float / double support in Python?

Andrew Bennetts andrew-pythonlist at puzzling.org
Wed Feb 12 03:18:02 CET 2003


On Wed, Feb 12, 2003 at 01:47:11AM +0000, Brandon Van Every wrote:
> 
> Points taken.  But my point is, the leaner Python is, the closer you can get
> to low level before having to switch over.  Ergo, it makes sense to support
> all basic machine types.

The problem being that the more low-level bits Python directly supports, the
less productive it is, because it forces you to care about the low-level
bits.  That probably suits you; but for the bulk of Python programmers the
difference between float and double is just an annoyance that would get in
the way.

I use Python *because* it abstracts away these details, because mostly they
don't matter.  If do they matter for something I need to do, I'd try to find
or write an extension module for Python that does the details for me, and
then continue to use Python for the rest.

There is no one programming language that's perfect for everything, of
course.  But Python is excellent for most things, and for those it's not it
can often be extended in C/C++ for anyway.  But saying that Python should do
something like distinguish between floats and doubles suggests that you're
missing the point of Python.

-Andrew.






More information about the Python-list mailing list