Haskell was used in the first programming class at university and that was excellent. Some people that already know how to program had to throw their preconceptions out the window. So everyone was either a novice at programming entirely or at least a novice at functional programming. It wasn’t problem at all. I’d argue most students wrote better code in that class than they did in the imperative/OO classes that followed. Especially the students that weren’t (and likely never did become) programmers.

The thing about imperative and OO programming is that it’s hard to do well. I honestly havent seen more than 1/10 developers write ”good” OO code even after 10 or 15 years as professionals. Large scale OO is a cognitive load that requires extreme focus, skill and talent. I prefer functional (or OO using an extreme functional discipline like all immutable types etc) because I don’t have that talent, focus and skill.

My point isn’t that Haskell should be the language of choice. I think it’s a great language but I think e.g laziness makes it too hard to reason about performance and behavior. Today I’d recommend F# I think.