1. Bump. Completely broken tests.
2. Extremely cryptic description.
3. Worst kata experience ever (after solving 6500 ones...)

Thanks man

Ahh, I wish I read this before I attempted to hammer the solution with randomness.

fixed by B4B

The solution can't overcome a time limit without an index. Therefore, an index is not useless here.

It's not a performance kata, but it deserves at least a 6 kyu in my opinion. I solved it in JS and it demanded "more advanced" functionalities not to time out. I encountered 6 kyu katas much much easier than this one.

Your task is to find the largest d

-5 > -10

It's either performance Kata or math Kata. Either way it's not 7 kyu Kata.

If you think that this is a performance Kata, then you are in trouble.

Done in python

If you don't pass a case, but resubmit with the same code and pass, it is not an issue

Uh, how is such situation not an issue? It means that either tests are incomplete and sometimes let invalid solutions through, and sometimes not, or opposite: that some valid solutions are sometimes recognized as invalid. If one and the same solution sometimes passes and sometimes not, then it definitely is kata issue, isn't it?

mmm? Dunno, I didn't read the whole thread. But seems like this user found a kind of input where his code fails. So just grab the input, feed your ref solution with it and extract the expected result?

Or am I missing something?

Not following:

So, what should be the answer?

@tony: why not just add that as a fixed test in all versions?

@FArekkusu, @scannning

I don't think it is an issue. If you don't pass a case, but resubmit with the same code and pass, it is not an issue. You don't give enough information. What is the input? What is the language? What is your output? I know you cannot replicate this, but there is just too little information to find out what is the problem.