Quite possibly a solution problem, not a kata problem. 441 Submitted solutions did not have this problem.

You didn't, explicitly or implicitly, call .toString() on an undefined value somewhere in your solution, did you?

Closing.

In JavaScript all the Random test gives me the next error:

   at simplify 
    at Test.it._ 
    at begin 
    at Object.it
    at Test.describe._
    at /runner/frameworks/javascript/cw-2.js:159:11
    at Promise._execute
    at Promise._resolveFromExecutor
    at new Promise
    at Object.describe
`            at Object.handleError
`        at ContextifyScript.Script.runInThisContext 
`    at Object.exports.runInThisContext`
    
  The regular tests work fine and tyring one of the random the algorithm also work... So it must be something about the random tests generator.

Ah, I actually just realized the main reason why when you mentioned cheat. To translate a kata, you must see the solution in another language to do so - which is a security hole for cheating. You would have to forfeit the kata in order to see the solutions - which we also wanted to avoid.

Hi, fair points but both are debatable.

Ad 2: I'm not dreaming about a perfect system as there are no perfect systems :) If I suppose that one doesn't want to cheat, then he has to solve the problem in the translated language. This is even more true, when there are random tests.

Ad 1: Many katas are in JavaScript. The fact that I have to produce a low-quality JS solution doesn't help me much in Python translations. But this could be different for other people.

A few reasons.

1. Quality - You probably shouldn't be translating a kata that you haven't solved before. I get that this is debatable but the feeling was that forcing people to solve the kata in another language will result in higher quality translations overall even though it may not always make a difference.
2. Honor/Ranking - Some kata aren't that hard to translate but are hard to figure out. To say, "here is exactly how to do the kata in one language now translate it to another language" - without actually having to solve it yourself means that you would get credit for completing the challenge when that credit may not be deserving. Again, not a perfect system. I don't think perfect is possible here - but viewed as an aggregate its likely to solve the problem better than not forcing the requirement. a

Hi, I have a question. Why is this: "Once you have unlocked the solutions in any of the existing languages" required? Why should one e. g. know JavaScript, Haskell or other language to translate a kata to Python (or vice versa)? In many cases a translation can be made based only on description.

I understand the technical side (i. e. when one edits a kata he can also see the solution and test cases in original language and the existing translation), but this is more a limitation than a reason.

Thanks,

suic

Thanks,

suic

We may have to add that assembly reference to the runner, I think right now we only have a handful of assemblies being including within the compilation process. We will add that to our list of things to update.

Once you have unlocked the solutions in any of the existing languages, you can then add a new language to the kata by clicking on the + icon next to the existing languages. From there you can create a kata translation. Kata translations need to be approved by moderators.

p.s. Something to remember when creating a translation is that the description is shared between all languages, so you need to add language examples to the description in addition to the ones that are already there. Also if a kata has language examples, then you must add examples for the new language. This tends to hold up some translations from being approved.

How do I help translating existing katas to other languages?

NM. It was an issue with all custom test cases. Should be good now.

Was an issue with custom test cases. This has been fixed. Thanks.

I have test cases in my current Kata and I get the error 'Test didn't pass: Unknown error'

Do you have any test cases? We need to clean up the error messaging caused by some recent changes, but it looks like people are getting that error because they are submitting code that doesn't have any test cases, which is why it was working when you submitted (because you were using the kata author's test cases instead).

Test engine is down