ANNOUNCE: Python Compiler (No spell errors)

n89553 at stud.upb.de n89553 at stud.upb.de
Thu Feb 24 14:27:31 EST 2000


Andreas Otto wrote:
> > Does a requirement exist for a Python compiler and are your ready to pay
> > for it ??
[...]
> Our compiler is
>  a) not free
>  b) optimizes the source code in an aggressive manner,
>     you will loose all the stuff you don't need.
>  c) produces one stand alone Python/C source-file,
>     with every command change to his c-command
>     ( no big String-Array )
>  d) you will find most of the errors at compile-time
> 
> my english is not as good as it should be

Having read your weg pages, I suppose your german language 
profiency can't called perfect either.
I hope that your customers won't turn down your offer because
of your inattention for these petty details.

I can't say that I've understood the kind of 'compiler' you offer, 
but I guess it'll sum up to something like a combination of 
the freeze and the py2c-utilities. Further code optimization 
requires object declaration.

Please correct me if I´ve got it wrong.


I guess something like that won't be reasonably applicable
for a rapid prototyping language, because you end up
correcting all these compiler errors.
But the status of a software project may change and other 
people have other uses of Python.

What about developing a backtracking algorithm which analyses the
flow of data from the program's interface to internal objects
and applies considerate restrictions based on several conversion rules
under consideration of the objects' __coerce__() methods?  

What about concentrating on special applications which do
usually not require some critical features of Python? 
What about optimizing (the correct term might be: "mutilating")
the language for embedded systems?

I think there aren't many political reservations against you 
wanting to get paid for your product. 
Said product should be worth the euros, though, and I'm not
convinced by a badly written Usenet posting or a clumsy web page. 


Friendly regards,

Mirko Liss

-- 
M Liss,  mail <n89553 at squid.upb.de>



More information about the Python-list mailing list