Well, mathematically speaking, "having one solution" means "having at least one solution", i.e. being solvable. So are the solutions always unique, is that what the kata wants to say?

I just found it funny. Great, the other might be the shortest solution possible.

I had the same problem, very obscure that you cannot name your function hex... I mean, I get the problem, but how should one know...?

Of course, I was just commenting because for me, as a mathematician, it was on the first glance confusing that this actually works :D

I just thought this sentence could be reformulated a bit: "It must be possible to follow the line with no ambiguity (lookahead of just one step, and never treading on the same spot twice)."

We misinterpreted it like, if you have two options to go from a corner, but only one leads to a valid path, then the line would still be valid.

I'm no native speaker, but what about something like: "It must be possible to follow the line without ever treading on the same spot twice, and by going step by step without ever having to choose between two adjacent and still unvisited tiles to go next. (The latter condition might hold only in one direction. The line will still be considered valid in this case.)"

whyyyyyy :O

Logically it would make more sense to set isOposite "" "" = True, as it is an empty condition, which is always fulfilled.
I also suggest you to include testing cases, where the two strings do not only differ by upper or lower case, but also by letter.

There were test cases (in Haskell) without a king although you stated that there would always be one... (Not a big deal though, I just return false if there is no King)

I used Haskell

There were test cases without a king although you stated that there would always be one...

The trace of the 0x0-Matrix [] is usually defined as zero because it is an empty sum.
(I would never use the notation [[]], but you could interpret it as a 0x1-Matrix, which is not square, hence it might make sense to return 'Nothing' here.)

Thank you, it worked. Apparentely I also misunderstood your notion of 'ambiguity'.

My solution failed a single test (the "more edge cases"), can I see the corresponding mazes? I cannot find my mistake.