.

shits happen... ;) (especially when coding XD )

Well, pick one randomly (a direction), just as you're asked for? ;) (it's related to the ? command)
Your code is run a lot of times and there is a statistical check that you effectively picked randomly a direction.

ok good.

The thing you just learnt: global vars are evil... ;)

then there is 2 options:

• either you're facing a "real" performances trouble, meaning your time complexity is too highg to solve the task and you'll have to change you're algo (...see foot note!)
• or you're still in an infinite loop and you didn't print enough stuff to saturate the buffer before the time out.

note: I hope you're not telling me that your ide "works fine with the example tests but not when clicking attempt"?? -> if so, keep in mind that the sample tests are just what they are called for...: SAMPLE ;)
So your next move would be to get the inputs to see what the problem actually is.

I'm sorry I don't understand what you're saying. But sounds like you're telling that you cannot get feedback when the tests time out on cw. For that, there is trick: print stuff to the console from inside your while loop(s) (whatever you want). If the time out is caused by an infinite loop, you'll saturate the buffer before the time out, this way, and so you'll be able to have feedback in the console. Up to you to disentangle what's wrong and where, from there.

it runs perfectly for the examples I have from the output of the execution.

what do you mean by that, exactly?

In the Kata description it says the output should be "Coordinates of the bishops in lexicographical order". So ["b5", "d8"] would make sense, because that's lexicographical order.