Lisp for the 21st Century | Hacker News Reader

Lisp for the 21st Century

lambdassociates.org

Starter Pack is another effort that addresses some of the noted deficiencies in Common Lisp.

Technically, the whole project including GUI installer only works on LispWorks. However, the selection of libraries makes a good approximation to a CL standard library and it's not too hard to figure out how to install all of them in SBCL or whatever other CL implementation you have. This gets you much closer to the starting point of a "batteries included" language like Python.

If every CL implementation in the world simply bundled all of the libraries in Starter Pack, a lot of the legitimate criticisms of CL would be solved.

(If you are one of those people who just say "too many parentheses", however, there is no help for you.)

Heh... From reading the first part of this, I rather expected it to be an implementation of a Lisp-like language (complete with first-class functions, macros, the whole deal, maybe sans some of the bloat) on top of some other runtime environment like Java or Python. That would give the developer the power of Lisp with the libraries of some other language.

It quickly became apparent that it was just an attempt to "dumb down" Lisp.

Incidentally, does anyone know of a (mature) language that does the above? Preferably derived from Scheme? I have really wanted to use this for some time, but I've resisted implementing it myself, thinking that "Someone MUST have done this by now..." As it happens, people have, but it's typically more of an "I hacked this together last weekend; please don't actually use it for anything!" type project.

Qi is far from being a 'dumbed-down' Lisp. It's got a turing-complete type system (IIRC something is of a type if it satisfies /full code/), it unifies a couple of paradigms - it's an advanced language, not a dumbed down one.

Well, go figure -- not an hour after I post this, I find Kawa: http://www.gnu.org/software/kawa/

I remember finding this before, but for some reason deciding it was not what I wanted. I can't find anything important missing, though... Cool.

I was about to post saying you must not have looked very hard. I've been using Kawa successfully for years. Its biggest fault is that there are so few bugs that I'm not motivated to keep up with new releases.

There is at least one other JVM-based Scheme implementation out there, and it's stable and well maintained too.

There is at least one other JVM-based Scheme implementation out there, and it's stable and well maintained too.

The next LispNYC meeting (Tues., Nov. 13) is going to be on this subject:

"Rich Hickey Presents Clojure - A Dynamic Programming Language for the JVM" (http://lispnyc.org/home.clp)

I just remembered the name of the other implementation: SISC, an interpreter. (Kawa is a compiler, but it does have an interactive REPL.)

I remember looking at Bigloo a long time ago, but I haven't kept up with the development.

Bigloo is a Scheme compiler that supports the JVM as one of its back-end targets.

Kawa is a JVM compiler that supports Scheme as one of its front-end languages.

SISC is an interpreter that runs on the JVM. There are other Scheme interpeters for the JVM, but SISC is the most prominent and said to be the fastest interpreter. In some tests it even beat the Kawa compiler.

Looks like pg hurry things up with getting Arc out or Qi will beat it to become the 21st century 100 Year Language. :)

Wow - what a great effot. I'm definitely looking forward to the 2008 version which includes web programming.

And #FFFFFF on #000000 for the 20th.

Do efforts to save Lisp just detract from the growth of functional programming in general?

(1) Lisp isn't (just) a functional programming language.

(2) Why should lisp advocates care about "functional programming in general"?

I agree with (1), but most "functional" languages aren't just functional.

As for (2) -- "A rising tide lifts all boats."

Reply on news.ycombinator.com