[Catalog-sig] Package comments

[James Bennett]

On Wed, Nov 4, 2009 at 5:01 AM, René Dudfield <renesd at gmail.com> wrote:
> People with commercial packages on there would be right to not like
> comments in some respects.  Since commercial organisations often like
> to control their PR as much as possible.  So in this way, the comments
> are not such a good thing for PIL.  However also, commercial
> organisations pay a lot for feedback, QA and market research... so in
> a way it is also good for commercial packages too.
>
> The python community likes openness I would say, and comments go
> towards more openness.  Comments move the communication on pypi from
> entirely author based, towards letting users speak as well.  Should
> openness be valued more than valuing an authors wishes?

Except this isn't "openness", not by a long shot. The current
implementation gives package maintainers *no voice whatsoever* --
openness, discussion, communication, etc. must be able to go both
ways.

More important, though, is the question of what PyPI is really
supposed to be, and what features are part of its scope. Personally I
lean toward keeping the scope narrow and letting other services fill
in additional features as desired, but it's instructive to note that
the only times I've ever seen the broader approach work well are in
communities where it's assumed and understood from the outset that
you're buying into a pretty large stack of development and
project-management technologies (e.g., Debian's packaging system,
Perl's CPAN, etc.). PyPI is not and never has been such a thing.



-- 
"Bureaucrat Conrad, you are technically correct -- the best kind of correct."