Kata trial: Array.diff hero

[ejini6969]

Array.diff hero should be retired due to the following reasons:

• Solution can be directly retrieved from Array.diff
• JS only
• Over-ranked at 5kyu
• Still using Test.expect
• Stuck in Node V.10 to V.14 update

React 👍 to keep or👎 to retire.

[ejini6969]

Opinion from Kayleigh on retirement -> https://discord.com/channels/846624424199061524/1021730541034283018/1030493915566055464

[error256]

There's no open kata issue to notify the author and whomever else about this trial process, is this normal these days? Additionally,

Solution can be directly retrieved from Array.diff

This by itself doesn't require retiring, what about fixing the tests?..

Over-ranked at 5kyu

Since when is a rank a reason to retire?

Still using Test.expect
Stuck in Node V.10 to V.14 update

The fix is trivial: https://www.codewars.com/kumite/634b6c67d225c000237a9beb

[Kacarott]

Fix has been approved, so the last two points of original post are irrelevant.

This by itself doesn't require retiring, what about fixing the tests?

Do you mean that the "standard" solution to the harder kata is not the expected solution? Or that the easier kata should reject solutions which are too fast?

[error256]

In this kata 0 ≤ x ≤ 25, that's the point of the kata. Solutions with a Set are about 2 times slower than the most naive of specialized solutions and much slower than better ones. But random number generation in the tests is slow as well, so input generation for the random tests will probably have to be changed in some way to test performance. And since the difference in performance isn't huge, using very big arrays may not be possible, so the number of tests should be changed. I don't have any estimate of how difficult it would be to change everything to make it work.

[Kacarott]

Oh interesting, then the point of the kata had been lost on me (and presumably, many other users). I don't think I even noticed the number range restriction. In that case, I would support fixing the kata rather than retiring, if possible.

[JohanWiltink]

That point had been lost on me as well. Yes, there are performance gains to be had by micro-optimising. I'm a bit torn if that's worth a kata though. If so, it needs very discerning tests, a rather specific description, and a reference solution and a solution that Should Not Pass ( possibly several ).

[ejini6969]

Sorry about that! I too did not notice the different solutions that users can get if the tests were adjusted to be more performance-wise! Then, it might be worth to stay and be fixed sooner or later.

Since when is a rank a reason to retire?

Only applicable to old katas which cannot be sent back to beta. I'm not sure if this one is part of that category, anyhow it's not relevant now in the current context!