Prev: comp.lang.lisp charter
Next: Casio Collection Men's GOLD LABEL Edifice Watch - Replica Watch Fake
From: Sean T Allen on 23 Apr 2008 20:25
On 2008-04-22 03:01:33 -0400, Scott Burson <FSet.SLB(a)gmail.com> said:

> Just came across this...
>
> http://www.regdeveloper.co.uk/2008/04/21/bray_ruby_rails/
>
> You'd think there would be some way for Lisp to fill this gap.

i've been thinking about this after going a pass at checking out cl-python.

the rubinius folks are working on a vm for ruby and a class library where
everything is in ruby. if someone did a cl-ruby it could leverage that work
by the rubinius people.

there is a lot of interest in the ruby community in improving the performance.
there are numerous attempts at it. running on .net, on java, new
interpreter for
the standard ruby, rubinius... if a cl-ruby implementation could show obvious
speed or other gains over rubinius and use all those pure ruby libraries.. then
headway might actually be made.

From: Jason Nielsen on 23 Apr 2008 23:01
On Wed, 23 Apr 2008, Sean T Allen wrote:

> i've been thinking about this after going a pass at checking out cl-python.
>
> the rubinius folks are working on a vm for ruby and a class library where
> everything is in ruby. if someone did a cl-ruby it could leverage that work
> by the rubinius people.
>
> there is a lot of interest in the ruby community in improving the
> performance.
> there are numerous attempts at it. running on .net, on java, new interpreter
> for
> the standard ruby, rubinius... if a cl-ruby implementation could show obvious
> speed or other gains over rubinius and use all those pure ruby libraries..
> then
> headway might actually be made.
>

The path of least resistance is to take the output of parse_tree
(http://parsetree.rubyforge.org/) which outputs sexp's and write a
translator from the this "ast" to common lisp. I am sure that a
performance boost will be had with little effort! Lisp compiler magicians
have been figuring out how to squeeze performance out of a dynamically
typed language for... say a few years. If you are sufficiently motivated
you could likely do some type inference from your sexp ast tree to common
lisp and decorate with type info... and wizz-bang easily 20-50 fold
performance improvement! What you don't know won't hurt you but what you
do know makes you realize how painfully unknowledgeable you are...
re-educate and repeat until sufficiently disheartened by the recursiveness
of it all ;-).

Jason
First  |  Prev  | 
Pages: 1 2 3 4 5 6 7
Prev: comp.lang.lisp charter
Next: Casio Collection Men's GOLD LABEL Edifice Watch - Replica Watch Fake