#include <iostream>
#include <cstdint>
#include <vector>

using namespace std;

template<typename T> static inline constexpr T chmax(T& variable, const T value) noexcept
{
	if (variable < value) return (variable = value);
	else return variable;
}

template<typename T> static inline constexpr T chmin(T& variable, const T value) noexcept
{
	if (variable < value) return variable;
	else return (variable = value);
}

static inline constexpr int64_t solve(const uint32_t N, const vector<int32_t>& A)
{
	vector<int32_t> S(N), max_sum_from_left_by(N), min_sum_from_left_by(N);
	S[0] = A[0];
	max_sum_from_left_by[0] = S[0], min_sum_from_left_by[0] = S[0];
	int32_t max_csum = S[0], min_csum = S[0];
	for (uint32_t i = 1; i != N; ++i)
	{
		S[i] = S[i - 1] + A[i];
		max_sum_from_left_by[i] = max(max_sum_from_left_by[i - 1], S[i] - min_csum);
		min_sum_from_left_by[i] = min(min_sum_from_left_by[i - 1], S[i] - max_csum);
		chmax(max_csum, S[i]), chmin(min_csum, S[i]);
	}

	vector<int32_t> max_sum_from_right_by(N), min_sum_from_right_by(N);
	S[N - 1] = A[N - 1];
	max_sum_from_right_by[N - 1] = S[N - 1], min_sum_from_right_by[N - 1] = S[N - 1];
	max_csum = S[N - 1], min_csum = S[N - 1];
	for (uint32_t i = N - 2; i != UINT32_MAX; --i)
	{
		S[i] = S[i + 1] + A[i];
		max_sum_from_right_by[i] = max(max_sum_from_right_by[i + 1], S[i] - min_csum);
		min_sum_from_right_by[i] = min(min_sum_from_right_by[i + 1], S[i] - max_csum);
		chmax(max_csum, S[i]), chmin(min_csum, S[i]);
	}

	int64_t ans = INT64_MIN;
	for (uint32_t i = 1; i != N; ++i)
		chmax(ans, max(static_cast<int64_t>(max_sum_from_left_by[i - 1]) * max_sum_from_right_by[i], static_cast<int64_t>(min_sum_from_left_by[i - 1]) * min_sum_from_right_by[i]));

	return ans;
}

int main()
{
	cin.tie(nullptr);
	ios::sync_with_stdio(false);
	
	uint32_t N, i;
	cin >> N;
	vector<int32_t> A(N);
	for (i = 0; i != N; ++i)
		cin >> A[i];

	cout << solve(N, A) << '\n';
	return 0;
}