Subject: Re: What open source implementation of Lisp do you prefer and why?
From: rpw3@rpw3.org (Rob Warnock)
Date: Wed, 04 Mar 2009 04:32:53 -0600
Newsgroups: comp.lang.lisp
Message-ID: <06udnf1h1ZRIxTPUnZ2dnUVZ_uqdnZ2d@speakeasy.net>
Javier  <javuchi@gmail.com> wrote:
+---------------
| What open source implementation of Lisp do you prefer and why?
+---------------

CMUCL, because it's solid [IME], has a good compiler, and its
interpreter's design means that it's good for "scripting", too:
it starts up *fast* [slightly faster than CLISP, actually, on
the machines I use], and its interpreter does lazy analysis of
DEFUN bodies [which is good if you have large scripts with a
bunch of DEFUNs only a few of which get called on any single
script invocation]. Its "green threads" are also well-integrated
with standard CL stream I/O, and give you automatic non-blocking
I/O when running multiple network connections at the same time
[e.g., when using it as a web application server].

It has some downsides, but none of them currently affect me:

- Only has "green threads" [user-mode coroutines], so threading
  only uses a single CPU, and that only on the x86 platforms
  (Linux & BSD, mainly). [But all the platforms I currently
  run it on are single-CPU, x86 Linux or BSD.]

- Doesn't (yet) run on Windows. [But I don't do Windows.]

- Rebuilding is a bit tricky. [But I tend to run the "-RELEASE"
  binaries without ever rebuilding.]


-Rob

-----
Rob Warnock			<rpw3@rpw3.org>
627 26th Avenue			<URL:http://rpw3.org/>
San Mateo, CA 94403		(650)572-2607