[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Chicken-users] Numbers anywhere?

From: Joerg F. Wittenberger
Subject: [Chicken-users] Numbers anywhere?
Date: 15 Jan 2003 19:13:21 +0100
User-agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Common Lisp)


Felix Winkelmann <address@hidden> writes:

> Joerg F. Wittenberger wrote:
> > BTW: I used the flush enabled output ports to improve the
> > chickenlib/socket.scm; it's still not perfect but faster on average.
> > Anybody at the list interested?
> Well done! Where's Tony...? ;-)


Slow down.  Another turn and everything breaks.  Right now I see the
chicken version take up at most 60% of CPU and deliver at a 15th the
speed of the rscheme original.  Profiling is turned on.

* when I switch profiling off and all but -unsafe optimization on, I
gain approx. 30% speed.  Is this normal range?

* any numbers how rscheme and chicken are supposed to compare speed

* does anybody have some fast code to multiplex i/o around select(2)?
  I really doubt there must be a problem with the chickenlib code, but
  I can't find it.

Also of high interest, but maybe not for the list, so please reply
privately, if you have following info:

* anybody here, who knows how much faster apache is for plain file in
  comparison to apache/mod_perl, php etc. for higly dynamic pages?
  more like 10:1, 5:1, 2:1

* How fast would you expect an all-Scheme web server to be in
  comparison to apache?

best regards


The worst of harm may often result from the best of intentions.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]