If you remove monadic IO, laziness and immutability from Haskell, then you're a lot closer to OCaml than you are to Haskell.
But more importantly, immutability is used everywhere in the industry.
Re:Haskell for a New Decade [pdf]
by FullyFunctional on Wednesday June 24, 2020 @02:06AM
(#23624625)
(It's a HN rule not to discuss down votes)
However the problem with the parent post of that it presumes Haskell isn't useful because of X, when clearly it's useful for many and changing X would make it not Haskell. That sort of comment isn't useful.
I think for as long as I have known Haskell (1992-) there has been interest in and a desire for a strict variant of Haskell. In 2020 it seems there are a lot more options for that, which is great, but they aren't Haskell.
Re:Haskell for a New Decade [pdf]
by mlthoughts2018 on Wednesday June 24, 2020 @06:50PM
(#23634752)
> “ changing X would make it not Haskell”
this seems like some version of a No True Scotsman argument.
"Favor Immutability" - Effective Java, circa 2001
Should not be controversial at this point.
Re:Haskell for a New Decade [pdf]
by mlthoughts2018 on Wednesday June 24, 2020 @06:51PM
(#23634756)
Immutability is great. What does that have to do with Haskell?
This to me, is the only reply thus far that addresses what I also perceive to be the fundamental problem with Haskell - it doesn't provide any new convenient tools to solve real world problems, while requiring one learns a bunch of useless abstractions that aren't used anywhere else in industry as a default (monads, laziness, immutability)