There will be four tasks on each competition day.

The run-time CPU limit is 10 seconds and the run-time memory limit is 256MB for every test.

The compile-time CPU limit is 10 seconds and the compile-time memory limit is 256MB.

Submitted programs must not do any input or output, must not use the network, must not do any system calls, must not attempt to gain points other than by solving the task as specified, and must not attempt to interfere with the operation of the grading system.

The use of standard mathematics, string and algorithm libraries is permitted.

Unless otherwise stated, the word number means non-negative integer.

Unless otherwise stated, numbers will not exceed 1,000,000,000. That is, they may be represented easily as the type int (C/C++) or longint (Pascal).

The phrase between N and M shall be interpreted to mean all numbers not less than N and not greater than M, where N<M..

Contestants may submit at most once per task per minute. More frequent submissions will result in submission failed.

Contestants may submit at most 1MB per task. Note that the submit command submits a zip file containing all files in the current folder with the name *.c *.cpp *.pas or *.txt, whichever matches the file named on the submit command. This zip file must not exceed 1MB in size.

Contestants may view the scores for at most two submissions per task in a 30 minute period, using the release tokens of the web grader interface.

The final score for each task will be the maximum of the following: rounded to the nearest integer:

• The sum of the scores of the subtasks for any released submission
• The sum of the scores of the subtasks for the last submission, whether the last submission is released or not

The first subtask of every task is designed to be solvable by all contestants.

Each subtask is has either a fixed score, or a score determined by a formula given in the subtask.

The total score for most tasks is 100 points. For some tasks, it is possible (but not likely, in the opinion of the Scientific Committee) that a contestant may score up to 110 points.

Detailed test results will include one of the following:

• OK
• OK subtask
• OK score score
• Wrong result
• Invalid library call
• Too many library calls
• CPU Time limit exceeded
• Forbidden system call
• Run-time error
• a RunC failure message

Additional RunC output may also be shown in the detailed test results.