This is a major issue and a game stopper when you have a tens of thousands of folders and notes with Norwegian or other Scandinavian names or characters in the name.
And if it’s true for Norwegian and Islandic, it will also be true for Swedish, Danish, and most likely most other languages that use double byte char set where the characters should sort after Z.
Why do you not just confirm to ISO and UTF-8/UNICODE in all part of your software?