I read those two links. If I ignore some of the points that were valid 10 years ago but not anymore (e.g., there's now stack), and if I ignore some of the perso

I read those two links. If I ignore some of the points that were valid 10 years ago but not anymore (e.g., there's now stack), and if I ignore some of the personal anecdotes, I come away feeling like much of what is said is not necessarily wrong but is nevertheless an exceptionally negative take. I think you can make almost any tool sound broken and unusable if you use nasty enough and strong enough language about a few examples. Also, there really is very little substance behind the actual language criticism. It's hard to make a blanket statement that typeclasses are "overkill"; they seem to be showing up all over the place in new languages. Similarly, type annotations are incredibly useful and are also showing up all over the place these days outside of Haskell-land. The point on non-strict evaluation is by far the strongest and most legitimate.