Large-scale Python (was: Language Niches (long))

[Paul Prescod]

Cameron Laird wrote:
> 
>...
> 
> Without contesting any of Paul's larger points, what
> are the facts here?  I believe:
> * Python is good for team-work
>   and big projects; and
> * There is increasing recogni-
>   tion and use of Python in
>   this role.
> What role has "language change" played in this, though?
> I don't think Python's syntax or semantics support
> modularization and packaging significantly better than
> in '96 (Paul, do you have a Jython point in mind?).

I can't remember that far back. One feature I depend upon for
large-scale programming is threads and they are certainly the sort of
thing that you can imagine would be added in the transition from
"scripting" language to "general purpose OOP" language. And the
scalability of Python threads could also be improved.

I also think that the deprecation and assertion features are very useful
in large projects. 

Nevertheless I concede that Python was always designed as if it were
meant for "real work" so it didn't have to change a great deal to be
considered appropriate for large projects.

> Does performance improvement in hardware which makes
> raw speed less frequently a constraint constitute a
> "language change"?
> ...

Not really. I have a sense that there are certain project-size and scope
threshholds that push languages to mature and Python smoothed over the
first one pretty easily because it was designed to scale in the first
place. I would not be surprised if future ones require more changes such
as interfaces, multi-processor threading and optional static type
checking. And of course performance improvements! Or maybe it just
*seems* like bigger programs need those extra features and when we get
there we will find that they do not.

-- 
Take a recipe. Leave a recipe.  
Python Cookbook!  http://www.ActiveState.com/pythoncookbook