Here is a follow-up to the discussion we had during the last meetups regarding simple Haskell.
If I understand correctly, the proposition was that effect systems may be too challenging for beginners and that sticking to IO is a viable path to learn and practice Haskell. Another proposition was made about how far can IO be used before it becomes an issue, for example: not above 10k line of codes.
I wanted to thank you for the discussions, as a beginner I am having trouble to understand effect systems, and learning that you don't have to use them to be productive was a great discovery. I read about IORef and MVar, and that enable me to get started with some practical code. This already feels like a big improvement over my Python, and I'm curious to see how far I can go without using more complex stuff.
Yea, IO is good enough to "get started". You can always "climb the abstraction later" later on as you feel more comfortable. This makes me think that perhaps we need a "progression ladder" of some sorts, to recommend to beginners. By the way I don't like the phrasing "simple haskell" -- I think it should be "easy haskell" or even "beginner haskell" -- anybody who has watched Rich Hickey's talk knows what I mean.
Here is a follow-up to the discussion we had during the last meetups regarding simple Haskell.
If I understand correctly, the proposition was that effect systems may be too challenging for beginners and that sticking to IO is a viable path to learn and practice Haskell. Another proposition was made about how far can IO be used before it becomes an issue, for example: not above 10k line of codes.
I wanted to thank you for the discussions, as a beginner I am having trouble to understand effect systems, and learning that you don't have to use them to be productive was a great discovery. I read about IORef and MVar, and that enable me to get started with some practical code. This already feels like a big improvement over my Python, and I'm curious to see how far I can go without using more complex stuff.
Cheers
Yea, IO is good enough to "get started". You can always "climb the abstraction later" later on as you feel more comfortable. This makes me think that perhaps we need a "progression ladder" of some sorts, to recommend to beginners. By the way I don't like the phrasing "simple haskell" -- I think it should be "easy haskell" or even "beginner haskell" -- anybody who has watched Rich Hickey's talk knows what I mean.
(What you think to be 'simple' today can end up becoming an unwieldy beast 2 months later)