This is a pretty bad writing compared to classic writers (Hudak, Hutton, Bird or Thompson). It begins with nix, which is a cancer (solving of non-existent problem by attention seeking narcissists - stable interfaces and semantic versioning will do).

Then it proceeds with a subpar code, which mixes styles without justification, is cryptic and unreadable. You don't have to zip and then unwords, case is cryptic and redundant, etc.

And it is not some rant, it is an adequate peer-review as it should be.

Simplification (reduction) to perfection (when there is nothing more to remove) both in abstraction and in code, is not optimal with Haskell.

Piling up of esoteric bullshit such as lenses and monad transformers is a non-goal.

Xmonad, cabal and, of course, GHC itself are still the gold standard. No bullshit dependencies.


No comments.