“…In R n , for n ≥ 2, Hardytype inequalities first appeared in the paper of Nečas [31] in the context of Lipschitz domains. However, it has been well-known since the works of Ancona [3] (p = 2), Lewis [29], and Wannebo [34], that the regularity of the boundary is not essential for Hardy inequalities, as it was shown in these papers that uniform p-fatness of the complement suffices for a domain to admit the integral p-Hardy inequality (2). Uniform n-fatness of the complement is also necessary for the n-Hardy inequality (in R n ), see [3,29], but this is not true for p < n.…”