It is not just a matter of rearranging letters of one solution to invent another solution; the letters have to be from known words and they match the numbers.

e.g.

1-800-CODE-WAR
1-800-2633-927

So, 1-8000-AND-WARD does not meet that criteria because the letters and numbers do not match.
So, 1-800-COD-EWAR works for the numbers, but is disqualified because EWAR not a known word.

Why unclear? The next value after Z is [ because it is the next symbol in ASCII table (as well as in Unicode table). There is no mention in kata that only capital Latin letters are valid.

Thanks :-)

25 is the limit because that's the length(alphabet) - I. After Z it's unclear what the next value should be. I don't mind leaving the lower limit up to the implementer (I believe the tests go down to 3x3 or 4x4 boards as a min), is there value in adding a limit?

I think this is an issue on the kata description instead of the test itself. At least in the Python sample test:

test.it("32x32")
test.expect_error("Should throw an error. Board cannot be larget than 25 by 25", lambda: Go(32))

The kata does not mention anything about the valid range of board size. While a board size of a non-positive number must be invalid, how about 1 by 1? Yeah, it does not make sense to play on such a board, but is it valid? No idea.

And if we are to rely on our common sense and make our own assumptions in cases where the kata description does not mention anything, then how should we know a board must not be larger than 25 by 25? (By the way, typo in the sample test case shown above: "larger" instead of "larget".)

not even close to an issue as long as you do not provide:

• the language
• the input
• the expected output (done)

(and did you at least check that the problem is not something else than the saize...?)

Nice to hear you like it!