[Foundation-l] Making wikimediafoundation.org more open to contributions

[Erik Moeller]

2011/1/29 phoebe ayers <phoebe.wiki at gmail.com>:
> Having many wikis is an ongoing source of irritation for many, and it
> would be great to resolve this issue. Are there good arguments *for*
> having separate sites?

Yes, and I think most people generally underestimate the complexity of
the issue. The reasons for WMF to spin up separate sites have varied,
but to try to put it as simply as possible, a dedicated wiki, in all
technical and social respects, focuses collaborative activity, which
can enhance productivity and reduce barriers to participation. In the
case of e.g. StrategyWiki, it also allowed us to try some radical
changes (like using LQT on all pages, or receiving hundreds of
proposals as new page creations) without disrupting some surrounding
context. I have absolutely no regrets about our decision to launch
StrategyWiki, for example -- I think it was the right decision, with
exactly the expected benefits.

Meta itself has grown organically to support various community
activities and interests that had no other place to go. It has never
been significantly constrained by its mission statement. The "What
Meta is not" page only enumerates two examples of unacceptable use:

1. A disposal site for uncorrectable articles from the different
Wikipedias, and it is not a hosting service for personal essays of all
types.
2. A place to describe the MediaWiki software.

Its information architecture, in spite of many revisions, has never
kept up with this organic growth, making Meta a very confusing and
intimidating place for many, especially when one wants to explore or
use the place beyond some specific reason to go there (vote in an
election, nominate a URL for the spam blacklist, write a translation).

So, let's take the example of OutreachWiki as a simple case study to
describe the differences between the two wikis.

1) The wiki's main page and sidebar are optimized for its stated purpose;
2) As a new user, you receive a welcome message that's specifically
about ways you can support public outreach (
http://outreach.wikimedia.org/wiki/Template:Welcome )
3) All special pages remain useful to track relevant activity or
content without applying further constraints;
4) Userboxes and user profiles can be optimized for the stated purpose
(e.g. http://outreach.wikimedia.org/wiki/Languages_and_skills )
5) There's very little that's confusing or intimidating -- the content
is clean, simple, and organized.
6) If the OutreachWiki community wants to activate some site-wide
extension, it can do so, focusing only on its own needs.

On the other hand:

1) Activity is very low;
2) The wiki is largely in English;
3) Meta has a long tradition of hosting outreach-related content, and
many pages still reside there or are created there.
4) The existence of yet-another-wiki brings tons of baggage and
frustration (more dispersed change-tracking for users who want to keep
up with all activity, more creation of meta/user page/template
structures, more setup of policies and cross-wiki tools, etc.).

It's not a given that 1) and 2) are a function of having a separate
wiki. As we've seen with StrategyWiki, activity is largely the result
of focused activation of the community. The small sub-community that
cares about public outreach on Meta is ridiculously tiny compared with
the vast global community that could potentially be activated to get
involved through centralnotices, village pumps, email announcements,
etc. So the low level of activity on OutreachWiki is arguably "only" a
failure of WMF to engage more people, not a failure of a separate
wiki. (It certainly makes all the associated baggage much harder to
justify.)

But, I think the disadvantages of working within a single system can
be rectified for at least the four most closely related backstage
wikis (Meta/WMF/Strategy/Outreach). I do think working towards a
www.wikimedia.org wiki is the way to do that, importing content in
stages, with a carefully considered information architecture that's
built around the needs of the Wikimedia movement, a very crisp mission
statement and list of permitted and excluded activities, a WikiProject
approach to organizing related activity, etc. But it also would need
to include consideration for needed technological and configuration
changes, in descending importance:

- namespaces (e.g. for essays, proposals, public outreach resources,
historical content)
- template and JS setup to support multiple languages well (e.g.
mirroring some of the enhancements made to Commons)
- access controls (e.g. for HTML pages)
- FlaggedRevs/Pending Changes (e.g. for official WMF or chapter information)
- LiquidThreads (e.g. for a movement-wide forum that could
increasingly subsume listservs)
- Semantic MediaWiki/Semantic Forms (e.g. for event calendars)

To simplify security considerations, we might want to have all
fundraising-related content elsewhere (e.g. donate.wikimedia.org).

An alternative strategy, of course, is to focus on making the
distinction between different wikis as irrelevant as possible by
vastly improving cross-wiki tools, but the former approach seems more
viable in the not too distant future. I don't think "just move it all
to Meta" is the correct answer.

-- 
Erik Möller
Deputy Director, Wikimedia Foundation

Support Free Knowledge: http://wikimediafoundation.org/wiki/Donate