HN Explorer

What Would be a Good Lisp Project?

9 points by lambdaman ↗ HN
I've been going through SICP and I'm currently on chapter 4. I was wondering if there are any projects (open source) or new ideas that could keep me busy for 3-4 months?

I'm also contemplating if I should learn another lisp like Clojure since it seems that it's a bit impractical to build something big with Scheme alone.

7 comments

[ 3.1 ms ] story [ 25.7 ms ] thread
stray
A lisp version of matplotlib wouldn't suck...
melling
People are using Racket Scheme.

http://racket-lang.org/

Build a website. Instant gratification and you can always grow it into something big.

CyberFonic
Considering that the lack of a well implemented GUI is holding back broader adoption of Scheme, how about a windowing solution that leverages ubiquitous browser technology. You might want to take a look at Lively Kernel for some ideas. But since then, HTML5 support for Canvas in both 2D and 3D has greatly improved so might be better option than SVG.
gruseom
When used as advertised, SVG is slow. But there's a way to use it that's actually much faster than Canvas for some purposes: batch up everything you want to render into a single blob of SVG text and then splat it on to the screen in one go using innerHTML. This is faster than drawing the same thing with a series of Canvas API calls and far faster than making incremental modifications to the SVG DOM.
zephyrfalcon
There are several Scheme implementations that are well suited to building "something big". Someone mentioned Racket, but there's also e.g. Chicken Scheme or Gauche, both of which come with a sizable collection of libraries to do real-world stuff. Then there are implementations like Scheme48, which seem to be tailored to building something "on top" of it.
IvarTJ
I personally dream of making a web community application with Scheme in which users can program bulletin items. Inspired by LambdaMOO, but with a focus on typical web facilities such as threaded discussion and wiki articles.