riston, that helped quite a bit. Also the title update are more to the point now. I'm glad you emphasized what you mean by "sorting" - though I got it immediately when I understood the "bit vector construction" issue, the term lead my thoughts away from the problem.

Yes, I updated the description, if anyone has better proposals on the description post here.

Exactly, I left out the final looping because this more like the homework for user.

True, actually I missed the second part of this kata which would be returning the sorted array. This exposes the problem, that the user can use any sorting algorithm's and I can not be sure if the bitwise operations were used or not.

Yes, I'm not really sure if it should be marked as a spoiler or not, so I was cautious, though I forgot that you wouldn't be able to see it :-/

I agree that perhaps a better description is needed, as I was also a little confused as I read through it.

I can't figure out what is expected. Are we meant to be encoding multiple integers into one? If so, what bit size should each number have? Are we meant to handle nuegatives? What on earth are we meant to be doing here?!

You posted this as spoiler - when I stated I failed to understand the description, implying the I had not even solved it. So I counldn't read it before having it solved ;-)

True, it's a "spoiler", but a better description of the issue than the one given.

Description need a brush up!

The term "sorting" does apply but is not really a good guide to understand the issue.

Also, the restriction of one-variable-only does not make much sense. Is a function parameter a variable (in my view, yes)? Then you need at least two - and counting the argument we hit three. Don't know how you will enforce it either, I suggest you skip this restriction.

The description doesn't make much sense. How is this even sorting?

The description is a bit confusing. The exercise is quite trivial, but classifying this as "sorting" is quite misleading. This is constructing a bit vector set.