3 ms·

My two cents is that anything is hard to learn without an application you can test your knowledge against. It’s always hard to learn language X until you go ahe

My two cents is that anything is hard to learn without an application you can test your knowledge against. It’s always hard to learn language X until you go ahead and build something in it. Functional programming is hard because in a lot of cases, especially ones beginners encounter, the imperative solution is simpler. Purity and types are things I think you only truly appreciate when you’re writing large or complicated programs. I wasn’t able to really grasp FP (beyond using things like map() in Python) until I took a compilers course which used OCaml, and the ability to pass around and destruct these very complicated immutable trees was a very natural problem to tackle in the domain

throwaway81523 5y ago
> My two cents is that anything is hard to learn without an application you can test your knowledge against. It’s always hard to learn language X until you go ahead and build something in it. If it's something really new and different, like Haskell is to an old school imperative programmer, I think the opposite approach is best. Treat it like a topic in math, start from zero, work out small problems to exercise the basic concepts, then start putting them together. Immutable data structures are another new and shocking thing, but less complicated than fancy typed FP is. Start with seeing how you can "update" the first element of a linked list by making a new first element and linking it to the existing tail. Then see how AVL or red-black trees let you do something similar with tree nodes in log(N) time, so you can use those instead of hash tables without a monstrous slowdown. That's probably all you need, but the next thing after is probably Chris Okasaki's book Purely Functional Data Structures. It is pretty readable once you've seen some basics.
mesarvagya 5y ago
> I wasn’t able to really grasp FP (beyond using things like map() in Python) until I took a compilers course which used OCaml, Which compilers course is it?
- legobmw99 5y ago
  It was two courses actually, the first was taught out of https://www.cs.cornell.edu/courses/cs3110/2019sp/textbook/ https://www.cs.cornell.edu/courses/cs3110/2019sp/textbook/ and the professors personal notes on the history of programming language design, and the second was taught primarily out of Andrew Appel’s tiger book with some content from the dragon book

throwaway81523 5y ago

> My two cents is that anything is hard to learn without an application you can test your knowledge against. It’s always hard to learn language X until you go ahead and build something in it. If it's something really new and different, like Haskell is to an old school imperative programmer, I think the opposite approach is best. Treat it like a topic in math, start from zero, work out small problems to exercise the basic concepts, then start putting them together. Immutable data structures are another new and shocking thing, but less complicated than fancy typed FP is. Start with seeing how you can "update" the first element of a linked list by making a new first element and linking it to the existing tail. Then see how AVL or red-black trees let you do something similar with tree nodes in log(N) time, so you can use those instead of hash tables without a monstrous slowdown. That's probably all you need, but the next thing after is probably Chris Okasaki's book Purely Functional Data Structures. It is pretty readable once you've seen some basics.

mesarvagya 5y ago

> I wasn’t able to really grasp FP (beyond using things like map() in Python) until I took a compilers course which used OCaml, Which compilers course is it?

legobmw99 5y ago
It was two courses actually, the first was taught out of https://www.cs.cornell.edu/courses/cs3110/2019sp/textbook/ https://www.cs.cornell.edu/courses/cs3110/2019sp/textbook/ and the professors personal notes on the history of programming language design, and the second was taught primarily out of Andrew Appel’s tiger book with some content from the dragon book