1. 17

  2. 9

    This post would make a stronger case if it had a few examples.

    1. 5

      Yeah I have no idea what it means at all.

      1. 1

        This example comes to mind. You can model async with monads, but that’s incompatible with the rest of Rust’s design constraints.

        The author also seems to have done a lot of work on parsing and apparently thinks poorly of monadic parser combinators. He goes into a little more detail on that here.

      2. 3

        I dunno, I’m pretty sure many in the research community are well aware of the problems of monads - this is why many are moving towards algebraic effects as an alternative.

        1. 2

          Sure, but the whole FP community’s response is: we already have monads! That included me for a whole while. It’s an issue.

          1. 2

            The post is clearly taking aim at researchers:

            One of the great ironies of programming language research is that language researchers are both the most frequent victims of the Turing tar pit mentality, as well as its most fervent perpetrators.

            Often I find it’s the researchers who are the most aware of these issues.

            1. 1

              From my experience, yes they’re aware, they just don’t do much about it. Everybody is comfortable working on their more and more advanced/obscure applications of monads.