Wet Dream--python to native compiler

[Aaron K. Johnson]

In message <W5qL9.1984$LY2.101577 at newsc.telia.net>, "Fredrik Lundh" wrote:
> Aaron K. Johnson wrote:
> 
> > > 1. Universities aren't impressed by Compiler Writing any more. (Been
> > >    done before, they say).
> > >
> > > 2. There are too many languages, so any given Language gets a fraction
> > >    of what they would have got when 'there was Lisp ....  and there
> > >    was everything else'.  <Smiling warmly at the Lispers who are saying,
> > >    huh? But _nothing_ has changed ...>
> >
> > It seems that if a relatively underused language like ML could have a
> NATIVE
> > CODE compiler, than so could python, with its large user base.
> 
> users don't write compilers.

not all users write compilers. some users do. the larger the user base (one
would think) the larger the subset of those who are compiler writer/computer
scientist/hardcore users there would be.

or do you actually think that people who write compilers for a language never
use the language after writing it?

Anyway, most people tend to think where you need speed in python, re-write to
C. Most people don't see the need for a native compiler because of this. But
wouldn't it be great never to have to use { or } syntax ever again? Maybe its
just me, but I wonder why it took so long for such a beautiful and simple idea
as whitespace based syntax to catch on (or re-catch on--Fortran uses
whitespace, apparantly). It think if python could fix its speed issues, or at
least approach java efficiency, there would be no need for C/C++. I dream of a
day where one could write a device driver in pure Python!

cheers,
aaron.