On 10/13/05, Giovanni Pensa <[EMAIL PROTECTED]> wrote:
>
> While in a "housing" solution, how does Spiffy go?  Any experience,
> benchmarks, opinions about it?  How could it compare with a
> fastcgi/scgi solution?  I see two mod_ solutions (Perl and PHP),
> losing popularity against fastcgi/scgi solutions (PHP, Ruby, Python)
> and only a few HTTP solutions (Lisp/Araneida, Python/Zope).  I
> understand a Common Lisp server, but maybe scheme is better suited for
> "scripts". But Chicken can also be compiled, so I'm a bit confused.
> Ideas?
>

I can't really say much about all those different methods of interfacing
to a http-client, but what I like about the "direct" approach with spiffy is
that I have full control over dynamic content generation and the fact
that I don't need to install and setup a separate tool. Directly running
your code (interpreted, compiled and interpreted on the fly in .shp
pages) in the server allows quite elegant solutions.

But in the end it is a matter of taste, of course. Some things are
much better handled via a cgi.

Spiffys performance is acceptable (IMHO), some very simple benchmarks
I made a while ago showed it to be halfway between the PLT server
and the S48/SUnet server (serving a tiny static page). For complex
dynamic content generated in compiled code, Spiffy should be faster
than the PLT server, though. Chicken as the underlying implementation
has very effficient threads - it would be interesting to compare the PLT
server with Spiffy under very high load.


cheers,
felix


_______________________________________________
Chicken-users mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/chicken-users

Reply via email to