Sorry, I have no experience with Ruby, and it is indeed weird that the solution produces wrong results.

Could it be some limitation of large integers in Ruby?

Hopefully the author of the Ruby translation could give us an answer :)

If we're being picky, then 0 is neither positive nor negative and also sort of depends on context. For purposes of this kata though, we include it in the range. Consider following application: rounding time given in minutes to 5-minute intervals. Numbers ending in 1-5 produce a number ending in 5, while numbers ending in 6-0 produce numbers ending in 0.

Either way, examples in the description provide necessary context to determine how input should be mapped to ouput.

Thanks! CoffeeScript looks like an interesting language :) Approved!

More test cases wouldn't hurt :)

Thanks! Approved it.

Approved! Thanks :)

Thanks! Approved.

Love this, such elegant solution. :)

Could add some test cases.

Instructions are a bit unclear. Problem description asks to print numbers from 1 to n, while test case expect a list as return value.

A couple more test cases wouldn't hurt.

Thanks!
I've considered that possibility as well, but it requires better understanding of musical theory, so I might make it a separate kata in the future. :)

Typo in test cases that raises an error.

Description doesn't say anything about implementing len() method, while test cases use it.

Thanks!

Nice kata! Forced me to learn sorting by more than one key in different orders.

You're absolutely right!

I must have been too tired to have missed that.

I've now corrected the description, solution code and test cases.
Thanks for catching that!

edit: While I was on it, I also noticed that random tests didn't inlcude
testing with flat notation input. Fixed that as well.