fk FFT February!

maybe this is some motivation for me to add editorial to H

Hello Codeforces Administration,

My Codeforces account ([email protected]) has been disabled, and I am receiving the message "User is disabled by administrators."

Could you please let me know the reason for this restriction and whether my account can be reviewed for reinstatement?

I would appreciate your guidance regarding the next steps.

Thank you for your time and assistance.

Looking forward!!!

Hello, since AI is not allowed this year, are we still allowed to run code to solve problems or has that also been disallowed? Also, will all 3 team members have to be present together during the exam if there is any proctoring to monitor AI usage?

about that part where it took you 2 years to get to specialist. what do you think would be a better path than the relentless headsmashing

Subject: Request to review disabled account (handle: kavyasanger) – skipped submissions

Hello Codeforces Team,

My account kavyasanger (email: [email protected]) has been disabled, and I would like to request a manual review.

In my recent contests, my submissions were marked as Skipped. I believe this was a mistake, for these reasons:

A flagged contest where I made no submission. In ED round 195, I did not submit any solution, yet that contest was also flagged on my account. I can’t see how a contest with no submissions could involve any violation. The problems I solved were simple. In the other flagged contests, the problems I solved were straightforward, low-rated ones. They use standard approaches, and I am fully capable of solving them on my own. Because the solutions are short and common, similar code from other participants is natural. I have not cheated. I did not share my code, and I did not use anyone else’s solutions. I write my solutions myself in C++, in a plain, simple style.

I have been an active participant for a long time and want to keep competing fairly. I’m happy to explain my approach to any of these problems, or to provide any other information you need.

Could you please review my account and, if the flags were a mistake, restore my account and my skipped submissions?

Thank you for your time.

Regards, Kavya Codeforces handle: kavyasanger Email: [email protected]

Subject: Request to review disabled account (handle: kavyasanger) – skipped submissions

Hello Codeforces Team,

My account kavyasanger (email: [email protected]) has been disabled, and I would like to request a manual review.

In my recent contests, my submissions were marked as Skipped. I believe this was a mistake, for these reasons:

A flagged contest where I made no submission. In ED round 195, I did not submit any solution, yet that contest was also flagged on my account. I can’t see how a contest with no submissions could involve any violation. The problems I solved were simple. In the other flagged contests, the problems I solved were straightforward, low-rated ones. They use standard approaches, and I am fully capable of solving them on my own. Because the solutions are short and common, similar code from other participants is natural. I have not cheated. I did not share my code, and I did not use anyone else’s solutions. I write my solutions myself in C++, in a plain, simple style.

I have been an active participant for a long time and want to keep competing fairly. I’m happy to explain my approach to any of these problems, or to provide any other information you need.

Could you please review my account and, if the flags were a mistake, restore my account and my skipped submissions?

Thank you for your time.

Regards, Kavya Codeforces handle: kavyasanger Email: [email protected]

uocj nhu a Khai

damn, best contest announcement and birthday invitation ever.

Good idea! Why didn't I expect to write like this.

Alternative sqrt-decomposition approach (i.e. $$$O(n\sqrt{n\log{n}})$$$, runs under 1.5s on my PC):

#include <bits/stdc++.h>
using namespace std;
#define int long long
#ifndef yoshi_likes_e4
#define endl '\n'
#endif
#define problem ""
#define multitest 0
#define debug(x) cerr << #x << " = " << x << endl;
void init() {}
vector<int> p;
constexpr int B = 1000;
template <bool REVERSE> struct Voda_SIXSEVEN {
  int32_t p[200000]{};
  int32_t pmx[200000];
  int cnt[200000];
  int sum[200000];
  int ps[200000];
  int32_t lzs[200];
  int32_t lza[200];
  bool clean[200];
  int N;
  void apply_lazy(int BLK) {
    if (lzs[BLK] != -1)
      for (int k = BLK * B; k < BLK * B + B; k++)
        p[k] = lzs[BLK];
    if (lza[BLK])
      for (int k = BLK * B; k < BLK * B + B; k++)
        p[k] += lza[BLK];
    lzs[BLK] = -1;
    lza[BLK] = 0;
  }
  void rebuild_block(const int BLK) {
    if (clean[BLK])
      return;
    clean[BLK] = 1;
    int32_t pr = 0;
    int ss = 0;
    for (int k = BLK * B; k < BLK * B + B; k++)
      ps[k] = (ss += p[k]), pmx[k] = (pr = max(pr, p[k]));
    int ct = 0, sm = 0;
    for (int k = BLK * B + B - 1; k >= BLK * B; k--) {
      cnt[k] = (ct += (p[k] == pmx[k]));
      sum[k] = (sm += (pmx[k] - p[k]));
    }
  }
  Voda_SIXSEVEN(vector<int> &Xp) {
    N = Xp.size();
    if (REVERSE)
      reverse(Xp.begin(), Xp.end());
    for (int i = 0; i < N; i++)
      p[i] = Xp[i];
    for (auto &i : lzs)
      i = -1;
  }
  void UPDATE_SINGLE(int i, int x) {
    if (REVERSE)
      i = N - i - 1;
    apply_lazy(i / B);
    p[i] = x;
    clean[i / B] = 0;
  }
  void UPDATE_RANGE(int l, int r, int x, bool SET = 0) {
    if (REVERSE)
      l = N - l - 1, r = N - r - 1, swap(l, r);
    apply_lazy(l / B);
    if (l / B != r / B)
      apply_lazy(r / B);
    if (SET) {
      if (l / B != r / B) {
        for (int i = l; i < (l / B + 1) * B; i++)
          p[i] = x;
        for (int i = l / B + 1; i < r / B; i++)
          lzs[i] = x, lza[i] = 0;
        for (int i = r / B * B; i <= r; i++)
          p[i] = x;
      } else {
        for (int i = l; i <= r; i++)
          p[i] = x;
      }
    } else {
      if (l / B != r / B) {
        for (int i = l; i < (l / B + 1) * B; i++)
          p[i] += x;
        for (int i = l / B + 1; i < r / B; i++)
          lza[i] += x;
        for (int i = r / B * B; i <= r; i++)
          p[i] += x;
      } else {
        for (int i = l; i <= r; i++)
          p[i] += x;
      }
    }
    clean[l / B] = clean[r / B] = 0;
  }
  int query(int l, int r, bool CNT = 0) {
    if (l > r)
      return 0;
    if (REVERSE)
      l = N - l - 1, r = N - r - 1, swap(l, r);
    int ans = 0;
    int32_t pm = 0;
    int ls = lzs[l / B];
    int la = lza[l / B];
    int rs = lzs[r / B];
    int ra = lza[r / B];
    if (l / B == r / B) {
      if (ls == -1) {
        if (CNT)
          for (int i = l; i <= r; i++)
            pm = max(pm, p[i]), ans += (pm == p[i]);
        else
          for (int i = l; i <= r; i++)
            pm = max(pm, p[i]), ans += pm - p[i];
      } else {
        return (CNT ? r - l + 1 : 0);
      }
    } else {
      if (ls == -1) {
        if (CNT)
          for (int i = l; i < (l / B + 1) * B; i++)
            pm = max(pm, p[i]), ans += (pm == p[i]);
        else
          for (int i = l; i < (l / B + 1) * B; i++)
            pm = max(pm, p[i]), ans += pm - p[i];

      } else
        ans += (CNT ? (l / B + 1) * B - l : 0), pm = ls;
      pm += la;
      if (CNT)
        for (int i = l / B + 1; i < r / B; i++) {
          rebuild_block(i);
          int32_t is = lzs[i];
          int32_t ia = lza[i];
          if (is != -1) {
            ans += (pm <= is + ia ? B : 0);
            pm = max(pm, is + ia);
          } else {
            int32_t *tmp = lower_bound(pmx + i * B, pmx + (i + 1) * B, pm - ia);
            if (tmp != pmx + (i + 1) * B)
              ans += cnt[tmp - pmx];
            pm = max(pm, pmx[i * B + B - 1] + ia);
          }
        }
      else {
        for (int i = l / B + 1; i < r / B; i++) {
          rebuild_block(i);
          int32_t is = lzs[i];
          int32_t ia = lza[i];
          if (is != -1) {
            ans += (pm <= is + ia ? 0 : B * (pm - is - ia));
            pm = max(pm, is + ia);
          } else {
            int32_t *tmp = lower_bound(pmx + i * B, pmx + (i + 1) * B, pm - ia);
            if (tmp != pmx + (i + 1) * B)
              ans += sum[tmp - pmx];
            if (tmp > pmx + i * B)
              ans += (long long)(pm - ia) * (tmp - pmx - i * B) -
                     ps[tmp - pmx - 1];
            pm = max(pm, pmx[i * B + B - 1] + ia);
          }
        }
      }
      if (rs == -1) {
        pm -= ra;
        if (CNT)
          for (int i = r / B * B; i <= r; i++)
            pm = max(pm, p[i]), ans += (pm == p[i]);
        else
          for (int i = r / B * B; i <= r; i++)
            pm = max(pm, p[i]), ans += pm - p[i];
      } else
        ans +=
            (CNT ? (pm <= rs + ra ? r - r / B * B + 1 : 0)
                 : (pm <= rs + ra ? 0 : (r - r / B * B + 1) * (pm - rs - ra)));
    }
    return ans;
  }
};
int ST[800000];
int lzs[800000];
int lza[800000];
void down(int i) {
  if (lzs[i] != -1) {
    ST[i * 2 + 1] = lzs[i];
    ST[i * 2 + 2] = lzs[i];
    lzs[i * 2 + 1] = lzs[i];
    lzs[i * 2 + 2] = lzs[i];
    lza[i * 2 + 1] = 0;
    lza[i * 2 + 2] = 0;
    lzs[i] = -1;
  }
  ST[i * 2 + 1] += lza[i];
  ST[i * 2 + 2] += lza[i];
  lza[i * 2 + 1] += lza[i];
  lza[i * 2 + 2] += lza[i];
  lza[i] = 0;
}
template <bool SET = 0> void updateR(int i, int u, int v, int l, int r, int x) {
  if (v < l || r < u)
    return;
  if (l <= u && v <= r) {
    if (SET) {
      ST[i] = x;
      lzs[i] = x;
      lza[i] = 0;
    } else {
      ST[i] += x;
      lza[i] += x;
    }
  } else {
    int mid = (u + v) / 2;
    down(i);
    updateR<SET>(i * 2 + 1, u, mid, l, r, x);
    updateR<SET>(i * 2 + 2, mid + 1, v, l, r, x);
    ST[i] = max(ST[i * 2 + 1], ST[i * 2 + 2]);
  }
}
void update(int i, int u, int v, int p, int x) {
  if (u == v)
    ST[i] = x;
  else {
    int mid = (u + v) / 2;
    down(i);
    if (p <= mid)
      update(i * 2 + 1, u, mid, p, x);
    else
      update(i * 2 + 2, mid + 1, v, p, x);
    ST[i] = max(ST[i * 2 + 1], ST[i * 2 + 2]);
  }
}
int query(int root, int l, int r, int lowerbound, int k) {
  if (ST[root] < k)
    return -1;
  if (r < lowerbound)
    return -1;
  if (l == r)
    return l;
  down(root);
  int mid = (l + r) / 2;
  int res = -1;
  if (ST[root * 2 + 1] >= k)
    res = query(root * 2 + 1, l, mid, lowerbound, k);
  if (res == -1)
    res = query(root * 2 + 2, mid + 1, r, lowerbound, k);
  return res;
}
int get(int i, int u, int v, int l, int r) {
  if (v < l || r < u)
    return 0;
  if (l <= u && v <= r)
    return ST[i];
  else {
    int mid = (u + v) / 2;
    down(i);
    return max(get(i * 2 + 1, u, mid, l, r), get(i * 2 + 2, mid + 1, v, l, r));
  }
}
void build(int i, int l, int r) {
  lzs[i] = -1;
  if (l == r) {
    ST[i] = p[l];
  } else {
    int mid = (l + r) / 2;
    build(i * 2 + 1, l, mid);
    build(i * 2 + 2, mid + 1, r);
    ST[i] = max(ST[i * 2 + 1], ST[i * 2 + 2]);
  }
}
void Yoshi() {
  int n, q;
  cin >> n >> q;
  p.resize(n);
  for (auto &i : p)
    cin >> i;
  build(0, 0, n - 1);
  Voda_SIXSEVEN<0> VODA_1(p);
  Voda_SIXSEVEN<1> VODA_2(p);
  while (q--) {
    int t;
    cin >> t;
    if (t == 1) {
      int u, v;
      cin >> u >> v;
      VODA_1.UPDATE_SINGLE(--u, v);
      VODA_2.UPDATE_SINGLE(u, v);
      update(0, 0, n - 1, u, v);
    } else if (t == 2) {
      int u, v, x;
      cin >> u >> v >> x;
      VODA_1.UPDATE_RANGE(--u, --v, x, 1);
      VODA_2.UPDATE_RANGE(u, v, x, 1);
      updateR<1>(0, 0, n - 1, u, v, x);
    } else if (t == 3) {
      int u, v, x;
      cin >> u >> v >> x;
      VODA_1.UPDATE_RANGE(--u, --v, x);
      VODA_2.UPDATE_RANGE(u, v, x);
      updateR(0, 0, n - 1, u, v, x);
    } else if (t == 4) {
      int u, v;
      cin >> u >> v;
      int A = get(0, 0, n - 1, --u, --v);
      int pos = query(0, 0, n - 1, u, A);
      cout << VODA_1.query(u, pos - 1, 1) + VODA_2.query(pos, v, 1) << '\n';
    } else {
      int u, v;
      cin >> u >> v;
      int A = get(0, 0, n - 1, --u, --v);
      int pos = query(0, 0, n - 1, u, A);
      cout << VODA_1.query(u, pos - 1) + VODA_2.query(pos, v) << '\n';
    }
  }
}
signed main() {
#ifndef yoshi_likes_e4
  ios::sync_with_stdio(0);
  cin.tie(0);
  cout.tie(0);
  if (fopen(problem ".inp", "r")) {
    freopen(problem ".inp", "r", stdin);
    freopen(problem ".out", "w", stdout);
  }
#endif
  init();
  int t = 1;
#if multitest
  cin >> t;
#endif
  while (t--)
    Yoshi();
}

(I realized this probably is a lot less elegant than the solution above.)

Yes, this idea is more easy to understand. I just do it like this.

You might as well tell your friend to refer you with their link so they get the reward instead of some person on codeforces

I hope to provide translation options for problems on the programming competition page, similar to the interface of a math competition, which would be more convenient and further constrain the use of AI by participants in the rules.

referral links on cf now? next post: free macbook if you sign up with my link >_>

KAN MikeMirzayanov Vladosiya geranazavr555 Any updates on np254.

Please one look can help it out.

brotherman, six seven eight

or else.

what, no cake? fine i'm registering xD

orz chi biet uoc

Why is G rated *2100? The more I work on it, the weirder it feels.

happy early birthday!!!

As a former cwOIer, Getaway_Car orzzzzz!

Thanks for the detailed feedback, and sorry for the trouble.

1- Agreed, that verdict was misleading. We're fixing it so time-limit kills on interactive problems show as TLE.

2- You guessed right: "Run on samples" fed the sample input directly and didn't run the interactor, so it can't judge interactive problems correctly.

We'll fix that too. We'll do our best to make sure these issues don't happen again in upcoming contests.

You're right, and sorry for the confusion. On interactive problems a solution killed by the time limit was shown as WA/RE because the interactor's verdict was used instead.

We're fixing it so time-limit kills always show as TLE, and I promise this won't happen again in upcoming contests.

The problem statement says "It is guaranteed that each column of the input grid contains at least one frog."

Is AtCoder necessary while practicing CF daily?? Please help, I am a beginner:)

About problem G:

I've seen there is an unofficial editorial that talks about a solution using Edmonds' Blossom algorithm. I don't understand how the author managed to make it run in 96ms, because I did exactly that on contest but my solution TLE-ed on 2 test cases. chuanyu.fan would you mind sharing your code or explaining any kind of pruning you did to make it run that fast? Thanks.

It can be proven that we only need to check (first and last maximum element) and first and last element in general, we don't need to check the rest tho.

apparently the HXH narrator didn't like it

I liked the problems (at least ABCDE, because I didn't read the other ones), but I have a couple of suggestions based on my experience today:

1. I feel that today's problems D and E had very strict TL and ML. So basically in problem D you can't store more than two [N][N] arrays, which one can argue is a skill this problem checks among others, but I'm not sure those details were intentionally left for the participants to deal with. Also, 1 second is too tight for both of those problems because they admit many volatile implementations. Again, one could say it was a skill issue, but for me it is kind of like why.

2. As pointed out above in the comments, I believe that TLE in interactive problems may lead to weird verdicts such as RTE because my solution was trying to write to empty stream or whatever. Very misleading anyway.

3. I don't know how "run on samples" button works (does it check the output byte-by-byte? Does it ignore whitespaces? Does it run the checker?), but my now-passing solution for E doesn't pass the samples according the "run on samples" button. My guess is that it tries to just feed the full "sample input" to my solution and check its output, whatever the word "check" means here, ignoring what queries my solution even does.

Anyway, my screencast (where I talk but in Russian) for those who wants to witness my struggles

Ur other account is?

I'm confused on problem B. I had a much more complicated algorithm, and I think this test case requires k = 2.

1

6

fofofo

fooooo

in the top row, no matter where you move the two right frogs, it creates an imbalance. You can cancel one frog on the top with the bottom, but that still leaves two unsolvable frogs at the top. Can someone explain how k = 0 or 1 works with this test case?

forgetting FFT february

D can be done with Ternary Search. It's like killing an ant with a bomb, but who would say killing an ant with a bomb isn't fun?

Will all topics be taught in the order the blog mentioned?

What about Funny FFT February. (FFF)

We're really sorry about how long submissions sat in the queue today and for how much that took away from the contest for you. We've rejudged all submissions and we're fixing the cause of the "Judging failed" errors.

We'll do our best to make sure this doesn't happen again.

Thank you to the authors and testers for organizing this contest! I really enjoyed participating. Great job!

first contest let me fumble got cooked by A

3

Auto comment: topic has been updated by Artix (previous revision, new revision, compare).

I really liked the 2278E - Swap and Negate. The transformation idea $$$b_i=(−1)^{i−1}a_i$$$ is simply generous! I wish to practice more of this kind. Could anyone find me some similar problems?

Hello again, Codeforces community! After some thought, and after not being able to understand the editorial for Div 1B/2D (Robot championship), I have come up with this following solution:

https://codeforces.com/contest/2273/submission/397045971

I have included a semi-formal analysis of the problem in my comments, in hopes of making the solution easier to understand for other people. Please check it out if you are interested, and you can reply to this comment if you have any doubts.Thank you!

Oops I failed 13 hours in 397045309

Auto comment: topic has been updated by Jayanth25 (previous revision, new revision, compare).

finally a contest where i could solve stuff :D

I think that some submissions for Problem E that should receive TLE are being judged as RE instead.

Killed
Wrong answer: contestant output ended before a final answer

why it got so much downvote?

That was my first time participating in a div 3 after becoming an expert. Stop crying now

you are feeling like vinicious isn;t it, 2 years ago he was the best footballer, now he is a flop. So don't be like vinicious

Being expert and asking this question lol..Ask this one to LLM too

It's hard to give a concrete advice because every one is different and the burnout cause differs from one to another but it might be helpful to remember the goal that made you start programming in the first place and try to achieve it again slowly without any pressure, or the things that you had fun doing like projects, problems, reading about CS topics etc..

I am sad that these issues with the grading of submissions keep happening. For me, it reduces the fun a lot, because I am just feeling frustrated and cannot fully focus on the next problems because of it. After solving C and D, I tried for a bit on E, but already was annoyed enough to just stop.

The website does not make it better (it was hard to see the global announcements, no pop-up or anything,)

Auto comment: topic has been updated by Artix (previous revision, new revision, compare).

Bad quality problems, too much heuristic problems. Allowing AI for some tasks is a bad idea because people used it for the other tasks.
I'm pretty sure among the top 1000 teams there were at least 900 cheaters.
They fixed some problems mid contest.
I really didn't enjoy it :(

Lunar Giant gave me nightmares

its free

peak

Is this free or paid

How many problems is an Expert expected to solve?

You've registered unrated.

Hello, I participated in Codeforces Round 1131 (Div. 3) with the handle Radin1.

My rating before the contest was 1333, and I solved 5 problems. I registered as Rated, but my contest history shows no rank or rating change, while other participants have received their rating changes.

Could you please check why my participation was not rated?

Thank you.

Sorry for the necro-post.. (and my bad English also)

The time complexity is $$$O(sqrt(K))$$$, with K = (number of edges at the beginning) + (number of edges added from calls of add_friend(u, v)) (i.e. $$$K \lt = M+Q$$$). This can be referred from these two lines below:

if (adj[u].size() > adj[v].size()) swap(u, v);
adj[u].push_back(v);

Denote deg[u] = degree of u.

Imagine $$$deg[u] = 0$$$ (i.e. $$$adj[u]$$$ is empty), and you want to make u the one with smaller degree when connecting with $$$v_1, v_2, ..., v_i$$$ in order, which means making edge $$$u \rightarrow v_1, u \rightarrow v_2, ... , u \rightarrow v_i$$$.

• When making edge $$$u \rightarrow v_1$$$: $$$deg[u] = 0$$$, thus $$$deg[v_1] \gt = 0$$$ must satisfies.
• When making edge $$$u \rightarrow v_2$$$: $$$deg[u] = 1$$$ ($$$u$$$ is connected with $$$v_1$$$ before), thus $$$deg[v_2] \gt = 1$$$ must satisfies.
• ...
• When making edge $$$u \rightarrow v_i$$$: $$$deg[u] = i-1$$$ ($$$u$$$ is connected with $$$v1, v2, ..., v_{i-1}$$$ before), thus $$$deg[v_i] \gt = i-1$$$ must satisfies.

So, to make edges $$$u \rightarrow v_1, u \rightarrow v_2, ..., u \rightarrow v_i$$$, these much satisfy:

• $$$deg[v_1] \gt = 0$$$
• $$$deg[v_2] \gt = 1$$$
• ...
• $$$deg[v_i] \gt = i-1$$$

And to make those satified, $$$0 + 1 + ... + (i-1)$$$ edges in total must be made before to $$$v_1, v_2, ..., v_i$$$.

Thus, it takes $$$i + (0 + 1 + ... + (i-1)) = \frac{(i-1)i}{2}$$$ edges in total to finally make edges from $$$u$$$ to $$$v_1, v_2, ..., v_i$$$.

But only up to $$$K$$$ edges can be made from the input. ($$$K$$$ is denoted at the beginning)

So $$$\frac{(i-1)i}{2} \lt = K$$$

$$$\rightarrow i \lt = sqrt(2K)$$$.

$$$\rightarrow adj[u]$$$ can only stores up to $$$\sqrt{2K}$$$ other nodes..!

Now, in detail to time complexity of each functions:

• $$$offline()/online()$$$ and $$$cal()$$$ takes $$$O(\sqrt{2K})$$$ (each function has a loop through $$$adj[u]$$$).
• $$$add()$$$ takes $$$O(1)$$$.
• $$$del()$$$ takes $$$O(\sqrt{2K})$$$ (has at most two loops through $$$adj[]$$$).

$$$\rightarrow$$$ Time complexity: $$$O(Q\sqrt{2K}))$$$.

It's still quite suprising that someone could think of a solution this unique. I'm wondering if there are some sort of patterns or ways to think of this kind of solution..

longest queue ever seen

The idea I had for E was a bit different (and I think similar to what others have done), but the editorial solution is quite pretty, I have to admit

Oh ok I think you just opened my mind

Coding.

Anyone had same thoughts ?

At contest time, there were bad test cases, after contest there is hacking phase right, if people think the test cases are not good enough they'll add better test cases, so if they can find hack(a testcase) which suppose doesn't work with your code they'll add that test case and re run everyone's code kind of

hi lol

Coding.

Ignoring the absolutely ridiculous amount of cheating that happened (as expected), I think quite a high percentage of the problems were hard, but I enjoyed some of them ngl.

it was good ig, the one who deserves execution is the author who used "rock scissor paper" instead of "rock paper scissors"

Your idea was to calculate the distance of all numbers from the minimum to the maximum. You asked about an error in the code – take a look at where you’re calculating the distance and operations from all the numbers to the minimum; you’re multiplying by i + 1. Presumably, your logic was that you would cover the elements in this prefix, but the trick is that you can cover the suffix up to i + 1 and get a lower score; I wrote the exact same algorithm at first, but then came to the same realisation.

    int ops = a[n-1] - a[0];
    int total = 0;
    for(int i = 1; i < n; i++) {
        int gpa = a[i] - a[i-1];
        int w = min(i, (n - i));
        total += gpa * w;
    }

btw your idea with diff arr is right

I beg your pardon — I was wrong. I misunderstood the complexity of nth_element() and my earlier reasoning was off. You were right to call it out. Sorry about that and thank you for your patience.

fr me too

the random $20 draw is the only prize i can realistically aim for xD

bro i didnt get a notification and couldnt farm contribution wtf is this :(

*public execution.

something...

I felt that G was easier than F

AI edited

Difficulty jump from E to F was insane

Yea I eventually got that, but my main concern was why and how would it even pass the pretests in the first place, since this thing is not even input dependent. Went from a generational run (like a +100 predicted delta) to a -42 finally (because of probably the most silly mistake).

Reaching expert is legit imposs :wilted_rose

I thought bro was gonna share some gambits and ideas with moves

AI generated?

sth meaning?

I experienced the same thing, maybe they have rerun those tests for problem you've solved and now it's WA. Or maybe you just imagined it?

Bro I was writing code for problem 'A' for 1 HOUR :sob:, which means — more practice!