NEW LISP

They should combine the major lisp languages, CL, Racket, Clojure, Scheme, Arc, etc, to form M-lisp

The names inspired from String Theory being shaken out to form M-theory, the most complete science theory developed.

What do you think?

It would be lazy, multiple cored, anything oriented, macro/small definition, and if you really wanted, a simple parser could be written to translate any language.

It wouldn't be lazy, it could* be.

Laziness by default is not a virtue :(

How would you combine that? Racket is Scheme with stuff, Arc is newly-specified so it may have better conventions and naming but less stuff, but Common Lisp? Common Lisp is a Lisp-2, old and ugly, and can't be combined with the others without fucking them all over.

The best you could do is to have Arc (perhaps, I don't know much of it) with Racket's environment and library minus classes, and a CLOS-like object system.

I'd suggest a different name, because of M-expressions.

>>1
NewLISP and MLISP are already things. Stop being so clueless.

Common Lisp is a Lisp-2, old and ugly, and can't be combined with the others without fucking them all over.
Cry more.

Great, another idiot who doesn't actually know any of the languages he worships, but still feels the need to jump on a bandwagon.
Stop making threads, >>1.

Lisp is nothing more than a repl loop that can be written in one page of code. Lisp is written in itself so youre only using someones interpretation of how Lisp should be implemented. Most lisp constructs that we think as part of lisp were actually just functions that lisp programmers wrote and got added to the "language" (even though there is no standard lisp language, there are only standard implementations).

>>5
Cry more.
Wah wah wah. CL is still dirty.

What really is the essence of list processing? Is it CLOS? Is it automatic memory reclamation? Surely it must be the late binding of variables?

Take away the garbage collector and dynamic type information and you will get something layered not far above assembly language. This would merely be a stripped down abstract syntax tree where only the order of evaluation and data passing is expressed. Not a merry world to play in, where you can't tell a symbol from a pair! Yet, this Lisp would still bear with it something in the spirit of Lisp, namely macros. As such, it would be very suitable as the language of a code generator, created by a compiler.

Add static type information and you get C and Pascal with funny syntax. Reunite it with a garbage collector and you get ML or an efficient version of Java. Add dynamic type information and you get Perl, Python and Ruby. So why put up with the parenthesis if its all the same anyway? Well, being able to transform the code with macros is like crafting your own compiler, tailor made for your problem space. In Lisp, there is no such thing as boilerplate code.

Truly Lisp is every language, especially Lisp! And its essence is its syntax!

>>8
Implying "dirty" even applies to programming languages.

I'm actually currently working on a specification for a personal language which I call *Tangle*, I am going to try and include the following features:
    - apl/j-style syntax
    - builtin PCRE
    - dynamic typing
    - no OO
    - macro definition if possible
    - hyperoperators
obviously the default datatype would have to be arrays

Don't change these.

Name:		Email:
Entire Thread Thread List