"I’ve excluded functional languages such as Lisp, Scheme, Haskell and F# as first languages; for people who aren’t specialists...they’re not suitable."
This is objectively wrong, but we must take it upon ourselves to improve our messaging & outreach #fp
Conversation
this is so tiresome. haskell is not some esoteric specialist language. probably neither is Lisp tbh, i just haven't learned much Lisp yet.
1
13
Oh man. I hear you. You are doing as much as anyone to fix these misconceptions, so you at least can feel no guilt
1
5
As someone who's been using haskell for work for over a decade, who's set up dozens upon dozens of Haskell meetups, and who is a strong proponent of Haskell, I agree that it is by far not some esolang, but saying "this is objectively wrong" is just stroking your own ego.
2
2
I'll be honest with you, this kind of stuff alienates and scares people away. You're not doing a great job of improving the message and outreach of fp.
2
2
Hmmm. Thanks, food for thought. I stand by "objective": countless beginners have found it perfectly suitable, in no small part thanks to the good work of people like you. There's no wiggle room there. I don't see how my ego's fingerprints are on this one, though
1
3
The form is part of the message.
1
Help me out here, what's the problem - "objectively"? More hedging or softening required? I'm not trying to attack or persuade the author, I'm trying to encourage us (the FP community) to improve our comms, which I think is what you're saying too; perhaps my eye has a log in it
1
1
It positions you as an Infallible Judge of Things which is probably not what you wanted.
1
1
Thanks for your feedback, in any case. I'll have a think
2
1
I'd probably agree with the above sentiment that expressed. But I think the critique could have also initially been worded in a more friendly, empathetic way.
Ultimately the axioms and notation we choose to base our work on are based on experimentation and intuition. There is probably a landscape of 'better' and 'worse' formal systems that could be tested, but we're a long way from being able to design experiments for that.
1
Show replies




