GCD in standard library?

[A. Lloyd Flanagan]

Gareth McCaughan <Gareth.McCaughan at pobox.com> wrote in message news:<slrnb70gip.12d5.Gareth.McCaughan at g.local>...
> Tim Peters wrote:
> 
> For what it's worth, I'm all in favour of an enormous
> standard library provided that
> 
>   - it's really well designed
>   - it's efficiently, portably and robustly implemented
>   - all of it is actively maintained and can be relied on
>     remaining so for a good while
> 

Absolutely.  GCD, for example, might be easy to code, but where's the
harm in creating a fast, industrial-strength version and putting it in
a library?

I think the biggest reason Java (which has a lot of problems) has been
so successful is the huge number of standard libraries available for
it.  We should be encouraging the same for python.

(As an aside, we should also be upgrading them independently, not
tying all the standard libraries into a particular version of python
and only releasing them when a python release is made.  That's one of
those Java problems I alluded to earlier.)