[pydotorg-www] Final draft of redesign RFP

[anatoly techtonik]

On Wed, May 9, 2012 at 2:20 PM, Michael Foord <michael at voidspace.org.uk> wrote:
>
> One of the big problems we have right now is the complexity of the toolchain and the work required to get new contributors. I'm really hoping the new system solves this problem.

(disclaimer: the text below may hurt feelings, as usual, nothing personal)

The main problem is not the toolchain complexity, but the process,
people, and as a result - time required to get approval for changes. A
lot of open source Python projects have everything OK with the web
part, and python.org is a central for many of them. I doubt anybody
would resist from testing your skills against python.org challenge.
Unless discouraged.

Toolchain is the last reason why no volunteers step up to improve the
things. It starts with the pydotorg crew. If it was more inclusive for
new contributors, more open and thought about new volunteer's needs
instead of teaching "their own ways" or at least try to understand the
problems new people may have with the whole process, then the
situation can be different. I'd say that pydotorg crew should make the
first step to welcome new contributors and proactively attract them.
If pydotorg crew doesn't do this - it is dead. By attracting I mean
giving at least one talk about pydotorg stuff on each local PyCon,
listing current problems and reporint status. Right no there is no
even tracker - I have to open my own 'personal' tracker for python.org
just to list things I like to be fixed [1], [2].

Ideally, python.org could be an exemplary project where "eating your
own dogfood" principle could lead to the evolution of Python on the
web. A place, where design decisions could be weighed, argumented and
protected. There could be a book listing all historical approaches to
evolution of request processing in different frameworks. There could
be a DB listing user stories and problems encountered to match them
against one framework or the other. But, of course, there should be a
place first to discuss the development process and the ways to improve
it.

pydotorg development team is dead. It doesn't have name, doesn't have
members, goals and public place. From outside it looks that the team
is Martin alone, and his sole goal is getting stuff fixed - not
community support or making the stuff more attractive for newcomers.
There are some people who maintain python.org services in private
pydotorg mailing list (sorry guys, don't remember your names), and
while I suspect that they should be doing something important, I don't
really know what exactly they do. Last time I've tried to remove damn
/moin/ suffix from wiki.python.org, I have not only to persuade people
who don't use wiki as much as I do, but also teach how to do this
without any access to configuration files. I've tired.

There is no guarantee that buying python.org will make it alive. It
may have an opposite effect. I'd say the starting point is to make the
process open. Take OpenStack as an example. Ask them what is to be
open and how to make it work. It is not that I am not going to
persuade anybody myself - it just seem pointless from the point a
well-known troll. Why am I not doing anything myself? Well, until
`pydotorg` list is a closed list and public volunteers are not looking
more important than some "regents collective" (even if it is not so in
reality), my OCD doesn't allow me to concentrate on anything else.

I hope this raises some points.

1. http://psf.upfronthosting.co.za/roundup/meta/issue340
2. http://code.google.com/p/pydotorg/issues/list
-- 
anatoly t.