Olimpiade Sains Nasional 2026 Bidang Informatika
We are pleased to invite everybody to our TOKI Indonesian NOI Open Contest 2026!
Summary
- TOKI Indonesian NOI Open Contest 2026 is an open competition mirroring Indonesian NOI 2026 (OSN 2026 Bidang Informatika in Bahasa Indonesia).
- The problemset is the same as the problemset in the original competition and will be given in both Bahasa Indonesia and English.
- The contest will be held on TLX (you may need to register for a TLX account if you do not have one).
- If you have any questions please contact sc-osn-2026@googlegroups.com.
Rules Overview
- There will be 2 (two) competition days.
- There will be 3 (three) problems on each day to be solved in 5 (five) hours.
- Each contestant may submit up to 50 submissions for each problem. This limit may be increased for some problems.
- The supported programming language is C++.
- During the competition, each contestant can only see their own scores.
Schedule and Technical Details
- The contests will be held at the following time ranges:
- Each contestant can select any 5-hour window within the time range to do the contest for each competition day.
- Both contests are available on TLX: Day 1 and Day 2. Please register in those contests.
- Each contestant may start the contest any time within the time range, by clicking the "start" button.
- There will be no additional time given if a contestant starts the contest in less than 5 hours before the contest ends.
- The open contest does not support clarifications, but all important announcements from the original contest will be broadcast as well.
Problems Types
- There are 3 possible problem types. Not all problem types might be given.
- Batch problems:
- The most common problem type.
- Given an input, contestants write a program that reads input from stdin and writes output to stdout.
- The program must run within the given time and memory limits.
- Example: Bogor Defence, OSN 2023
- Interactive problems:
- Contestants write a program that interacts with judge's program: judge's program writes to stdout and is given as the stdin for contestant's program, then contestant's program writes to stdout and is given as the stdin for judge's program, and the interaction continues until a given goal is reached.
- Contestant's program must output the answer within the given time and memory limits.
- Example: One-Day Delivery, OSN 2023
- Output-only problems:
- Contestants will be given all the input test cases.
- Contestants submit output for each input test case.
- Contestants do not have to write a program; the outputs may be solved manually.
- Example: National Defence, KSN 2020.
Problem Scoring
- For batch and interactive problems:
- A problem consists of multiple subtasks with various points.
- A subtask consists of multiple test cases.
- A submission gets a subtask's points if it solves all of its test cases.
- A problem may have open subtasks where the test cases are given to contestants.
- For output-only problems:
- A problem consists of multiple test cases with various points.
- A submission gets the test case's points if it produces the correct output.
- For each problem, the score can also be rewarded based on a given scoring formula.
- The scoring formula will vary for each problem and will be explicitly stated in the problem description.
- Example:
- Batch: Duck History, OSN 2024.
- Interactive: National Treasure, OSN 2024.
- Output-only: Monochrome Mosaic, OSN 2024.
Contestant Scoring
- For batch and interactive problems:
- On creative scoring, for each submission, the score of a subtask is the minimum score among all its test cases.
- A subtask final score is the maximum of its score among all submissions.
- Contestant's score for a problem is the sum of all its subtasks final score.
For example, if the first submission got 30 points for the first subtask and 0 points for the second subtask, and the second submission got 0 points for the first subtask and 40 points for the second subtask, then the contestant's score for the problem is 70 points.
- For output-only problems:
- A test case final score is the maximum of its score among all submissions.
- A contestant's score on a problem is the sum of all its test cases final score.
- A contestant's total score is the sum of all of problems score on both competition days.
- Contestants will be ranked first by the total score (descending).
- Then, two contestants with the same total score will be ranked by time penalty (ascending).
- Time penalty is defined as the sum of:
- the time (duration) to reach the final score on Day 1 for the first time
- the time (duration) to reach the final score on Day 2 for the first time
- Time penalty is defined as the sum of:
Grading System
- TLX will be used as the grading system.
- Source code limit for each submission is 300 KB.
- For each test case, the grader will output one of:
- Accepted: program solved the problem within the given time and memory limits.
- Wrong Answer: program stopped within given time and memory limits, but produced the wrong output.
- Runtime Error: program crashed or exceedeed the memory limit.
- Time Limit Exceeded: program exceeded the time limit.
- Skipped: test case is not graded because there was a test case in the same subtask that was not solved.