You are totally right.
I changed the description.
The intent of the unique radius point was to ensure that there are at least k different radii in the set. Otherwise it would not be possible to find a solution.
I replaced that with: There is a valid solution.
That should be less confusing

And I changed the limit to fit to the tests: 0 < k <= N <= 1000

Done

Hi, I'm glad to hear that you liked it.
I'm a bit short on time right now, but I will incorporate your suggestions on the weekend.
Have fun solving katas!

I'm sorry. I realise now, that I should have been more clear and that this was the wrong place for a complaint like that.
I realize now, that this should have been discussed in the beta phase of this translation and it probably was. I'm not yet familiar with the process of releasing katas ;)

Anyways, here is why I thought that there is an issue.
Writing down a given equation is in my opinion worth 7 or 8 kyu.
My assumption( based on the tag and the comments) was, that the 4 kyu status was given because in other languages than c++ you have to find some optimated algorithm to pass all tests in the given time limit.
Since c++ is rather new at CW, I came to the conclusion, that it was translated later on and got the same 4 kyu rating although it is really easy in c++.
And I believed, that this translation is too easy, and therefore the translation itself is an issue.
Therefore my solutions would be either downrank this translation ( is that even possible?) or prevent the translation to c++.
But then again, if I understand it correctly, it is kind of too late to complain about that now ;)

From the tag "optimization" and the other comments I come to the conclusion, that a simple brute-force should not solve this kata.
In c++ it does...
I typed in your formula using two for loops and passed all tests. That was disappointing.

Sorry for that. I never created a kata so I didn't know what to look at and saw no comment on this issue.

As written in the description only the leftover is rounded:
...and 766 is the nearest integer to '766.158...' .

In C++ the public testcases don't use 'namespace std' and therefore throw errors.

Unclear instructions. I did some try and error until I managed to treat the empty Notes correctly.

No it wont. And if the return statement is replaced with 'return names[0]' it will pass some more tests.
So I guess something is wrong on server side or I completly missunderstand the kata.

I solved the kata, but I am still not sure, if I should create new Nodes and fill them or use the source Nodes and therefore destroy the source list.
Is this explained in the other linked katas?