Yeah!
Real Python programmers like Python so much they want to be able
to use Python for everything - including things Python wasn't
originally designed to do.
Also, the fact that Python is so well written, simple, (mostly) logical
and easy to understand makes hacking it a tractable temptation.
Also, as good as Python is, it isn't perfect. But trying to figure out
how to make it better takes one into some very educational territory.
Python has done most of the simple stuff right. Most of the problem
areas get you into some of subtle problems that are still research
areas, or at least, not trivial. ( See past discussions of "overly
pedantic type checking" for example, or more recently: how to integrate
hypertext + literate programming + safe distributed agents & modules.)
But, onto another thread:
I once noticed that Guido had the start of a Python to C translator.
I assume the intent was that you could prototype code in Python, and
if you needed better performance, you could compile the Python into
C and the C into a loadable python object module.
[ I don't know how far from ready-for-prime-time that code is. ]
Has anyone considered Python to Modula-3 ( and figuring out how to
integrate modules written in Modula-3 into the Python image. ) ?
That would seem like an easier mapping. ( Maybe not - I also assume
that the python to C translated modules would make heavy use of
the modsupport and other Python runtime support - i.e. they wouldn't
be a translation to "raw" C . )
- Steve Majewski (804-982-0831) <sdm7g@Virginia.EDU>
- UVA Department of Molecular Physiology and Biological Physics