Jump to content


From Meta, a Wikimedia project coordination wiki

Wikipedia, like any large organization, has a complicated organically-grown set of interlocking, explicit and implicit structures that guide its growth and self-awareness. this page is a long way from maturation.

New pages of note: Organization chart, Corporate sponsorship (by Raul654), and its kin...

Internal Links[edit]

These are pretty random right now; I haven't read most of what's on meta, and some of these (like Liars) are just rants with one or two kernels of useful commentary. Feel free to edit this page, if you come across it, and add your favorite discourses/policy pages/rants from the Wikipediaverse.

+ Wikipedia and corporate structure - How the user net and WikiMedia will work together, preserve principles, make contingencies. Stale since Larry left.
+ Wikimedia - descriptive. vaguely takes over where the above left off.
+ More heat than light - disc of present structure for editorial, social, and software decisions; proposed alternative.
+ Is Wikipedia an experiment in anarchy - ancient LS essay on preserving vandalism.
+ What to do with www.wikipedia.org - Growth of official redirection/main page policy [??-12/2003]; stale.

+ Source of conflict - theories of sources and resolutions of conflict shared by many active members
+ Liars - One user's perception of power trouble  

+ Wikimedia budget (Hardware 6/2004, ...) and projections
+ Wikimedia servers ( Current: all squid hits + ganglia reports; Dated: Hardware order 1/04 + Upgrade discussion 4/04 + Hardware order 5/04, new network proposals as of 06/04; ... )



1. Formal opinion pages (note how effectively the pre-WP Wiki philosophy is shared by the majority of super-active contributors, who are also active on other wikis. It seems to have been a big deal for people to start dating their comments; other change from tradition meets resistance)
2. Formal structures. Committees (arb'n, med'n), Board (of Trustees), restricted user privs (steward/admin), restricted orgal privs (owner/developer).
2a. See WM Budget, Local chapters, ...
3. Informal key channels. Community pages (maintenance, VfD/ViP/RfC/QP, Ref/VP/RfA...); high-vis pages (in conflicts; Main/CP sections); User Talk: pages; timely policy pages.


Growth is generated organically, helped along by software improvements (related changes, ^rv powers, proxy blocking), division of labor (project pages, audit pages & squads), ... here are some key orgs for handling/promoting growth.

+ Volunteer Fire Dept (rarely used today)
+ Volunteer Squads (RC, New Pages)
+ Welcoming Committee (Angela et al)
+ Embassies (little developed)
+ Stewardship (none elected since initial burst)


Everything has gaps and peaks in its self-awareness. For instance, WP is very aware of its vandals, while quite unaware of many of its greatest contributors [there are many ways to flag a noticed edit as vandalous, but no ways to flag it as superb]. Other peaks and gaps:

+ Potential legal and privacy issues. (A carryover from the personal societal concerns of its most vocal editor subcommunity. F: encourage high-interest people to form a writing group, lay out the key problems and branched solutions; teach them to write concisely. Extract these discussions from other fora.)
+ Inclusivity, "free-for-all" consensus. (Active vandals attract attention and counter-rants from so many, they can consume 100 Wikipedian-hrs of effort for every hour they devote to vandalism. F: have small rotating group that handles aggressive, emotional, or otherwise disruptive contributors, and writes publicly about why; limit the duration of their remedies. extract discussions from other fora.)

+ Lack of multilingualism. Meta is 95% english. (and barely used! F: WM should work hard to attract some language experts; have a high-vis page of bi- and tri-lingual WMans; have a special squad of translators in a publicity dept for disseminating announcements/policy changes; develop/buy/acquire software to speed translation; have press contacts fluent in 10 languages)
+ Lack of connection to users; lack of deep stats. "how many users haven't read or edited the site in a month (estimate)?" "how many of those can we reach to tell them about recent improvements?" "what pages do users tend to leave from?" "what are common clickthrough paths of new visitors?"
+ No distinction b/t random, minor edits and focused, deep, or requested ones. Imagine tracking the filling of requests (from WikiMoney and 'most requested' pages) over time; letting long-time editors note occasional awesome contributions they run across (just add a 3-col 'nice work!' table of art_id, editor_uid, reader_uid); ...


Connection b/t power structure and external bureaucracy[edit]

  • corporate entity for legal, financial, and operational purposes.
  • concise decision-making entity for large-scale, far-out planning.
  • concise contact for official press interviews; titled positions for comfort of nonwiki world with this broadly-controlled organization.
  • social empowerment for acting boldly in WM's name and outside text-editing (hard for many in this group of editors)

These are roles of the most visible/offical power structure (the board, named positions related to wikimedia, official orgs/Vereins/etc under various external jurisdictions), and there are separate entities/bodies/pages on Meta for each role.

Connection b/t power structure and scaling / quality improvement[edit]

These connections are realized via a parallel, internal power structure -- stewards, admins, developers of mediawiki, supporters of the machine rooms. The most active people in these structures people spend much of their time actively monitoring and safeguarding the daily working of WM projects.

Low-impact power structures to empower all users[edit]

By distinguishing different levels of devotion and reliability in users, a weak friendly heirarchy supporting scaled quality control and social structure can be encouraged, allowing for this thousand-person group to grow to a million active participants.

  1. reliable devoted users,
  2. reliable sporadic users,
  3. unreliable devoted users (emotional, trouble with NPOV, contentious),
  4. unreliable sporadic users
  5. casual trolls
  6. casual vandals
  7. active trolls
  8. active vandals

The first two should be recognized (via adminship/other) as soon as possible, to distinguish them quickly in heated situations. The last two should also be recognized & dealt with as soon as possible, to limit the amount of community effort expended on interactions with them.

Ways traditional Wiki is broken[edit]

  1. Key namespace chunks taken up with stale content, rather than overviews of related content
  2. Long discussion --> comprehensible summary/overview/archives : no canonical method for it
  3. Stale content: no canonical way to mark it so. Disowning authorship leads to disowning responsibility for pages; "shared authorship" happens automatically, but "shared responsibility" does not.

Possible solutions[edit]

  1. Reclaim namespace (Translation, article count reform, etc.). Archive actively, overview actively.
  2. Replace massive stale discussions with brief-summary (users & ideas involved) and archive-link.