Gibbs entropy and Shannon entropy

The two formulations seem identical to me:
$H(x) = sum p(x) log(1/p(x))$
why would the equation be attributed to Shannon rather than Gibbs (in the context of information)?