pro python paper

Peter Hansen peter at engcorp.com
Tue Aug 10 10:03:48 EDT 2004


beliavsky at aol.com wrote:
> ccdetail at gmail.com (ccdetail) wrote in message news:<305aa162.0408091905.5bce8f9a at posting.google.com>...
> 
>>http://www.journyx.com/pdf/PythonAtAGlance.pdf
>>
>>the above paper is a response to some of our prospects complaints
>>
>>they were asking us why we use python
>>
>>thought you guys might be interested
>>
>>http://journyx.com/clf
> 
> I think the paper exaggerates the virtues of Python relative to other
> languages, reducing the credibility of the authors. 

I agree with this to some extent.

> The paragraph in the Python myths section
> 
> "It has no compiler to native code, which means Python programs run
> slower.  True, it does not compile. That does not mean, however, that
> Python programs run slower. If you are tracking the speed of quarks or
> landing the space shuttle, you might want to write in Assembly or some
> other very low-level language.  Other than that, a standard business
> application written in Python will run every bit as well as one
> written in any other language."
> 
> pretends that the main alternative to Python is Assembly, or some
> other
> "very low-level language". 

Actually, it does't really do that.  What it says clearly is that for
certain types of application (time-critical ones), Assembly or another
low level language would be a better alternative to Python.

This is not at all the same as the general statement "the main
alternative to Python is Assembly".


> Have the authors not heard of C++ (with the
> Standard Template Library) or Fortran 95 (with multidimensional arrays
> and associated intrinsic functions)?

I'm quite sure they have.  Does anyone really think of C++ or FORTRAN
for *time-critical* software (this is different than software where
performance is the ultimate goal, by the way).

> It is more honest and credible to admit that a Python program often is
> slower than one in a compiled language, but that reduced programming
> time in Python often outweighs this deficit.

That's true though.

-Peter



More information about the Python-list mailing list