Hello Codeforces!
On April 10, 14:35 UTC Educational Codeforces Round 42 will start.
Series of Educational Rounds continue being held as Harbour.Space University initiative! You can read the details about the cooperation between Harbour.Space University and Codeforces in the blog post.
This round will be rated for Div. 2. It will be held on extented ACM ICPC rules. After the end of the contest you will have one day to hack any solution you want. You will have access to copy any solution and test it locally.
You will be given 7 problems and 2 hours to solve them.
The problems were prepared by me and Alex fcspartakm Frolov.
We'd like to thank Nikolay KAN Kalinin, Grigory vintage_Vlad_Makeev Reznikov, Vladimir vovuh Petrov for the testing and help in preparing the round.
Good luck to all participants!
UPD: Some information from Harbour.Space U
The third Hello Programming Bootcamp finished as teams from around the world learned, competed, and got to know each other in the cities of Kollam and Moscow.
“I'm positively surprised with the level and number of Indian teams. I believe when following a systematic and regular training schedule — they will certainly attend top world records in programming” says Mike MikeMirzayanov Mirzayanov.
The results of the India boot camp are as follows:
We would like to congratulate the top teams from the India side in both divisions A and B — and especially recognise division A’s winner: Team SDV — Vasily platypus179 Alferov, Vaibhav gvaibhav21 Gosain, Motasem Motarack AL-Kayed sponsored by Social Discovery Ventures, and winners of the Educational Codeforces Round 39.
The division A winners are: Team SDV (platypus179, gvaibhav21, Motarack), TheVindicators (IIIT Kharagpur) (pranjal.ssh, Chenghiz), DomiNUS (NU of Singapore) (vodanhna, darknsux, minh141198).
The division B winners are: BEU Programmers (Baku Engineering University) (mahmud2690, Mamedov, fuck_coderinuse), Penza State University (Fortin, WooDeND, Key_J) and Sab_lite_hai (IIT(BHU),Varanasi) (Enigma27, Dsingh_24, nitishk24).
All results from India side
All results from Russia side
From the ever present in ACM World Finals city of Moscow, to the new-comer on the world stage city of Kollam — Hello India x Russia Programming Bootcamp showcased heart, passion, and a universal drive towards a better competition.
“It's the first time I participate at Hello ACM-ICPC Bootcamp in Asia and I admire local hospitality and students motivation. Everyone seems to be eager to learn, participants stay very late upsolving the contest problems and revising the topics they were thought during the day” states Gleb GlebsHP Evstropov, The Coordinator of the Programming Committee.
The latest in world finals simulation training would not be possible without the bronze sponsorship of VTB, the institute with the most extensive international network of any Russian bank, boasting more than 30 banks and financial companies in more than 20 countries.
Fill out this form and we will keep you up to date on our next boot camp!
UPD: Tutorial
MikeMirzayanov as problemsetter :D
Yes, sometimes I write problems: http://mirror.codeforces.com/contests/writer/MikeMirzayanov
WOW!! Is there anybody in codeforces involved in more problem setting than you? Wish to see you in contest as contestant just once.
Thanks! I like to take part in contests for now. But each time I can take part in a contest, I think about time I can dedicate to development of Codeforces. You see that there are a lot of ideas around Codeforces I want to be implemented. Each time I think like "it is better to improve Codeforces than have fun taking part in a contest".
Yeah, you "sometimes" write problems :)
I think there is a better word than sometimes :) You write so many problems.
I think you meant April 10 in the post instead of April 14
Thanks, fixed. I was testing your attentiveness.
We somehow got used to your "I was testing your attentiveness" phrase during bootcamp. Thanks for perfect lectures and taking care of each student.
is it same as you check our attentiveness during camp?
Dreaming of purple after this contest! Good luck to everyone
Good luck to you! I'm hoping for huge rating jump too.
Nice
Just became Purple again... But I will still take part in the contest.
Good luck to everyone!!
You won't be affected in this round by any rating change. It's only rated for Div 2 contestants :)
Contests are good for practice, you shouldn't just do them to increase your rating.
And I only replied because I thought he may have forgot about it or doesn't know.
Thats the spirit .
Dear contest setter MikeMirzayanov, please Give thanks to MikeMirzayanov for this awesome codeforces platform.. :D
I express my gratitude to myself for ten years of efforts to develop Codeforces and Polygon!
dedication for users by "MikeMirzayanov"*
Good luck to everyone!
Just 20 minutes before the contest can we see 6000+ participation :)
typo : its IIT Kharagpur not IIIT Kharagpur
i think you have to stop rounds until you update servers
Today's E is a copy-paste of New Year and Rainbow Roads from Goodbye 2017.
Why are you posting this before the round ends?
I didn't know that pointing out a question was copied from CodeForces contest itself, that too, a very recent one and the one with the most registrants ever made me dumb.
Thank you for your insight. Maybe have the balls to say it from your main account instead of hiding behind a fake one.
Wow great logic -_-
Ah, I'm sorry. The contest was almost at an end, and I thought of leaving for dinner..
So to me seems like a notorious coincidence :)
Two questions:
Can we just copy paste during contest (since it was a past round and we can use the code)
And will the round be unrated because of this?
Yes, you can. You can read the rules about using another's code.
This is educational round, so I don't think it's such a big deal about copypasted problems.
Isn't it that the ideas of the problems can be well known (or something like that) and not the whole problem?
Well it is kind of a big deal, it is rated for Div 2.
Guys, I apologize for the coincidence of tasks. It was a really unexpected surprise for me. It seems that Lewin and I read the same articles.
Unrated?
Edit. It should be, given that despite copy-pasting the exact same problem, there were still some issues with it at the beginning...
Please don't make it unrated. Will be very unfair for the majority.
Define "unfair" The way I see it, people who copy-pasted the code had an "unfair" advantage and people who didn't copy-paste the code implicitely had an "unfair" disadvantage. Since the category "people who didn't copy-paste the code" represents the majority, it follows that the majority had an "unfair" disadvantage.
I"m not saying that the round should be unrated (I'm not saying the opposite either), but "will be very unfair for the majority" seems like a misleading statement to me.
I received a mail about violating the rules of the contest, by copying the code.
I really copy other solution but the solution was published before the contset. It was the solution of the 908F.I don't know whether i violated the rule.
My solution : http://mirror.codeforces.com/contest/962/submission/37173319
The code I took it from: http://mirror.codeforces.com/contest/908/submission/33845527
Please nullify the claim that I collaborated with someone else during the contest. It is pure coincidence that they took the same code.
Probably someone else copied the same solution so these solutions are same which is violating the rules
Thanks for your reply. But int this passage [Your text to link here...]` ,it says that we can use the code which is published before the contest . QAQ
You can use it but system will find out if there are two simillar solutions which is forbiden. You legitametly copied solution but someone else did too
Thank you !
Fuck me, spend an hour coding F with wrong solution, i extracted nodes in simple cycles not edges.
Was G "Weiler Atherton" polygon clipping?
In problem A, what is the correct answer for: 4 1 2 1 3
3
Runtime error on test 15 in problem D. Can anyone help?
I think this might be the problem:
Mapa contains compressed indexes, so it shouldnt have more than 2 milion element.
mapa[2*idx[i]] will exceed 2000010?
Why would it? There wont be 2 milion elements in mapa. You see that mapa[val] = compressed_val
Oh so you are storing compressed values, i didn't look at it
Yes. Anyway, thanks for trying to help me!
I think kapillamba4 is right idx could have a huge value since its based on elements in vector b:
Yes but nums [ mapa [ 2 * idx [ i ] ] ] is not a problem because idx, has huge value, but mapa is map<int,int>, so mapa[ huge_value ] is not a problem.
I think its possible you are generating more than 2000010 different values, since your inner loop runs around ~59 times
and there could be up to 150 000 different values you will end with more than 2M elements.
thank you. it may be the problem.
Try Creating a test case for all numbers 2 to 150000. This test case will give segfault related to mapa
Replace 2000010 by 3000010 everywhere and you won't get segfault anymore
I think 3000010 wont be enough :( because as he said, it can be about 60 operations in while loop.
Does anyone have any idea about pretest 10 in B?!
Try 5 3 0 ...*.
problem G seems to be exactly https://main.edu.pl/en/archive/oi/5/okn
Does O(N^2) solution passes problem G? I don't see how can you check whether a line segment completely lies within a polygon quacker .... quicker than O(N) time deterministicly.
Did someone solve E using MST?
how to solve D ?
I have runtime error on test 15, but i think my idea is correct. Index compression and keep indexes for every value. go from 1 to compressed_max_val, sort indexes and do what problem tells exactly.
make a map
map<ll,set<int>>mp
in which you insert the index of each numberfor each pair in
mp
you need to remove the first two elements in the set. before doing that, you need to insert the second item of the set intokey*2
you will have to repeat this step while the size of the current set greater than 1take a look at this code
I have solved this problem by using Priority_Queue . Thanks all :)
Problem E Test Case 11 anyone?
Don't know about your solution, but mine also failed on test 11 and I've found this counter-test:
Answer seems to be 12
Thanks.
why is it 12 not 10?
How can it be 10? You need to connect 1-3, 2-3, 3-4, 3-5
6 0 P 1 B 2 R 998 R 999 B 1000 P
Answer: 1006
Mine was failing with this
Problem E, Why is the output for this case 10 and no 7? 4 1 P 2 R 6 R 8 P
Got it, lol
Even though it is Educational round, is it legit to just copy&paste previous Codeforces contest problem? I don't think it is a good idea.
Please don't make the queue even longer.
Oh you mean when people copy paste past codes in under 2 minutes is not legit?
Then what should the contestants do? Should they have to write all the code again, when knowing that there is already an accepted code in somewhere else? What's the difference between copy&pasting segtree/MaxFlow/FFT algorithm and copy&pasting exact same code on exact same problem?
It's perfectly legit to copy & paste previous code. It is open to everyone (On previous contest's submission history), it is his own written code. We can't blame contestants to use such kind of source codes.
But it seems too unfair for everyone, isn't it? That's why problemsetters should avoid same problem to the contest.
Actually my sarcasm was in the fact that authors shouldn't copy past problems so that we would be able to copy past codes. Of course once we know the problem we should paste it but what i mean was that there shouldn't be the opportunity to copy something in the first place. So i have the same opinion as you. :D
I misunderstood your opinion a lot. I'm sorry for it.
Really yes, contest should not be copy & paste contest.
No problem XD
I hadn't solved the older version of the problem. After the round I submitted my code from this E to the old F from Goodbye only increasing the constraints and it got Accepted. Complete copy-paste of a problem
test 11 for D has given WA for quite large part of submissions, what's so special about this one?
what is it? Even I can't debug.
I got WA11 in my first submission, and it was because I was adding the new element to the first position of the two old ones, not the second
In problem C
In one operation you can delete any digit of the given integer so that the result remains a positive integer without leading zeros.
In case of 101 why the answer is 2 not 1 (remove first 1 and result is 1).
means that leading zeros must also be deleted.
01 != 1
I think it means after do an operation the result will remains a positive integer without leading zeros. My bad ;).
According to the problem's statement, one cannot delete a digit in a way that leads to the incorrect number, i.e. 101 → 01 is prohibited. Hence, one should make the following operations: 101 → 11 → 1 (two operations).
i think this round should be unrated, Problem E is literally copy paste from goodbye 2017 F, it is unfair for someone who hasn't solved it before
goodbye 2017 F http://mirror.codeforces.com/contest/908/submission/37175943
Educational Codeforces E http://mirror.codeforces.com/contest/962/submission/37171865
it won't affect my rating anyway
Making this contest unrated will also be unfair to more people, won't it?
I think there is a mistake in Institute name of TheVindicators. It should be IIT Kharagpur instead of IIIT Kharagpur
there will be manyyyyyyyyyy hacks in problem A
Someone please hack my C and D.
halyavin
You can do it by yourself, afaik.
I am not good at hacking.
An update of this comment : (sorry for my editing skills)
I do all my hacks in problem 962A - Equator with this test case:
PD: And I was hacked too...
MikeMirzayanov, what article did you saw? (I'm just curious, and I want to know about the theory behind)
+1. Btw where did he mention that he got idea from an article.AFAIK I asked a clarification and he answered me.Did he mention it publicly??
http://mirror.codeforces.com/blog/entry/58845?#comment-424747
https://arxiv.org/pdf/1603.00580.pdf
Can somebody confirm that %lld started working on Codeforces? What is the earliest C++ version for which it works?
it works with c++14
Please don't make this round unrated, otherwise it would be unfair to me who deserve a massive downfall of rating.
Contest should be rated IMO. Would affect the majority of the people very heavily, moreover not many people solved E from div 2 so let it just be a coincidence and move on from this, but no need for the contest to be unrated.
Loooooooooooooool I submit the same code of problem E in goodby 2017 F only I change only 'P' to 'G'
goodbye 2017 F http://mirror.codeforces.com/contest/908/submission/37179332
Educational Codeforces today http://mirror.codeforces.com/contest/962/submission/37172686
this round should be unrated it`s not fair I spend more than 1 hour to solve it and others take it copy & paste !!!!!!!!!!
Educational round learns us how to make a copy and paste from previous rounds !!!!!
How many people solved E and how many of them was Div2(and maybe how many of those saw the problem before) ?
well there is a lot of people participating in the Goodbye 2017.
30 or 50 Div2 copy-paster could affect a lot of near-purple guy.
and even "if" there is less than ten people that copy paste it. Updating the rating when there is a flaw in the contest isn't "competitive".
Are there official rules/guidelines for when contests are changed from rated to unrated? At the moment, it seems like it's at the discretion of MikeMirzayanov (not that there's anything wrong with that).
If you have code for a previous problem that can be applied to a current problem, and it is allowed to copy + paste, then why wouldn't you? It similar to having a template on hand for segment tree, or max flow. It's not something that should be learned, but should be obvious for a competitor.
Furthermore, Goodbye 2017 happened on CF and after the contest ended all submissions were made public so everyone have the opportunity to upsolve and make their own solutions for E.
It was unfortunate that a problem happened to be exactly the same as one before, but invalidating the round would be too much of a takeaway from the time and effort the problem setter put in, and the competitors as well. If you feel like you placed lower this round due to an unfair advantage by others, then just do better than them next round when it won't be so unlucky (hopefully).
By your logic, it would be entirely OK to have a round consisting solely of problems which were previously seen in other contests on CF. At that point, the contest no longer tests programming skills, but who can search through previous contests (and editorials) better and more efficiently.
Also, "If you feel like the round was unfair, just do better next time" to me sounds like a very superficial (and somewhat insulting) statement.
P.S. I'm not taking sides here. Not saying the round should be unrated (not saying that it should be rated either). Just pointing out your arguments towards keeping it rated are not very consistent.
On your second point, I'm not saying that he has to do better (as if it was his fault that he got "cheated"). I said that he can "just do better than them" next contest. If you're the better competitor but someone else used an unfair advantage to get ahead, then that sucks, but you have opportunities (future contests) to show that you are better and earn back any undeserved rating (and it is possible for them to lose it as well). Things will return to what they belong in the long run.
On your first point, I'm not saying that a contest should be kept rated if every question was a previous question with a new story. I'm just saying that in this case where one question happened to be a repeat that the contest should be kept rated. I didn't think about what should be done in the extreme case.
However, Googling solutions and bruteforcing small cases then searching OEIS is already a "technique" that people use to solve problems. It wouldn't be such a big jump to search old editorials for hints/solutions to similar problems if there was an efficient way to do so.
So hack period was limited to 12 hours and extended again to 24?
And now it's back to 12? This is weird.
How to solve E? :c
Here
To add on, it's the New Year and Rainbow Roads question.
How to solve B ?
i understand ,how i can solve .Thanks!
When does the System Test begin?
In problem D, my solution was working on my local compiler(Ubuntu), but gave me a different answer on the CF judge on the first test case itself (thank god :P)
I tried using the custom invocation command, and realised the output was different (3 2 8 1 instead of 3 8 2 1). Eventually AC'd it by changing the way I played around with the iterators, though could anyone help me out in understanding the reason for this?
Submission link: http://mirror.codeforces.com/contest/962/submission/37165193
I also faced a similar error when i erased 2 elements(first 2) from set but when printing the set, they were not actually erased.
Submission link
Edit: works on custom invocation but fails on local.
You've got UB.
Take a look of set of ones. It is {2, 5, 6}. What iterator is? Is some object pointing to object in container and iterator knows what iterator is next or prev or something else.
What you did? itr2 points to 2, itr3 points to 5. You did
Cool. 2 was removed from set. But what itr2 is pointing to now? It is undefined. So itr2++ is undefined too as goog as result of whole code after. And for different compiler there will be different result.
How to do it right?
If you pass iterator to erase function it will remove element iterator points to and return iterator points to next element.
So you must do something like:
Yes, cleared now, thanks!
Is system testing waiting until 12 hours later?XD
It has started now.
IMHO. This is the slowest system test I've seen
lol yes
Waiting for system testing when system testing is not started Waiting to change rating after system test xxxxD
So... Why are people saying that it's unfair to the majority if the round is unrated?
A significant but still minority portion of participants solved the problem and got a free problem count from it. Isn't this unfair to those who didn't solved/copied it? (Heck, even if I spend 20mins solving it I am still tanking time penalty.)
Hope that the round goes rated.
I think G is the same as this problem.
UPD : Sorry I didn't see the above comment.
http://mirror.codeforces.com/blog/entry/58845?#comment-424700
Do you think O(n^2) is good enough for G? I have yet to find a solution which works faster than it.
Is there any additional score for a successful hacking attempt?
can someone please give me some hint on how to optimize simple O(M^2) solution of F. Thanks!
What is your simple O(M^2)? I used DFS-tree, and would check for each back-edge if it makes a legitimate cycle (without any extra edges). A break; makes it linear.
if you don't mind can you please look into it : http://mirror.codeforces.com/contest/962/submission/37210161
Rated or Unrated ,that is a question.
I really need answer for this question :D
Hope that the round goes rated...
you forgot the semi-rated option
Yes,you are right,and now we can see that the round goes rated XD
MikeMirzayanov please tell us ?? Rated or Unrated
Whether this round goes rated or unrated doesn't matter to me.Bcz I really enjoyed it.
I don't know why pepole are asking is this contest will be rated o not.
In this blog, it was clearly mentioned that the contest is Rated for Div. 2.
If for any case, it will be unrated, the post will be also updated. So, I don't think commenting same question is necessary.
Problem E was exactly the same as a past problem F:
http://mirror.codeforces.com/contest/908/problem/F
I think most of the participants remember this problem because it was a Good Bye 2017 contest where almost everyone participates. When I remembered that I saw this problem before I got there and copied one of the already accepted codes (The contest was unrated for me anyway). Now I get a message saying that my code coincides with someone else's code (Apparently we both copied Petr's code xD) and that this can lead to my account being blocked !!
MikeMirzayanov Is this really a violation? Is it really my fault that the problem came only a few months before? and on CODEFORCES?
The contest is unrated for me anyway, but I don't want to be treated as a cheater and getting my submissions skipped, in addition to getting a note that this can lead to my account being blocked !!
they shouldn't even think about banning on this particular problem. Bcz almost everyone did that copy-paste on that problem. lets hope that the message was just for formalities.(I am in div-2 and managed to solve only A,B,C)
I hope so. The rules specifically say that it's ok to use codes that were written before the start of the contest, and the code I copied was written about 4 months ago (Thus before the start of the contest)
This is ridiculous. In my opinion the contest must be unrated without any doubts.Also seeing the comments in this blog .I think you must respond with some reply to the people .The main thing as per me that differentiates codeforces and other platforms (specially csacademy) is that codeforces accepts its their mistake and gives an apology compared to just ignoring.Hope you dont change..Added upon that doing this adds a lot of frustation.Look into it MikeMirzayanov KAN fcspartakm vintage_Vlad_Makeev
I don't think many div-2 contestant solve this problem.
In addition it's rated only for Div.2
Make contest un-rated for people with copy-pasted solutions on E and rated for others. If it is possible think it will be a good solution for the problem.
The rule clearly states that you can use any code publicly distributed before the contest. Why should it be unrated for those who copy-paste?
Some people may have used GB2017's editorial without copying a solution so we can't know who used the editorial... Declaring people who solved E unrated would not be "a good solution" either.
TB to the educational round Educational Codeforces Round 37 (Rated for Div. 2)
920E - Connected Components? was a case of 190E - Counter Attack and the contest wasn't get unrated (copy & paste code will work)
it's totally unfair that the contest get unrated ... :|
hope the contest being rated & it'll never happen again :)
Yes you're right.
Still no rating changes :( Like previous educational round it's taking too long
I received a mail about violating the rules of the contest, by copying the code. But I have conclusive evidence that I had taken the code from a AC code of another problem, 908F. If u want I can link the solution here, from which I took the code. However, it was present BEFORE this contest began. So I am not sure how this can be a violation of the rules. It's the setters fault that the problem was exactly the same.
My solution : http://mirror.codeforces.com/contest/962/submission/37168611 The code I took it from: http://mirror.codeforces.com/contest/908/submission/33800403
I too had a AC code to this problem:908F (http://mirror.codeforces.com/contest/908/submission/33824171) But that got some error, so I decided to take another code.
Please nullify the claim that I collaborated with someone else during the contest. It is pure coincidence that they took the same code.
MikeMirzayanov It seems that my rating has not been updated for this round. Could you please look into the matter?
Can anyone explain how to solve F?
PS:When will the editorial for this round be posted?
Calculate all the double connected component by tarjan (which means the component is connected after any vertex is remove).Let v = (the number of vertexes of the double connected component),e = (the number of edges of double connected component).If u == v then this double connected component is a sample circle.
Thanks I'll try it.
I received a mail about violating the rules of the contest, by copying the code.
I really copy other solution but the solution was published before the contset. It was the solution of the 908F.I don't know whether i violated the rule.
My solution : http://mirror.codeforces.com/contest/962/submission/37173319
The code I took it from: http://mirror.codeforces.com/contest/908/submission/33845527
Please nullify the claim that I collaborated with someone else during the contest. It is pure coincidence that they took the same code.
This is really funny. Anyways, here is the proof that you haven't broken any rules. You used code which was available to the public before the contest which is clearly allowed.
From the link:
We never really got an answer to the 5 year old first comment asking whether it's ok to use other people's previous submissions.
When i see five "skip" , i am so sad. I want to know whether i violated the rule.
I can't put my finger exactly but there was an offline contest(LAN) which was later held as an online round here on cf. This was from some training camp. In that case people who had participated in the local event were asked not to participate. In that case this will be considered cheating but in yours in my humble opinion definitely not cheating.
Thank you! I hope MIKE can reply me. But i got some lessons,i will be stronger and do the problem myself!
My Rating is going to increase , but still i think so making this contest Rated Will be Unfair for many people and even for me , because I_Love_Mohan_Pyare .
The decision is made. This round is rated.
Can any body explain to me why I'm unrated for both ECR41 and ECR42. I don't think I'm against the rules.
Copyforces should atleast reply something now.
Only one problem was similar to a previous problem and "copyforces" the word you are using??? this is so lame. And Mikemirzayanov had already explain in this comment.( http://mirror.codeforces.com/blog/entry/58845?#comment-424747 )
Great! This contest is was held in the suitable time!I could watch the UCL quarter-final 2nd legs:AS Roma vs Barca.It really a history match.(sorry for my poor english).
Does this round have ranklist and hacker ranklist, and when will they be updated?
Anyone knows why i get WA 11? I am trying to do MST. 37176128
In problem E.The answer of 3 0 P 1 B 2 P is 2 or 3?
Oh,I know.
And the editorial is gone...
Where is the editorial???
Here
thank you
Ur welcome
Please link editorial to problems, would be helpful to find editorial from problems in future.
Can someone help me figure out how to solve the problem B? I've already read some solutions, but I couldn't come up with my own answer, and I am not sure how to solve it by greedy. Thanks in advance!
nhakhoa you may have a look at my solution. I replaced Dots(.) with "A", and "B" such that there is no occurrence of "AA" or "BB". And if number of A is zero. Then replace remaining dots with available B such that no "BB" occurs. Same is when number of B is zero. Submission
Then count number of AB in the string.
I think this one is much simpler. http://mirror.codeforces.com/contest/962/submission/37165159
For all of the users with skipped solutions because of used public codes of 908F: for sure, I'll revert the punishment, it is not a violation (see http://mirror.codeforces.com/blog/entry/8790).
Currently I just arrived to Beijing and Internet connection in my hotel is really bad. Is it good idea to buy local sim-card to have stable Internet? What provider/tariff can you suggest?
Buying local sim-card is ok, when my internet is not stable I always use sim-card to connect CF and it always work.
Welcome to China. Unfortunately, the delay in accessing codeforces and other foreign websites is relatively high. However, the local SIM card I used has always worked well. You may need a VPN or SS. In addition, in China you cannot directly access Google or YouTube's website because of the existence of GFW.
My phone with China Unicom's plan can visit codeForces properly. It may be slow while first access, but it will work fine later. Good time in BJ.
If you connect the Internet in China (like www.baidu.com) bad,you should buy local sim-card. Otherwise,you should use a vpn.
Maybe daily card is suitable for you (1G/day 1yuan/day)
Who can tell me how to solve the problem F?qwq
Why have the ratings decreased?