[Haskell-cafe] Haskell-Cafe Digest, Vol 217, Issue 16

[Anthony Clayden]

>>* Haskell's big problem right now is adoption.
*

Heh heh, this sentence is multi-ways ambiguous. As in the 'problem' is
that there's too many adopters/the wrong sort of people trying to
adopt Haskell. The 'problem' is that Haskell is not keeping to its
objective to "avoid success at all costs".

> I'm not quite sure this is a problem. Haskell and its libraries are designed around math concepts, as opposed to CS concepts (semigroup vs IConcatenable). Learning it all but requires learning some algebra as well, thus a lower adoption rate is expected.


Thank you for saying that out loud. I've been feeling for some time
that GHC is getting harder to use. You're telling me that because I
don't get Category Theory you want to throw me out of using Haskell.
People who are allergic to Lambda-calculus are already excluded.

So the ostensible difficulty on these threads: that the documentation
is unhelpful, is not the problem at all. In fact it's deliberate that
the docos are exclusionary. As part of this grand plan, Viktor is
deliberately making the Libraries docos impenetrable. (By which I
mean: the style is clear, but the material is unlearnable. Apparently
I need to know what is an endomorphism before I can write a Foldable
instance)

And that explains a paradox: I learnt Haskell over a decade ago. There
was far less intro material; and certainly no store of StackOverflow
questions/answers. It should have been more difficult to learn than
today. But I don't remember learning being as difficult as the claims
on this thread. (Memory can deceive, I guess.)

Back then Haskell was a much smaller, simpler language. There was less
that could go wrong at the type level; and the error messages didn't
keep recommending to switch on bamboozling features that (as it turned
out) made the difficulty harder. Don't go claiming that if I switch
off those features today it's as if I was still using H2010. Simply
not true.

I know that's not true, because I can still go back and use Hugs, or
use GHC v7.10 (vintage ~2015), and have a much more pleasant/less
bamboozling experience.


Then "Haskell's big problem right now" is (depending on your point of
view) either GHC or too many of the wrong adopters. And with Dependent
Types coming along GHC is going to get a lot worse. (Maybe GHC will
eventually get through the fog, throw out a load of cruft and become
on average as usable as v7.10. I'm not prepared to go through the pain
for 5 years to find out. I'll stick with v7.10.)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/haskell-cafe/attachments/20210918/144e99fc/attachment-0001.html>