothing about what the predicate LOGICALLY means -- the
// same set of facts is true either way, and a query like
// parent(tom, X) still finds exactly the same two answers, just
// possibly in a different ORDER (which itself usually doesn't matter
// for correctness, only for which answer comes back "first").
//
// A predicate relying on a red cut is genuinely different, using
// classify/2 from Challenge 2 as the concrete example:
//
//     classify(X, negative) :- X < 0, !.
//     classify(X, zero)     :- X =:= 0, !.
//     classify(X, positive).
//
// Swapping the ORDER of these three clauses -- say, moving the
// positive clause (classify(X, positive).) to be FIRST instead of
// last -- would completely break the predicate's actual meaning.
// classify(-5, R) would then match the (now-first) positive clause
// immediately, with no guard to reject it, reporting the WRONG
// answer (positive) for a negative number, before ever reaching the
// negative clause at all. The correctness of this predicate
// genuinely depends on trying the guarded clauses BEFORE the
// unguarded catch-all one, in that specific order -- exactly the
// "correctness tied to execution order" cost the chapter names as
// the real price of a red cut. Pure fact/rule reordering is safe;
// reordering clauses that rely on red cuts and fall-through logic is
// not.

WHY THIS WORKS AS AN ANSWER
------------------------------
This contrasts pure, reorderable facts against classify/2's own real
cut-dependent clause ORDER, giving a concrete example of exactly what
breaks (the wrong classification for a negative number) if the
clauses were reordered, directly demonstrating the chapter's own
claim that cut ties correctness to execution order.