#include <iostream>
#include <cstdio>
#include <cstdlib>
#include <algorithm>
#include <cmath>
#include <vector>
#include <set>
#include <map>
#include <unordered_set>
#include <unordered_map>
#include <queue>
#include <ctime>
#include <cassert>
#include <complex>
#include <string>
#include <cstring>
#include <chrono>
#include <random>
#include <bitset>
#include <fstream>
using namespace std;
#define int long long
#define ii pair <int, int>
#define app push_back
#define all(a) a.begin(), a.end()
#define bp __builtin_popcountll
#define ll long long
#define mp make_pair
#define x first
#define y second
#define Time (double)clock()/CLOCKS_PER_SEC
#define debug(x) std::cerr << #x << ": " << x << '\n';
#define FOR(i, n) for (int i = 0; i < n; ++i)
#define FORN(i, n) for (int i = 1; i <= n; ++i)
#define pb push_back
#define trav(a, x) for (auto& a : x)
using vi = vector<int>;
template <typename T>
std::istream& operator >>(std::istream& input, std::vector<T>& data)
{
for (T& x : data)
input >> x;
return input;
}
template <typename T>
std::ostream& operator <<(std::ostream& output, const pair <T, T> & data)
{
output << "(" << data.x << "," << data.y << ")";
return output;
}
template <typename T>
std::ostream& operator <<(std::ostream& output, const std::vector<T>& data)
{
for (const T& x : data)
output << x << " ";
return output;
}
ll div_up(ll a, ll b) { return a/b+((a^b)>0&&a%b); } // divide a by b rounded up
ll div_down(ll a, ll b) { return a/b-((a^b)<0&&a%b); } // divide a by b rounded down
ll math_mod(ll a, ll b) { return a - b * div_down(a, b); }
#define tcT template<class T
#define tcTU tcT, class U
tcT> using V = vector<T>;
tcT> void re(V<T>& x) {
trav(a, x)
cin >> a;
}
tcT> bool ckmin(T& a, const T& b) {
return b < a ? a = b, 1 : 0;
} // set a = min(a,b)
tcT> bool ckmax(T& a, const T& b) {
return a < b ? a = b, 1 : 0;
}
ll gcd(ll a, ll b) {
while (b) {
tie(a, b) = mp(b, a % b);
}
return a;
}
struct Line {
ll k, b, i;
Line() : k(), b(), i() {}
Line (ll _k, ll _b, ll _i) : k(_k), b(_b), i(_i) {}
ll getVal(ll x) {
return k * x + b;
}
};
struct Hull {
vector<Line> lines;
vector<ll> borders;
Hull() : lines(), borders() {}
void addLine(Line L) {
while(!lines.empty()) {
if (lines.back().getVal(borders.back()) >= L.getVal(borders.back())) {
lines.pop_back();
borders.pop_back();
} else break;
}
if (lines.empty()) {
lines.push_back(L);
borders.push_back(0LL); //leftmost query
return;
}
if (lines.back().k <= L.k) return;
ll x = div_up(L.b - lines.back().b, lines.back().k - L.k); //must work for negative!
lines.push_back(L);
borders.push_back(x);
}
int getMinVal(ll x) {
int pos = upper_bound(borders.begin(), borders.end(), x) - borders.begin();
if (pos == 0) return -1;
pos--;
return lines[pos].i;
}
};
const int N = 1e5+7, K = 207, INF = 1e18;
signed par[K][N];
int dp1[N], dp2[N];
signed main() {
#ifdef LOCAL
#else
#define endl '\n'
ios_base::sync_with_stdio(0); cin.tie(0);
#endif
int n, k;
cin >> n >> k;
k++;
vi a(n);
cin >> a;
vi p(n + 1);
FOR (i, n) {
p[i + 1] = p[i] + a[i];
}
FOR (i, N) {
dp1[i] = INF;
}
dp1[0] = 0;
FORN (t, k) {
Hull h;
FOR (i, n + 1) {
int j = h.getMinVal(p[i] * 2);
if (j != -1) {
dp2[i] = dp1[j] + (p[i] - p[j]) * (p[i] - p[j]);
par[t][i] = j;
}
else {
dp2[i] = INF;
}
if (dp1[i] != INF) {
h.addLine(Line(-p[i], dp1[i] + p[i] * p[i], i));
}
}
FOR (i, N) {
dp1[i] = dp2[i];
}
}
cout << (p[n] * p[n] - dp1[n]) / 2 << endl;
vi cut;
int i = n;
for (int t = k; t; --t) {
cut.app(i);
i = par[t][i];
}
reverse(all(cut));
cut.pop_back();
cout << cut << endl;
}
