As I mentioned the part in this kata that encourages you to write your own tests is good. I agree with that. But still I wrote my tests and everything was ok. But then when I tried to attempt the solution it gave me errors and the message was that It could not recognise the Ghost class even though I imported the class and passed the tests. And by the way I was using Java to solve the problem.

Yeah, I guess you are referring to the C# translation. I'd like to just remove it, but it was created several years ago (unfortunately after the kata itself was out of beta), and any changes to it now would potentially break solutions that have already passed.

I would not select an option to allow translations for any kata that I made going forward.

oh okay. Out of curiousity, which language(s) did you try?

When I first made this kata, I thought it was better not to include tests in an attempt to encourage people to write their own, but that might not be what the community considers "best practices" now.

As for the errors, did they give any error messages?

Tests were not there. I also wrote my own tests and wrote "Ghost" class and tested them and they all passed but when tried to "Attempt" them, I received errors for no apparant reasons. To be fair I learned about tests and how to write them a little bit and that was good but still I couldn't get it passed for no apparent reasons.

Too easy, or...?

Not the best kata that I know of.

Will do!

I don't think it should ever have had a getColor() method in C#, but somebody else created the C# translation, and this has been out of beta for years, so making changes may cause issues for already accepted solutions.

@ogryzek, I guess this is now resolved, if so, update and mark it as resolved once the issues are fixed.

Testing in C#: "'Ghost' does not contain a definition for 'GetColor'".
The Details don't mention such a method.

Test cases for the java are wrong. it checks strings equality with == instead of .equals , please fix it.

Cool feedback. I'd be curious as to what sort of background someone might have where they are more familiar with data structure and algorithm puzzles than creating instances of a class with a default value in an OOP language.

As far as I know, the level of each kata is determined through feedback during its beta process, which for this particular one was more than 9 years ago. It's possible that criteria and other things have changed. The language versions and directions for testing, etc. seem to have.