The ASPN compiler

Russ Salsbury clpy.NOSPAM at russellsalsbury.com
Tue Jul 15 00:18:09 CEST 2003


noah at noah.org (Noah) wrote in message news:<c9d82136.0307120941.63540143 at posting.google.com>...
> Gerhard Häring <gh at ghaering.de> wrote in message news:<mailman.1057969337.5637.python-list at python.org>...
> > Fuzzyman wrote:
> > > What's the score on this (the ASPN python compiler) 
> > Read the whitepaper, available on the ActiveState site.
> > ...
> > -- Gerhard
> 
> The short answer is that it looks like a dead project.
> 
> Actually, you are better off going to Mark Hammond's .NET page here:
>     http://starship.python.net/crew/mhammond/dotnet/

Well, folks, I actually went off and read the paper.  In spite of the
declaration of success at the end of the paper, the project appears to
have been just an academic proof of principle exercise.  The compiler
is too slow to be of practical use and not much of the much of the
runtime and modules were implemented.  You should read it; it doesn't
look like it could have ever succeded.

The paper suggests that a way around the mismatch between the dynamic
typing of Python and the static typing of .NET would have been to
introduce advisory type declarations into Python.  Think Guido would
buy that?  It also suggested maybe type inferences could solve the
problem.

>From this paper it's evident that this was not a class effort like
Jython.  Was Active State suckered by some marketing ploy of
MicroSoft?  To be fair maybe they all expected a follow-on project to
do a real implementation, but I bet it's all part of VB-forever.

I also noted there is a dead-in-the-water Perl.NET implementation.

-- Russ




More information about the Python-list mailing list