Not enough Python library development [was PEP scepticism]

[Bruce Sass]

On Wed, 4 Jul 2001, Paul Prescod wrote:
> Bruce Sass wrote:

> > I think this needs a PEP from the core Python developers,
> > for four reasons:
>
> I have only a meta-comment. After years of discussions about Python
> CPAN, I can see two things that are guaranteed to kill any effort:
>
>  1. The requirement that it be sophisticated, comprehensive, logically
> consistent from the start.

Even if there are working schemes that meet those requirements?

I can see your point being a problem if it is necessary to start from
scratch and think up all the sophistication and consistency.

>  2. The requirement that the core team be deeply involved in its
> creation.

Ok.  I think `well represented' is good enough for the creation...

...and see it as a matter of convincing Guido and the gang that there
would be clear advantages to moving towards a package based
distribution... how else do you have a big standard library and not
fill up peoples harddrive with stuff they don't want or never need,
or upgrade pieces of the system without doing a full release.


> > 1) anything less is "just" third-party (no matter how well done)
>
> Well, that's better than what we have today. And the Vaults (what we
> have today) is better than what we had before it. Incremental steps are
> the only path that will lead to success.

I agree.

The next incremental step seems to be packaging up Python modules.


I should have added a fifth point...
5) if any packaging scheme comes to fruition and is adopted by the
   community, the developers would have to live with it (by either
   using it themselves, or dealing with people who are)


- Bruce