3 ms·

I really wanted to dive into Racket a few years ago and still really like Racket. But this is when they switched focus to the Rhombus project, which killed a lo

by bmitc 3y ago

I really wanted to dive into Racket a few years ago and still really like Racket. But this is when they switched focus to the Rhombus project, which killed a lot of interest and personal momentum I had going for Racket. It has also become clear that Racket's maintenance core is too academic. I don't think Racket is too academic, but by the very nature of Racket's primary community being professors and graduate students, it means there is a very small core team with a lot of the interest and use outside of that core being shortlived as students go through PhD programs. There are a few nonacademic core members, but there are few.

fn-mote 3y ago
IMO this part is FUD: > when they switched focus to the Rhombus project, which killed a lot of interest and personal momentum I had going for Racket The language has been "complete" for a long time. If you're actively trying to get something done, I think it's unlikely that some core language work is going to stop you. It isn't clear what domain the parent wanted to work in; I won't say it's impossible they had problems but ... details are everything at this point. The academic part of Racket gave rise to `syntax-parse`, a truely fabulous improvement over the standard Lisp/Scheme/whatever you used to work with way to write macros.
- bmitc 3y ago
  Why is it FUD? It's my own feeling. In this specific case I had just spent a week of my and my work's time and money learning Racket at an official Racket school only to find out at the end that they're creating a new "surface" syntax. The split of do I learn and participate with Rhombus or Racket is a concern. The idea that the core maintainers, or really the core maintainer (singular), is busy with Rhombus is a concern. The explanations given at the time for the Rhombus project were primarily dubious and unclear. The uncertainty and doubt were not created by me. It was enough to make me move on. Other languages like F# and Elixir suit my needs. > The academic part of Racket gave rise to `syntax-parse`, a truly fabulous improvement over the standard Lisp/Scheme/whatever you used to work with way to write macros. That's not the academic part I am referring to. My comment is about who the maintainers and users are and how their profession or studies affect how Racket is interacted with. (This is not a complaint or judgement. It's an observation.) > The language has been "complete" for a long time. There are a fair number of bugs in Racket's libraries, including core ones. They aren't discovered because academic users are not touching bits that more industrial-oriented users would find. In particular, I am thinking of the GUI toolkit, and in my experience, only one person is capable of fixing those.

by bmitc 3y ago

fn-mote 3y ago

IMO this part is FUD: > when they switched focus to the Rhombus project, which killed a lot of interest and personal momentum I had going for Racket The language has been "complete" for a long time. If you're actively trying to get something done, I think it's unlikely that some core language work is going to stop you. It isn't clear what domain the parent wanted to work in; I won't say it's impossible they had problems but ... details are everything at this point. The academic part of Racket gave rise to `syntax-parse`, a truely fabulous improvement over the standard Lisp/Scheme/whatever you used to work with way to write macros.

bmitc 3y ago
Why is it FUD? It's my own feeling. In this specific case I had just spent a week of my and my work's time and money learning Racket at an official Racket school only to find out at the end that they're creating a new "surface" syntax. The split of do I learn and participate with Rhombus or Racket is a concern. The idea that the core maintainers, or really the core maintainer (singular), is busy with Rhombus is a concern. The explanations given at the time for the Rhombus project were primarily dubious and unclear. The uncertainty and doubt were not created by me. It was enough to make me move on. Other languages like F# and Elixir suit my needs. > The academic part of Racket gave rise to `syntax-parse`, a truly fabulous improvement over the standard Lisp/Scheme/whatever you used to work with way to write macros. That's not the academic part I am referring to. My comment is about who the maintainers and users are and how their profession or studies affect how Racket is interacted with. (This is not a complaint or judgement. It's an observation.) > The language has been "complete" for a long time. There are a fair number of bugs in Racket's libraries, including core ones. They aren't discovered because academic users are not touching bits that more industrial-oriented users would find. In particular, I am thinking of the GUI toolkit, and in my experience, only one person is capable of fixing those.