Any form of streaming that would be worth watching would also give away some information about problems, which is not desirable. I don't think there's a way to make streaming work for public rated rounds without ruining the viewing experience. Screencasts work fine, and if you really want to watch live there's contests like Blitz Cup.

First time to see Problem A divided into subtasks:)

Delayforces

As a tester, ubfeiaeiuabfeiabfieauifenhliuofajs

read the comment again and apologise when you realise the mistakes in your comprehension

As conqueror of mishai, even I cannot conquer how great this round will be

Using ai to write a request for a second chance for using ai is crazy work

I am currently finishing the editorial, it will be posted in about 2-3 hours. Sorry for the delay.

This game took 5 years off my life i saw something black. The last time I saw something black I was walking by the police station

Don't try to play the rating. Your goal should be to get better at cp, not to get a bigger number on your profile. Read this: https://en.wikipedia.org/wiki/Goodhart's_law .

As a

VIP

tester, I'm very important.

Update: the contest will contain 7 problems.

I can't be the only one who just assumed that $$$\sum{n} \leq 5000$$$ in E, and spent half an hour trying to optimise my $$$O(t \cdot 5000^2)$$$ solution T_T

As a tester, I lowkey tested two weeks after getting sent the invitation :)

$$$C$$$ was harder than everything else.

i guess we are in NXN(No XOR November) now

As a tester, I ❤️ interactive and communication

yes i agree screencasts serves same purpose

usually if someone has reserved power to also explain their solutions they’re not taking the contest seriously which removes anything exciting about watching a live performance

Well I got around +80 in fact. And another fun fact: I posted 2 bullshit blogs asking for helps for a problem (which both get around -10 lmao). Then, I posted a blog about cheaters, which, my intention is not to farm contribution but to expose them. I ended up getting a lots upvote and got a j*b as a (contribution) farmer lol

If you want to get contribution, contribute to the community, not shit posting like this. And if you ask why your blog saying hi to the community gets downvote; we already saw so many blogs like that and it's annoying.

I genuinely found D so much easier than C but then again I hate any bitwise problems

so this is also discontinued then:( my dream of getting that tshirt remained a dream

Your brain really did get TLE

Destroy Data Structures December implies that DSU, a data structure, will be destroyed as well

how about non stop number theory november?

I don't know whether I'm stupid or not, but I genuinely believe that C is much, much harder than D and E. Yet somehow, C got significantly more ACs. I'm honestly very confused.

Here is my solution, which took me an hour of thinking and brute-forcing to arrive at this observation. .-.

By doubling the i-th element, its value becomes 2i, which changes k to k XOR (i XOR 2i).

Now, we can observe that i XOR 2*i consists of exactly two set bits.

For convenience, we can choose elements of the form 2^k, since doubling such an element simply shifts its set bit one position to the left, resulting in two adjacent set bits in the XOR.

With this observation, we can process the bits from the highest to the lowest and use the operation described above to construct the desired k.

Wait what?

For problem G, the $$$O(n2^n + qn\log t)$$$ solution works out to only about 2e8 operations, yet with a 7-second time limit, it got TLE!?

For problem D, the $$$O(n\log n)$$$ LCA solution got TLE — and $$$n$$$ is only $$$3\times 10^5$$$, even with a 2-second time limit!?

nooooooo!!!!!

send my rating back T_T.

Sooooo 2 xor problems?

Someone might be using the judging machines to mine Bitcoin.

This submission is one hundred percent AI-generated

$$$0$$$ IQ solution for problem C: you can just toggle a random number until you get a big enough xor.

See 397353027

Unless someone can hack it :D

Screencast with commentary

Both speed up DP transitions that would normally be O(N^2), but they work in different ways.

D&C relies on the fact that the best split point only moves right as you go forward. So you solve the middle position first by checking all candidates, then solve the left half using only candidates to the left of that best point, and the right half using only candidates to the right. This brings each layer down to O(N log N). It works for layered DPs like "split an array into K groups" where the cost is some function of a range.

Convex Hull Trick relies on the formula itself. If the transition can be rewritten so each old state is a straight line, then computing a new state just means finding the lowest (or highest) line at a given x. You only keep the lines that matter, and with sorted slopes and queries it runs in O(N).

nice...this is basically to humble the guy who's higher rated than u but u know he's sus

what is this score dist

5 seconds away from solving G ;_;

This was a competition testing how fast you can code

I think I spent the most time thinking on C :)

397395985 ok Bro

You are right, I knew somewhat ditching all the English lessons in school would hit me :)

Interesting scoring distribution:)

Am I the only feeling that time constraint on D is too low ?

I was lacking time so I went straight to recursion (in python), and I got TLE (even after using bootstrap)

I know recursion is slow in python, but see, I increased the time constraint by moving it to a private contest, and the runtime was 2.8s.

Please tell me if I should ask for rejudging :)

another argument against: streamer might get help from chat (e.g. your bug is on line XYZ) — seems similar to the rule: You are forbidden from using AI to diagnose or resolve errors based on system feedback (e.g., after receiving a rejected verdict like "Runtime error on test 1" or similar, you are prohibited from asking AI-based systems to help you fix the issue).

So many times when I'm watching a video of someone doing a sudoku for example, I see something that they miss

thanks for the solutions sir!

F can be solved just using lazy segment trees 397403664.

Is C always solvable by doubling only 1 or 2 numbers? I just doubled N and if the xor wasn't high enough, I iterated from 1 to N-1 and tried doubling one of those.

Seems like that worked?

Damn the editorial approach for D is so much better i used LCA + difference tree (idk the name i just wrote difference array for tree) the algo itself was simple just a lot of preprocessing the color tree thing is way more beautiful

I have a greedy approach for Problem C.

Greedy: Process numbers from large to small

1. Key observation The bitwise OR of all integers from 1 to n is always a number consisting entirely of consecutive 1-bits. Let L be the highest set bit of n. Then k = (1 << (L + 1)) — 1, and 2*k has all bits from position 1 to L+1 set to 1 (using 0-based indexing). Also, every chosen value is at most 2*n, so the final XOR is smaller than 2^(L+2). Therefore, we only need to make bits 1 through L+1 of the final XOR equal to 1. The result will automatically be either 2*k or 2*k+1.

2. Greedy strategy Process i from n down to 1, maintaining tmp, the XOR of all processed elements. For each i:

• If replacing i with 2*i makes tmp larger, choose 2*i.
• Otherwise, keep i unchanged.

In code:

if ((tmp ^ (2 * i)) > tmp) {
    p[i] = 1;
    tmp ^= (2 * i);
} else {
    tmp ^= i;
}

1. Why does this work? Group the numbers according to their highest set bit. For a group containing numbers in the range [2^q, 2^(q+1)-1], doubling any number in this group makes its highest set bit exactly q+1. The comparison tmp ^ (2*i) > tmp has a simple meaning:

• If bit q+1 of tmp is 0, doubling i makes the XOR larger, so we double it and turn this bit into 1.
• If bit q+1 is already 1, doubling i makes the XOR smaller, so we keep i unchanged. Since we process numbers in descending order, the first number in each group will set bit q+1 to 1 if necessary. All remaining numbers in that group are kept unchanged, so they cannot affect this bit. After finishing a group, its corresponding bit is guaranteed to be 1. Moreover, all numbers in subsequent groups are smaller. Neither their original values nor their doubled values can affect any previously fixed higher bit. Thus, we can greedily fix all required bits from high to low, guaranteeing that the final XOR is at least 2*k.

Implementation:

void solve() {
    int n;
    cin >> n;

    vector<int> p(n + 1, 0);
    long long tmp = 0;

    for (int i = n; i >= 1; --i) {
        long long x = 2LL * i;

        if ((tmp ^ x) > tmp) {
            p[i] = 1;
            tmp ^= x;
        } else {
            tmp ^= i;
        }
    }

    for (int i = 1; i <= n; ++i) {
        cout << p[i] << " \n"[i == n];
    }
}

Here, p[i] = 1 means choosing 2*i; otherwise, we choose i.

You make a very good point. Actually, I've been thinking about this issue too. For people who need it, watching others' contest livestreams is also a good opportunity to broaden their thinking and train, just like watching others livestream games.

The main reason this isn't allowed is the potential cheating it brings, but the rise of AI makes cheating easy. Even more, current online contests' prevention of AI relies almost entirely on morals. Seen this way, allowing livestreaming seems like a fairly reasonable choice, because watching a livestream isn't even as convenient as AI.

Overall, I support allowing livestreaming.

https://clashforces.duckdns.org/

Similarly, no one wants to watch a chess player plays chess in silence. So, why is GothamChess still famous? Because he explained his thought, make jokes. So I would say that cf contest stream would be interesting too if you make it interesting.

"many" newbies have quite a lot of upvotes, for example duhminh had 50+ contribution when he was newbie,

so js stop larping and post good+helpful content

getting downvotes is your own fault, not ratism!

Noooo why 10 min delay

Why it is delayed?

nice problems, but too tight TL

LOL that's what I think about and seriously, C and D should be swapped, but both are not really a hard problem. Just need some observation.

As a tester, this is a 2x tester comment combo

That CSES problem is more generic; you can solve problem D with that technique, but problem D also allows a simpler ad hoc solution that uses a fairly straightforward O(n) algorithm.

D is very basic, while C requires more thinking though you don't need to know hard algos

get my rating back!

7 problems -> 5 problems

rank 40 -> rank 500

Imagine A is an interactive problem ::skull

As a tester, I haven't tested any other rounds except this for the last 9 months

i did not cheat you are a cheater if you don't have the ability so why blaming others?

Bro please don't just flood recent actions with this nonsense. First of all what on earth do you mean by competitive programming documentation ??? Then the timestamps are enough of a proof of cheating. Since, I am unrated myself I can't really review the code, but the timestamps are enough for me to tell that you have cheated. How on earth can you just solve C and D within like 10 minutes and then not solve A it is clearly very stupid that you created another account to restore the original account, since I believe this account could be banned as per one person one account rule on cf no?

Thanks for sharing your concerns. I understand why the timestamps and problem-solving order may look unusual. However, I genuinely solved the problems independently using my own code and did not copy or use AI-generated solutions. By documentation, I meant official C++ and STL references. My second account was only used to submit an appeal because I couldn't access my original account, not to bypass any restriction. I respectfully request a fair review of my submissions by the Codeforces team.

idk about others but me and my bros train like this: topic->contest->editorial->upsolving and it takes around 8 hours with breaks in sum,I found it very useful though

An opportunity to provide any necessary verification.

ok give us the recording you redo the problems on your own

Thank you for giving me an opportunity to verify my work. I'm happy to solve the problems again independently and record the process to demonstrate my problem-solving approach. Please let me know which problems you'd like me to attempt and your preferred recording format. I'm willing to cooperate fully.

all problems in youtube or whatever i'll watch

It depends on you,during my summer break i was doing 5 problems daily,something like 2-800,1-1000,2-1300,1400,1500),(5hrs daily),and kept learning from every question which i couldn't solve ,mostly doing questions of desired rating from problemset(recent ones).(I ain't top just sharing my experience)Also,try to maintain a streak which motivates you to continue further(when classes start),so that daily one problem becomes your habit which helps in your long run.

is problem rating distribution for educational rounds not announced beforehand?

Auto comment: topic has been translated by adramabama233 (original revision, translated revision, compare)

problem E...how am i even supposed to start thinking in the direction of making a new array

I was stuck just trying to find if i can somehow do something with even and odd positions the whole time.

Please give any advice.

ok

in Div3,4 and educational rounds, the contest format is ICPC alike, which is similar to each problem having 1 point

Sure, I'm willing to attempt all the problems independently. Please share the original problem links or specify which problems you want me to solve. I'll avoid watching any solutions beforehand and record my problem-solving process transparently.

May I ask why you stopped doing CP last year? I mean, I'd like to understand what made you quit, so I can (maybe) learn from your experience and try to think of a way overcome if I ever face them.

prepare your video, you clearly cheated, don't cry on it....

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

Convex hull trick (and LiChao)
DnC trick
They are completely different.

bruh

I mean the code is one hundred percent AI-generated

1. If you only used the new account for an appeal, why did you take a contest on it and use a completely different handle and photo.
2. The only counterproof you said is "I'm not cheating". You didn't even bring up smth like "yeah i code wrong" or "yeah i use comments to make it clearer for me" No you just said Nah I'm not cheating Trust me

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

Lets define the time will take is X so X = X+1 so the time has been passed is equal to 1second after the time has been passed so that means the actual number is the far past and far future (+-972597246592745672465742967562479679524796567492579626562475967952697567269576297567624965972976529765265295627495629756972465247959247659245... Equal to Pi digits!) so we stay Forever and travel through time

Bro fym top 80 Team isn't qualifying for Amritapuri but fucking random ass Teams qualify just because they are best in their college

Agree,

There should be some min no of problems that should be solved to get into regionals,

This no should not be fixed, depending upon the contest difficulty it should change.

This 3 questions were very basic, so they should have considered teams who have solved 3 or more problems.

Love to know other's view : )

Dạ thầy ơi cho em hỏi mình được dùng hàm startswith không thầy

i think it is super surprising to me when i know i got admission in vss.

Maybe they explain their thoughts along the way, which is interesting for many people.

Although I think publishing the screencast after the contest would serve the same purpose without the downsides.

Even watching umnik screencasts where he just listens to music is quite fun, I usually skip to the parts where he gets a wa due to some weird bug and rages afterwards, its one of my hobbies.

perhaps it`s worth at least reading some articles about these algorithms before asking questions about them

I mean they both have the optimal point is non-decreasing?

the catalog tap in codeforces has a lot of blogs full of advices that might help you, also keep in mind that some of these advices are 10+ years old but they should still be useful.

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

really frustrating to see the deserving candidates being left out while people get selected due to quotas, just ruins the day >:

The code for problem G doesn't implement the corresponding solution.This may cause readers to be confused.

Thanks

thanks

How many levels it has? Only level 0, or something more? And also... what is the rules of the game? What I should do?

I just wonder sometime how you guys can come up with these solution with so optimal logic. I dont even know if I can ever go to that lvl lol