Lisp web frameworks [was: Lisp user group culture]
- From: Robert Uhl <eadmund42@xxxxxxxxxxxxxxx>
- Date: Sun, 11 Mar 2007 15:01:14 -0600
Brian Adkins <lojicdotcomNOSPAM@xxxxxxxxx> writes:
The "halo effect" of Rails is pretty significant. I think a clear
message about how to be productive developing web apps with Lisp (even
if some frameworks/libraries have to be created, polished or
repackaged) could spark some interest. I admit I was impressed enough
with the Rails screencasts showing how to create X in Y minutes to
check it out - and it actually worked as advertised!
I see no reason why a framework that makes web development in Lisp as
easy and enjoyable as Rails can't be put together.
I've been kinda playing with this for my Tasting Notes app
<http://latakia.dyndns.org/tasting-notes/>. I've been very happy with
Django on the Python side of things, and thought that it'd be cool to
try making a Lisp Django to help run my stuff. Right now it's very
basic, but I hope to improve it considerably.
At the moment, I'm just happy that I was able to replace my Python app
with a Lisp one. My next bit of work is to (finally) finish my port of
Blosxom, and get my blog working on it.
Maybe eventually I'll release skeletor to the public. It's a pretty
basic framework right now (hence the name), but it has made a few things
Even if Lisp were *only* comparable to Ruby vs. superior with respect
to productivity, the performance gain is a significant factor. Yes, I
know you can scale Ruby/Rails; however, a 10x or more performance gain
definitely got my attention.
Believe it or not, Lisp isn't particularly faster than Python for my web
app. It was actually slower, until I optimised two functions. But
then, I rather suspect that the bottleneck in a database-backed app is
the database, not the front end.
I admit the same question I've heard from others has crossed my mind,
"If Lisp is so great why don't libraries, etc. exist for it like they
do for Ruby, Python, ...".
I think the problem is that thousands of libraries exist: little
one-offs that a fellow made to solve a problem, and never really got
around to releasing. It's almost too easy to write a library in Lisp!
I do feel that the lack of a clear open source Lisp leader (as far as
I can tell) could be a hindrance compared to Ruby/Python/Perl/PHP &
now Java, but I don't have enough knowledge/experience to judge that
at this time.
I've placed my bets on SBCL: native code, simple build process, decent
extensions. But no doubt CMUCL and CLISP will be around for years to
come. I don't think that their existence is too much of a hindrance,
any more than the existence of Jython and friends hurts Python.
Robert Uhl <http://public.xdi.org/=ruhl>
Apple's original usability studies contradicted the Xerox ones. The
difference? Xerox studied people who were used to the idea of computers
and user interfaces. Apple studied random lusers. --Peter da Silva