From Wikimania 2011 • Haifa, Israel

I used to mostly live at English Wikipedia. Lately, I've gone meta.


  • Study of WM editor trends across WPs to construct a measure of inter-language and inter-project connectedness. These efforts allow us to identify vulnerable language communities that are more 'linguistically isolated' from the rest of the global community.
  • WM's "Sapien Switchboard"-- the ever-growing network of human translators in WM, how that network can be mapped, made more efficient, and understood in terms of latency and bandwidth.
  • Machine translation, and its reliability for different types of english text.
  • Understanding whatever trends may be found in election2011 data
  • Preparing for Election August 2011
  • Preparing for Election2012/3

Summer of Manifesto 2011

Where I write a host of verbose posts waxing about my dreams for the future
***DRAFT VERSION***, check back later for the real thing, don't rely on this one, opinions are still being formed.

Virtually none of these ideas are 'mine'. I didn't create them, I just read about them somewhere and summarized them in a way that made sense to me. Ideally, I would cite all the influences, but human memory doesn't work that way. My apologies to all the people who actually generated and shared these brilliant ideas. Readers welcome to help with this-- if you've seen an idea before, shout out where you saw it.

Philosophical Overview: glasnost and perestroika, openness and reform

In the Summer of 2011, the Wikimedia movement is beginning to be ready for another burst of explosive growth. As part of our new openness, we stand poised to reach out to the vast multitudes of people still unconnected to our projects. Our movement, its projects, and its communities are on-track to become very open and very diverse-- demographically, linguistically, culturally, and geographically.

Some of the first steps to this new openness are already in place. New offices in the Global South will help us expand to new populations, while usability projects are working to reach those excluded by our existing technology. In connection with that, we recognize that new populations of new users are going to be coming and we, as a foundation and as a movement, will have to rethink and reform our policies and ourselves in order to welcome these new populations and embrace them into the existing movement.

Mere calls for courtesy and openness are ineffective-- openness and reform must go hand-in-hand. We need to make sure that new users get those same kinds of freedoms that we had-- the latitude to form their own projects, customize their own software, and crafting their own policies, taking us in exciting new directions we can't even imagine now.

Going forward, I see some of our focuses being:

  • Empowering and enabling the creation of "New WM Projects", whether they are WMF-hosted or not.
  • Reaching out to existing projects to form partnerships and "welcome them into the movement" in meaningful ways, even if they aren't WMF-hosted.
  • Improving interoperability across all projects, including those not hosted by WMF, making information sharing easier, while making our communities more connected.


Reduce barriers between WM projects
  • Unified template space. Just as Commons is our unified media file host, we need to designate a template host. This could be Commons, Meta, or its own project. All WMF projects need to be able to pull Template files from it, just as we can currently pull image files from Commons. Transwiki extension ScaryTranscluding iwtransclusion
  • Unified user space. , I speak one language but have SIX different user pages to customize, six different talk pages to check up on. (and that's just counting WMF wikis). For our massively multilingual community, the problem's substantially worse. Let's just designate a single project's userspace to be "The" user-space-- Meta or Commons are the logical choices, but we could allow the global userspace to be its own project should Meta and Commons object to the designation. With time, we'd develop the ability for all projects to, if they want, 'pull' user space files from this global user space, just as we do with images from Commons.
  • Unified user_talk space. If I work on six different projects, I don't want to monitor six different talk pages. I may want one talk page that will alert me no matter which project I'm viewing.

Reduce barriers between WM language communities
  • Language Census Identify every single active editor's language skills. Issue a "call to action" for those with special skills.
  • Nurture Multilingual Communities Nurture important bilingual communities, devote special attention to them
  • Organize Communication Foster translation and cross-language communication, even for the mundane.
  • Who's Online? Search for users recently active who have a needed language skill set. + chat
  • Request for Global Comment. Implement a process where an editor can request global input from users across languages and project-types.
Reduce barriers between Mediawiki communities
  • Decimate the barriers to new project creation.
  • Shared user accounts across mediawiki installs.
  • Shared templating across mediawiki installs.
  • Distributed Projects-- The Next Big Thing.
Hosting ≠ Endorsement ≠ Financial Support
  • Don't conflate hosting, official affiliation, branding, endorsement, and financial subsidization. Those are each completely separate issues.
  • Allow official projects to be hosted by third parties if they really want.
  • More importantly, allow third-party-hosted sites to become 'official' projects even if they want to keep their third-party hosting.
  • Some projects have a mission that is "not inconsistent" with ours, but their mission may not be "a priority" for our movement. Let those projects join us, provided they can do so at a net plus to our resources and mission.
  • By the same token, we should at least consider subsidizing any future projects that would be a "mission priority" for us, even if they aren't hosted with us.
  • We need a movement name and logo anyone can use without any prior approval from anyone. Sympathizers and WM-style projects could us it, but it would be more like a flag and a rallying cry than an organization. Anyone could use these marks, even people we don't like.
  • We need a name that means "informally allied"--- meaning someone at some unofficial level has decided the mission and values of a project look 'not inconsistent' with our movement.
  • We need a name that means "Officially allied"-- meaning that the foundation itself has signed off on the alliance.
  • At the highest level of alliance, we would have "Official WMF Project"-- namely, those that have always been with the WMF, have a long time relationship, and thus are are projected to be viable and tied to us for the foreseeable future.
  • "Reclaim" the many projects that logically belong in our movement but, for historical reasons, were created apart from the foundation. Encourage these presently-unaffiliated projects to ally with the WM movement. List of Potential Wikimedia Movement Partners
  • Consider Re-naming WMF as part of brand expansion.
New projects + Be Bold
  • Restart new projects process ASAP
  • Let interested parties create 'specialist wikis' so they can optimize their mediawiki install/culture for their subject matter's needs.
  • Let interested parties create 'style wikis' that have a similar mission to existing projects but a slightly different style. Simple English, for example.
  • Experiment with "signed content"-- content that is not owned, but is generally controlled by a user.
  • Accept projects will fail, just as edits are reverted and articles are deleted. This is okay.
Socializing is educational and productive
  • More social features are good.
  • Make a global semi-social space
  • Nurture bilingual communities
  • Build a global community.
  • Make elections a priority. Hold regular yearly elections, even when there are no seats to fill.
Scope and hosting
  • All information is educational, but different mediums do vary greatly in their 'educational content per dollar of hosting/bandwidth'.
  • Choose hosting based on pragmatic concerns-- is it educational enough to merit the cost we pay to have it?
  • Embrace Art. Non-explicit forms of learning, like Art and Music, can be just as 'educational' as explicit forms of learning, like Math or Science. Artistic works and their study are, indeed, educational, not recreational-- it's just hard to describe the 'truths' being learned in Art, and it's harder for geeks like us to understand Art.
  • Embrace Oral History. Among industrialized nations, a recently deceased person was most likely born sometime in the 1920s. Their personal recollections of the events of the 20th century are an invaluable, even 'priceless', source of knowledge. At a minimum, these contributions should be "Very Highly Welcomed" at Wikimedia. Think bigger, a huge "Record your Grandparents for Wikimedia" campaign could bring us global goodwill and invaluable information that might otherwise be lost forever.
  • Embrace Data and Data Analysis. Large datasets are, objectively, the most informative content we could host. In the past, large datasets weren't a priority simply because of the work required to 'translate' mathematical insights into human languages that are understandable by a lay audience. Opening Science is a mission priority, and scientists would love a large, open, non-commercial entity to host and analyze large datasets.

  • When social innovation fails, try technical innovation.
  • Transwiki transclusion needed
  • "Average Quality" isn't a good measure our projects' quality. Instead, use "Total high quality content* or "Time-weighted quality".
  • Using "Average quality" implies that it would be possible to lower our quality by adding to our quantity. This is not true. Properly run, some projects like Wikisource will always get better with more in-scope content.
  • For some projects, we need technological innovation to improve quality.
  • Call Referendums.
  • Show us the debates.
Increase Participation / Openness
  • We need a no-judgment space in which new users can control how much they want to collaborate. New users could write in the first person, they could user poor grammar, they could have an 'authorial voice'.
  • Deletion demoralizes. Except in cases of legal issues, don't delete, just tag as officialy deleted.
  • Let people 'come as they are', and let us figure out how to incorporate their contributions into the core projects.
  • Eternal Editor Survey Let our editors tell us as much about themselves as possible, provided they do so in good faith. Automatically quantify and analyze this data in near-real time. Incorporate public domain psychometric measures into the survey.
Wikify development
  • Create a strong "call to action" for readers who are developers.
  • Lower barriers to participating in development, advertise the need for developers.
  • Safely and responsibly allow developer autonomy to the extent possible. Make it easy to go from idea to new feature on a WM project.

The Site Architecture You Can Edit Newbies Incubator Opening up Wikipedia's Data: A lightweight approach to Wikipedia as a platform

The "No Judgment in Brainstorming" section
  • Bilingual wikis. We have a lot of people who speak, e.g. en and de. Why not let them write in both for the audiences that understand both. Perhaps using en-style syntax but with the de vocabulary added to the mix. This is possible for all language pairs with large, diverse communities.
  • Long Now-- an 'artistic' copy of a wikimedia data dump designed to survive a global catastrophe, last for 10,000 years, etc. Symbolic, reminding people we're something very special. cf. Voyagers dics, Pioneer1 disc, Areicibo signal, Nevada nuclear storage facility art
  • Make a "TED"-like conference for our movement that's more than just a WMF-focused Wikimania.


Motivating "crisis-tunities"
  • The global community engagement problem -- there's little investment in the foundation beyond its role to host a project.
  • The Mediawiki problem-- we use MW even when we shouldn't.
  • The quality problem-- on quality-based projects like WP, some articles aren't good and don't seem to be getting better.
  • The deletion problem -- deletion of legal, good-faith content is inherently demoralizing.
  • The conflict problem -- edit wars are inherently demoralizing
  • The geographic diversity problem -- our editor population does not reflect the geographic diversity of the "internet-connected world" to the extent desirable.
  • The demographic diversity problem -- our editor population does not reflect the demographic diversity of "internet connected world" to the extent desirable.
  • The content diversity problem-- a few reference books isn't "all the world's knowledge"
  • The new project problem-- the new project process has stalled.
  • The technological innovation problem-- considering our overwhelming appeal to computer-savvy populations, our technological evolution has been relatively slow.
  • The 'hard line' question-- what kinds of projects and content are never ever appropriate to be Foundation-hosted?
  • The language problem-- our global community must learn to intelligently form consensus without a single common language
  • The catastrophic collapse problem-- infrastructure instability could still lead to severe disruption under some scenarios
Personal Beliefs , Values
  • Communication leads to world peace.
  • All information is educational to some extent.
    • At some point soon, we will have the power to host all available human-generated text created prior to 1900, at virtually no cost to us.
    • This moment will happen much sooner than we imagine. This moment may have already occurred.
    • This knowledge about our near-future technological powers should guide our vision. An encyclopedia is just the tip of the iceburg.
  • Art is highly educational to its devotees. The distinction between "High" art and "Low" art has traditionally been defined by social class rather than artistic merit. All communication mediums possess capacity for Art.


  • alter Extension:Interwiki so it can use site-logos or images or other graphic cues to indicate source

Proposed Roadmap to becoming a WM project

for existing projects not already hosted with WMF
  1. Self-identified membership. Display marks on your project's site that tells your readers "We consider this project to be part of the Wikimedia movement". Use provided standardized text , logos, or marks on your pages once they're developed.
  2. Informal membership. The global Wikimedia community can affirm that a project is, in its opinion, part of the Wikimedia movement. Some discussion and consensus process could approve the alliance, vetoable by board vote.
  3. Formal membership. The Foundation can affirm that a project is, in its opinion, part of the Wikimedia movement.