C's syntax (was Re: Python Formatted C Converter (PfCC))

Alex Martelli aleaxit at yahoo.com
Wed Oct 25 06:59:00 EDT 2000


<jay.krell at cornell.edu> wrote in message
news:mailman.972458596.11769.python-list at python.org...
    [snip]
> >> I like the C++ template style:
> >>
> >pointer<array<int>> a;
> >
> >is illegal because it will be parsed as a
> >right-shift... "implementation before design" seems to be C++'s motto.
>
> Stroustroup admits he could've/should've fixed this in Design&Evolution,
but
> it is apparently too late.
> But, please, "implementation before design" is the overwhelming pattern in
> programming. Just look at Java and Python imho. Stroupstroup does lay out
> his design goals in The Design&Evolution of C++, a book I strongly
> recommend.

I second the recommendation.  Regarding the theoretical primacy of
design -- that can only exist for artefacts that do not _evolve_...
in practice, as you deliver artefacts to users, their perceived
needs change (partly as a result of getting experience with your
artefact!), so, you must evolve your artefacts _within the
constraints of by-now existing usage patterns_.  One think I like
about Extreme Programming is how it faces up very frankly to
this inevitable fact, and deduces procedures from there (by sharp
contrast, the now-discredited "waterfall model" assumed you _could_
do all of your design up-front, without delivering any usable
code until your design was full and complete...).


Alex






More information about the Python-list mailing list