As was said, limiting the tape length wouldn't really change the difficulty at all, it would just slightly limit the range of solutions available.

Yep

Completely agree with you !!

Fixed by Chrono79

I've fixed the arguments being flipped.

For the original issue, I've spent way more time than I had planned trying to see it it's an issue. I'm going to mark as resolved because I've tried different solutions from different languages and the accepted solutions can handle that test. Not agreeing with other implementation doesn't make this an issue.

According to the CW wiki, "Best Practices" are solutions that you would expect to see in production code. Maintainability and readability are favored over performance and elegance. Of course, in the real world, production code standards will vary. Regardless, this solution reads well, and would be easy to manage due to the granularity.

Okay no, there's a specifically laid out proper procedure for dealing with numbers. If you aren't using the proper procedure, it's bad code (difficult to maintain, change, understand, etc) as proven by people having difficulty getting test cases to work.

That's your personal opinion;-)... For me, it's enough...

@smile67 Not a good enough reason for testing against string and double, string culture is a thing!

My feelings exactly.

if (size > 2 && size%2 == 0) { answer[size/2][size/2-1] = 1; }

The test cases are wrong for all even sizes > 2. You should use the above code just before you return the answer.

Presumably because of the comment describing how the pattern was derived

Resolved.