Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!rutgers!apple!bloom-beacon!DACTH51.BITNET!FM
From: FM@DACTH51.BITNET
Newsgroups: comp.lang.scheme.c
Subject: Oaklisp
Message-ID: <8904160055.AA05556@rice-chex.ai.mit.edu>
Date: 15 Apr 89 22:24:00 GMT
Sender: daemon@bloom-beacon.MIT.EDU
Distribution: inet
Organization: The Internet
Lines: 15

A while ago somebody, I think one of the authors,  K. Lang or B. Pearlmutter,
posted a arcticle saying that a new release of OakLisp is available.  I don't
have this article anymore, but anyhow it would  be  pretty  useless  since  I
can't ftp.  Does somebody know a way  I  can  get  Oaklisp  over  BITNET ?  I
really like the ideas in Oaklisp and I can't wait to play with it.

In that same article the  author  claimed  they  have  a  fast  long  integer
arithmetic.  The multiplication works in O(N^1.59).  As an  example  he  used
the naive way to compute factorial(1000).  That seems a little bit strange to
me.  The only O(N^1.59) multiplication that  I  am  aware  of  is  Karatsubas
method.  If you look at it closely, it is  a  O( N^.59 M )  algorithm,  where
N is the size of the smaller number and M is the size of the  larger  number.
Thus you don't gain anything if you multiply a large number by a  small  one,
which is what you do all the time if you  compute  factorial(1000).  Comments
anybody ?