3 ms·
> I get answers about idempotence and list comprehension and strong typing which are great tactics but I never get the sense that they fit into an overarching s
by e12e 7y ago
> I get answers about idempotence and list comprehension and strong typing which are great tactics but I never get the sense that they fit into an overarching strategy for how my life will be made easier by using Haskell.
First, I suppose those three things fit together in the sense that they make list comprehensions safe and fast.
But as you say, Haskell isn't the only language that gives you that.
I think the killer feature of haskell is that the community went through ridiculous contortions and pain to get rid of side effects.
Which means that Haskell programs exclude certain kinds of bugs - so it's not just that you can program in a functional style, but that you can be fairly certain there's no "shifting sands" under your functions.
On the other hand, I see many haskell programmers say that the powerful type checker forces them to think a bit differently, and perhaps harder, when writing - with the result that once the type signature fits, the function tends to "just work".
Anyway, I never did really enjoy Haskell - not as much as StandardML anyway. Sadly there doesn't seem to be any sml with good real-world libraries, so I've not really been using that either, outside of university :/
Perhaps looking at one of the few popular haskell utilities, and see if you feel Haskell is a good fit? I'm honestly not certain myself.
https://github.com/jgm/pandoc/blob/master/src/Text/Pandoc/Readers/CommonMark.hs https://github.com/jgm/pandoc/blob/master/src/Text/Pandoc/Re...
- codebje 7y agoHaskell's type system works best for me in allowing fearless changes, even significant ones to the core of a code base. The compiler won't let me overlook something. A comprehensive test suite would achieve the same thing, but I'm rarely confident that my test suites are truly comprehensive.