Should a command line argument for increasing memory limits count towards byte/character count?

Question

Most Java interpreters have a predefined memory limit. However, it's a -X setting - an implementation specific setting.

With C, the call stack is usually heavily limited. There's usually a compiler setting to increase this limit. This is implementation-specific and as far as the C spec is concerned, there's no call stack limit.

Should implementation-specific options to increase the memory limit count towards the byte/character count? In other words, should limitations of the physical world count towards the byte count?

Answer 1

It depends.

We usually require submissions to work only up to the limits of the compiler/hardware, as long as the underlying algorithm works for arbitrary input.

Unless the challenge specifically says otherwise, if a solution requires a wider integer type, a higher recursion limit, a bigger stack limit, more bytes of RAM than there are atoms in the observable universe, etc. for some/most/all of the test cases, you don't even need the -X flag, so you don't have to add it to your byte count.

However, if the challenge spec requires your solution to work for specific test cases, the extra flags should count, as would the required import statements to work with arbitrary-precision integers.