[Catalog-sig] Do we really need this list?

David Lyon david.lyon at preisshare.net
Wed Nov 18 00:36:47 CET 2009


On Tue, 17 Nov 2009 18:10:56 +0000, Chris Withers <chris at simplistix.co.uk>
wrote:
> That may be, but discussions of one seem to invariably involve 
> discussion of the other. It seems not many people know about catalog-sig 
> anyway so the discussions end up on distutils-sig or, worse yet, 
> python-dev...

Well, anybody is welcome to start a discussion on some relevant topic.

For example, I don't mind seeing discussion on PEP-345. It was
created in 2005 so that means it's been open for four years.

Also, I've noticed that there are two distinct political camps with
python packaging. 

One camp is the traditional pythoners who were there
from day one, know all the tools, know the tools that they like, and
know how to install them, manually if necessary. Generally speaking
they are python centric and not interested in any other languages or
even try them. That's ok. Let's call them specialists.

The other camp are the ones who have come from the outside python
and are used to perhaps alternate packaging solutions. Maybe this
camp go from place to place and are exposed to different tools and
have to make whatever they are given work. Lets call them corporate
software contractors.

Moving on...

As some know, Guido posted a week or two back on getting python
more functionally compatible with CPAN. Outsider's "know" that 
python lags perl (and other languages) in third-party package 
loading capability.

It's not always possible to discuss those issues on distutils-ml
because everything must be realistically implementable within the 
legacy code framework of distutils.

A lot of PEP discussion should be high level and it should be 
driven from catalog-sig instead of distutils-sig imo. Especially 
PEP-345.

Catalog-sig should be about the real world needs of packaging and
distutils-sig should be about the implementation imho.

within




More information about the Catalog-SIG mailing list