From: William Djaja Tjokroaminata <billtj@...>
Date: 2002-10-03T12:04:40+09:00
Subject: Re: thoughts on typelessness

Hi batsman, holy python!

My goodness!  It looks that they are already doing (or finishing) *the*
right thing to do, while Alan Chen and I were still discussing
it.  Hopefully Rite will incorporate a similar concept as Pysco.  However,
I really doubt its goal:

    "High-level languages are faster than low-level ones!"

no matter how much Pysco may use gazillions of bytes for memory.  For one
thing, assembly is (in general) always faster than C.

The Psyco web also contains several links; one of them is Pyrex.  In
Pyrex, we can write

 def primes(int kmax):
  cdef int n, k, i
  cdef int p[1000]
  result = []
  if kmax > 1000:
    kmax = 1000
  k = 0
  n = 2
  while k < kmax:
    i = 0
    while i < k and n % p[i] <> 0:
      i = i + 1
    if i == k:
      p[k] = n
      k = k + 1
      result.append(n)
    n = n + 1
  return result

which just exists in my dream in Ruby.  Ahhhh.......

Bill
============================================================================
Mauricio Fern?ndez <batsman.geo@yahoo.com> wrote:
> I knew it ;-)
> In fact the exact wording of the paper is 

> "Preliminary results show potential speed-ups of a factor 100, which
>  means than we can hope execution speeds closer to fully optimized C than
>  to conventional Python." (sic)

> BTW, more info at http://psyco.sourceforge.net/