1. 27

with online REPL

  1.  

  2. 10
    1. 7

      Why do they consider this a new language and not just a Haskell compiler targeting the JVM?

      1. 5

        I think it’s partially a marketing choice but Typelead says they are going to add things to the language. One example is PureScript’s row polymorphism.

        1. 4

          There’s an entry in the FAQ that says something pretty close to that. A few of the other FAQ entries are also on that subject.

      2. 6

        The description reads, “a powerful language for building scalable systems on the JVM”

        What kinds of scalable systems have been built with Eta?

        1. 5

          Eta is a fork of GHC. The code generator and the runtime have been replaced. Some very scalable systems have been made with GHC but nothing with Eta yet.

        2. 6

          Sweet landing page, very polished.

          Funny opening with “A powerful language for building scalable systems on the JVM” followed by (can’t say for sure without checking the compiled output) what looks like a terribly unscalable implementation of “quicksort”.

          1. 3

            It’s perfectly reasonable idomatic Haskell. It’s got a bug, mind you (it de-dupes), but it’s idiomatic and performant when compiled by ghc. Since Eta is based on ghc, there’s a very good chance it’s also performant.

            1. 3

              Doing it in-place would be ideal. Here’s an example of an efficient variant of quicksort:

              http://hackage.haskell.org/package/vector-algorithms-0.7.0.1/docs/src/Data-Vector-Algorithms-Intro.html#sort

              But I don’t mind seeing the naive quicksort implementation. It’s not algorithmically optimal but it’s very easy to understand.

          2. 5

            is the logo intentionally a reference to llvm’s?