This is taken from Open Logic Project, p. 214, Release 2020-06-25.
Extensionality is formulated as:
$$ \forall x \forall y ( \forall z (z \in x \leftrightarrow z \in y) \to x=y). $$
Now, by the definition of implication ($\to$), $\forall z (z \in x \leftrightarrow z \in y)$ can be false while $x = y$ is true and the implication is true. But this is absurd, as that means that the statement: "Two sets which do not have all their elements in common are the same set," is valid, though it is clearly not. I must be missing something big here. Where is my reasoning faulty?