Hello, Codeforces!
We are pleased to announce Codeforces Global Round 26! Thanks to XTX Markets for supporting the initiative! In 2024, we are holding 4 such rounds. The series results will take into account the best 3 participations out of 4.
On 09.06.2024 17:35 (Московское время) we will host Codeforces Global Round 26. 
Codeforces Global Round 26 marks the second round in the 2024 series of Codeforces Global Rounds. These rounds are open and rated for everyone.
The prizes for this round are as follows:
- The top 30 participants will receive a t-shirt.
- 20 t-shirts will be randomly distributed among participants ranked between 31 and 500, inclusive.
The prizes for the 4-round series in 2024:
- In each round, the top-100 participants get points according to the table.
- A participant's final score will be the sum of the points they earned in their 3 highest-placing rounds.
- The top 20 participants across the series will receive sweatshirts and placement certificates.
We extend our gratitude to XTX Markets for supporting the global rounds initiative in 2024!
The problems were written and prepared by null_awe, flamestorm, le0n, and buffering.
We would also like to thank:
- Our amazing coordinator TheScrasse for his quick and thorough review.
- Our last-minute Russian translator Alexdat2000.
- Our testers bronze_coder, conqueror_of_tourist, Ra16bit, Flamire, Esomer, PurpleCrayon, Ritwin, awesomeguy856, Tekor, Sugar_fan, htetgm, sum, lunchbox, Ste, liympanda, CSQ31, oursaco, Edlue, eggag32, Tsotne533, serichaoo, Mononoke001, redpanda, janY_, tiagodias, pablopalou, S2speed, cry, Marckess, madlogic for their rigorous feedback.
Round Information:
- Duration: $$$3$$$ hours
- Number of problems: $$$8$$$ problems
- Score distribution:
$$$500 - 750 - (750 + 1250) - 2500 - 3000 - 3000 - 4000 - 5000$$$
We eagerly anticipate your participation!
UPD: Editorial! https://codeforces.net/blog/entry/130252
UPD2:
Winners!
1. tourist
2. jiangly
3. Benq
4. jqdai0815
5. Nachia
First Solves
A: Benq
B: Thienu
C1: nwn
C2: Benq
D: Be_dos
E: Tom66
F: kiwihadron
G: tourist
H: rainboy
Woah!! Global Round again... Let's Go!
Expect great and unique problems! Wish everyone good luck!
I expect to achieve pupil rank, good luck to everyone.
as the #1 buffering fan and a testuwuer, this round is fire asf
As a tester, I really enjoyed the problems! Also, Ritwin orz
As a tester, bronze_coder and Ritwin orz
As a tester, I admit I'm orz.
As a tester, this round goes hard
Thanks for your information!
why (1000 + 1000 ) ?
you mean that there are c1 and c2 ?
Yes they mean that problem C will have subtasks
thanks
As there is no cyan tester, I hope that I can get out of cyan to blue!
I already made it out of cyan but after rating rollbacks hope this time before rating roll back.
As a problem, I liked the testers.
As a global round, I hope I will enjoy the contestant
me when i see arul hii arul
me when i see arul hii arul
me when i see arul hii arul
me when i see arul hii arul
no more of this nonsense
me when i see 123gjweq2 bye 123gjweq2
me when i see arul hii arul
me when i see arul hii arul
me when i see arul hii arul
no more of this nonsence
me when i see lomienyeet bye lomienyeet
i LOVE Global rounds hoping in another -150!!
damn $$$-150!!$$$? you'd be the lowest-rated user by far
Not really, there are some people who have -40 rating somehow
Surely $$$2199 - 150 \cdot 148 \cdot 146 \cdot \, ... \, \cdot 2 < -40$$$
it'd be actually 2199 — (150!) * (150! — 1) * ... * 1
nope. https://en.wikipedia.org/wiki/Double_factorial
Imao, couldn't even suspect there was such thing
Yes, waiting for another -150.
As a tester, I tested the round
Like contest has problems?
flamestorm Orz!
flamestorm orz!
CP > IND vs PAK. :D
i was waiting for this comment XD
What does CP stand for?
Competetive Programming
Hahaha!
CP(pretest passed) >>> Any other Enjoyment
The round feels more solid with sum involved as a tester.
hoping for + 3 contribution :3
Get a cloth first after think about contribution!
-_-
How are Global Rounds different from any other div2 and div3's?
Everyone can participate officially and it will be rated for everyone
score distribution changed!
What's your choice Ind vs Pak or this contest?
Ind vs Pak !! after 2 yrs T-20 World Cup ...
Okay, choice is good, but not good enough to post in CF community :XD
finish the first 4 problems as soon as I can then go back to watching the match :)
Is this contest suitable for beginners also?
Yes, It's for everyone
Hoo ok,thanks for the reply
is it a rated contest?
It is
8 questions! I hope there is something for me as well.
Dragon Boat Festival Health！
Will try to improve and upsolve questions from this round, best of luck everyone!!!
Speedforces
I'm in
good luck !!!
the number of participants are going to decrease because of IND VS PAK match
Will there be fewer contests because of Euro 2024?
gl & hf for everyone!
Why do E and F have the same score?
as a contestant i hope to solve the problems
Hoping to become Expert
nicee
YESSS
English or Spanish?
why is cry crying (crying emojis)
let's go! good luck to everyone
all the best guys
is this contest is rated?
yes
click here
I hope I will become pupil (if I got +209 delta and become specialist it will also be not bad lol :')
Literally 1984
literally 1984
Nineteen Eighty Fources
Literally 1434 (I lost the game)
ABC1C2 speedforces
I was not prepared for this 'Dashboard'!
reading statement of E be like
I think I've caught a group of cheating noobs spreading wrong solution of B and giving +100 points to lucky participants.
How you could find them?
In my room
Animated Video editorial for problems [A-C2]:
https://www.youtube.com/watch?v=hS8Z3k57f6Q
I was asked to test this round a couple days ago and the idea of making some manim animations about the round came to my head. As the round was still being altered I waited until the last day to try to make a video editorial (I would be surprised if there are no mistakes throughout the video lol). I've worked on it basically non-stop for the past 24 hours, but i am not super happy with the result as there was a big time crunch. I am totally exhausted and will go to bed soon but it would make my effort 100% worth it if even one person benefits from it. Thank you authors for the problems and null_awe for asking me to test!
How could you do the animation? I'm really curious about that a long time ago, because I want to use that to draw the explanation too.
I use the library Manim and the source code for this video and my other stuff can be found on my github although there are probably better people to learn manim from as my code is messy.
Thanks a lot. I have to find it for a long time.
P/S: Firstly I wonder why you could do the animation so fast in the contest, and at last I find out that you are one of the tester, which is make sense :v.
wow that's so cool! good work sir
Nice video. But why your voice is slowed? Is it your natural voice or it's being slowed?
I think a mix a both, ig I usually talk slower than average but this video was worse because I had no time to rerecord parts that didn't sound good
Can the authors tell us what is the secret of 1984 ?
George Orwell has never been as strong as he is today!
I have been told that 1984 is when you get arrested for using the display toilets in home depot
In E, the answer is n- min vertex cover + 0/1 (1 if we have a node of deg 1 in min vertex cover), right? if yes, how to handle the 2nd part?
You can use $$$f_{i,0/1,0/1}$$$ represent in $$$i$$$'s subtree, whether $$$i$$$ is chosen or not, and whether we have choose at least one leaf or not. The minimal cost for this case.
Rerooting technique with the original dp works too.
Could you please tell me what technique are you refereeing to?
This is the topic. We basically calculate the classic min vertex cover dp, then do a second dfs and change dp's so that when we go from vertex v->u, the dp value's are as if u is the root. When we come back we do the same thing too. This might not be possible with certain tree dp's but in this case it is.
thankss
Oh right, i am dumb, I kept trying to get a relation for $$$f_{i,0/1}$$$ where i is the subtree, and whether we have choose at least one leaf or not.
It's a bit difficult for me. But these problems look very interesting!
C2 hint anyone? :)
Solve C1 by DP, and C2 is just a counting version of it.
yeah i solved c1 with dp but could not solce c2, can you tell me how to do it ?
void solve(){ cin >> n; vector v(n+1); vector dp(n+1,0); vector dx(n+1,0); f0r(i,1,n+1)cin>>v[i]; f0r(i,1,n+1){ dx[i]=dx[i-1]+v[i]; dp[i]=max(dp[i-1]+v[i],abs(dx[i-1]+v[i])); } cout<<dp[n]<<endl; }
thanks a lot ;)
Next time, please format your code properly before commenting to ensure readability. ;)
From what I'm understanding here, in C1 you are keeping $$$dp_i$$$ as the maximum value obtainable, and $$$dx_i$$$ being the minimum one.
We can add $$$dpc_i$$$ and $$$dxc_i$$$ to supply this process with counting. At first, $$$dpc_0 = dxc_0 = 1$$$. For any $$$i > 0$$$:
Sorry for the very late reply, it was midnight at my place when you asked this, and I wasn't really mentally sane to read and parse unformatted code. Now that it's morning time here and I'm a bit more conscious that I could actually read you well enough to provide an explanation catering to your own codes.
Code references (upsolved, I overcomplicated myself in-contest so I won't show you those), in case my words were overloading:
Oh right! Thank you very much for explaining it to me, and sorry for the unformatted code, I will make sure to format it correct the next time ;)
in reverse order calc variants to get current value from previous
Thanks for this round!)
i'm too good at writing buggy code spent 2hr for simple hashing implementation :)
How to solve D?
Ignore all preceding
ain the string. We'll get to that later, and we'll call the new string after deleting that prefixss. It's certain that any string being the answer must be a prefix ofss(can have a fewapreceeding it, we'll get to that).Iterate the prefixes in descending order, for each length, add all position in
ssthat matches the prefix — this part can be sped up using Z-Algorithm. Brute force from the start ofssto end to see if the pattern covers the string fully (and only leaving strayaat most). If not possible, ignore. Otherwise, denote $$$k$$$ as the maximum number ofato precede the patterns without sidestepping on each other, add $$$k+1$$$ to answer. $$$k$$$ can be calculated through the bruteforce, and it's better to calculate each value of maximum preceedingacount of each index before starting this step.Time complexity being roughly $$$\mathcal{O}(A \cdot n \log(n))$$$, with $$$A$$$ being the complexity of whatever DS you use to keep track of the indexes (I used
setso one more $$$\log(n)$$$ into the mix, still works).Oh hell
So smart!!!
Thanks a lot
I had a different approach from the solution above, it seems easier to me, maybe it will be useful:
look at all not 'a' — let's say there are k of them. You will cover them with t-strings. So each t-string contains some divisor of k non-'a' letters. Iterate over all divisors d and check if strings
0-th non-'a' -> (d- 1)th non-'a',
dth -> (2*d — 1)th,
(2*d)th -> (3*d — 1)th are the same (I used hashing)
If so, we have found a good t, but it can be extended by additional a's on the left and right — this needs to be carefully calculated using amounts of a to the left from first t, to the right from last and between ts.
Yeah its easier but a bit slower
Nice aproach tho!
I had a more different approach, you can find the count of all char except $$$a$$$, now you just need to find the common divisor of the counts of all characters except $$$a$$$, iterate on the string for each of these common divisors and then find out greedily if we can divide the strings into $$$a$$$ and a string $$$t$$$ such that count of each char except a in $$$t$$$ is $$${count/divisor}$$$ where $$$t$$$ has no preceding and trailing $$$a$$$.
Now, Replace the substrings equal to $$$t$$$ with the character t and you have a string consisting of $$$a$$$ and $$$t$$$ only. We need to distribute the $$$a$$$'s into $$$t$$$. Time complexity is $$$O(A*n^{4/3})$$$.
Woah, I actually thought of this but gave up and did the systematic one above coz' that was the easiest way to think and not break my brain down in the process. Glad to see that the one I gave up on still worked nicely :D
Yes, it does have a heavy implementation
Can you tell why hashing with binary search does not work:
Fix the first not equal to 'a' character and its position L and iterate over the right border R of a substring that has only a's beside it then for each right border binary search the left border of extended substring that now has some a's as a prefix: inside the binary search check if it is possible to divide s into substrings by iterating over the positions j of a first not equal to 'a' characters and checking if the substring that starts with j has more or equal a's beside it than binary searched number of a's and if the substring that starts with j and has a length R-L+1 coincides with the substring s[L:R] (this can be done with hashing), after all j's are considered delete all j's that are not the starting positions of some strings in the partition.
Store the indices of all character in $$$s$$$ which is != 'a' in an array $$$v$$$. Fix the number of character we will use to form a substring in $$$v$$$, let's say it is $$$k$$$. Then the substring formed by $$$v_1, v_2, ..., v_k$$$ and $$$v_{k + 1}, v_{k + 1}, ..., v_{2 * k}$$$ and ... must be the same.
all possible $$$t$$$ will be of the form $$$aaax...$$$
where $$$x$$$ is the first non $$$'a'$$$ character
let pos be the first position where $$$x$$$ occurs iterate over all possible length $$$len$$$
let $$$t \ = \ s[pos..pos \ + len \ - \ 1]$$$ check if t is valid or not with hashing
if $$$t$$$ is valid $$$ans \ += \ 1 \ + \ no. \ 'a' \ can \ be \ placed \ before \ t$$$ '
tasks were good btw
best D i have seen, orz!!
C2 almost killed me, spending nearly 2 hours on it. Anyway, I think the problem in this round is pretty nice.
sem
This contest may not have been my best performance, but I enjoyed it a lot.
is problem E , pseudo independent set?
for every leaf, delete it and solve biggest independent set, answer is the maximum value plus 1. idk why it works.
Here is what I thought: ignoring the very first root, no other subsequent roots can be leaves. (I don't count an isolated remaining vertex at the end as a root.) If a vertex (v) becomes a leaf, then clearly all its neighbors except one would have withdrawn edges from it by becoming roots and attaching that edge to the next upcoming root. This implies that no two edge endpoints can be leaves simultaneously (ignoring the first root, which can be a leaf, and its neighbor could also be made a leaf).
On D I used prefix function for the initial string ignoring (if existing) a prefix just of "a". Now, a prefix to be a good string need to appear in the string such that all non "a" characters are used. This is an easy check using a frequency array but the remaining problem is how to identify good strings that starts with "a". I brute forced this part cause I did not find any good solution on this. I am close enough to the real answer?
very close
You already know how to find the good strings that don't start with 'a'. Call such a string G.
Then you use the prefix function to decompose S into S = "[a] G [a] G [a] G..." where [a] is a string of consecutive 'a'.
So G is a good string. How many 'a's can we add to the beginning? Anywhere from 0 to the minimum size of all the [a] blocks.
So basically, there is 1 candidate G for every integer 1-n, and we can check how many strings with suffix G work in O(n/|G|) time, which works.
I understand now. I think my brute force failed when the G contains some "a" at the end and I am also considering those "a" as the new prefix for next good string. Thank you!
Hoping to get editorials that are easy to understand.Many a times these codeforces people give stupid non understandable editiorials
what is wrong in this code https://codeforces.net/contest/1984/submission/264959968 i'm not able to find the error
Dpforces >>>>
I nearly got F and G :(
My solution for G uses roughly $$$\frac12n^2$$$ operations, and I will put my submission here later.
Judge solution also uses around $$$\frac{1}{2}n^2$$$ operations most of the time, and I don't think it goes over $$$n^2$$$ for any of the tests.
We made limit $$$5n^2$$$ because we didn't think it would allow unintended solutions, and it made it slightly more flexible for contestants.
how to do C1 ?
.
Actually every task is very cute and contains something new. But I feel B and D are difficult for it's scoring or position.(And actually I go panic during the round, oops)
Anyway thanks for good problems!
I actually spent like 30 minutes panicking at D, heh. I kept rolling my chairs thinking "can this bruteforce of all string prefixes TLE?" and then completely forgot that sum of reciprocal would only growth at $$$n \log n$$$ rate instead of $$$n^2$$$, hehe.....
B made me panic as well
Fantastic contest! Enjoyed solving A, B, C1, C2 for 3 hours!
Thanks a lot for the round null_awe flamestorm le0n buffering and all testers!
cheatforces
The B from this round will haunt me forever. I took almost one hour to solve it, haven't felt so stupid in a long time. C was such a cool problem even though I only managed to solve C1, dp is not really my strength.
Thanks for the problems and I hope everyone did well.
I know right, B was a nightmare! It took too long to realize that if zero appears within x (not in the last digit), then there's no answer.
I didn't do C1 with dp, but instead applying "Option 2" ONLY whenever the current sum is most negative; otherwise, apply "Option 1"
yeah same I took too long to figure out the condition with zero. I did a pretty weird thing on C1:
and I was very surprised that it worked xd. C2 was a bit too much for me and I didn't really have time to solve it because of B.
Me too buddy, me too. Shit happens, don't stop believing in yourself!
wow, How did u thought and wrote that C2 in 13 minutes. Godly recovery of the contest 😲
thanks! It's not too bad, I only lost a few points so I'm good, at least I can do Div. 4 on Tuesday :).
For problem E. Why is the answer 6 for example 4. I figured the answer is 5. If you pick v to be node 1, then you can shuffle 10-6-4 and get 5 leaf nodes.
I think I still don't understand the problem statement of problem E actually. So it is saying repeat step 1 and 2 recursively I believe for the subtrees. But I still don't know how to get 6 from the 4th example.
Problem G: https://codeforces.net/gym/103414/problem/L
Good tasks, enjoyed solving C1 and C2 (I solved them without DP). Went blank on B, hence implemented a DP solution there.
On problem D, how is the answer to the string "abbaaaabbb" 3? I understand that t="b" and t="abbaaaabbb" work, but I can't figure out the third string.
This was a really common question asked, "bbaaaabbb" also works.
Thank you
Actually I was stuck thinking about the same testcase. Eventually after a lot of thinking (in hindsight I have no idea why I couldn't spot it immediately) I got why it was 3. But I was considering asking you problem setters during contest.
This makes me wonder, is it allowed to answer such doubts during contests? To understand why the answer for the samples is such? I usually don't know what can be or what can't be answered and don't wish to waste the problem setters' or my time by asking such questions and then waiting for the answer.
We are not allowed to answer such questions (unless the statement is unclear and it's our fault). We received around 50 questions about that specific test case and answered all of them with "check carefully".
That makes sense. Thanks for the clarification.
Cool round!
Being stuck in C1, then turn to D, then fail many times until 2 hours passed :(
Then I discovered that C1 and C2's implementation is way more easy than D :(
Luckily a master in my room made some weak hashes, and only changed his hash into double hash, then multi-hash, all of them had fixed bases and modulos, and I gained extra +250 points :)
But still to have a negative delta :(
can you take help of others in your room? I dont understand how it works can you explain?
Er, it doesn't mean that the master made weak hashes delibrately to give me points. I locked my D problem so I can see their solution to D, and then I can make hacks.
Maybe he just didn't know the importance of randomizing his hashes so he made mistakes again and again, which enabled me to gain points many times. In the end he still couldn't pass D because I was hacking him.
upd: the above seems the reason of following
My curiosity, but are there any antihash test in D originally or is it caused by hacking? It seems many D gets FST strangely.
is my observation correct for C-1 you only have to use operation of type-2 once or never?
yeah But how you observe? could u explain me yet i didn't understand
Never understood problem statement of E in the competition.
Most of B's hacked submissions have several identical but illogical pieces of code. Is that really a coincidence?
somebody leaked a solution which included two bugs (checking if the sum of digits was 35, checking if length of number was 10). most who copied did not realize that these were incorrect which resulted in all the FSTs and hacks.
hopefully this is a lesson to not cheat!
All these guys need to be perma banned. Cheating is very annoying for those of us working hard for our rating.
In D question ,I used p=31 for hashing in solution 264942064 and I got wrong answer on testcase 29 and then I only change the value of p to 27 and It get accepted 264969257,
As you can see in above submission.What is the mistake in this?
Does anything can be done about it?
Most probably it is an intentional hack. I can advise you to use a random p next time.
I took the value 31 from cp algo article
No.
Sadly no. I hope you could learn that in an environment like Codeforces where exploiting is possible, if you wanted a heuristic method to pass, you should at least cover its constants well so that it would not be reverse engineered for counter-testcase production.
Original comment up above already showed you one way of doing so. You could also randomize and/or add more prime modulus into the hashing.
Hacked!
You should randomize your hash, random modulos or bases are all OK, simply changing anothor fixed base can't prevent you from being hacked.
rainboy solved problem H first
rainboy finally hit gm !
Problem 'H' first solved ecnerwala ?
Auto comment: topic has been updated by buffering (previous revision, new revision, compare).
Whoever has leaked the solution to B, must have intentionally put a bug in the code so that everybody gets hacked.
I thought that as long as none of the submitted problems passes the pretest 1 (samples), the round will not be rated.
Did I misunderstand, or has this changed? (I just lost 166 points because I left the round after a few minutes...)
Relax. Why you worried about the rating changes?
You are right, I don't care much about the points, but still want to understand the rules. Now I learned.
It has always been like this and nothing has changed. You should have absolutely no submission to be unrated.
i didn't manage to debug my D solution in time, now i finally got AC, and want to understand why it works. After cutting the a-parts from the sides, i am looking only at prefixes, that have at least one symbol of each non-a type that is in the string and also checking for all of them, that their number in prefix is a divisor of total number, and if it's all satisfied, i just check for O(n) with z-func etc. Is it true that i check no more than O(√n) prefixes?
In the problem D, why does 'abbaaaabbb' have 3 valid string t?
Valid strings for $$$t$$$ are
abbaaaabbb(obviously),b(the rest are justa) andbbaaaabbb(valid partition beinga+bbaaaabbb=abbaaaabbb, notice that $$$t$$$ only needs to exist at least one, not necessarily at the beginning of the partition).Took me a while there during contest as well.
Got it. Thank you!
I really like problem H, it is possibly the best geometry problem I've ever seen.
However, I do think G is a bit implementation heavy, and the difficulty gap between F and G is a bit too big. Was this intended or the writers overestimated the difficulty of problem F?
We overestimated problem H (see the rating predictions), maybe it would have made sense to give G and H the same amount of points.
(This is not a complaint) You underestimated G more than you overestimated H, so the F G gap became large
Yes, I think G was a bit too hard for its position (or maybe F was too easy). We were considering having to solve F in n log n (as a subtask or just on its own), maybe this would have helped -- but it would have also increased implementation.
i dont think adding a mostly implementation subtask helps, like difficulty is always a secondary concern and quality should be primary.
I wasnt complaining, just informing Scrasse that he got the wrong issue. H wasnt underrated by a lot(it shows 3250 on predictor) and 7 AKs is very reasonable. Nice contest
Why would you be aware of a somewhat interesting optimization and not set it for F :(. Hate when authors only leave the iq test and not the more interesting part
Overall this was one of better recent contests to me tho :)
From the predictions and the actual difficulties, it seems that G was underestimated too much. If G was indeed as 3000 rating problem, that would be much more balanced.
Just curious why all setters and testers think G is that easy. All my friends around my level agreed that G should be at least of difficulty around 3200 to 3300.
We initially solved this problem for odd $$$n$$$ and extended it for even, so it might have been that for the setters it was easier to extend the logical line of reasoning for the solution. I think more testers could have been beneficial (at least higher rated ones) to get a better sense of G, we received some comments about hard implementation but nothing that made us worried about the difficulty. We’ll try to get a more wide sense of difficulties next time.
I was unable to solve E or F but found G pretty easy.
264891491 264900499
How this solutions is equal??
it's same idea but not the same code
It is normal to have similar ideas on the same problem
and if i cheat C1 from this person why I didn't cheat C2 and D????
LOL, I got the same warning. Please MikeMirzayanov have a look at this false positive.
Indeed, there was an error, it has been fixed now.
Thanks bro
Hello dear Codeforces and MikeMirzayanov. I want to tell you that I got a message where it was said
But I don't know one of them and task B was easy. I mean 8532 people solved task B, there can be solutions that are similar to each other. I want you to check my code again please. Check everything that you know, please!
Thanks you in advance!
Seems quite same: 264929515 264902138
Yes it seems quite same, but I wrote it myself. I can prove you with everything. I don't know either of them. And there can be a reason when I blocked my solution. After someone in my room could block his solution, and just shared or gave my code to others, because he knows that if he will share his code his solution can be ignored.
Please check this! I think you comprehended me, so thanks in advance!
.
deepak0003
Appealing for a review on https://codeforces.net/contest/1984/submission/264889510 and https://codeforces.net/contest/1984/submission/264917529 . The only matching parts are a for loop from 1 to n — 1 and the use of intitializer_list variants of std::min and std::max (which I think is pretty usual for this case). I did not post this solution anywhere during the contest and I have not copied it from somewhere else either.
I was trying to login to my cf account that is kxhitz. But it doesn't login then i commented on the last contest then I got to know by a cf candidate informed me that my account has been banned. But I haven't cheated anyone's code in any contest. I have written the code by myself in every contest I had given till now. Even Sir I haven't received any warning message on cf account or on my registered gmail id.
Please reopen my account so that i can participate on further contests and also practice the questions.
Your thinking process in extremely slow
not bad!
null_awe, buffering, were the winners of t-shirts announced?