Jump to content

Wikipedia:Wikipedia Signpost/Single/2011-02-14

Add links
From Wikipedia, the free encyclopedia
The Signpost
Single-page Edition
WP:POST/1
14 February 2011

 

2011-02-14

Foundation report; gender statistics; DMCA takedowns; brief news

Contribute  —  
Share this
By Tilman Bayer

Foundation's report for January: Anniversary impact, Brazil and India travels

The Wikimedia Foundation's monthly report for January has been published, much of it related to the celebrations of Wikipedia's tenth anniversary on January 15. The Communications Department observed that although journalists and other commentators have often been very critical of Wikipedia since its inception, "as the 10th anniversary approached, the international media seized the opportunity to reassess: this resulted in hundreds of stories around the world that were overwhelmingly positive" (an observation that had similarly been made by Sue Gardner and some Wikipedia critics, see Signpost coverage). The report called the shipment of "more than 80 Wikipedia 10 celebration kits" (with T-shirts, buttons and stickers) from the WMF to event organizers worldwide "an important pilot for the Wikimedia movement: new data about customs, logistics, and postal services for a wide range of nations has been gathered, and new processes for soliciting orders from chapters or other groups for timely delivery have been developed." Preparations to set up a Wikimedia merchandising webstore are underway.

The Human Resources Department reported a downside of the celebrations: "A large percentage of the staff in San Francisco was out for at least a week with the 'WikiPlague', a variant of the RSV virus that we seem to have caught at the 10th Anniversary Party." The department also reports that it has "started tracking metrics for new hires and the Wikimedia Foundation as a whole, and will start compiling anonymized data regarding diversity and other internal characteristics so that we stay mission-aligned."

Staff members of the Global Development department spent time in India and Brazil in January, and progress with the "Catalyst Projects" for both countries was reported.

Among the visitors to the Foundation's office in January, the report records representatives of IT firm Trivad, Inc, three consultants from communications firm OMP (a former employer of Chief Community Officer Zack Exley) attending a "Wikipedia brainstorming", the CEO of Paymentwall (a company offering ecommerce solutions) and the CEO of Charity Navigator.

Wikipedia's gender gap examined further

A graph of decreasing bars from females occupying 15% initially to less than 5% on a logarithmic scale.
Very active users more likely to be male? Male/female percentage among English Wikipedia accounts who state a gender in user preferences, sorted by their edit count
A line graph with three lines: Unspecified oscillating at 5000 registrations per day, Male at 260 per day, and Females at 60 per day. Text overlaid in the gap reads: About 5,300 accounts are registered every day. Only 37% make an edit afterwards in any namespace. Males (31%) are less likely to edit afterwards than females (33%).
No noticeable effect of the New York Times article: gender stated in user preferences of newly registered accounts on the English Wikipedia in recent weeks

After the widespread discussions about the small proportion of women among Wikipedia editors had brought increased scrutiny of the study where the widely quoted "13%" estimate had originated (see last week's Signpost coverage), attention turned to evaluating the gender statement that users on Wikimedia projects can provide in the "preferences" for their account. While this information is optional and the majority of accounts do not state it, it can produce some interesting numbers.

On the Wikitech-l mailing list, DaB. provided statistics about some large Wikipedias and Commons from the toolserver database, which were used by Wikipedia researchers Joseph Reagle and Paolo Massa to produce tables.

Dispenser collated API and Toolserver data to do some further analysis, relating the statement of gender to the user's registration date (within the last month), edit count, and whether or not the user gave an email address.

English Wikipedians' gender by edits
Edit count up to Users Percentages
Male Female Unknown Male Female Unknown F/M
0 5531 1191 95588 5% 1% 93% 0.22
1 1199 246 28024 4% 1% 95% 0.21
3 1153 216 23302 5% 1% 94% 0.19
7 1105 164 17695 6% 1% 93% 0.15
15 1210 155 13684 8% 1% 91% 0.13
31 1385 158 11764 10% 1% 88% 0.11
63 1453 135 10433 12% 1% 87% 0.09
127 1377 121 8642 14% 1% 85% 0.09
255 1305 103 7059 15% 1% 83% 0.08
511 1139 82 5363 17% 1% 81% 0.07
1,023 1032 60 4243 19% 1% 80% 0.06
2,047 847 41 3233 21% 1% 78% 0.05
4,095 781 42 2409 24% 1% 75% 0.05
8,191 725 38 1752 29% 2% 70% 0.05
16,383 572 37 1255 31% 2% 67% 0.06
32,767 413 26 788 34% 2% 64% 0.06
65,535 227 14 343 39% 2% 59% 0.06
∞ 104 5 214 32% 2% 66% 0.05

DMCA takedowns of fair use and US-Gov-PD images

The Foundation complied with two more DMCA takedown requests last week, continuing the recently established custom of making copies of them available on its website (cf. previous Signpost coverage).

The first request came from the US Department of Health and Human Services, concerning photos on Commons that apparently had been mistakenly designated as public domain by publishing them on the government's own websites: "Although the images had been posted to the public NCI/NIH Websites in the past, that posting was done in error. ... The photographs are protected by a license agreement and none of the parties involved ... has ever intended for the image to be in the public domain." Last month, the photographer had contacted the NCI, who took down its own copies of the images and notified the Wikimedia Foundation.

Another DMCA takedown request last week resulted in the deletion of a photo that had been illustrating the article about 1960s style icon Talitha Getty. According to the image description page as still available in Google's cache, the image had been uploaded on 15 November 2010, copied from another website with a resolution of 344 × 457 pixels, with a standard non-free content rationale as it is frequently used for portraits of deceased persons, which includes a fair-use claim.

Briefly

  • Spanish chapter: The Wikimedia Foundation has approved the founding of Wikimedia España. In a post on the Foundation's official blog ("Welcome Wikimedia’s 30th Global Chapter, Wikimedia España"), Jorge Sierra (User:Lucien leGrey), the chapter's president, said that its formation had begun in September 2007.
  • "Movement roles" work meeting: The notes for a recent meeting of the Foundation's "Movement Roles" workgroup in Frankfurt have been published.
  • Expert survey: The Wikimedia Foundation's Research Committee has launched a survey "to understand why scientists, academics and other experts do (or do not) contribute to Wikipedia, and whether individual motivation aligns with shared perceptions of Wikipedia within expert communities." It grew out of a discussion started on Wikiversity in November. (Notes about the survey's design)
  • Community fellowships: The Foundation's Community Department has posted a job opening for the "Head of Community Fellowship Program", a new position to "coordinate the recruitment and support of Wikimedia Foundation community department fellows". Introduced last September (Signpost coverage), the program temporarily employs community members to "lead intensive, time-limited projects focused on key areas of risk and opportunity". It was recently suggested "to give the community more of a voice in both proposing and selecting individuals and projects" as the program is being scaled up, and fellowships outside the Community Department are being created (Signpost coverage: "Wikimedia fellowships discussed and clarified"). In other news, according to the Foundation's January report (see above) the fellowship of Maryana Pinchuk (m:User:Buickmackane) "has been extended for an additional 12 months, so that she can coordinate the writing of histories of several more Wikipedias." Together with Victoria Doronina (User:Mstislavl), she had started an eight-week research fellowship around the end of September to develop methods for writing such project histories (Signpost coverage), and had published an (English-language) "attempt of a social history" of the Russian Wikipedia (see last week's "News and notes").
  • GLAM fellowship plans: Liam Wyatt (User:Witty lama) has outlined the priorities of his recently begun 12-month Wikimedia fellowship, whose goal it is to "help create systems and processes to make outreach partnerships [with cultural institutions] more efficient and effective." They include a system of "Wikimedia GLAM ambassadors" (similar to the Foundation's "Campus ambassadors"), improving how-to documents and case studies, technical metrics and tools, and better communication about GLAM partnerships. His posting was published shortly before embarking on a trip to India, where he is currently meeting with various Wikimedians and representatives of cultural institutions, as already reported (about Mumbai) by the Hindustan Times ("Wiki to tie up with city galleries, museums").
  • October Board minutes: The minutes for the October 2010 meeting of the WMF Board of Trustees have been published. Among other topics, they shed some more light on the Board's three-hour discussion about the recommendations of the 2010 Wikimedia Study of Controversial Content, which had been expected to result in a resolution at that meeting (Board member Phoebe Ayers had already described some of the discussion in November, see Signpost coverage: "Controversial content and Wikimedia leadership"). According to the minutes, "an exercise to capture the level of agreement around the appropriateness of each recommendation" showed that "half of them were uncontroversial, while others required further discussion" (it was not stated which recommendations belonged to which half). A working group on the issue was "to make an initial report to the Board by November 6, 2010".
  • Global Development IRC hours: The logs for two IRC office hours by the Foundation's Global Development department have been published, featuring the department's head Barry Newstead and its two new employees, "Head of Global South Relationships" Asaf Bartov, and "Chapter Relations Manager" Moushira Elamrawy. Asked about the plans to set up a Wikimedia office in India, Newstead said, "We will make an announcement on India early next week." About the topic of subnational chapters, he said that "WMF has no issues with subnational chapters in general", but that fundraising might be "trickier at a subnational level". Asaf Bartov expressed "pain" at the "Global South" moniker that has been used to describe geographic areas with respect to the WMF's development, calling it an "infelicitous term. Please consider it temporary; we're hoping to involve the community in picking a better term." He said that he would "mostly be dealing with non-chapters."
  • Wikimedia Hungary report: The Hungarian Wikimedia chapter published their (English-language) report for January 2011 last week. It records the chapter's fundraising results (HUF 5,7 million or about US$28,600), describes celebrations of Wikipedia's tenth anniversary, including a conference that featured a keynote by the Foundation's Head of Public Outreach Frank Schulenburg (available on Youtube), noted that the Hungarian Wikipedia had to deal with "a copyright violation case where one user has uploaded 5000+ articles from copyrighted encyclopedias over 5 years" but that the resulting media coverage had been mostly positive, and mentions the chapter's plans to set up its own toolserver this month.
  • Wikimedia Netherlands report: The Dutch Wikimedia chapter published their report for January 2011 this Monday.
  • UK university recruitment tour: Six members of the Contribution Team held a Recruitment Drive at London's Imperial College last week, which was declared a "roaring success" by Wikimedia UK. It involved handing out leaflets, speaking to people around campus, discussing a possible Wikipedia collaboration with the Imperial College librarian, and "hanging out at the Student Union with our laptops, ready to answer questions, invite people to edit and persuade people that Wikipedia can be fun!" The team will be hosting similar events at the Universities of Nottingham, Sheffield, Leeds and Manchester between the 1st and 4th of March (believed to be the first such week-long "tour"). For more information or to sign up as a volunteer or attendee, see the events page.
  • MIT OpenCourseWare videos: Board member Samuel Klein (User:Sj) reports that he has been working with MIT OpenCourseWare to make some of its teaching videos available on Wikipedia (example), releasing them under the CC-BY-SA license instead of the initiative's more restrictive default CC-BY-NC. He describes some usability concerns about video on Commons that were put forth by OCW.
  • Egyptian Wikimedians and the revolution: After the Egyptian Revolution of 2011 led to the resignation of president Hosni Mubarak last week, Wikipedian Mohamed Ibrahim (User:Mido, known to many Wikimedians as lead coordinator of the Wikimania 2008 conference in Alexandria) responded to messages of concern and support on the Foundation-l mailing list: "Wikimedians I'm aware of [are] safe and were part of this great revolution, and we all thank you for your support." Mido proudly noted the Wikimedia connections of activist Wael Ghonim, "who was widely credited as one of the main people who moved the protests in 25th Jan, he was also leading the Health speaks initiative from Google - enriching Arabic Wikipedia medical content- and was discussing over with the Wikimedians in Egypt the Wikipedia 10th anniversary event in Cairo sponsored by Google in part just few weeks before the revolution. I don't know yet the status of that but no matter what I promise that we will work even harder on spreading free knowledge to help building our country." Ghonim had publicly expressed his support for Wikipedia when donating during the recent fundraising campaign, calling it "a noble project to educate humanity". In a web-only outtake of his interview on 60 Minutes, Ghonim stated that "Our revolution is like Wikipedia... Everyone is contributing content, [but] you don't know the names of the people contributing the content. This is exactly what happened."
  • Trustee looks back: WMF Board member Phoebe Ayers (User:phoebe) has posted a summary of her activities in "the past three months" on her personal blog, many of them Wikimedia-related.
  • African language Wikipedias: Ian Gilfillan (User:Greenman) has posted an "African language Wikipedia update" on his personal blog, noting that the efforts to found a South African Wikimedia chapter are nearing completion, and that a bid to host Wikimania 2012 in Stellenbosch is underway. He compared the development of the article count of several African language Wikipedias since 2007, and remarked: "Of all the official South Africa language Wikipedias, all but Afrikaans, and of course English, are extremely low traffic, where a single edit is a noteworthy event! ... There’s been a flurry of activity in the rest of Africa though. A number of languages have shot past 1000 articles, and two are even growing faster than the relatively stable Afrikaans."


Reader comments

2011-02-14

Wikipedia wrongly blamed for Super Bowl gaffe; "digital natives" naive about Wikipedia; brief news

Contribute  —  
Share this
By Tilman Bayer and Tom Morris

UK tabloid wrongly blames Wikipedia for US national anthem gaffe at Super Bowl

Performing at this year's Super Bowl (an American Football game on February 6 that was the most-watched US television program in history), singer Christina Aguilera mixed up the lyrics of the US national anthem, causing boos from the audience and subsequent media outrage. An article by the UK's Daily Mail claimed on the following day that she had been "singing botched lyrics found on Wikipedia". As proof, the tabloid presented a screenshot showing the wrong lyrics sung by Aguilera in the 23:52, 6 February 2011 (UTC) version of the article The Star-Spangled Banner (taken from the diff view of an edit correcting them at 23:59). It claimed "the mistake on the website, as seen hours before the Super Bowl and since fixed by a user, matches the mistake she sang". However, as was quickly pointed out by Wikipedians, the article's revision history indicates that the wrong lyrics had in fact first been inserted at 23:51 on February 6, i.e. after Aguilera's performance (in fact the immediately preceding edit at 23:50 consisted of removing a statement about the incident).

Several other news publications cited the Daily Mail's claim, including The Guardian [1] and The Age [2]. Wikimedia UK has requested a correction from the Daily Mail.[3] Jimbo Wales remarked: "I wonder how often we link to the Daily Mail as if it is actually a source for anything at all? The number of times we should do so is really quite small – for most things they are just useless".

The New York Times also mentioned Wikipedia in its coverage of the incident, but more correctly, highlighting Wikipedia's timeliness instead of its alleged unreliability: "Aguilera’s flub was heard by tens of millions of viewers. Twitter was immediately abuzz with talk of her mistake, and by the third quarter her Wikipedia page was changed to include the incident."

Students largely unaware of talk pages, version histories, NPOV and verifiability

A blog post from The Chronicle of Higher Education reports that "Wikipedia’s editing process is still a mystery to students", based on a study ("Young adults' credibility assessment of Wikipedia", appearing in this month's issue of Information Communication & Society) that had 210 US college students carry out information-finding tasks, such as: "You are helping your nephew with his homework. He needs a map of Charles Darwin's voyage around the globe, the entire voyage. Help him get such a map".

While 77% of the participants used Wikipedia at least once during the tasks, and most students appeared to know that Wikipedia content "comes from other regular Internet users like them", the study's authors (Ericka Menchen‐Trevino and Eszter Hargittai from Northwestern University) observed that

The CHE quoted Menchen-Trevino's "surprise" about these results for members of what is often called the "digital native" generation, and Hargittai as stating that "students learned what they did know about Wikipedia from professors and peers rather than from information available on the site itself", and that many of them increasingly "approach Wikipedia as a search engine."

Briefly

  • Deletion of programming languages sparks controversy: The deletion of a number of programming language related articles (Nemerle and Alice ML, both of which are currently going through deletion review) caused controversy on the programming section of Reddit and Hacker News. It also prompted the widely read programmer Zed Shaw to write a post titled "Wikipedia's Notability Requirements And The Slash", criticizing Wikipedia's deletion policy and calling on people to not donate to the Wikimedia Foundation and to try and fix what he sees as problems with the notability requirements. He also announced that he had registered the domain notnotable.com and suggested using it to host deleted articles (similar to Deletionpedia).
  • Against demonization of Wikipedia: A comment titled "Wikipedia – not the devil", by the editorial board of The Current (a student newspaper at Green River Community College in the US) objected to "anti-Wikipedia policies" by teachers and lecturers: "It's time to end the rampant misconceptions surrounding the popular collaboratively created encyclopedia known as Wikipedia, and whether it's a citable source of information. It is a qualified source for generalized information – and can be cited as such – but, like any legitimate encyclopedia, it should never be used as a primary source."
  • Wikipedia compared to Asimov character: A comment in The Hindu ("An empire without kings") praised Wikipedia as "today's de facto standard for fast and fairly reliable information on the Internet", and compared it to the fictional supercomputer Multivac that answers mankind's questions in Isaac Asimov's science fiction stories.
  • Wikipedia's deceptions exposed?: Christwire.org published what it describes as an interview with a Conservapedia editor, who accused Wikipedia of deceiving the public, e.g. by hiding the fact that it was "secretly supported by anti-US atheist Muslims" and by describing Christwire as a parody site.
  • Public Policy Initiative: An article titled "Public policy students publish work on Wikipedia" in The Reporter, a news publication by Western Carolina University, proudly reported that the institution "is one of 21 universities stretching from Harvard to Texas Southern to Berkeley selected to participate this spring in the Wikimedia Foundation’s Public Policy Initiative."
  • Using Wikipedia to drive unflattering images away: In an article on SEO news site Search Engine Land, a marketing consultant described how he had successfully used Wikipedia in "updating Google image results for online reputation management" for a model who wanted to remove "unflattering and outdated" photos from the top Google image search results of her name. The process involved securing rights to an image that the client liked, so that it could be donated under a free license, and replacing the existing image in the Wikipedia article (after finding out that "merely taking down the offending image" didn't work).
  • Wikipedia acknowledged but not cited: Lorcan Dempsey from the Online Computer Library Center mused that "we still don't appear to know what to make of Wikipedia", describing two cases: First, he noted the mention of Wikipedia in economist Edward Glaeser's new book "Triumph of the city" ("Following common practices, Wikipedia is not listed in the bibliography or citations, because any Wikipedia fact was verified with a more standard source. But I still have a great debt to the anonymous toilers of Wikipedia who made my research much easier at many points in time.") Secondly, he reported that his daughter had been given an assignment in high school to insert errors into Wikipedia (she "had her changes corrected almost straightaway, to the extent that it was not possible to complete the assignment as given. In fact, she ended up being barred from editing pages as her behavior was seen as unacceptable").
  • Quora and Wikipedia: In an article title "Why Quora Is Not Wikipedia", Sébastien Paquet (professor of Computer Science at Université du Québec à Montréal) explained the differences between Wikipedia and Quora, a question-and-answer site that has gained prominence in recent months, by noting Wikipedia's principles against original research and about citing reliable sources. "Wikipedia is past-bound: it offers knowledge of what has been known. ... By contrast to Wikipedia, Quora is not past-bound. It is future-oriented."

    Reader comments

2011-02-14

Gaining accreditation as Wikimedia photographers at sports events

Contribute  —  
Share this
By Tilman Bayer
Wikimedian photographer Inisheer at the Portugal–Argentina match in Geneva last week
Snowboarder at the Champs Leysin championships

On February 9, two members of Wikimedia Switzerland, Inisheer (Fanny Schertzer) and Ludo29 (Ludovic Péron), were officially accredited photographers at the international football "friendly match" between Portugal and Argentina, held at the Stade de Genève (Argentina won 2–1.) Their photos are being uploaded to Commons.

Inisheer shared with The Signpost via IRC what the two had learned about gaining accreditation as Wikimedia photographers, drawing from their experiences with other events; these include a 2008 tall ships show in Brest, the 2009 World Cycling Championships in Mendrisio, ice hockey games, a Prix de Lausanne ballet competition, concerts and, last week, the Champs Leysin snowboard championship. (See earlier Signpost coverage of another of their photography activities: "Paris to Cape North 'raid': 300 nordic images for Commons").

Esteban Cambiasso (right) during the Portugal–Argentina "friendly match" in Geneva
Cristiano Ronaldo during the Portugal–Argentina match
The goal-keeper of HKm Zvolen, a professional Slovak ice-hockey team.
Ballet dancer at the 2010 Prix de Lausanne in Switzerland

Inisheer explained their tried-and-tested procedure to persuade organizers to admit them as officially accredited photographers: "First of all, we look on the event website to see if there is a form for accreditation requests, and, if so, we just use it. If not, we write a short email, presenting ourselves as photographers and writers for Wikimedia Switzerland, the support structure ('structure de soutien') for Wikipedia in Switzerland", using an @wikimedia.ch e-mail address (the Swiss Chapter also provided them with Wikimedia business cards). They have adjusted the wording learning from their "past mistakes... In short, we banned the words 'association' and 'volunteers' from our mails, as they are synonyms of 'amateur'. The only critical step is when, sometimes, you are asked about your press card number - then we have to mention volunteering to justify the fact that we don't have a press card. But as our communicating skills improve, it's less of an issue now".

Their mail would then continue by mentioning very briefly that Wikipedia is the 5th or 6th most visited website in the world, and then "we kindly ask for two accreditations so that we can work in the best conditions to illustrate the related articles. It's important that your recipients knows at first 1. who you are 2. what you want. They don't care about anything else." Their emails include an illustrated PDF pressbook of about six pages, describing the Wikimedia movement and mentioning their own past work and collaborations. "The hard part is to get the first contacts, to build your reference list. Then, you can say 'they accepted us and they're bigger than you, so you have no reason to reject us'." Another "big mistake" they made early on "was to write our requests in a way that let them think we were asking for a favour, instead of 'you get the chance of being featured on wikipedia with nice photos' ". This disqualifies you at once, she says. Rather than talking to organizers in terms of being given a chance, the photographers take the line that they are part of the media, and as such they can gain accreditation.

Inisheer explains that most pictures of footballers on Commons were obviously taken without accreditation from audience seats, which is often allowed (and organizers of sports events generally appear to be more tolerant than concert organizers – with notable exceptions, see Signpost coverage: "International Olympics Committee issues legal threat over Creative Commons photography"). However, one has to be lucky or wealthy enough to get a very good seat and there are still quality benefits in being able to shoot pictures directly from the sidelines. Of course equipment makes a difference, too – the Swiss Chapter financially supported their purchase of a 300 mm f/2.8 lens.

Inisheer also regards accreditation as a valuable outreach tool: "As we ask for an accreditation, we're getting in contact with people in the name of Wikimedia CH, saying 'hi, we exist, we do this and this and this, that makes us a noteworthy media', and most of them like to meet some of the real people who build Wikipedia... Initially, we thought that organizers would be interested in getting free pictures of their events, but it's rarely the case. Sometimes, it even scares them. So we don't emphasize the license."

Asked if she shared the concerns of some other Wikimedia photographers about commercial reuse (see Signpost coverage: "Making money with free photos"), Inisheer said that "I think it's quite clear that Wikipedia is available for commercial purposes, the help pages keep nothing secret about this. So if someone feels uncomfortable with that, and I fully understand that POV, they should stop contributing." For herself, she didn't consider it much of a concern: "When I google my name, I find tons of reuses of my pictures, mainly on personal blogs and reports of NGOs, public services and so on. The only commercial reuses I noticed were by newspapers, I've seen nobody make a thousand million bucks with my work. My only concern would be to see one of my pictures used in a way I would strongly disagree with, like propagating racist opinions, and in such cases, I could invoke my moral rights, which stay attached to the author under European laws." Regarding violations of the requirements of the CC-BY-SA license, she recalled a television station that reused one of her photos without credits: "I wrote a mail and they apologized. The license is often not credited, though, but as long as the reuser doesn't claim his own copyright on it, it's ok for me."

See also related Signpost coverage: "Wikimedians accredited as photographers at royal wedding"

Reader comments

2011-02-14

Articles for Creation

Contribute  —  
Share this
By Mabeenot
WikiProject news
News in brief
Submit your project's news and announcements for next week's WikiProject Report at the Signpost's WikiProject Desk.

This week, we interviewed WikiProject Articles for Creation which offers unregistered users an opportunity to create new articles with help from the Wikipedia community. Unregistered users have not been able to create new articles since the Seigenthaler incident in 2005. The Articles for Creation page was created to allow unregistered users an opportunity to contribute new articles after being reviewed by a registered user. WikiProject Articles for Creation was initiated in 2007 to deal with the growing backlog of unreviewed articles. The project maintains a showcase of featured, good, and DYK content that began as submissions by anonymous users.

We interviewed project "participants" Someguy1221, Sven Manguard, Chzz, and Graeme Bartlett. Someguy1221 came across AFC by accident in December 2007 after spending time patroling recent changes with VandalProof. He primarily reviews AFC submissions and helps with formatting, sourcing, and neutrality. Sven Manguard has "no clue" how he ended up at AFC, but he says "I know why I stayed... I was welcomed by a great group of people." He gives kudos to Chzz and Fetchcomms for being warm and friendly. He says AFC serves as a great niche for people who are not "prolific content generators," because he can feel "a certain level of vindication from being able to serve as an intermediary, helping people who are content generators get their content onto Wikipedia." Chzz doesn't think of AFC as a WikiProject, thus he never bothered to "join" it despite being one of the project's central figures. He considers AFC "part of a fundamental Wikipedia process" and has been answering {{helpme}} requests and giving feedback since June 2009. He tends to push a lot of submissions through Did You Know. Graeme Bartlett is an admin who also works on Files for upload, the multimedia version of AFC. He likes to start near the end of the alphabet when burning through backlogs.

How many submissions does the project receive in an average day? Are most submissions accepted or declined? What are some ways an unregistered user can improve an article's chance of being accepted? What are the fastest ways to get a suggestion declined?

Someguy1221: The statistics are easiest to keep track of only after September 2008, whereafter each submission was given its own page. There have been 25,807 submissions in that time, of which 7864 have been accepted (or roughly 30%). There are on average 30 submissions each day. These statistics can be found at Wikipedia:WikiProject_Articles_for_creation/Submissions. There is actually some undercounting of the number of declined submissions, as many have been deleted entirely (although those are often ones made in bad faith, which should not be counted anyway), and sometimes multiple declined submissions are made under the same title. The best way for an any user to give his submission a high chance of being accepted is to provide reliable sources for the article, and write it in the same tone as those sources. In fact, the fastest way to get one's article declined (beyond the trivial stuff like making a blank submission, or not writing it in English) is to not include any sources. Sometimes the reviewer will find sources himself, but more often we simply won't. This is why we remind the submitter at multiple points in the submission process to cite their sources. A less common, but very quick way to get an article declined is to assert the non-notability of the subject. I've seen multiple submissions contain something along the lines of "This band hasn't gone on tour or released any albums, but they are working very hard in Dave's garage on their first single." Particularly relevant to this question is the "quick fail criteria" section of Wikipedia:WikiProject_Articles_for_creation/Reviewing_instructions.
Chzz: I believe others have given you stats; my guess would be 50 a day, but it does vary greatly. The vast majority are declined - at least, the first draft, and that can be quite demoralizing for us reviewers. But if we work carefully with the users, helping them add references, then they not only improve the submission, but also they learn more editing skills, and thus get more interested and involved in Wikipedia.
Fastest decline would be copyright violations, or blank, offensive or other silly submissions. Beyond that, it'd be writing about something that doesn't have significant coverage in reliable sources - and very often, the subject is autobiographical, corporate, or about the author's band, book, website, etc.
But the three most important things are references, references and references. If I had a penny for every time I've had to explain the basic need for verifiability, I'd need very big pockets.

What are some common issues with image submissions? How do they differ from article submissions?

Sven Manguard: Understandably, a large number of people that come to FFU to submit images do not have an understanding of copyright law, of what constitutes fair use, or of the licenses that Wikipedia does and does not accept. As a result, users often submit images that they do not own under the {{PD-self}} public domain tag for own work, or provide links to images on Flickr or Picassa accounts that are licensed as "all rights reserved" on or one of the "non-commercial" creative commons licenses. I consider this to be a problem for two reasons. First of all, it is rather disheartening for a user to submit for upload a file and have it rejected out of copyright concerns. If Wikipedia can use an image under fair use, I will upload it and fill out the fair use rationales myself, but in many cases, I am forced to decline. I try to explain the reasoning behind these declines, but I doubt that dampens the blow to the submitter, and I often wonder how many potential repeat contributors we lose because of this. The second way this knowledge and communication barrier causes issues occurs when contributors submit their own work through a photo sharing service, but license their work "all rights reserved" or with a "non-commercial" license. The contributor is indicating that he or she wants to contribute the image, however Wikipedia is legally bound to decline the request unless the release license is changed to a Wikipedia compatible free-use license. When we ask the contributor to change the license, and they are watching the requests, they do so almost every time. What makes this frustrating, however, is that only a very small number of contributors watch their submissions and respond to our requests. While the struggle for clear communication with new users and concern over copyright are issues across AfC and Wikipedia as a whole, they seem especially concentrated and prevalent at files for upload.
Graeme Bartlett: The problems with images are as Sven Manguard states, almost all due to copyright, either the wrong license or license unknown. The rules to include images are much stricter than the average blog or facebook page, so new people have a bit of trouble understanding.
Chzz: Invalid, or incomplete, declaration of copyright. People freely (and yes, illegally) copy images from the net all the time, so they find it hard to understand the need to get clear, unequivocal permission to use images. And the intricacies of nonfree content confuses the bejesus out of all of us, of course.


The project had a sizable backlog until a drive in 2008 eliminated the backlog. What strategies did the project use to accomplish this? How has the project prevented a new backlog from forming in the past three years? Do you have any tips for other projects attempting their own backlog elimination drives?

Someguy1221: Under the original AFC system (pre-September 2008), all submissions in a day were made to the same page, which was archived every 24 hours. This made it very easy to forget or ignore old submissions, which would simply be shoved away in an archive, never to be seen again. And indeed, that is what happened for over two years. In 2008, a barnstar was offered to anyone who assisted in clearing the backlog (I actually helped, but never asked for one). Lists were assembled of all the archives by year and month, and very gratifying checkmarks applied whenever a day or month was cleared. The new system of a new page for each submission has greatly assisted us in keeping the backlog to only a few days at most. With this system, a single page, CAT:AFC, shows us every outstanding submission, and they don't leave that page until they are reviewed. Martin created a date parameter for the AFC templates that causes the oldest submissions to be listed first, or the exact opposite of what we had before. Someguy1221 (talk) 05:16, 1 February 2011 (UTC)[reply]
Graeme Bartlett: Tips would include a barnstar, discussing the project in places outside the normal project pages to attract new interest, and adding the backlog to the list of backlogs. Naturally the backlog should look to be attractive to work on, and some may attract only those with the most gnomish personality. When we meet new people here we can suggest projects to join.
Chzz: I was not there in 2008. Since 2009, it has frequently backed-up, and when it does, new users become disillusioned with the lack of responses. I've often had my own mini-elimination-drive, enlisting help from other users I know, but mostly I just clear it down myself. For example, on 20 Jan with help from Bsherr (talk · contribs), we cleared it from over 100 to less than 10 (pending). Two days later, it was back over 100, and I processed all of them, in a manic 5-hour session of reviewing. Since then, I've had several similar sessions, clearing 50 or so. It's like painting the forth bridge.
This is a significant and increasing problem for Wikipedia as a whole; that a large portion new users simply want to write their article, and few people want to help others write articles; thus, the ratio of helpers to helpees is not good, and getting worse. The helpees can get pretty disillusioned with the dearth of COI and spam, too.

The project funnels submissions through the Article Wizard. How intuitive is the wizard for new editors? Are there any features you wish were added or changed?

Someguy1221: I always had the impression the Article Wizard was too much, too many rules, too many instructions. The problem is that... well, that's the nature of Wikipedia. I thought about pruning the Wizard once, but I came to agree that users need to be told everything that's in there. We have a page for submitters to provide feedback. Some say it's too much, while others liked it just fine. Regardless, I can't think of a better way to do it.
Chzz: I have a strong aversion to wizard-style submissions of any kind, ever since the horrors of clippy. I believe the basic wiki interface should be vastly simplified, instead. The Wizard, if we need one, should probably just ask the title of the subject, a short summary of what it is about, and then just have a series of lines with box to put a sentence, then a compulsory box to say where that sentence comes from (the reference). It'd add appropriate info to the references, collate named references, and format things neatly. And make the tea for us, while we wait.
I don't think a wizard, or indeed AFC, is the answer to simplifying things for new users editing Wikipedia; I think we need more fundamental simplification of the entire process.

How difficult is it to bring a new article proposed by an unregistered user all the way up to Featured or Good Article status? Does the unregistered user or a member of the project typically do most of the legwork in getting the article up to higher assessments?

Graeme Bartlett: Very few articles from AFC make it to the higher levels of assessment, that is because we are mostly dealing with people's first steps in Wikipedia. However there are a few anonymous experienced users around using the system too. These people create pages at the C or B class, whereas most are stubs or start class.
Chzz: So rare, that in processing well over 9000 submissions, I've never seen it happen. I don't often keep in touch with submissions I've accepted, beyond basic formatting when live and maybe a few follow-up questions on my talk page. The vast majority don't bother to continue editing after getting their one article live; the few who do are usually clueful enough by that time to work out Wikipedia for themselves.
As, sadly, only about 0.1% of articles are FA and 0.3% are GA, it is unsurprising that few AFCs rise to those exalted classifications.

What are the project's most pressing needs? How can a new member help today?

Sven Manguard: I've been helping out at files for uploading, the image uploading component of Articles for Creation, for about three months. In that time, I can't recall seeing more than a half dozen people in total helping to review the submissions there. While FFU does not see the same volume of submissions that the other sections of AfC do, I do believe that it is a tad understaffed at the moment. Helping out at FFU is relatively easy and does not require a large time commitment. I would be more than willing to show any interested users the ropes, and would encourage them to contact me if they're interested.
Chzz: Please help out! - you'll have my everlasting gratitude. Anyone can help; it is very simple. Unfortunately, I think the instructions are a bit complex, so I'll try to simplify;
  • Go to CAT:AFC
  • Look at the ones under 'P' for pending
  • If it is 'acceptable' - that is, if you think it'd be likely to survive deletion (CSD, PROD and AfD) then 'accept' it by moving it live, and tidying it up. (It would be nice to add it to categories, tell the user it was live, and all those other trimmings)
  • If it is not 'acceptable', decide if it can be fixed up - ie, is the subject notable? (Do a bit of googling)
If so, either fix it, or explain to the author what needs fixing. (On the page itself, and/or on their talk page)
If not, 'decline' it. Edit the page, and you'll see a template at the top, something like...
{{AFC submission|||ts=20110206020642|u=Example|ns=5}}
You need to add a letter D for 'decline', the parameter 'reason', and some reason;
{{AFC submission|D|reason|This fails to meet our [[WP:GNG|notability requirements]], because it lacks [[WP:RS|reliable sources]]; please see [[WP:BAND]].|ts=20110206020642|u=Example|ns=5}}
A note to people not afraid of .js:
If you add importScript('User:Timotheus Canens/afchelper4.js'); to your your js page; it makes it very easy to 'accept' and 'decline' giving standard rationales, using a drop-down thingy.

Anything else you'd like to add?

Graeme Bartlett: A related page is the WP:Drawing board that gets a few proposals per month.
Chzz: Yes, please - a short-as-I-can, opinionated, bit of a rant;
As Wikipedia matures, we shift focus from "quantity" to "quality". This makes it harder for new users to join in. It is no longer a reasonable expectation for any truly new user to write a new article, from scratch, and for it to conform to basic policy and guidelines without help. I'm sure that less than 1 in 250* first articles are - without considerable help to develop them - really, genuinely meaningful additions to the project. It is no longer a matter of mere competence; learning to edit takes effort.    *The figure is a very rough estimate; we desperately need to analyse this information.
A few people will learn things for themselves, but largely these are from the type we already have in abundance. The system lends itself to attracting the younger, more technical adept, mostly male, mostly American editors. To widen the demographic, it is essential we make the whole system more friendly and accessible.
There is a disparity with the way new users are treated if they make a live article, as opposed to if they use AFC. Most (truly) new editors creating a live article get a CSD-tag, and/or other template warnings. In AFC, they (mostly) get informed help. I think that creation of new articles needs general reform - and I'm sure it will happen, one day. The system is nasty, to new users - and really, that is why we don't get retention of new editors, and why the editorial mass becomes more and more of an entrenched community.
However, it is not a mere volume of new blood we need, it is quality editors. AFC is most frequently used for Conflict of Interest, so we may be focusing resources on the wrong people. Most new editors are not here to edit an encyclopaedia, they are here to have their article up.
We're approaching the stage where anything with good coverage in RS is already on Wikipedia, particularly for Western culture. We're not lacking in articles on e.g. (ex-) Kings, a (non-trivial) USA town, or some invention (that is known to a few thousand). So naturally, most new articles are either junk, (few) genuine new discoveries or - and this is the big one increasing - conflict of interest cases. At least, this applies to Western culture topics; the drive for new editors from other places can shift that, but raises other issues, as they often need considerable help with the language itself, and there are the difficulties in sourcing information in other languages.
The ratio of helpees to helpers is shrinking.
I think we will require a 'quarantine' for all new articles, similar to AFC, changing auto-confirmed requirements to e.g. 100 edits / 1 month, and thus force all new users to create articles in a 'holding pen', with a checking process. This would inevitably create AFC-type backlogs, but would considerably reduce new page patrol, speedy deletions, proposed deletions, requests to userfy, and annoyed new users, shouts for help, people blocked, and so forth. So instead of productive new users getting horrible nasty CSD template warnings, they'd get helpful advice - "Sorry, can't accept that because of x, y, z - if you can address those concerns though, sure, it'll be fine - ask again".
New users are the lifeblood of the project; we need to treat them much better.


Next week, we'll search for the equation that determines a WikiProject's success. Until then, calculate how many interviews you've read in the archive.

Reader comments

2011-02-14

RFAs and active admins—concerns expressed over the continuing drought

Contribute  —  
Share this
By Tony1 and Dabomb87
Jim Devlin, c. 1873, from the new featured list Philadelphia Phillies all-time roster (D), which lists all players with surnames starting with the letter D. The Phillies are the oldest continuous, one-name, one-city franchise in all of professional American sports, dating to 1883.
Red Dooin was the Phillies' catcher for 13 seasons (1902–14). This baseball card showing his portrait is held in the Library of Congress. The age of the team and its history of adversity have earned it the distinction of having lost the most games of any US sports team.
This week's "Features and admins" includes Saturday 5 – Friday 11 February. Our regular coverage of new featured content appears under the special story.


RfAs and the decline in active administrators

RFA trends January 2010 – January 2011

The number of successful RFAs has continued to fall over the past year, despite a temporary spike last August after The Signpost's story on the RFA drought, and a sudden (but not statistically significant) upswing at the start of this month. Over the past four years, the fall in numbers is significant: 408, 201, 119, and last year just 75 promotions. The author of that Signpost report, WereSpielChequers, maintains a page of information on RFAs. We asked him to comment on the trends over the past six months: "RFA is perhaps not as gloomy a story as the raw figures might suggest. A good trend is that self-nomination, once seen as a bit of a negative, is emerging as almost the norm at RFA; and I'm happy to see the drop in unsuccessful RFAs, as I hope it's largely the consequence of changes such as semi-protecting WP:RFA and other measures to dissuade the WP:Notnow candidates with only a few hundred edits from running."

Possible reasons

However, WereSpielChequers is concerned about the declining number of successful RFAs, and his experience has led him to an original view. "RFA has been in a deepening drought since the unbundling of Rollback in early 2008. Having spoken to editors on a few other projects, I think we could operate with far fewer active English Wikipedia admins than now, but only if we are prepared to make adminship a much bigger deal than it once was, shifting from the model of admins as ordinary editors who mostly do non-admin stuff, to the more full-on model that one of my Spanish Wikipedia contacts described as 'appoint a new admin, lose a good editor'. I think this future would work, though we wouldn't be the sort of community that I'd like us to be."

WereSpielChequers asks what he sees are six critical questions: "(1) With total edits broadly stable at 200,000 a day, how long can we afford to have our number of "active admins" dwindle by 1% a month? (2) What would the trend be and how much of a safety margin would we have by defining active admins rather more strictly than one edit in the past 90 days? (3) Active, uncontentious editors with more than a year's experience can usually get through RFA without difficulty, so why are so many editors waiting three years or more? Should we do more to persuade editors with between one and three years' experience to become admins? (4) WP:Admin coaching seems to be moribund; would it be worth reviving? (5) Does the community want adminship to become the big deal that the RFA crowd are making it, or would people be happier with adminship being a less exclusive club? (6) Would people prefer a large number of active editors who also happen to be admins and occasionally use the mop, or a small number of admins who have little or no wikitime for non-admin activity?"

Bureaucrats respond

Our bureaucrats are primarily responsible for managing the RFA process. The Signpost asked their opinions on the issues raised here. For Dweller, the relevant issue is not the definition of active admins, but "workload and queues; the stats are just one way of measuring problems, [and anyway,] bots do a lot of work previously done by admins." He thinks "a lot of people have been scared off RfA, with good reason. Others are simply waiting to be asked, because they're too modest to self-nom." On WSC's last point [6], he says that obviously everyone would prefer a mix between those two extremes, "which is the status we currently have". How much of a big deal should adminship be? Dweller believes the current balance is about right. "The problem lies with the small number of RFA participants." He emphasises that above all, "RfA's biggest problem is that too few long-standing editors frequent it."

Anonymous Dissident says: "The promotion rates over the past few years are concerning, and the issue merits immediate attention by the community. If the number of active admins continues to diminish while Wikipedia continues to grow, we'll have major problems on our hands. What strategies might be considered? I don't think a reformatting of the system or a paradigm shift are in order; the RfA process is not "broken", as has frequently been asserted. Instead, we need to look for ways to adapt the culture to these dire conditions. We need to be more open when evaluating candidates, especially on matters of time served and experience in particular areas. We need to dispel the notion that adminship is for a privileged and powerful few, so that more editors will have the courage and inclination to lodge candidacies. In short, we need to revive the attitude that adminship is not a big deal, clichéd though its mention may appear."

WJBscribe says RFA is currently failing in its role of creating new admins. "Either attitudes at RFA need to change or we as a project need to consider other ways of appointing new administrators. The cost to the community in terms of too few active admins doing too many tasks is hard to see but I believe it is there. In particular, I worry that there is now too little capacity for admins to double-check the work of others and ensure that 'routine' blocks/deletion/protections are being done correctly."


New administrators

The Signpost welcomes two editors as our newest admins.

At the time of publication there is one live RfA, for ErrantX, due to finish 18 February.


From the new featured article, English composer Frederick Delius, aged 45, photographed in 1907
From the new featured article, User:Edgar181's adaptation of the crystal structure of a rhodocene derivative, after Donovan-Merket et al. (1998). The element rhodium is shown in purple, and the characteristic two rings of five carbon atoms on either side of it can be clearly seen.
From the new featured list, Grammy Award for Best Rock Song: Bruce Springsteen, 1995 recipient, performs in 1988 in the most unlikely location, East Germany.
Fourteen articles were promoted to featured status:
  • Frederick Delius (nom) (1862–1934), an English composer in the Romantic tradition. (Nominated by Brianboulton and Tim riley; picture at right)
  • Battle of Towton (nom), fought on Palm Sunday 29 March 1471, has been called the "largest and bloodiest battle ever fought on English soil", in which King Edward IV "chased his opponents out of England or killed them", according to nominator Jappalang.
  • Adelaide leak (nom), the revelation of a dressing-room incident that occurred during the third Test match of the 1932–33 Ashes series between Australia and England, more commonly known as the Bodyline series. As nominator Sarastro1 puts it, it's the story of "the people involved, and how much they hated each other".
  • Tales of Monkey Island (nom), a 2009 graphic adventure video game. (S@bre)
  • 1950 Atlantic hurricane season (nom), the first year in which hurricanes received names. The season saw many very strong hurricanes. (Hurricanehink)
  • Rhodocene (nom), a molecule containing an atom of rhodium – a transition metal – bound between two planar "ring" systems of five carbon atoms. The hydrogen atoms are omitted for clarity. (EdChem; picture of a derivative at right)
  • Adenanthos cuneatus (nom), a shrub native to the south coast of Western Australia and originally described by French naturalist Jacques Labillardière in 1805. It is grown in gardens in Australia and the western United States, and a dwarf and prostrate form are commercially available. (Casliber and Hesperian)
  • Stark Raving Dad (nom), the premiere of the third season of American animated television series The Simpsons, originally aired in 1991. (Scorpion0422)
  • Startling Stories (nom), an American pulp science fiction magazine published from 1939 to 1955, with characteristic cover art by E.K. Bergey. (Mike Christie)
  • Thomcord (nom), a seedless table grape variety developed in 1983 by Californian grape breeders working for the US Department of Agriculture. The article has received an expert review by one of the men who created the variety, Dr Ramming. (Visionholder)
  • U.S. Route 30 in Iowa (nom), a major east–west US highway that spans 330 miles (530 km) across Iowa, from the Mississippi River to the Missouri River. (Fredddie)
  • Allegro (musical) (nom), one of the most eagerly awaited Broadway musicals ever. Nominator Wehwalt says "it just could not live up to expectations and closed after exhausting its advance sale and a bit more. Perhaps ahead of its time, perhaps fatally flawed because of a poor plot?"
  • Entoloma sinuatum (nom), a poisonous mushroom found across Europe and North America that causes primarily gastrointestinal problems that have been described as highly unpleasant. (Casliber).
  • Monticolomys (nom). Nominator Ucucha says, "In 1929, someone caught a nondescript little brown mouse on Madagascar, which landed in a museum full of nondescript little rodents. However, this particular little brown mouse turned out to be unknown to science."


Eight lists were promoted:


Quentin Matsys's The ugly duchess (c. 1530), a satirical oil-on-oak painting possibly based on a drawing after Leonardo da Vinci.
An écorché (flayed figure) of a horseman and his horse, prepared by anatomist Honoré Fragonard and on display at the Musée Fragonard d'Alfort in Paris.
Six images were promoted. Medium-sized images can be viewed by clicking on "nom":
New featured picture: death masks are wax or plaster replicas of a person's face made after the person's death. Here, two workers, circa 1908, use plaster to create a mould of the deceased person's face. This mould will then be used to make the death mask.
Information about new admins at the top is drawn from their user pages and RfA texts, and occasionally from what they tell us directly.


Reader comments

2011-02-14

Proposed decisions in Shakespeare and Longevity; two new cases; motions passed, and more

Contribute  —  
Share this
By Ncmvocalist and Lord Roem

The Committee opened two new cases during the week. Four cases are currently open; two of which have posted proposed decisions.

Open cases

Monty Hall problem (Week 1)

Opened on 12 February 2011, this case involves allegations of problematic behavior relating to the Monty Hall problem article. During the week, one editor submitted under 2 kilobytes of content as on-wiki evidence. A deadline for evidence submissions has not yet been set.

Kehrli 2 (Week 1)

Opened on 11 February 2011, this case involves allegations of disruptive editing to the Kendrick (unit) and Kendrick mass articles. The case is following on from the 2006 case concerning Kehrli (talk · contribs). During the week, two editors submitted over 10 kilobytes of content as on-wiki evidence. A deadline for evidence submissions has not yet been set.

Shakespeare authorship question (Week 4)

On 10 February 2011, drafters Newyorkbrad and SirFozzie posted a proposed decision for arbitrators to vote on. Proposals which are being voted on include "standard discretionary sanctions"(see Signpost coverage: 20 September 2010) and rulings concerning two editors. Yesterday, arbitrators Cool Hand Luke and Elen of the Roads added a new principle to the proposed decision; the principle builds on the proposals made in the workshop. A motion to close was then adopted on 14 February, ending the case.

Longevity (Week 12)

On 7 February 2011, drafter Kirill Lokshin posted a proposed decision in the workshop; the proposals, consisting of 14 kilobytes, attracted comments from arbitrators, parties, and others. During the week, more than 50 kilobytes of content was added to the workshop, of which more than 15 kilobytes was contributed by a single party. On 12 February 2011, the drafter submitted a proposed decision‎ for arbitrators to vote on. Proposals being considered include "standard discretionary sanctions", an evidence subpage remedy, rulings concerning two editors, a specific ruling that affiliation with the Gerontology Research Group does not in itself constitute a conflict of interest when editing longevity articles, as well as a proposal that urges WikiProject World's Oldest People to seek experienced editors as mentors to the WikiProject.

Motions

Date delinking

As reported last week, arbitrator Newyorkbrad proposed two motions to amend this case. The motions were passed this week with two recusals:

Other matters

The Committee conditionally suspended the indefinite ban of Lyncs (talk · contribs) (formerly Justanother (talk · contribs) or Justallofthem (talk · contribs)). The conditions are such that Lyncs is subject to a single account limitation, an interaction ban, and a Scientology topic ban.

Reader comments

2011-02-14

Bugs, Repairs, and Internal Operational News

Contribute  —  
Share this
By Theo10011 and Waldir

MediaWiki 1.17 deployment failed, postponed

The planned update of MediaWiki as the underlying software which forms the basis of WMF wikis to version 1.17 failed last week (Wikimedia Techblog). The original deployment was expected to begin 07:00 UTC on February 8 (see previous Signpost coverage), but preparations took longer than anticipated and actual deployment began at around 13:00 UTC.

Several issues became apparent almost immediately. The parser cache miss rate almost doubled with the new deployment, at which point the Apache servers, which are responsible for delivering content to users, became overloaded and started behaving unpredictably. The increased load culminated with multiple issues across the project from increased lag to even outage for some users. At this point, the deployment was rolled back to the previous 1.16 release. The tech team investigated and prepared for another attempt after resolving some technical issues. A second attempt was made at 16:27 UTC, but this ran into similar performance issues and had to be called off 90 minutes later. Further attempts were put on hold.

Danese Cooper, Wikimedia's Chief Technical Officer, blogged about the failed deployment and explained what the Foundation had attempted to deploy:

After further investigation and several fixes to the release, Rob Lanphier, a developer with the WMF, added that "some of the unsolved issues are complicated enough that the only timely and reasonable way to investigate them is to deploy and react". As a result of this, he said, a new plan had been drawn up in which 1.17 will be deployed on "just a few wikis at a time". The tech team believes the problem was located in the configuration of the $wgCacheEpoch variable, which caused a more aggressive culling of the cache than the servers could handle (Wikimedia Techblog).

The team decided on a two-stage deployment for their next attempt (reviving some old code for project-wise upgrading). The first phase took place 6:00–12:00 UTC on Friday, February 11. This was limited to the Simple English Wikipedia and Wiktionary; the Usability and Strategy Wikis; Meta; the Hebrew Wikisource; the English Wikiquote, Wikinews and Wikibooks; the Beta Wikiversity; and the Esperanto and Dutch Wikipedias.

At the time of writing, the deployment had been completed on all but the last two projects. The Hebrew Wikisource, included after a request from a community member, gave a chance to observe the deployment on a right-to-left language wiki. The team also reported some localization issues which triggered ParserFunction bugs on both nl.wikipedia.org and eo.wikipedia.org. The traffic from nl.wikipedia.org was enough at the time to cause a noticeable spike in CPU usage on the web servers, including some time-out errors; thus, deployment onto nl.wikipedia.org had to be delayed. After these issues are resolved , the second wave of deployment is expected to start on Wednesday, February 16 (see the current list of WMF wikis that are already running 1.17).

An IRC office hour Q&A was held on matters related to the ResourceLoader, which is expected to cause compatibility issues with some existing Javascript code. Trevor Parscal and Roan Kattouw, the main developers of the ResourceLoader, were available on IRC on February 14 at 18:00 (UTC) to answer queries related to the new feature.

In brief

Not all fixes may have gone live to WMF sites at the time of writing; some may not be scheduled to go live for many weeks.

  • RefTools default: Through a poll at the Village Pump, it was decided by near unanimous consensus to make the refTools gadget a default feature on the English Wikipedia. Implementation discussion is proceeding at MediaWiki talk:Common.js, and the switch is currently on hold until MediaWiki 1.17 is deployed (see also last September's Signpost dispatch on RefTools/RefToolbar and other references-related tools).
  • New wikitext parser? After the last attempt at creating a wikitext parser generated excitement and lengthy discussion on the wikitext-l mailing list, a new attempt, by AboutUs.org developer Karl Matthias, has been brought to light under the name of "Kiwi". Discussion is ongoing in the post's comments and on wikitext-l.
  • Distributed wikis: One of the applications of a defined parser grammar would be the ability to have alternative, non-MediaWiki interfaces to interact with the content. The full video was released of "Distributed wikis", a talk at linux.conf.au by Brianna Laugher (User:pfctdayelise; abstract) that touched on the subject. She proposes the use of the write function of the API and custom interfaces to create WikiProject-sized "forks" as the key to the long-term survival of the project, arguing that community decentralization would create smaller, more manageable and more welcoming groups of editors. Laugher compares these to the editing groups that existed at the beginning of Wikipedia, and those that can still be found in various specialized wikis (most notably wikias). Similar ideas have been floating around for a while; see meta:Versioning and distributed data.
    The talk also examined the possible application of the concept of distributed revision control to wikis, to allow easy forking and merging (a concept explored in various earlier proposals), and compared Flagged revisions to release versions in software development.
  • Call for skins: A call for skins was made on the wikitech-l mailing list.
  • Public status-monitoring page: The WMF announced that a public status monitoring page has been set up at http://status.wikimedia.org to allow staff, community members and the general public to keep an eye on the servers' uptime and status (Wikimedia blog). The service, powered by Watchmouse, has been in the works for a while, having been mentioned in the last three WMF Engineering Updates. Its relationship with previous monitoring tools Nagios and Ganglia is still unclear.
  • Bug fixes: Several long-standing bugs have been fixed:
    • The successful close of bug #17160 (opened January 2009) allows for the display of namespace aliases in userpages to be gender-specific (revision #82029). As The Signpost reported last week ("Widespread discussions about the low participation of women in Wikipedia"), the blanket use of masculine namespace forms—such as Benutzer (male user) rather than Benutzerin (female user) on the German Wikipedia—had come up in the recent discussions about Wikipedia's "gender gap", where it was criticized as "hardcoded discrimination" and Sue Gardner had called it "awful".
    • Bug #16013 (opened October 2008) was resolved in revision 81573, allowing the direct linking to sections (tabs) of the preferences page.
    • Bug #8680 (opened January 2007) was closed successfully, after the introduction of $wgFooterIcons allowed easy customisation of MediaWiki sites' footers.

      Reader comments
If articles have been updated, you may need to refresh the single-page edition.
Get the latest headlines on your user page – just add {{Signpost-subscription}}.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Wikipedia_Signpost/Single/2011-02-14&oldid=1183872578"