I am Mike Mirzayanov. AMA!

But why not!

Many people here seem to know me. Or maybe not quite? I am ready to answer your questions. Ask me anything! I plan to spend no more than 3 hours on answers. Sorry if I can't answer all the questions.

Let's more or less follow the rules from here https://www.reddit.com/r/IAmA/wiki/index

Below is actually a quote from the rules.

The interviewee begins the process by starting a post, describing who they are and what they do. Then commenters from across the internet leave questions and can vote on other questions according to which they would like to see answered.

The interviewee can go through and reply to the questions they find interesting and easily see those questions the internet is dying to have the answer to.

Because the internet is asking the questions, they're going to be a mix of serious and lighthearted, and you'll find yourself sharing all kinds of things you won't find in a normal interview.

Comments will be removed under a few circumstances:

• Abusive or harassing comments.
• Requests for personal favors from the OP (For example, "OP, can you send me a signed autograph").
• Top-level comments must ask a question.
• Comments where there would be no possibility of a real answer, especially where it is deliberately creepy or offensive.
• "I bet OP won't answer this"-type responses, which usually come after the OP has finished responding to questions.
• "Fluff," non-contributing responses from users, responding to all of the OP's comments for karma/attention.
• Repeatedly asking the same question.

Please, don't attempt to bypass the rules by adding a ? to a nonquestion.

Questions must be directed toward the individual(s) doing the IAMA.

Under our policies, astroturfing is the practice of an individual or group of individuals who plant questions in an IAmA post for a particular purpose. This kind of behavior is forbidden.

UPD 1: Oh, I'm kind of tired of doing this. Please keep asking questions. Tomorrow I will answer again. Don't forget to vote for the most interesting ones. I will not have time to answer everything, I will choose something. Thanks!

UPD 2:

• 8 days ago, # ^ |

  As a top contributor, do you enjoy my blogs?

  Only partially. I like more the posts that develop our community in terms of CP domain. But as an entertaining contest, it's quite funny. I didn't read everything, but some of the pictures amused me. I think that I have upvoted some of your posts.

• 8 days ago, # ^ |

  As someone who was not on CF during Monogon's contribution crusades, did you upvote?

  Honestly, I did not understand what your "who was not on CF" means. But yes, he has a lot of cool posts. And I upvoted them a lot. I am grateful to him for his contribution to the community!

• 8 days ago, # ^ |

  What are your views on anime?

  This is not my genre. Maybe for now? I haven't even watched classic anime films. A year ago, I was in a movie with my daughter watching https://en.wikipedia.org/wiki/Weathering_with_You It was good, I didn't fall asleep!

  • 5 days ago, # ^ |

    Ah, Weathering with You! Really a nice film! :)

• 8 days ago, # ^ |

  Do you play any video games?

  No, but years ago I loved this thing. The last thing I remember, about 20 years ago, I played Diablo 2 in between solving programming problems. I allowed myself to play after solving the problem. It seems that I quickly realized that I like programming more.

• 8 days ago, # ^ |

  As a top contributor how do you feel being ratioed

  • 8 days ago, # ^ |

    ← Rev. 3 →  

    -82

    The comment is hidden because of too negative feedback, click here to view it

    • 7 days ago, # ^ |

      -69 contribution

  • 7 days ago, # ^ |

    Could you please explain why so many dislikes on this guy's comment?

    • 7 days ago, # ^ |

      I think that they are angry with Mike didn't answer their question?

• 6 days ago, # ^ |

  The comment is hidden because of too negative feedback, click here to view it

  • 5 days ago, # ^ |

    how exactly does this comment contribute to the community

• 8 days ago, # ^ |

  ← Rev. 2 →  

  +98

  Any thoughts on a new CF rank?

  It seems to me that I would rather adjust the boundaries of the ranks than introduce a new one. I don't think we need to increase their number. But I wasn't thinking strongly in this direction.

  • 8 days ago, # ^ |

    Since we have your attention on this topic, have you considered implementing something like "hysteresis" feature for rank promotion/demotion? Right now some people reach their new top rank and stop competing, presumably because they are afraid of losing their new shiny title right after the next contest. Some of these people even start creating alt accounts to have a risk-free fun. This behaviour is probably undesired.

    I think that a simple fix would be to require an account to stay in the new rank zone for 2 contests in a row before the new title is locked in. And in the same way, don't demote someone until they stay in the lower rank zone for 2 contests in a row.

  • 8 days ago, # ^ |

    any update on updating the ui of website

  • 7 days ago, # ^ |

    ← Rev. 4 →  

    -40

    The comment is hidden because of too negative feedback, click here to view it

    • 7 days ago, # ^ |

      300 gap is too huge

    • 7 days ago, # ^ |

      ← Rev. 2 →  

      -10

      Just remove IM and add +100 to everything below it. Also, there is no hard limit that everything should be 200 rating points wide just because they look good. Boundaries should be based on rating distribution and skill set rather than some fancy figures.

      Expert is 300 because there is not much difference in skills between 1650 and 1850.

      When these titles were introduced Master and IM was 100 rating points wide. Just revert to that and it's fine.

• 8 days ago, # ^ |

  Also do you prefer cereal before milk or milk before cereal?

  I remembered everything that I had eaten lately and do not remember the cereal there. But I would rather add milk to the cereal than vice versa.

• 8 days ago, # ^ |

  Hi! I always had this question, Where did you learn to create websites like codeforces or polygon? And also do/did you participate in any competitive programming contest?

  I love programming contests and have participated in them a lot around 1999-2006. Probably, my main achievements can be found at the link https://cphof.org/profile/topcoder:Mike%20Mirzayanov It seems there is no information there, but as part of the team we won twice at IPSC https://ipsc.ksp.sk/

  I was also always interested in inventing, organizing and creating something. Around 2000, I was involved (and did most of the work) in building the site acm.sgu.ru

  How did you learn? Somehow I didn't think that I studied especially. I perceived all this as an implementation problem at the contest. Well, you need to write something, you write it. Yes, before that it is worth reading how smart people do this.

  But I've always had an interest in technology. Around the 2010s, it was my favorite pastime — to take some new framework and make a minimal project on it with my own hands. I've tried a lot of things and figured out how it works.

• 8 days ago, # ^ |

  would you ever add the feature to see the people who friended me?

  No, I think these are the basics of privacy. Do not wait for such a feature. It is possible that I will rename friendship as subscribers or similar, idk.

  • 8 days ago, # ^ |

    I think something like watchlist (idk) might be more appropriate, since "subscriber" implies that we are subscribing to something.

    • 8 days ago, # ^ |

      Or maybe you can "favorite" users instead. Or even if we want to stay similar to what we have now, maybe to be able to "star" a user.

  • 8 days ago, # ^ |

    Any chance to also implement custom tags for this list? Right now my codeforces friend list includes different categories of people, such as "rival with a similar rank", "rust language user", "famous frequently reported cheater", etc. People are added there for very different reasons and surely not all of them are "friends".

    • 8 days ago, # ^ |

      There is a "lists" feature on cf

    • 8 days ago, # ^ |

      Do you mean that the friend standings in a contest should have all of these multiple categories? This feature is there in Codechef, but I find it a bit complicated.

  • 8 days ago, # ^ |

    Follower is a better term than subscriber I think

  • 7 days ago, # ^ |

    Follower is what I was thinking of too.

• 8 days ago, # ^ |

  What work in cf you find to be most tedious and uninteresting?

  It's boring and tedious to deal with cheaters. It takes a lot of time and effort. I also do not like the bureaucracy (documents, etc.), which sometimes have to be done.

  And what do you find the most interesting?

  My favorite thing is to go into a flow state and write something valuable and complex. It happened that I was in this state for weeks — I ate and slept little, but I wrote a huge amount of complex code. Unfortunately, nowadays it is so rare.

  It is also nice to solve some complex technical problem that didn't allow the system to work correctly for a long time.

  Of course, it's interesting to come up with something new and implement it. Here, for example, diagnostics for C ++ on the website — it seems like no one in our community has done this. It's a funny thing.

• 8 days ago, # ^ |

  Do you secretly compete in codeforces rounds for which you haven't tested or set problems?

  No, sometimes I read problems and come up with solutions. But rarely. More often than not, I just monitor different metrics and just how it goes. Sometimes I suggest something to coordinators or even help something if something goes wrong. But this is rare (and the less often the better). More often than not, I do some other work in parallel.

  Do you know which rounds I think are the best for me? There were several rounds that I completely forgot about. There was a contest recently Educational Codeforces Round 120 (рейтинговый для Div. 2). At this time, I was getting a massage, and somewhere at the end of the massage (and the round) I remembered about it. I went to Codeforces and saw that everything went well. Wonderful feeling!

• 8 days ago, # ^ |

  ← Rev. 2 →  

  +156

  I believe that Nikolay KAN Kalinin makes a huge contribution. He has a slightly invisible job to you, but he's just a monster. Many thanks to him!

  In general, the greatest contribution in my eyes is work of problem writers and coordinators. Many here underestimate the complexity and importance of this work.

  • 8 days ago, # ^ |

    ← Rev. 2 →  

    +324

    "the greatest contribution in my eyes is my work"

    • 8 days ago, # ^ |

      a typo to be sure, but a welcome one

    • 8 days ago, # ^ |

      Common, it was a typo or incorrect translation of Google translate service. Most answers I preliminary translate in it, and do some manual fixes after it.

      • 8 days ago, # ^ |

        Oh, sorry! I actually didn't even realize it was a typo, since you do personally contribute a lot to problem-writing and coordination (at least that's my impression).

• 8 days ago, # ^ |

  When can we see you participating in a cf contest? Do you participate in the contests of other platforms?

  I do not exclude that someday I will start taking part in the rounds. But in that case, I will not make it secret. I'll let you know about it.

  Now it is rare, but sometimes I participate. This year I took part in several stages of Code Jam and FHC. It was fun! I'm surprised that for my standards I did something well.

  Sometimes, in a team, I helped with testing some contests. It was fun too.

• 3 days ago, # ^ |

  Dark Reader plugin for Chrome)

• 8 days ago, # ^ |

  Do you read all DMs you receive?

  No. Sometimes I can get tens of messages a day. But I somehow look through almost everything. And I respond to the most important ones. Sorry if you wrote to me, but I did not answer.

  For administrators, in order for the message to be marked as read, we must either reply to it or explicitly click the "read" button. Therefore, sometimes there are, as it were, unread messages that I actually looked through.

• 8 days ago, # ^ |

  If you did contests, what would be your rating?

  It seems to me that I have already answered this question somehow. It will be funny if the new answer is very different.

  It seems to me that I will start with a high purple rating and if I put in the effort, I will quickly remember how to solve problems and become a low red.

• 8 days ago, # ^ |

  How can one join codeforces headquarters?What's the procedure??I am asking this because in this cruel world it will be a blessing for someone if he gets a chance to work for the things he loves for life

  One way: come to St. Petersburg, start studying at ITMO. It will also be a plus to study well and get significant success in the competition. We have the practice to involve students in development. If you are interested and you succeed, then a long road together can await us.

• 8 days ago, # ^ |

  Top 3 competitive programmers according to you?

  For now, I will name two.

  tourist Нe is a living genius. Go to any top CP platform right now and look at the first line in the ranking. You will most likely see him there. But that's not all. I was fortunate enough to discuss problems and not only with him. I really like it. And I listen to Gennady's opinion not only on problems, but it is also important for me to hear his thoughts on various aspects of the system.

  Petr Peter. He was a legend when I competed. He sparkled brightly and for me, he is a model of the ideal top participant. He was top 1. He looks like a person who sees a lot of interesting things in life. He is successful at work. I was delighted to meet his child. He loves sports. In another life, I would like to be like Petr.

  • 8 days ago, # ^ |

    slightly corny message but something that needs to be said

• 8 days ago, # ^ |

  Is the Codeforces stuff large (excluding coordinators)?

  No, now it is about ~4 people: me, KAN, Una_Shem, geranazavr555. Help a lot for a long time kuviman, stingray and some others.

• 8 days ago, # ^ |

  What was the hardest feature to deploy?

  I do not know exactly. Perhaps organizing VK Cup 2012 was the most difficult for me. Then I put a lot of effort into making it go well.

  But in general, I often felt too tired. Excessively. In the first 5 years of the project's life, I worked 12-16 hours a day and often slept very little. I had no other interests (well, maybe, except for a little participation in my daughter's life). Well, I also worked at Saratov University and was engaged as a head of the programming competition center and coach. I guess I have never neglected my duties.

• 8 days ago, # ^ |

  What is your views on other CP sites? Have some of them inspired you?

  Nowadays it is rare. But I am following the development of other platforms and am happy about their success.

  I can't say that any platform inspired me exactly. Sorry, I got the most inspiration from Topcoder when I was active there as a participant and problem writer. It was about 15-18 years ago.

• 8 days ago, # ^ |

  ArnoldC was used in one of the 1st April's contests.

• 8 days ago, # ^ |

  They was used on surprise language rounds and April Fools Day Contests. Hi, Nickolas! And thank you :-)

• 8 days ago, # ^ |

  There are no such plans now, I see no reason.

  • 6 days ago, # ^ |

    Thanks for having no such plans. One of my best friends was a regular contestant and we lost him a few months ago. Now, sometimes I visit his profile. It's like a souvenir.

    • 6 days ago, # ^ |

      I'm sorry for your loss

    • 6 days ago, # ^ |

      sometimes I visit his profile. It's like a souvenir.

      Gave my goosebumps.

    • 5 days ago, # ^ |

      I'm sorry for your loss.

• 8 days ago, # ^ |

  Honestly, I find these kinds of questions pretty silly. Please do not be offended. It seems to me that they do not say anything about a person.

  Perhaps I cannot answer about the number and color. Ice cream I like, it is simple: good white sundae. I like to sprinkle it with chocolate, nuts, or eat it with berries. But I'm hardly a sweet tooth or a fan of any food at all.

  • 8 days ago, # ^ |

    I expected you to say "vanilla icecream" :(

• 8 days ago, # ^ |

  Every round I clean out hundreds of cheaters. After each round, I ban dozens of accounts. I believe that most of cheaters are punished.

  • 8 days ago, # ^ |

    The comment is hidden because of too negative feedback, click here to view it

    • 7 days ago, # ^ |

      That's the neat part — you don't.

      Cheaters who don't go to the length of copy-pasting solutions don't get caught. How do you tell them from normal participants, anyways? If they shared their ideas and re-implement it from scratch or near-scratch, then there would be no similarity at all with other solutions beyond the core idea, which is both extremely harder to check if they're similar (as in, I'm unsure if it's even theoretically possible), and have an extremely high rate of false positives.

      In general, any open competition without authority to monitor their participants and have some sort of reward must always expect rampant and pervasive cheating (this applies with even highly prestigious competitions with the ability to exercise heavy punishment on cheaters, actually). We can only prevent it's worst excesses, to hope to deal with every cheater is just childishly naive.

• 8 days ago, # ^ |

  Any plans of conducting long contests for educational purposes?

  There are no specific plans. But! Imagine, a well-known authoritative writer will come to me and say "I want to hold such a long educational contest." I will help and we will do it.

• 8 days ago, # ^ |

  Views about keeping an intermediate level contest between Div1 and Div2 (maybe some sort of Div3/2 XD )

  There are no specific plans. But! So far it seems to me that the div2 rounds are already complex enough to be interesting for many, who would you call div 1.5.

  If we start doing this, then most likely it will go to the detriment of the current schedule. It will say that there will be only 2 div2 and 2 div1 rounds per month (on average). Do you think it will be better?

  On the other hand, we will probably do it somehow. Or already? Well, it can out that the round is slightly easier as Div1, and there is no good difficult problem. It can be Div 1.5.

  In this question, I will strongly listen to the opinions of other coordinators, especially to KAN.

• 7 days ago, # ^ |

  ← Rev. 5 →  

  +8

  There was a long-time contest in the Gym last week Abakoda 2021 Long Contest, organized by NOI.PHI, The National Olympiad in Informatics, Philippines. It was aimed for beginners and novice competitive programming contestants.

  I would definitely vote for organizing more yearly long-time contests and/or monthly long-time contests in Codeforces, even if they would be unrated contests.

• 8 days ago, # ^ |

  I became interested in such problems at the age of 15, but only at the age of 17, I was able to intensively develop in this direction. Until the age of 17, I did not have a computer and did not have constant access to someone else's. It was a different time, my friend.

• 8 days ago, # ^ |

  Isn't free contribution enough ;) ?

• 8 days ago, # ^ |

  Oh sure. And this post motivated me to do this session. It turned out badly that time. We tried to record something, but to be honest, my English is too bad for such an activity. It's hard for me to speak freely on general topics so that it still looks normal in the recording. In short, I did not like what happened.

• 7 days ago, # ^ |

  I thought about such a thing. Perhaps it makes sense to hide uninteresting comments and comment threads in old posts by default (sort of like archiving). Of course, we must leave the opportunity to see everything. But in fact, there are a lot of comments that become uninteresting and even trash in terms of meaningful discussion.

  Sorting functionality sounds interesting, but this is a comment tree, not everything is so simple.

• 7 days ago, # ^ |

  I read almost all meaningful posts. But sometimes I miss something. Recently there was a post about a survey, I skipped it. I read about it only in the report.

  It seems to me that I spend about an hour a day on reading posts and comments in total. But I still do a lot on Codeforces besides this, so much more. For example, yesterday I spent a significant amount of time launching testing of the upcoming div3 round and working with problems, writers and testers.

• 7 days ago, # ^ |

  We can also support this field. This requires some effort, if it were quite simple, then I would have done it already.

  But the current API provides all the information (history of rating changes) so that the tools can be fixed. But that hasn't been implemented, has it? If so, then the authors or the community (if it is open source) did not make an effort to fix it.

  • 7 days ago, # ^ |

    ← Rev. 3 →  

    +38

    Predicting rating changes after a contest is at the moment almost impossible since that calculation needs the (hidden) true rating.

    Rating prediction used to work flawlessly, but with the true rating being hidden, every rating prediction tool out there that I know of is currently broken. Maybe in theory it is possible to find the true rating by going over the entire history of rating changes of all participants, but the sensible solution would be for CF to have trueRating in its API.

  • 7 days ago, # ^ |

    What you are suggesting is infeasible. For a browser add-on to provide a rating prediction it would have to spend around 11 hours querying the API, history of one user per 2 seconds for 20k users. This is too long, since we want live predictions.

    • 7 days ago, # ^ |

      Think wider: it can access it's own backend, not just Codeforces API. It will make many things much easier, faster and more reliable. And this backend can easily maintain number of rated events per year. Also such approach will significantly reduce load on Codeforces. As I know some plugins use similar approach. I don't see a reason why it is not a case here.

      • 7 days ago, # ^ |

        Again, infeasible. Codeforces has hundreds of thousands of users and rating changes are unpredictably rolled back. With great effort we can make something that would work semi-reliably but we want to do it properly, which we cannot.

        In any case we have to employ a heuristic to decide whether someone's rating is real or not. For example I thought maybe check is first rating above or below 1000 but look at qazsxdew!

        • 7 days ago, # ^ |

          Sorry, it seems I was wrong: I was sure there is a method to get rating history for a group of users (not only one). I confused user.info (which supports it) with user.rating (which doesn't support it). In this case I think I'll find a way to support something like trueRating in the nearest future.

          BTW, are you the author of the rating predictor plugin? I don't like it uses user.ratedList aggressively. Is it enough to return only users from a contest: I mean you pass extra parameter like ?contestId=1234 and we return only participants of this contest instead of all users? Does it work for this plugin?

          • 7 days ago, # ^ |

            I am a co-author of TLE (the discord bot). There are multiple rating predictor plugins. One of them is carrot, written by my fellow TLE co-author meooow. It hits user.ratedList only once.

            What you are suggesting can sort of be done already by querying contest.standings and then repeatedly querying user.info.

            Can you please create either a dedicated blog post or an issue tracker for codeforces API? Plugin developers have had many grievances over the years and while you have addressed some of them after a DM, we understand that you are drowning in DMs and consequently some of them were lost.

          • 3 days ago, # ^ |

            Hi, I've tried to make sure my plugin (which is not the only one of its kind) does not use the API frequently, but I'm happy to discuss the details and adjust things if need be.

            I agree with mblazev that a reliable means of communication between the Codeforces team and plugin developers would be really helpful.
            I could be wrong, but the API feels like the most neglected part of Codeforces today.

• 2 days ago, # ^ |

  After moving from Saratov to St. Petersburg (ITMO), I now have the opportunity to work on Codeforces almost full-time! I am only slightly involved in teaching at the university and am grateful that the authorities support my work on Codeforces. I don't do any research. I think I tend to be more practical than research.

• 7 days ago, # ^ |

  So that it's easier to see a GM SecondThread!

• 3 days ago, # ^ |

  "socket-style interactions" ... very interesting way to put it. I always thought, at the end of day a normal HTTP GET request also goes through a socket but when we call it "socket-style interaction" for some reason it intuitively sounds like streaming. IDK why

• 7 days ago, # ^ |

  Seconding the question--that said, I'm not a huge fan of the Edu/Div. 3 hacking system because I don't like waiting 12+ hours to see if I FSTed. I find it frustrating to discover that my solution to a problem is wrong long after I originally submitted my solution.

  Another issue I have with the current hacking system is that in the case where pretests are intentionally made weak (hacking is generally irrelevant when pretests are strong), it penalizes FSTs far more severely than failing pretests, when intuitively there's no reason one is a worse offense than the other. Secondarily, in order to determine your contest strategy (e.g. whether to attempt to find hacks, whether to stress test solutions that passed pretests, etc), you must essentially guess whether pretests are strong. In contrast, my understanding is that TopCoder's hacking system avoids these issues by making all pretests into sample tests. This means that FSTing is clearly distinguished from failing pretests (pretests are given to you as samples, system tests are not), and you can tell for yourself how strong pretests are (since you know exactly what cases were covered by the pretests).

  Unfortunately, I don't like TopCoder's approach either: FSTs generally introduce more variance than I'm comfortable with, and I find that TopCoder's system forces me to debug code to make absolutely certain there are no errors before submitting it, whereas on CF, once I'm reasonably confident my solution is correct, I can submit it and hope for the best, debugging only if it's wrong. This means that on CF, assuming pretests are strong, I can spend more time thinking about problems and less time debugging.

  For context, my preferred solution is to consistently make pretests as strong as possible, limiting them only as much as is necessary to prevent queue issues. Then, either eliminate hacks altogether or shift them to a short phase after each contest (much like TopCoder), in order to prevent users from viewing each other's code during the contest.

  Assuming this doesn't happen, though, I especially think that authors who do not intend to write strong pretests (i.e., those who are intentionally leaving edge cases, maximal cases, etc out of pretests) should say so in their contest announcements. At this point, the default assumption is that pretests will be as strong as possible, so contestants should know in advance that their strategy will need to change for these rounds. This is especially important because not giving everyone this information advantages anyone who does know in advance that pretests will be weak--for example, if I know an author personally and know that they like hacks and try to write weak pretests, then that information would give me an unfair advantage over other contestants in choosing my strategy (i.e., I'll know that I should debug my solutions particularly carefully, then lock them and look for hacks).

  • 7 days ago, # ^ |

    Edu hacking system is good if you want a break after the contest has ended.

    I usually assume that the pretests are weak.

  • 7 days ago, # ^ |

    ← Rev. 2 →  

    +35

    To be precise, even sample cases are not tested on TC

    • 7 days ago, # ^ |

      Thanks for clarifying!

      • 7 days ago, # ^ |

        However there is a "batch test" option in Arena which is notoriously known for having a first line result called "Success" which could be either "OK" or something negative, which actually doesn't tell you a thing about whether your output is correct! It just tells you whether your code executed properly (i.e. no RTE or TLE). To check the output correctness you need to look deeper, there is some other line telling you that in the middle of ten other ones. There have been multiple cases of people submitting their solution thinking that's ok on sample even though it wasn't cause they were deceived by this

• 2 days ago, # ^ |

  It seems to me that here it is impossible to say that something is strictly better or worse. There are pros and cons. It is possible that the rules will change.

• 7 days ago, # ^ |

  Usually, we try to have 3 div1 and 3 div2 rounds per month. Sometimes it turns out more, sometimes less. But on average, we manage to follow this schedule. I glad if we manage to conduct two Div3 rounds a month. Regarding sponsorship rounds, this is not the main goal for us. It is an important source of income, an indicator of business interest in the community, but nothing more. It's more important for me to understand that the platform contributes to the life of the community and the community is alive. Rounds, posts, trainings, championships and camps are all part of our community life. I plan to continue developing Codeforces so that it is useful, convenient, and irreplaceable for the community.

• 7 days ago, # ^ |

  Hi! It's not about money! It's a delivery issue. Since the start of the pandemic, there have been huge delivery problems. Many countries (including Brazil) don't accept parcels from Russia. A lot of packages are lost in the process or are returned to us after 4 (or more) months. In addition, the customs rules of the countries began to change a lot. The same problems affect manufacturers. This all complicates and slows down the work. :( We sincerely apologize for the delays and are working hard to remedy the situation. This is very frustrating for us too.

• 7 days ago, # ^ |

  And I’m sorry I answered instead of Mike. But it’s kind of more my thing. :)

• 7 days ago, # ^ |

  Rather yes than no.

• 7 days ago, # ^ |

  Let's wait. But the value 4000 is really very symbolic. It seems to me that such a success will really deserve something special!

• 7 days ago, # ^ |

  Yes, of course, the type of problems changes over time. This reflects the mood in the community. Both from the side of the participants and from the side of the writers.

  Personally, I am not so thrilled with the large number of math puzzles in our rounds. I am absolutely sure that sports programming helped (was crucial) what kind of developer I became.

  You know, in fact, in the 2000s, the type of problems was more biased towards implementation. And I loved it. For some reason, I knew how to enjoy the fact that again I quickly wrote non-trivial logic of some kind of modeling. Or implemented some parser again. Each time I did it a little differently. I wrote faster and more reliably. And it brought joy. Maybe it even looks a little like a dance: the music is the same, the movements are the same, but it's cool again and again!

  It is important to understand that fluency in coding is one of the most important things you can learn from our contests.

  On the other hand, coordinators have a much better understanding of today's problem needs than I do. I trust them. They can handle it!

• 7 days ago, # ^ |

  With a small team, we can't just develop the API, invest a lot of effort into it without a strong reason. Let's say there are some specific successful tools that help the community and are objectively is popular. And these tools need some API improvements. I think then their developers should try to discuss this idea in the form of a post here. Most likely, in this case, the post will have a lot of support. This can be an argument for investing efforts in this direction.

• 7 days ago, # ^ |

  Perhaps it is in vain that you downvote this question.

  Did the covid pandemic change your attitude to life?

  Fortunately, it was the pandemic that did not affect my attitude towards life so much. Although I was ill with covid. It would be better if this did not happen to me. Anyway, I cannot boast of excellent health. Unfortunately, in recent years this has been especially noticeable to me.

  Have you start thinking who would inherit Codeforces after your death?

  Well, that was an important milestone for me when I realized that my creation was probably able to outlive me. This is amazing! If I suddenly get hit by a bus, then a group of activists will have the opportunity to save everything and continue development. It seems to me that there are no secrets (passwords and keys) that I will take with me in case of an unexpected bus.

• 7 days ago, # ^ |

  At about 25 years old I started to think about them.

  • 4 days ago, # ^ |

    Wow it's amasing

• 7 days ago, # ^ |

  Why is Mike Mirzayanov?

• 7 days ago, # ^ |

  Why do you think UI has to be changed? :)

  • 7 days ago, # ^ |

    Not really, just a suggestion

  • 7 days ago, # ^ |

    honestly, it looks old (especially icons) and may cause rejection of new users

    • 7 days ago, # ^ |

      I can tolerate replacing the icons with newer ones but other than that, I don't want to see a redesign. I have seen very few UIs designed in the fashion of 2015+ that are even remotely usable.

      • 7 days ago, # ^ |

        The comment is hidden because of too negative feedback, click here to view it

        • 7 days ago, # ^ |

          We don't code at CF because its UI is old, CC is good compared to it"

          It's the other way around me . Watching a weird looking chef staring at my soul endlessly distracts me and ruins my contest experience. That's one of the reasons I suck at codechef.

        • 7 days ago, # ^ |

          Having used both Codeforces and CodeChef UI recently, I think that Codeforces is superior in every aspect. Fewer clicks are needed for almost any action, all important information is presented in a clean, compact and very readable way. If something isn't broken, don't fix it.

          That said, I'm browsing the Codeforces website from a desktop computer and this probably makes a big difference. Modern trend is to optimize website UI to be browsable from mobile phones with small screens and touchscreens. This happens at the expense of PC browsing experience, which becomes worse.

        • 7 days ago, # ^ |

          CodeChef UI looks UGLY AF, Codeforces is so much advanced compared to codechef.

• 7 days ago, # ^ |

  Look what happened to TopCoder. They fell into a loop of constant improvements of UI and every single update made it even more unusable than we have ever thought it is remotely possible

• 7 days ago, # ^ |

  I don't think there is a need for this. Indeed, the UI is outdated and just not very good. For example, icons, maybe I'll sit down and change them to more modern ones. But in the conditions of a small team, lack of strong expertise in this, and just the amount of work — I don't think it's right to drop everything and focus on this.

  I've witnessed a lot of redesigns that only made things worse in the end. It is challenging and deep work to redesign the look and feel of a working system with a large user base, history, and formed behavior patterns without breaking the UX. Big companies with world-class experts have failed this task.

  Rather, I would make a series of some incremental changes like updating icons, changing the font size, and unifying some of the non-uniform controls.

  • 7 days ago, # ^ |

    by modern icons do you mean a complete overhaul to minimalistic/modern or simply high definition versions of the current ones? i am sure the community, including myself, can help with redesigning something like icons.

  • 5 days ago, # ^ |

    Please don't make any changes . I love it the way it is . Simple and elegant .

• 2 days ago, # ^ |

  Of course not, but thanks. In this photo, I recently cut my hair. I need to get a good haircut more often!

• 2 days ago, # ^ |

  Is the top tab in the main menu exactly what you ask? It should top recent blogs (excluding which were published on the main page).

  • 61 minute(s) ago, # ^ |

    That's what I meant, yes. Didn't know there was already a feature like that until now.

    Hopefully something similar for comments too. (:

• 6 days ago, # ^ |

  EU individuals should be able to request GDPR-based "right to be forgotten" although idk how well this is enforced

  • 6 days ago, # ^ |

    EU means European union??

• 6 days ago, # ^ |

  The comment is hidden because of too negative feedback, click here to view it

  • 6 days ago, # ^ |

    ← Rev. 2 →  

    +24

    Why? If I recall correctly Codeforces is written in Java.

    EDIT: and supporting MSVC++ as a compiler isn't that useful, it's just strictly worse than GCC for competitive programmers.

• 2 days ago, # ^ |

  Yes, the website is slightly different for administrators. For example, there are moderation and administration tools. If I see that some trash in recent actions has appeared, then I can send a writer to read-only mode (and remove the trash). I also do most of the work on removing cheaters through the website. And administrators are shown some additional information (logs) in order to notice the degradation of the performance of some functions.

  

• 2 days ago, # ^ |

  But by the way, I don't remember the awkward ones. I remember lucky. Once I met a girl, we walked around the sights of St. Petersburg. And the guys came up, recognized me. They asked for a photo together. We had a great conversation. Then I found out that the girl for several years thought that I had specially set up the meeting in order to impress her :-)

• 2 days ago, # ^ |

  For now, I have no plans to open the Polygon or Codeforces source code yet. I don't think it will change anything for the better. It is wrong to think that as soon as you open it, millions of experienced programmers will immediately begin to make high-quality changes and improvements. Unfortunately, it doesn't work like that. Both projects are no longer so small (in total, they are ~20 megabytes of code). This is complex code that requires a lot of immersion to make edits in it. And I'm not just talking about the fact that it is very easy to break something (not only in terms of functionality but also performance). But in order for it not to turn into an unsupported mess, it is necessary to adhere to the laid down principles and share them. Implementing the development process in a mature open source project is a world of its own with a lot of work and effort. Most of the improvements to large open-source systems are not done by enthusiasts, but by employees of companies who are paid for it.

  And of course, it's a myth that open source itself improves quality or reliability. Heard about the log4j2 issue? Heartbleed? There are more such examples.

  It’s even more important to understand that you don’t need to fix what works. Polygon and Codeforces have been constantly developed for 13 years. They are getting better. We have a sustainable model for how to do this. Other projects may use a different model, but for me personally, the classic in-door development is closer to me.

  By the way, for example, the testlib code is open source: https://github.com/MikeMirzayanov/testlib This is an example of a simple project with microscopic code size. It is written in simple C ++ that almost everyone in the community knows. He is 17 years old and of course, there is a lot to improve now. Even so, the number of community improvements is minimal, and all commits require my careful review. In practice, any review took more time than fixing something yourself.

  By the way, I have contributed several times to improved open-source projects. I really love Far Manager. And this checkbox was implemented by me (it is a separate interesting story about how it all happened).

  

• 5 days ago, # ^ |

  The comment is hidden because of too negative feedback, click here to view it

• 2 days ago, # ^ |

  Thanks for the suggestion. Definitely, this page needs to be fixed.