#pragma GCC optimize("Ofast")
#pragma GCC optimize("unroll-loops")
#pragma GCC target("avx,avx2,fma")
#include <bits/stdc++.h>
using namespace std;
typedef long long ll;
int N, K, a[100001];
ll pref[100001] = {0};
ll dp[100001][2] = {0};
int last_split[100001][201];
bool check(int i, int j)
{
return (dp[i][0] - pref[i] * pref[i] >= dp[j][0] - pref[j] * pref[j]);
}
ll eval(int j, int i)
{
return dp[j][0] + pref[j] * (pref[i] - pref[j]);
}
int main()
{
ios_base::sync_with_stdio(false);
cin.tie(NULL);
cin >> N >> K;
for (int i = 1; i <= N; i++)
{
cin >> a[i];
pref[i] = pref[i-1] + a[i];
}
for (int k = 1; k <= K; k++)
{
deque<int> cht;
for (int i = k + 1; i <= N; i++)
{
dp[i][1] = 0;
cht.push_back(i-1);
while (cht.size() > 2 && eval(cht[0], i) <= eval(cht[1], i) && check(cht[0], cht[1]))
cht.pop_front();
eval(cht[0], i);
dp[i][1] = eval(cht.front(), i);
last_split[i][k] = cht.front();
if (dp[i][1] < eval(cht.back(), i))
{
dp[i][1] = eval(cht.back(), i);
last_split[i][k] = cht.back();
}
}
for (int i = 1; i <= N; i++)
dp[i][0] = dp[i][1];
}
cout << dp[N][0] << '\n';
int i = N;
do
{
i = last_split[i][K];
cout << i << ' ';
} while (--K);
cout << '\n';
return 0;
}
