Since you haven't solved the problem yet, I'm going to try to give two answers a more vague one without spoilers and a deeper one with spoilers.

To find a solution you need to think of the problem as it would be in reality. You have an array of disks, and parity information is stored as described. Any of those disks can fail, your system needs to be able to recover from the failure of any single disk, otherwise the failure of the last disk becomes a SPOF which is exacly what your algorithm tries to prevent. Try to think of the cleverest way to racalculate any array damaged without resourcing to special cases.

;-)

Done, I have iterated 50 times the random tests, 157 tests in total. Should be enough.

Will do

Translation to JS Kumited
https://www.codewars.com/kumite/57ed9755bd793ecdd0000e73?sel=57ed9755bd793ecdd0000e73

Fixed it. It was randomizing between 0 and length instead of length-1. Thanks again.

Thanks for the heads up. I'll look into this.

That shift rightwards is clever. I like it.

I did figure it out. Data won't be shown if you fully implement the Test.describe() and Test.it() strucutre. I added a large cluster test and I'll port the other tests to this strucutre when I have a few min.

Thanks, I'll have a look into that. I remember seeing Katas that don't give away much in terms of output. I'll try seeing if I find some of them and see if I can figure it out.

That said, I'm having trouble using large clusters because the data is output in the test reports and floods (even breaks) the console. Do you know if this can be tweaked so as the data is not displayed?

Thanks

Also, would you mind opining on the Kata's level?