Slovak Camp 2026 Blog
We were invited by the Slovakian Olympiad in Informatics to join their IOI selection from 29th of April until 6th of May. The Swiss delegation consisted of Albert Kovacs, Benoît Schmit, Máté Makai, and Laetitia Sophia Orglmeister, accompanied by Cheng Zhong and Ursus Wigger.
Tuesday, 28th April
We took the night train in the evening from Zurich to Vienna. The train ride can be broken down to:
- Culinary Water from both Austria and St. Gallen!
- Molotow, Tichu
- We managed to lock the door.
- No smartphones have been stolen (yet).
- A semi-successful attempt of sleeping
What will await the delegation this week? Will the phones be stolen? How many tasks will they solve? What has gambling to do with this camp? And most importantly, will they finally go to the beach?
Well, seems you need to read the blog posts to find out!
Wednesday, 29th April (Albert)
After having slept surprisingly good but also bad because the bed was about 30cm shorter than my body, we were woken up at 6 am on the train with breakfast. The bread was pretty good for train food, the coffee on the other side was basically just some water with some brown color and a bit bitter. Until 7:51, when we arrived in Vienna, we ate and played Tichu with Ursus being Jesus that literally sees all the cards from above (we had bunk beds).
The train ride from Vienna to Bratislava wasn’t anything surprising, except that the announcements always sounded as if the next halt would be hell because of some nice sound effects like feedback and what not.
In Bratislava we didn’t need much time to figure out, that Slovak people like to be vague. For example, sometimes they don’t tell you, from exactly which platform a train will be leaving. They only tell you, that 6 > platform number.
The buses took this to a next level though. Our bus was arriving in ~<6min . I still find it hard to understand what exactly it means, especially because later it switched to ~1 min, than to ~<1 min than to <1 min and about 8 minutes later, the bus eventually arrived.
Before the first contest we went on a walk and found many beaches! (fountains) And also learned the hard way that poles don’t move out of the way just because you’re looking at your phone.
In the evening we wanted to play some games, but the hostel has quiet hours between 22:00 and 6:00. We thought, that we could just play quietly but wanted to test, how soundproof the walls are. I thought, that I would leave the room and just listen, if I hear the others talking inside. Instead, I heard a loud WHOOOOOAAAAAA!!!!! The others understood the situation pretty quickly when I was laughing while entering the room…
Thursday, 30th April (Benoît)
We woke up at 7:50 (some of us at 8:29), ate breakfast, and then ran to catch the bus at 8:41. Once again, our bus was late and then got canceled. We arrived late at the contest, but it hadn’t started yet. The contest took place, as yesterday, in a very nice sky tower.
Hopefully, we didn’t have a geometry task, but instead a graph problem and a math problem.
At lunch, we had a buffet, and it was pretty nice. I got half a chicken and some other things. Then, five minutes after finishing lunch, we were asked to go back to the contest room.
In the afternoon, there was a task that was easy to understand but hard to implement. Since the grader was very slow and everyone submitted a lot in the last 30 minutes, we had to wait about ten minutes for each submission to be checked. I probably missed one edge case, which gave me 2 points on that task. Then there was a geometry task involving distances and odd cycles. The last one was a difficult dynamic programming task.
After seven hours of the contest, we were invited by the Slovaks to play StepMania, and we saw how much time they had spent on this game instead of working. They were crazy: more than 7 notes per second, and they still had a win streak of around 300.
Then we tried to find a place to eat, and Ursus really wanted to find a place with Kofola, a kind of “sugar-free” Coca-Cola, which, after checking, actually contains more or less the same amount of sugar as Coca-Cola. On the way, we found a sign that we tried to understand, but we still don’t know what it means.
We found a restaurant that looked cool, but unfortunately it was full. We still found a nice place to eat (with Kofola) and played Tichu.
When we came back to the hostel, we played some games on Discord, and since it was already past midnight, some people tried to solve the daily task, but I went to sleep.
Is there a correlation between being late and trying to solve the daily task? Read the May 4th blog post to find out…
Friday, 1st May (Laetitia)
As always we were of course all able to wake up on time* to go to today's contest (sleeping is overrated, solving the daily tasks right after release is way more important). After passing many beaches, we arrived again at the contest tower. So that begs the question ...
What will the tasks be today?
Indeed. In the morning contest there were two problems, both of which were geometry tasks. Once again you had to use convex hull, twice in the same task. Considering there were Geo problems in both of the previous days aswell, they surely don't have any more Geo tasks for all the other days (stay tuned to find out).
For lunch we got some pizza and just after that we had to go back for the second contest (it always says it starts at 15:00, but for some reason they then always move it forward to 13:00). Since there were already two Geo tasks today, they obviously cannot bring another one (actually you can never know in slovakia). So what will they find instead?
At least there was no geo this afternoon (Maybe there will be Geo tomorrow again? Hopefully not...).
The first task was quite nice, it was just an intuitive special type of binary lifting with some virtual nodes (some tried to do a binary search somehow, the intended solution used union find for some reason?). The third task was a bit of a Smiling Imp moment, as it required a lot of constant optimisation, resulting in the solution passes being random (of course we have experience from that due to the Artcollege task from the SOI Camp, with slower solutions sometimes passing while faster ones failed). Now the last (second) task: You had 10 input files and had to generate 10 mazes with as many dead ends as possible, no cycles, and everything being connected. You had 4 hours during which your program could run, but there was a twist: As you may know the computers on which we wrote the contests are 🥔🥔🥔 with an average of 3 fps while writing code. Once you started your maze generator the screen effectively turned into a painting. So while you had your program running, you couldn't do anything else (Brüdi).
So what was the solution for the maze generation? Well, first you pick some random tiles and add plus shapes, and when you two paths get close to each other, you start a random dfs in the wall (what does it do? no one knows) and if it finds at least 4 potential new leaves it is worth it to sacrifice a leaf (why 4? Smiling Mao.). But if it finds 7 or more, you should sacrifice two leaves instead (why? of course magic once again). Even the person presenting knew not why it worked (most of the time).
After being cooked by the contest we did a quadratic search to find a restaurant to eat dinner at. On the way Máté even found a clean way to find the largest element in a list in O(n2) by comparing every two elements with eachother. Eventually we went to eat burritos where we pondered the question what a "Jackfruit" is.
In the evening we played CrazyTime and some other games. Unfortunately CrazyTime is no longer being sold, which is why I can 100% confirm that SOI is definitely going to release CrazyStofl in the near future to give everyone the opportunity to experience something similar.
Saturday, 2nd May (Máté)
As on all the other days we had two contests: one in the morning and one in the afternoon directly after lunch. Because it was Saturday, the building was mostly empty, apart from us writing our daily two contests. In the evening we went to a sushi restaurant which was also surprisingly empty (we were the only ones there), but the food was actually really good though, so it didn’t matter. On the way back to the hostel, we saw some more of Bratislava than we had on the bus ride to the contest site. We walked through some parts of the university and its park.
Sunday, 3rd May (Albert)
Crazy things happened today…
Laetitia was not the latest, Máté arrived even later. It wasn’t that big of a problem though. Also, today at the contest, there were no geometry tasks which was truly shocking. Not that I wouldn’t be happy about it, but it wasn’t normal. Instead, we got a truly horrible tasks for the afternoon contest, but I will tell you more about that later. And obviously, in the evening it was a crazy time (we were playing Crazy Time for a long time. I’m sorry for this horrible dad joke, but this is how we get people to play it. We ask: “What time is it?” and surely, someone will answer: “It’s Crazy Time!”)
Laetitia is the only participant among us that learned HTML, JavaScript, and so on. And the others discovered the power of vibe coding! So now you can choose between my actual blog post which isn’t this short and Dancing Mice (surely, someone wrote about Stepmania on their blog post).
But before that, I will introduce a new thing to the blog posts from the Slovak Camp. The quote of the day: (Well, there won’t be one tomorrow, hopefully there will be one the day after tommorow. And after that, well I don’t know…)
“It’s illegal to draw Stofl. It’s like drawing god”
-Máté
Dancing Mice
Use Arrows or D-Pad
Mice
So, there was a task in the afternoon called "Ammeters". You could only get 15 points from it, so it should be trivial, right? Well, let me tell you something... If you understand the statement correctly, then yes. Otherwise, it's going to be hell. In the task, you get a graph with n nodes and m wires/edges. Following sentence is in the problem statement: "Each wire carries a current of some non-negative (possibly zero) magnitude in a certain direction." And of course, Kirchhoff's rule applies. And every wire has a cost, how much it would cost to put an Ammeter there (I hope, that is how it's called in English). The task is to be able to know for every wire, how much current is Following through it, with minimizing the cost.
I though, the sentence further above meant, that it was a directed graph. It did sound absurd, since wires tend to not care at all what kind of current flows in which direction, but whatever. Then, it made even less sense, where the current is coming from, and where it is going. And surely this means, that if a graph isn't a cycle, I don't even have to measure anything right? And in this case, I only have to measure one bridge, since surely no energy will spawn in a cycle, right? I was unsure, so I asked if this graph was a directed graph. The answer was less. I had some complicated solution, but I was pretty sure, that it was incorrect, because it was harder than 15 points. And it turned out to be incorrect.
After the contest, the other participants asked me, how I couldn't figure out, that it was a simple MST (not a hard algorithm). Well, I was solving the task for a directed graph, and it did not make sense until the end, when it turned out to be an undirected graph. Ach... I even asked them.
For dinner Ursus chose a burger restaurant which Reddit users described as a hidden gem. And it truly was hidden in some ways. From the outside it didn't look like it was a restaurant, but then it turned out to be actually pretty nice. (except that it was pretty full, so we had to sit at the bar)
One last side note: It was Sunday, meaning the people at The Spot (our contest location) weren't there. So once again, we ate cold pizza. But in my opinion, that's not that bad since it gets heavily compensated by the food in the evening.
Monday, 4th May (Benoît)
So we woke up again ate, went to the sky tower, and again we were late. There was one prefix tree and hash problem, and a DP in a circle problem. We ate lunch and for one time we had a free afternoon 🥳 we played games (once again step mania) then it was already time to go to the _beach_ restaurant. iItwas a nice place with Kafola (or Kafolo in the game). It was a big room full of decoration and I ate this:
And now there is a small résumé of what happened during the camp: For a better experience, put the sound on!!!
Tuesday, 5th May (Máté)
As usual, we had two contests today. In the morning, there was a regular contest, but this time the tasks were selected by the Swiss leaders and each task description featured either Stofl or Binna. The afternoon contest was a bit different, as it followed ICPC style more than IOI style. Unlike the other contests it had a live ranking, all-or-nothing scoring instead of subtasks and a penalty system as is usually seen in ICPC.
After that there was a rather odd sponsor talk held by a private high school called Skyro. They emphasized that they are a modern school that works with modern technologies and experts from big companies. They decided to give us a demo of making a functional cannon built in Unreal Engine. They didn’t seem to have much prior knowledge of the Informatics Olympiad, so they started asking us questions like:
“Do you know what a vector is?”
or
“Have you tried learning C++?”
As if we hadn’t spent the last few days programming in C++ and often trying to implement the convex hull problem (which is based on vectors btw). They also seemed very excited to tell us every now and then that what they were doing was “based on physics” or “based on math”, for example when an object fell to the ground in the game engine. However, they didn’t show us any formulas or even a single line of code. In the end we received some merchandise and ate pizza together.
Wednesday, 6th May (Laetitia)
Since we all know that the actual blog posts are by far less interesting than certain other things contained on this site I will keep this section short such that you may enjoy the second part.
Since this was the last day, and we had to check out of the hotel, we were actually able to all be on time, something which hasn’t happened this week before. But there still was one contest today. And what were the tasks? Of course, they were such 😈. One was a weird task about segments covering a line with chronology, which for some reason you just ignore. For the next one you had to calculate the median of some numbers on a grid. Why the median, and not the mean which would make way more sense? Just because. Even those who solved it had no idea why, and they just tried it and it worked somehow. 😈. The last task was truly 😈. I spent the entire contest debugging the Heavy-Light-Decomposition (which should work) and in the end it turns out you could do it just with a simple LCA… 😈😈😈. After the contest we went through the city in the search for beaches, and we even discovered a statue!
Afterward we went to eat dinner and subsequently went to the train station. This time however, the exact departure platform was written on the departure board! Even the journey to Zurich was uneventful, as not a single phone has been stolen!
Stofl Sweep
Now that all the unimportant stuff is out of the way, we can focus on the main part of this blog. The game!
- Defeat the Smiling Imp to win.
- If you go into negative quacks, you lose.
- In that case, click on Stofl to retry.
- If you collect enough cheese, click on Stofl to level up.
- Go to the beach to start!