4 ms·

I don't really understand. I thought this was an extremely interesting post, and it was very easy to follow. It's clearly not written for a complete beginner, b

• hardwaresofton 7y ago
  Yeah I clarified up top -- I think the title "Implement with Types, Not Your Brain" does a bad job of setting expectations. Personally, I clicked through expecting to find out about some new way haskell can help me use my types to verify how implementations work (the title is fantastic in this sense, and I saw the domain), but I was instantly hit with that bit of code that's somewhat inscrutable and doesn't do enough to set up the context of what's happening and why. Then the general point of the article basically was to use polymorphism and type holes to work your way to what you need. I felt like this wasn't what I signed up for, and wondered what others thought, only to go back and see upvoted comments bashing haskell based on this one example. This is what lead me to make my comment -- I think this article is too advanced to post here without some really good prose/writing. Not blaming sandy, as of course he doesn't control how his content spreads but I thought it just wasn't right for this context and sharing something that makes haskell seem harder than need be is damaging IMO. The case that "haskell will tell you what types it expects" is so much easier to make without the monstrous signatures in that post, or without type level computation (the ':).