I think that random test cases is not the best idea. It seems that some of the solutions willl be invalid. Because they will pass first test case.

I believe that each time someone submits a solution, the server allows the code to run for up to 3 seconds. And all code executes on the server, including the user's own tests (the ones kicked off by the "Test" button).

I definately like that idea, but how far can we push the VMs? I don't want to fail someone out in the even that their answer is sufficiently working on say, their computer, but the VM isn't given enough horses to power through it.

Could you provide an example of what you mean? This is my first Kata attempt, so I'm not entirely sure what's considered best practices, etc.

Thank you. Fixed

Well, in confidence, I was a bit stuck with the kata, feeling it should have an easy solution, but unable to catch the right turn. Once done, the solution is quite straightforward. I hope my new solution is more satisfactory. Thank you for the lesson :)

This should be fixed now

Oh-- also, there's a bit of odd grammar which makes one of the sentences hard to understand. In the tree shown before the function should return false needs a comma after "before", maybe?

Oh, okay. I see. In your examples, you only show the right-hand children having right-hand children of their own, so it looked like the list can keep going right or left, in a long line, rather than an actual tree. If you just tweak the examples to show more tree-ness happening, that should help things. ;P