3 ms·
I've read this article (http://bartoszmilewski.com/2014/11/24/types-and-functions/ http://bartoszmilewski.com/2014/11/24/types-and-functions/) , and the first a
by nkurz 12y ago
I've read this article (http://bartoszmilewski.com/2014/11/24/types-and-functions/ http://bartoszmilewski.com/2014/11/24/types-and-functions/) , and the first article in the series (http://bartoszmilewski.com/2014/11/04/category-the-essence-of-composition/ http://bartoszmilewski.com/2014/11/04/category-the-essence-o...), and the introduction (http://bartoszmilewski.com/2014/10/28/category-theory-for-programmers-the-preface/ http://bartoszmilewski.com/2014/10/28/category-theory-for-pr...). They've all gotten a significant number of upvotes here on HN, but no commentary about the contents.
Reading them makes me fear that I simply lack any proper capacity for abstraction. I liked thinking about the throwaway line that "an object should have more surface area than volume", but there hasn't been anything else that I've been able to grasp hold of. Nothing seems wrong, but so far, nothing seems particularly right either. He could just as well be discussing typography as type theory.[1]
Could someone who understands where he is going with this series jump ahead and talk about how a better understanding of category theory has improved their programming? Insights that wouldn't have occurred without it? I'm all for pure math as pure math, but I'm struggling to see how he's going to make the leap to practical programming. Or perhaps that isn't his intention?
[1] Come to think of it, his comment that Haskell accepted a double colon or the single unicode character equivalently was one more thing I felt I grasped. It struck me as a dangerous design decision, though.