The acronym problem description says to count dash as separator and ignore punctuation.
Then the tests start doing random other requirements like CamelCase.
Another example is the collatz tests throwing negative numbers at the function.
Is it policy that the descriptions not be enough to solve the challenges?
I really don’t think this is unique to the Haskell track either.
Secondly, there’s a bot closing github issues, which runs counter to the instructions provided at Frequently asked questions | Exercism's Docs - somebody should fix those.
Hi @lonetech 
Weclome to the Exercism Fourums!
Yes, it is policy. The instructions are intended to work with a given test file to serve as the complete specifications.
Descriptions and instructions are (mostly) shared across language tracks, so they leave out language-specific requirements. Language tracks also vary a bit in what they test for.
Secondly, there’s a bot closing github issues, which runs counter to the instructions provided at Frequently asked questions | Exercism’s Docs - somebody should fix those.
This is also policy, as there are not many Haskell maintainers active at the moment, so posting issues here in the forum allows cross-track maintainers to lend a hand. But we’ll take a look at the FAQ, doc and see if it needs adjusting. 
2 Likes
Weclome to the Exercism Fourums!