Thoughts about extensions to the Python language

Grant Griffin not.this at
Sat Mar 17 06:45:51 CET 2001

Carlos Ribeiro wrote:
> (flames off please ...)

If you want flames, you came to the wrong place.  (If you don't mind me
saying so. <wink>)

> What attracted me to Python was the simplicity and cleanliness of the
> language. Code written in Python is as clean as code can be.

Oh, except list comprehensions.  (Yup, I use 'em--but I always feel
guilty afterwards. <wink>)

> I have written
> systems using several other languages, including C, C++ and Perl, and none
> of these languages can even come close of Python in terms of readability.

Still, one can write FORTRAN in any language.  (But it takes so much
work to write FORTRAN in Python that lazy folks don't bother.)


> 6) Most important of all, it's clear that Guido had used a 'common sense'
> approach to design the original language.

Are you sure?  It appears to me that the strength of his sense is its
very _un_common-ness <wink>.  Anyway, I guess we just have to trust him
(at least until he leads us wrong--then we'll take him to meet Jimmy
Hoffa <wink>.)

As to the PEP process, it seems pretty good to me, except that it
sometimes gives the impression (perhaps mistaken) of "design by
committee".  In other words, new things seem to happen because somebody
thought of a good idea, and Guido agreed.  Now that's not all bad (after
all, it _was_ a good idea), but ideally, there would be some larger
vision for the future of the language.  (Guido has probably spelled that
out somewhere, but I guess I missed it.  Sling me a link, /F. <wink>)

Now, the usual manifestation of "growth" in software is new features
(often ones that merely clutter the forest with more trees), but Python
has painted itself into a corner in terms of its major feature being its
lack of features.  We love it for that, but that works against
"growth".  (Contrariwise, Perl has painted itself into the corner of
already having so many kitchen sinks that it's hard to find yet another
one that needs throwing in.)

So where does language growth come from?  Modules, Benjamin.  Modules. 
And documentation, Benamin.  Documentation.  I guess I'd like to see
more stuff added to the standard distribution.  For example, I think if
Numeric were in the standard distribution (or maybe in an alternate
standard distribution), that might make it easier for me to sell snake
oil to my friends.  (When I tell 'em they've got to download and install
six packages to run my software, they roll their eyes--or nod off.)

Oh, and the fact that Python doesn't have any single, compelling
data-plotting system (rather than several not-very-compelling ones)
makes it hard to sell as an alternative to Matlab.  Numeric users find
themselves pretty blind without that.

And from another angle, if Python ever shed its dynamic skin and came in
a static/compilable version (let's call the new one...oh, lemme
see..."P3K"), that would be a *big* step towards growth: scripting
languages are great, but there are very good reasons most of us still
use statically typed languages such as C/C++ for most of our work--and
will do so into the indefinite future.

embedded-real-time-ly y'rs


Grant R. Griffin                                       g2 at
Publisher of dspGuru                 
Iowegian International Corporation  

More information about the Python-list mailing list