Python in business programming [was Re: Beginning programming with Python]

[Gordon McMillan]

Nemeth Miklos writes:

[snip]
> The only thing I wanted, is to point out a potential shortcoming (if at all) of
> Python -- I am not sure if the lack of compile-time type safety is really a
> badly needed feature.

It is, but only if you're used to relying on it. That is, some 
people miss it dearly. Coming from a background of languages 
that did static type checking, I did at first. Adapting has meant 
big changes to my development style (the code, compile, test 
cycle is now rarely more than half an hour). Test suites are 
more important (but also easier). For large systems, 
decomposing the problem cleanly is very important - to 
preserve your sanity you need the smallest, cleanest 
interfaces possible between subsystems. But, if at first you 
foul it up, you can still rewrite most of it before lunch.

> I am neither a physicist nor a mathematician, I am a simple business software
> engineer. I like Python but I'd like to be convinced that Python is not only
> for scientific programming, but also for business programming. 

I can assure you that there are lots of business programmers 
who use and love Python. Probably around half of c.l.py 
contributers are involved in business, not math / science. 
Businesses tend to be pretty conservative, however, and 
exercise rigid control over the tools their programmers use. So 
acceptance at the mangerial level is slow.

In my business as a consultant, most of my clients want C, 
C++ or Java. That doesn't mean that I don't write it (and rewrite 
it until it's right) in Python first. And sometimes they accept 
delivery in Python (when they hear what the price tag would be 
to get it in C, for example).

- Gordon