Jump to content

User:John Vandenberg/WMF BoT candidature notes

From Meta, a Wikimedia project coordination wiki

This page contains my notes on issues currently facing the Wikimedia Foundation (WMF) and the movement. It does not highlight all the marvellous things the WMF has been doing, and the cool projects they are working on. It must be said that the WMF is a good organisation, with great people in it. I think there is room for improvement, and I believe I can ensure those improvements occur, but it is light years ahead of where it was five years ago, and its problems are largely due to its success.

The WMF has been entrusted with holding the assets created by the community; however the community maintains an active interest in these assets. The WMF should consult with the community on how those assets are used. When it was a smaller organisation, there was rarely a time when the volunteers thought the organisation was doing something other than what they expected of it: the organisation struggled to fulfil the 'core' tasks that were expected of it, and many core functions of the organisation were delegated to volunteers.

However, .. times have changed. Money changes everything. The WMF now has sufficient money that it defines 'core' to include projects that large cohorts of the community consider to be undesirable, let alone necessary. And the WMF is taking over parts of the movement that were previously run by volunteers, either intentionally or by disempowering or disgusting them. Undeniably this has resulted in professionalisation of some operations, where the WMF has properly resourced or managed the 'take-over', but at times the WMF has poorly planned or communicated the change of ownership. The community is losing confidence in the WMF as the custodian.

What is worse is that there are many instances of the WMF eroding our values, albeit in pursuit of the 'mission' via questionable means. It is the WMF boards responsibility to prevent this from happening, or take responsibility and rectify the problem when it happens, but unfortunately the board has not done either. Our community is full of idealistic people who have donated large slabs of their time to the mission; many have donated so much time that 'person-years' is the appropriate unit, with no money or fringe benefits in return. The board of trustees needs to ensure those volunteers are engaged and support the activities of the WMF – the staff time is paid for, effectively, by the free labour of the volunteers.

The board of trustees needs to initiate and oversee the next iteration of strategic planning, and they need to ensure the next set of strategic goals are reflective of an adequately engaged community. The WMF can't achieve any useful strategic goals without broad consensus from the editing community.

MediaWiki Foundation


Why should we have a decision about setting up a MediaWiki Foundation? The suggestion of setting up a MediaWiki Foundation has been raised many times over the years, without any resolution. I think now is the time to have a formal- and board-engaged discussion about it, to bring those discussions to a conclusion.

Board of Trustees


As part of preparation for the next strategic planning exercise, the constitution of the board of trustees in the bylaws needs to be revised to increase the presence of Wikimedia community members.

The Chapter-selected board seats have caused community-vs-chapter disputes that are harmful ("Are chapters part of the community?", so I believe they should be discontinued.

The entire board of trustees should be community elected so that the final endorsement of the strategy is made by a board of trustees that the community feels are champions of our values.

I would advocate for the Board of Trustees that is comprised along the lines of a 4+2:4 split between community and non-community seats:

  • four (4) Wikimedians with a two year term and an option to be re-elected for only one additional term,
  • four (4) non-Wikimedians people who have served on a board of trustees from a similarly sized non-profit (other than the Wikimedia Foundation or affiliates) with a two year term and an option to be re-elected for only one additional term, and
  • two (2) seats reserved for people who have been on the WMF board of trustees for two full terms (i.e. four years in total) and would like to continue to serve our movement in this capacity. There would be no limit to how many terms a person could be re-elected into these two seats, thereby allowing Jimmy to continue to serve on the board as long as he remains the community choice of representative.

Candidates for the four 'non-Wikimedians' seats' may have contributed to Wikimedia previously, but it would be their experience outside the movement that would be the criteria by which they were made eligible. Staff would be assigned to all of eligible candidates, providing neutral assistance as requested (e.g. uploading photos and posting their answers to questions, etc). The process would ensure that the winning candidates are up for the challenge of working closely with our community in a collaborative manner, and the community has chosen the candidates they think are most appropriate.

The only people who are eligible for the two additional community seats are people who have served on the board for four years. According to Template:BoardChart is limited to: User:Jimbo Wales, w:User:Michael Davis (User:Bookofjude?), User:Mindspillage/Kat Walsh, User:Jan-Bart, User:Stu, User:Anthere, User:Sj, User:Wing and User:Bishdatta when her term expires at the end of 2015. They are all Wikimedians, and the community would choose which two they want to re-elect. Of those, I am not sure about Michael Davis, and Bishdatta is the only other people who wasn't a Wikimedia contributor before joining the board, but I think she has earnt her right to be considered part of the Wikimedia community, and to stand as a candidate to be re-elected for a third term if the community believes she has done a good job. That list would grow over time. Jimmy would likely continue to be elected to the board by the community, but the proposed structure allows other experienced Board of Trustee members to fill his shoes when/if required.

Freedom redefined


The only non-free logos allowed on Wikimedia Commons are those owned by the Wikimedia Foundation. This use of copyright as an extra protection against trademark infringement may be text-book legal opinion, but Wikimedia is not trying to be a text-book corporation. Our values demand that we use the minimum protections possible to achieve our mission. It can be argued that we needed to take advantage of copyright law in the early years, but that time has surely passed. As the six most viewed website in the world, trademark law is now sufficient. The Wikimedia community regularly investigates the copyright status of logos of other organisations, and determines that they are not eligible for copyright, or the copyright has expired. e.g. commons:Category:Logos of Coca-Cola. Those companies need to rely on trademark law, which is the correct legal protection for trademarks. Wikimedia should play by the same rules. Wikimedia should not rely on copyright law, when our community is one of the largest groups of people declaring that copyright is not available to logos of other corporations. Last time I raised this on the wikimedia-l mailing list[1], my esteemed colleague SJ agreed.[2] It is possible (but not assured) that this stance may increase WMF legal costs, but at least then our lawyers will then be fighting on ethical ground rather than undermining our shared values of freedom.

The copyrights and trademarks of the Wikimedia logos are being used against Wikimedia affiliates and Wikimedia volunteers. I know of examples in many countries, but I will give some examples from Australia with which I am most familiar. In Australia we were told by the WMF that a QR plaque needed to be removed from within a church. This QR code was the excellent work of a Wikimedia volunteer operating on their own (i.e. no chapter involvement). I needed to beg with WMF for them to back down from this disgraceful request, which was essentially a kindly worded Cease and desist letter directed at a volunteer (and one of the women, I should add). Thankfully the WMF did back down in that instance, but the culture in the WMF of owning the 'assets' and using them against the community should never have been allowed to develop. After Gibraltarpedia, the WMF was using their ownership of assets to prevent other QR code projects from proceeding. I also had to intercede for the w:WP:GLAM/Freopedia project to not be terminated – again this was a volunteer led project, with Wikimedia Australia providing support. We could appreciate that projects like Freopedia may need to be put on hold for a while, but we wanted a public statement from the WMF to that effect. On several occasions I asked the WMF to publish their decision that projects like Gibraltarpedia should be put on hold, but they never did. The State Library of Queensland, a long standing partner of Wikimedia Australia, has been holding Wikipedia training sessions for their staff and their affiliates for two years – over a hundred of their staff have been trained. They include Wikimedia in their published strategic plan. They are part of our community. And yet, the State Library of Queensland does not feel comfortable using the logos except for joint projects with Wikimedia Australia. They should be able to use Wikimedia logos without a contract with an organisation in the United States, and the WMF prevents Wikimedia Australia from acting as a local representative of the movement. This behaviour from the WMF prevents adoption from GLAM and other government bodies in Australia, and elsewhere in the world. And the WMF staff are 'busy' creating these problems for volunteers and affiliates.

Many chapters, affiliates and volunteer initiatives have resorted to creating their own logos and brands because they believe in 'free' brands, and want brands they can use freely. e.g. printing shirts to give to their staff and volunteers, or that they can translate and sell. To this end, often the volunteers have created logos that are based on the 'Community Logo'.

Last year, the WMF trademarked the "Community Logo", which was released into the public domain by volunteer user:WarX (Artur Jan Fijałkowski), who contributing consistently to Wikimedia since 2005, especially creating logos for our community. The community explicitly selected this logo because they wanted a logo the community could use without fear of trademarks or copyright. WarX was kind enough to create a logo without copyright; however the WMF has now trademarked that against the express wishes of the meta community. It is scary that nobody in the WMF was aware that this logo had a special history, and it was not theirs to do with as they wished. WarX and others have created many logos derived from the Community Logo, assuming that it was "free". The Polish community has used these logos systematically. WMF didn't even bother to inform WarX when they trademarked his creation.[3] When the trademark of the community logo was pointed out to WMF staff and board members, the people who raised the issue were attacked and the issue was ignored. The only board member to show signs of caring was SJ, who I have great respect for, but unfortunately SJ has been ineffectual in resolving this trademark matter to the satisfaction of WarX and the community. Trademarking someone elses creation without informing them is immoral, even if they put their work into the public domain. The Wikimedia movement is one of the most vocal group of people standing up for the protection of the public domain. Creative Commons is the leader in protecting the Commons from exploitation of creative works by rogue organisations. WMF board members Jimmy Wales and Kat Walsh are board members of Creative Commons and the legal counsel for Creative Commons, respectively. They should view this to be a problem for which they must champion a resolution – apparently they don't, as they have been absent.

The WMF needs to withdraw its trademark of the Community Logo, and work with the community to create a trademark policy for the remainder of the trademarks that balances the rights of the community to use them against the real problems of the trademarks of being misused.

The Creative Commons community rallies behind other instances of trademark claims on public domain artwork, such as this trademark claim regarding File:Sharethis.svg.

Right to fork and long term planning


The 'Right to fork' has been a fundamental value since the beginning of Wikimedia. The WMF has done an excellent job of providing dumps; however the ability to easily fork is still elusive and does not appear to be a priority. The GFDL requires that it is possible to rebuild Wikipedia easily, using commonly available tools. I maintain that the WMF is not compliant with the GFDL as a result[4], but appreciate that some disagree[5] and overall nobody really cares anyway. (see also here) That said, most people agree that it is not possible to easily fork. For example, images are not provided in dumps, and they are not mirrored across the globe. This has resulted in images being lost forever when there was a software bug that destroyed old images. Wikix does include a method of obtaining the images. It is great to see that the recent funding round included funding for Wikix. However it, or similar software, should be elevated to become a core component of our infrastructure, and improved so that the spirit of the GFDL is observed. We should also perform a test 'fork' regularly to ensure it is possible and also reliable.

The most important challenge to the independence of Wikimedia is the U.S. jurisdiction. Wikimedia is currently inextricably tied to the United States jurisdiction, both in terms of the organisation and the content policies, especially copyright. While SOPA and PIPA did not pass into law, they should be a wake up call. If it was worth protesting against, we must also recognise the need to prepare against our biggest threat. Wikimedia needs to start contingency planning for the scenario of the US jurisdiction becoming undesirable for hosting of the Wikimedia projects. We need to identify potential jurisdictions that the servers could move to, understand the effect such a move would have on our content (e.g. which images need to be deleted?), and establish a nucleus operation in at least one of those potential jurisdictions to allow the migration of the projects to that jurisdiction to occur with relative ease and order. With this in place, our response to the next SOPA/PIPA will include the real possibility that Wikimedia will move if the introduced laws present a real danger to the Wikimedia volunteers or the site operations. This will cost money, but it will be in line with our values, as opposed to paying for lobbyists – which the WMF is now doing.

To guarantee long term independence, I believe the WMF needs to establish an Endowment. My preferred model for an endowment is one in which donors are explicitly donating to a narrowly defined endowment that ensures our basic infrastructure costs are covered, using professionally developed and well documented costing models. The donors should be unambiguously informed that they are donating to the very specific purpose of assuring the core technical and support operations, and the funds should only be used for that purpose. A more expansive endowment can be established later, when the basic operations have been assured. Determining our future costs to establish the desired endowment target involves long term planning, which is long overdue. Our GLAM partners want to know that Wikimedia has plans in place that ensure the preservation and availability of our content, and they mean 100 years rather than a few years. More urgently, the WMF needs a winding down plan that outlines how the WMF will ensure that the content and community will survive in the event that the WMF cannot continue to operate (may that day never come).

Supporting the ability to fork and establishment of an endowment should be elevated to Board of Trustees resolutions to ensure long term preservation and independence of the movement, the content and the community.

Regarding core values, two other programs that need to be brought into alignment are PediaPress and Wikipedia Zero. Our core values state that "[WMF does] not enter into exclusive partnerships, with regards to access to our content or use of our trademarks." However both of these are in practise exclusive partnerships (e.g. non-overlapping coverage). The WMF needs to convert both of them into open systems, document the process of setting up new partnerships into each program, and allow any provider to sign up on meta to become a partner. If these programs don't scale, the engineers need to document the roadblocks to scaling, and work needs to commence to eliminate those roadblocks.

Nurturing affiliates


I believe that a nation-based model is essential to the growth of Wikimedia, as offline activity and especially face-to-face socialisation mostly occurs within national borders. That said, the current model for chapters and other affiliates is far from optimal.

Rather than building a strong ecosystem of self-sustaining affiliates, the WMF has repeatedly pulled the carpet from underneath chapters, changing the model they operate within and then blaming them for not being stable enough to receive any funding.

Once an affiliate has been approved by the Affiliations Committee, the community should develop a "core" operational budget for it. A mature international organisation ensures its chapters either a) keep money raised within the country, or b) are provided with predictable levels of funding to support their core operations. The board has resolved that option (a) is only allowed for four countries, which means we need to employ option (b) for all others.

These "core" budgets will take a while to develop, as each country is different, and they will need to be revised periodically. As a global community we need to be asking "What tasks should be done routinely to support and build the Chinese community?", then "How many people need to be in an office doing those tasks that support the Chinese community?" and "Which affiliate can manage those staff effectively?" If there is no affiliate that can effectively manage staff, the appropriate strategic response is plan to support an affiliate until it is ready to effectively manage staff. The WMF has failed to do this in India and Brazil and the Arabic countries it has tried to penetrate, where the WMF has invested millions of dollars.

Baidu is one of our only 'competitors', and probably the most successful. They are extending their coverage to South East Asia, making it a real threat to our strategic goal of increasing readership and editors. They have serious money to throw around, and they are meeting with Wikimedians in S.E.Asia. If WMF continues to effectively ignore S.E.Asia, and forge ahead with a U.S.-controlled 'international' organisation that doesn't respect Asian culture, Baidu will take over large sections of this 'market'. The WMF fedex-ing t-shirts from the U.S. to China is just laughable – and a waste of donors money. The WMF needs to nurture the chapters that exist, build them up to be independent, and support the establishment of more affiliates. The current strategy has an abysmal record. It should be a snap decision to agree that it is strategically important to have at least three personnel supporting the Chinese language community, where we can see enormous growth in readership and participation. The only question is whether to have one staff member in each of the locations with a chapter (Hong Kong, Macau and Taiwan), or three staff in one of those countries, or any number of other arrangements. We should be able to have a mature and open discussion about this, with those chapters participating to help us find the optimal deployment.

As part of nurturing new affiliates, I recommend that the Wikimedia community establishes 'boilerplate' programs which are essentially drop-in solutions for new chapters, balancing the needs of the launching affiliate and the needs of the WMF and community in protecting the 'brand'. For example, after many years involved in chapters I think the first staff member of a chapter should be selected by a community election (e.g. on meta) rather than being a decision of trustees or of a small group of chapter members. The WMF can provide support in finding suitable candidates if necessary (especially for expert positions), and the candidates should also be identified to the WMF as is required for stewards. Most affiliates would jump at the chance of having their first employee essentially provided for them, even if it means that the trustees and members need to accept whoever the community selects, for good or ill. The core workload of the first employee should also be essentially provided by the boilerplate program, including sourcing merchandise providers and sending it to volunteers, manning stalls at trade conferences, assisting with organising local workshops, organising responses to press enquiries, managing community engagement including assisting with translations for important notices, facilitating cross-language communication, and the important task of reporting local activities back to the Wikimedia Foundation and Wikimedia community.

When the affiliate has a minimum number of members (e.g. 50?), and has a good track record of program activity, the organisation would be free to select its own permanent staff without input from the broader community.

With 'core' funding being collaboratively assured for all movement entities, the Funds Dissemination Committee would continue in its defined role of allocating funds for optional programmatic activities that target our strategic goals.



The greatest hurdle we face is quality. I think it is undeniable that the 'world' will forgive us for not having met the target of "increasing the number of total editors who make at least 5 edits per month to 200,000", and will even forgive us for not increasing the total editors at all, but they will not forgive us for having low quality content.

In 2011 the strategic target for quality in 2015 was "increasing the percentage of material reviewed to be of high or very high quality by 25 percent".

If I can briefly focus on English Wikipedia, the only stats I know of are the Good Article statistics which indicate that English Wikipedia appears to have already achieved this target, with the percentage of 'Good Articles' going from 0.303% to 0.412% in between 2011 and 2013 ('Featured Articles' has moved from 0.0893% to 0.0915%, but the Good Article percentage improvement absorbs the Featured Article target percentage improvement yet to be reached). However in real terms, this is a losing battle. On English Wikipedia, there are more articles being created each day (850) than are reviewed via the good article process each day (8.5).

Looking at German Wikipedia statistics, the percentage of good articles has decreased from 2.72% to 2.28% between 2011 and 2013. The Ukrainian statistics appear to tell a similar story through Google Translate. If there are statistics of Wikipedia in other languages, they are not linked to the English, German and Ukrainian statistics.

Out of curiosity I have briefly looked at the Wikisource projects, where the use of mw:Extension:Proofread Page was introduced by the community to provide objective quality. The Wikisource proofreading statistics indicate that it is now at 70% proofread quality of the "Page" namespace; however that only represents 20% of the mainspace content. As far as I know, Wikisource does not have a historical record of these quality metrics available online, but my guess is that there has been a 25% increase of quality on many of the Wikisource projects.

The WMF needs to start regular measurement and reporting of quality statistics of projects, large and small, including sister projects.

I also believe that we need to set a new target that is more ambitious for English Wikipedia, and the WMF needs to pull together content reviewers on each of the larger Wikipedias to determine what needs to be done to support them. This should involve teleconferences and symposiums. WMF puts a lot of funds into the software, flying our devs to be together for hacking events. The same needs to be done for content reviewers.

The second component of quality that the Wikimedia Foundation needs to prioritise is biographies of living people. The Qworty incident on the English Wikipedia should be a wake up call to the Board of Trustees. Having been an arbitrator on English Wikipedia for two years, and an active member of the community for many years, it is obvious that this is only one of many, who happened to get caught because they fell into the spotlight. There are many members of 'our community' who have 10,000+ edits and are maliciously controlling negative content about other people, and doing it anonymously. There are regular complaints being made on the talk pages, to OTRS, and in social media about these editors. The problem is made worse by our culture of protecting anonymity, resulting in an assumption that 'outing' anyone is a bannable offense, even if they are one of these problem editors. I predict that Wikipedia will increasingly become ridiculed for its current position of allowing anonymous edits to biographies of living people unless we can build better systems of identifying and preventing these problem editors. We need to innovate. We need to warn existing and future problem editors that malicious editing from an anonymous accounts is not safe, and the media is starting to undertake real investigative journalism of Wikipedia editors where they see problems.

The most important step in fixing the cultural problem is to introduce the ability for Wikipedia accounts to be voluntarily linked to identities in other systems, such as twitter, facebook, identica, etc. This could be included in the account creation process, provided that it is optional and the risks are clearly explained. With this in place, new accounts can declare up front that they are not trying to hide their identity, and do not mind their identity and COI being discussed publicly.

The next strategic measure that should be taken is the introduction of a complaint system designed for average Internet users with no wiki editing skills, and that complaint management system needs to support resolution and escalation of complaints (i.e. 'mark resolved', with 'mark unresolved' that ensures the complaint can't be 'resolved' again by the same person who marked it resolved the first time; an escalate function that feeds into a mediation process; rather than overusing the 'block' an account functionality, introduce a limit of one post per month to be used for people who complain about the same problem repeatedly despite the complaint being roundly rejected). It would also be a step forward to allow subjects to post a rebuttal (using a verified identity) that is visible in a separate tab for readers who want to read 'the other side' of the story.

The WMF staff and its board do not need to be part of the escalation process; however they do need to take responsibility for ensuring that it is working effectively. There will be times it does not work; we need to ensure they are not numerous. The board has a responsibility to ensure the community has the best tools possible, and that there are staff monitoring where there is high risk of unwarranted damage to living people.

I would like to see Liquid Threads implemented on the larger Wikipedias, as I believe that functionality would allow better content issue resolution. However the use of Liquid Threads (or similar) would need broad community support before being adopted. I would recommend it being trialed only on Biographies of Living People, and maybe limit its use to the subject of the article and other people who are identified to the WMF.