Not a issue.

idd

yes "Best Practice"

1. on else if 'l < w' not 'l > w'
2. addAll fails on NullPointer - needs condition with add 'w' if result is null
3. for zero and negative input crash with StackOverflow
4. what's happend if imput is valid but big number? Yes, StackOverflow again!
5. as a bonus is thousand times slower than solution without recursion

This is too many issues in few lines of code.

I've done a Haskell one but yeah, it seems the author is no longer active. :(

I reported the problem. Since more than 24 hours it's impossible to republish the Java kata in which Osmosis encountered surprising result with a buggy solution!

I have little to no experience in Java, but I encountered a similar problem; if you want to report it too, that was my issue's topic.

Unable to publish the new version. Codewars: "We are validating your code and publishing your changes. This shouldn't take too long."
but it takes hours!

@Abbe: see above Osmosis
@GiacomoSorbi: see above Osmosis
@Osmosis: I just saw that AssertEquals doesn't work well with List of arrays when the solution is wrong!... I completely modified the tests. Seems good now with the new version; unfortunately Codewars takes lot of time to put this newly published version so one has to wait until the new version is publicly available:-( As of now it's still an intermediate version where "Your test cases" don't work.
When you have time maybe later you can try your "buggy" version and see what happens.

@Osmosis: I modified the description with "...How many times will the mother see the ball either falling or bouncing in front of the window". Hope it is clearer. Thanks for the feedback!