As first reported Friday in Mediazona, a Belarusian court sentenced Wikipedian Mark Bernstein (User:Pessimist2006, of no relation to User:MarkBernstein) to three years of a type of house arrest for "gross violation of public order". He'd been in custody since March 11, previously sentenced to 15 days in jail for disobeying an official. After his release he wrote on social media, "I am free. Healthy physically and mentally. Thank you all for your support," according to Zerkalo.

The charges arose because Bernstein edited the Russian Wikipedia, giving information that appeared to violate a new law of the Russian Federation which limited how news on Russia's "special military operation" in Ukraine could be reported. His type of punishment is informally known under the odd name "home chemistry". ("Chemistry" is an informal term originated in the Soviet Union, where it originally meant incarceration combined with work at a place with health hazard, such as a chemical plant. In the current context, "home chemistry" means that the person lives at home with restrictions in freedom, and continues to work, with part of their salary withheld by the state.)

On June 6, Reuters (and many others) reported that the WMF was appealing a fine imposed by Russia on the foundation for similar alleged violations of the Russian law. See this issue's News and notes. – Sb

Stephen Harrison published a well-researched and circumspect summary of Tamzin's Request for Adminship in Slate: "Inside Wikipedia's Historic, Fiercely Contested 'Election'". Tamzin's RfA had been remarkable for having attracted 340 supports, 112 opposes and 16 neutrals – the highest-participation RfA in the project's history – and for its focus on the role administrator candidates' political views should (or should not) play in assessing their eligibility for the role.

Harrison commented on the fact that of Wikipedia's 1,034 administrators, only about 500 (465 at the time of writing) are considered active. Moreover, the number of successful Requests for Adminship per year is far lower than the number of administrators who die, leave the project, or otherwise lose the tools (voluntarily or otherwise). In part, this is a reflection of the fact that candidacies for Adminship have become a gruelling process:

Because it's a lifetime appointment, some Wikipedians have taken to treating RfA with all the seriousness and showmanship of the SCOTUS confirmation hearings—except that the Wikipedia version is all in written form, taking place on a dedicated wiki page.

Harrison concludes that something has to give:

Although there have been several calls over the years for RfA reform and proposals to make the process less corrosive, Wikipedia editors told me there has not yet been substantial progress in this area. According to a 2021 RfA inquiry hosted on Wikipedia, "Because RfA carries with it lifetime tenure, granting any given editor sysop feels incredibly important. This creates a risk-adverse and high-stakes atmosphere."

Then again, the notion that Wikipedia admins must have their powers for the rest of their days is certainly not an immutable law of the universe. With enough buy-in, that rule is just as editable as any wiki page. In Wikipedia as in life, we must pursue options for dialing down the heat.

For prior Signpost coverage on the 2021 RfA Reform initiative, see last October's Discussion report or the wrap-up at December's News and notes. – AK

Entrepreneur, which has often published articles promoting paid editing on Wikipedia, puts forward a more moderate view in Do Entrepreneurs Need a Wikipedia Page? The article dutifully explains that, to get a Wikipedia entry:

you need coverage in numerous reliable secondary sources, independent of the subject ... On the other hand, Wikipedia retains a reputation as being filled with a plethora of niche and un-noteworthy pages, or pages that contain incorrect claims based on dubious sources or no sources at all ... With a Wikipedia page, you may become a target of random trolls, ex-spouses, former business associates or disgruntled employees. Just because you could, doesn't mean you should.

It sounds like they've read the essay "An article about yourself isn't necessarily a good thing". They raise the specter of outing yourself as a phony – "Smart audiences can spot thinly disguised sponsored content" – and add:

Some (and perhaps many) Wikipedia editors are not just selfless nerds, but rather cold-blooded paid mercenaries. I have witnessed their work firsthand and was surprised at the audacity that editors-turned-paid-media-consultants exhibited in protecting their clients and dismantling substantiated, empirically factual information to maintain an entirely false narrative that their clients paid for. Once you have a Wikipedia page, make sure you allocate a budget to defend it from paid attacks or random vandals.

Despite covering most of the major points of why an entrepreneur should not hire a paid editor, they miss the main point. Whether there should be an article on a particular businessperson or company is for Wikipedians to decide, not entrepreneurs with conflicts of interest, or their paid flunkies.

In How to Edit your Law Firm’s Wikipedia Page: 3 Golden Rules on JDSupra, a newsletter for law firms, the founder of a legal PR firm says it all very simply:

• "The first rule of editing a Wikipedia page is NOT to edit a Wikipedia page," restating Jimbo's bright-line rule of only editing talk pages.
• "Be transparent", i.e. declare your COI.
• "Provide verifiable, third-party sources."

That should work just fine, unless too many Wikilawyers get involved. – Sb

The Australian identified a Wikipedia editor as a campaign staffer for an MP candidate prior to the last national election. The newspaper accused the Wikipedia editor of inserting disinformation into articles about candidates of opposing parties and removing damaging information about favored candidates. The main MP candidate involved won the election and said that the Wikipedia editor worked on their election campaign and edited Wikipedia, but the two activities were entirely separate.

When contacted by The Signpost through their user page email, the Wikipedia editor requested that they not be identified. They said, "there are currently defamation proceedings around" the article in The Australian. "It's "a pure piece of slander from beginning to end."

An investigation by The Signpost revealed that the editor admitted on-wiki to using multiple accounts, but said that he was unfamiliar with Wikipedia rules on sockpuppeting. An apparent autobiography of the editor has been nominated for deletion. – Sb

On May 28, timed to coincide with the start of this year's fundraising season in India (emails May 23 – June 20, banners May 31 – June 28), the Indian Express published a piece titled: "Raju Narisetti interview: 'Wikipedia is building trust with transparency'". The interview with Wikimedia Foundation board member Narisetti focused in particular on efforts to expand Wikipedia content in Indian languages and on the contributions of Indian editors to Wikipedia. Moving on to the topic of the Foundation's fundraising, the write-up of the interview continued as follows:

"More than 75% of the money we raise globally goes to two things. One is to give money back to the volunteer community so they can launch a new language. Two is about half of it goes to the infrastructure. You need to have databases and put it on the cloud and make sure it's reliable," he said. Although a lot of the money is raised in the more developed Western markets, most of it is actually flowing into the global south, where the growth will come in languages and users.

The statement that "most of" the money raised is flowing into the global south was queried by this reporter on the Wikimedia-l mailing list. In response, Megan Hernandez explained on Meta on June 2 that –

Raju was unfortunately misquoted, per a direct transcript of the interview. He more generally said "a lot of it is actually flowing into the global south" not "most of it." This is in line with our regional grantmaking in the "Global South" as well as other investments, including our technology support, which, as you know, ensures that Wikipedia is available in more than 300 languages globally. We have requested a correction to his quote.

A disclaimer was duly added to the article on June 3, and the passage now reads:

"Although a lot of the money is raised in the more developed Western markets, a lot of it is actually flowing into the global south, where the growth will come in languages and users."

For an analysis of WMF regional spending see this issue's News and notes. – AK

Reply All, the venerated podcast about the internet and society, used Annie Rauwerda's Depths of Wikipedia account as the hook for its second-to-last episode ever. They dove into three pages she's featured – cute aggression, the Pittsburgh toilet, and economist Guy Standing.

The last segment is perhaps of most interest to Wikipedians, as reporter Kim Nederveen-Pieterse sat down with (well, called) Standing himself. Shockingly, he claimed to be entirely unaware of his internet renown. But after Nederveen-Pieterse explained the legendary edit war over his photo and caption, he was asked to take a stand.

Standing wasn't too thrilled with the edit warring. "My goodness. What a waste of people's time, I'd have thought," he said. "It's sad." (We know.)

But he took no issue with the joke itself. "If it's a little aside that draws people to smile, that's great. Because we need a little humor in our lives, especially at this horrible time," he said. "But I hope that it draws people's attention to the serious messages that I've been trying to convey through my work" highlighting the feasibility of a universal basic income.

The Signpost has previously published an opinion piece which expressed appreciation for the Guy Standing joke and argued for the value of humor on Wikipedia. – S

• 40-minute Wikipedia documentary: Deutsche Welle presents Wikipedia – The pros and cons of the online encyclopedia, a documentary recounting a European perspective on the history of Wikipedia and some of its contemporary challenges. Also available in Spanish.
• Fundraising memes: indiatimes.com published a round-up of Twitter fundraising memes, noting that "people on the internet are having some fun with Wikipedia's humble request – no points for guessing that there are enough memes on it. There were also users who pointed out that Wikipedia does not require funds to begin with."
• Views of Nazi history: "Wikipedia acts as a check on Putin’s false view of history" says Noam Cohen in a Washington Post "perspective" piece. The sub-head is "Hitler myths are a particular source of vigilance for the editors of the Internet's research library" and it ends with praise for editors who hold "stubborn confidence that facts can guide us through the darkness". It touches on the efforts of our K.e.coffman, which you can read about in earlier Signpost coverage.

• Wikimedia Enterprise has a buyer: "Google is paying the Wikimedia Foundation for better access to information", says The Verge. There is more on this topic in this issue's News and notes and News from the WMF; see also our March, 2021 News and notes item titled "A future with a for-profit subsidiary?"

• The journal Nature shows us A cross-verified database of notable people, 3500BC – 2018AD combining biographies from several Wikipedias and Wikidata and comes up with "a cross-verified database of 2.29 million individuals (an elite of 1/43,000 of human being having ever lived), including a third who are not present in the English edition of Wikipedia." This may show us the way to increasing the number and diversity of bios on enWiki, but it is likely more important in its analysis of gender, cultural, economic and other real and potential biases.

This user essay originally titled "RFA trend lines" was started in 2020. You may edit it, but please do so on the original page and not The Signpost. – E

Trends in support percentage during a request for adminship are rarely informative, and these trends are difficult to interpret even when they might be informative.

As a first order approximation, let's assume there's an RfA where no new information comes to light over the course of the request and everyone !votes independent of each other. In this case, if we were to poll every Wikipedian, there would be some global, unobserved support percentage for the population; call it p. Given an RfA with n participants, each !vote in an RfA can be considered a Bernoulli trial with probability p. The number of supports, s, at any given time can be simulated by combining the results of multiple Bernoulli trials; this can be modeled as a binomial distribution of n trials and probability p.

RfAs run for multiple days and are among the most attended discussions on the project; this suggests that the final support percentage is a reliable stand-in for the population support percentage. By contrast, the trend line tells us almost nothing and may in fact be misleading. Our binomial model is the same we would use to model the ratio of heads to tails in successive coin flips. Imagine we are going to flip a coin for a contest and we want to prove that the coin we are flipping is fair. We flip it 150 times and track the number and order of heads and tails. After 150 coin flips, the ratio of heads to tails would be very informative: if it is far away from a 50% split then the coin is not fair. The order these flips occur in, however, is uninformative, and in fact, using it as evidence for an argument is logical fallacy known as the gambler's fallacy.

Our first order approximation of RfA trend lines represents a hypothesis regarding !voting behavior. Absent evidence to the contrary, we assume editors review the candidate and comment independently of others just like the result of a coin flip does not depend on prior results. But an RfA is not a series of independent tests. The amount of information available to a !voter includes not only other comments, but new question answers, and summary statistics like current support percentage. These can consciously or unconsciously affect how a participant !votes and justifies an alternate hypothesis: each !vote is related to the ones that came before it (and maybe even after it). If the population support percentage, p, doesn't change then this distinction is immaterial to our model.

Reconsider the coin flip example: if the probability of getting heads depends on the previous result such that getting a heads changes the probability from 50% to 50% (i.e., no change), then the dependent model and independent model will produce the exact same results. Differences only arise if the dependence changes the underlying probability. In statistical terms, we can say that the binomial distribution is robust against violations of the independence assumption as long as the sample size is much smaller than the population. For example, let's assume that getting a heads increased the likelihood of getting another heads. In that situation our independent trial model will be accurate at first but get more inaccurate as we have more trials since the non-independence will keep compounding making heads more and more likely. Bringing this back to RfA, the influence of prior votes on later ones is not a serious threat to the binomial (independent trial) model. It would only affect our model if there were thousands of !voters or if there was a major shift in the underlying probability.

Editors look at trend lines because they believe that (or want to evaluate whether) earlier votes influenced later ones to such an extent that a major shift occurred in the underlying probability. considering how !votes are non-independent, this intuition makes sense but is flawed. Essentially, this is a model selection problem, and the starting assumption ought to be the null hypothesis. As discussed above, this means that without evidence, we should assume that the order of !votes is not meaningful, just like the order of coin flips. Claiming that a coin is unfair because of the order of heads and tails is fallacious, so we cannot reject the null hypothesis on the basis of the trend line alone; we need some other kind of evidence. What is critical to understand in the context of RfA is that the trend line cannot tell us whether a change in the underlying support percentage occurred; they are only useful if we already assume that happened and even then can only help us determine when.

Like any hypothesis testing tool, a trend line is only useful if we already have a hypothesis. Unless there is an independent reason to believe the information available to participants has changed, the trend line is most likely to reflect randomness in the sample rather than a meaningful pattern. Without a rational argument as to why early !voters did not have the same information as late !voters, an argument from trend-line data is weak.

The accompanying image shows a trend line for the support percentage in a simulated RfA which ended within the discretionary range. It is a series of 150 Bernoulli trials, but at some point the underlying probability of support changed from just above the 75% threshold for an outright pass (76 percent) to well below the 65% threshold for outright fail (60 percent). The location at which this change occurred is difficult to determine from the trend line alone, and in fact the graph looks like other simulations where the underlying support percentage was above the discretionary range the entire time. The change in probability occurred after the 90th !vote, and despite that change, there is little evidence in the trend line alone to substantiate that. These simulations can be replicated (in spirit, since it's a random simulation) using the following R code:

When a new person steps up to a position, they inevitably want to make some changes. So, now that I've pulled Picture of the Day into my despotic regime, let's discuss how I'm going to ruin... um.... run it.

So, I'm Adam Cuerden, I've been a Wikipedian since around the start of 2006. I work a lot on restoring historic images. For example:

BEFORE	AFTER

And, yes, that was terribly self-indulgent. Just like this article is pretty much me indulging my sense of humour because I really hate talking about myself unless I can be horribly sarcastic and poke fun of myself the whole time. I'm sure that gimmick won't get old. Anyway! In late May this year, while I approached 8% of all featured pictures on English Wikipedia,^{[Note 1]} I was asked if I wanted to take over Picture of the Day, and apparently the mad cackling couldn't be heard all the way from Scotland, so no-one stopped me and it looks like I've gotten the position. Not that anyone else seemed to want it. It's actually apparently a lot of work and people tend to burn out on it in a few years, so, um... Thanks?

Given I have about one in twelve featured pictures, every June will be Adam Month, where every single picture will be by me. Or... we'll go more-or-less rigidly in order, to try and keep everything fair, changing around things only to celebrate holidays and avoid multiple similar images in a row. Probably the last one. Less controversial. And I like the other image creators.

All joking aside, I do plan to try to be rigidly fair. This means:

1. Insofar as is reasonable, images appear oldest to newest.
2. Anyone may suggest an exception for holidays, anniversaries, and so on.
3. I shan't undo any claims I've already made for dates, but shall severely limit myself for the future, giving others a chance to step in first. As long as "step in first" means "claiming dates in 2023", because when I came in we had empty spots a couple weeks ahead and I wanted a massive buffer so if I have a bad month, nothing breaks.
4. To avoid runs of very similar images, some rearrangement may be necessary. This will be done in such a way that maintains order as much as is reasonable.
5. Since it's easier to set up several similar POTDs in one batch, once a set starts, it will generally be scheduled once a month for the next X months . If this causes problems (like if it fills the schedule up with nothing but sets), I'll deal with it ad hoc. As an exception, since we have literally over a hundred featured pictures of money, pretty much all of which are older than anything else in the queue, the 28th of every month will be Money Day, where some coherent portion of these will be featured, until we clear them sometime in 2024 or so. Unless the June 28th Picture of the Day (Now with galleries!) goes horribly wrong. That's... still a possibility and would mean changing a lot of plans.

I also plan to stop a few practices I never liked. For example, occasionally sets of images would be put on the main page using a random algorithm to select one to display, meaning each image could have as little as the equivalent of a couple hours on the main page. This seems mean to our content creators, as it trivialises things that may have taken a lot of work. (Also, I'm still a little salty about the time it happened to me.) The templates are already set up to handle two images, even though this is, for some reason, not documented, and it's fairly trivial to include a small gallery where appropriate. It might mean breaking up sets into smaller parts, but that's better than dumping them in a way that hides most of the images unless people want to roll for a new random image over and over.

Secondly, I'm not going to censor POTD any more than absolutely necessary. Some things shouldn't be on the main page, but some of the decisions of the past feel rather arbitrary, especially when you consider what we've put on the Main Page without controversy. Consider these:

• 
  The Large [Breasted?] Bathers by Renoir
  Picture of the Day for 10 November 2019
• 
  The Taking of Lungtungpen by Archibald Standish Hartrick.
  Picture of the day for 8 August 2015
• 
  Confederate dead after the Battle of Chancellorsville
  Picture of the Day for 3 May 2015
• 
  "Hoarders and Wasters and Uncensored Penises" by Gustave Doré
  Picture of the day for 20 October 2009

As such, it's rather surprising to see images had been kept off the main page which arguably aren't nearly as bad. Both of the below appeared in Wikipedia:POTD/Unused, which was meant as a place to put images that could not possibly be put on the main page, and I fail to see why these arguably tamer images would fail to make the cut.

• 
  Seagull expelling guano
• 
  "Indecency" by Isaac Cruikshank

I feel Picture of the Day should be a balancing act between potential harm and potential educational benefit, with a bias towards running the image, and I'm not seeing any significant harm from running these images. However, I do see harm in censoring them, as it sets weird precedents, makes odd value judgements (seagull poop is worse than dead people?), and just generally feels wrong. There were originally a couple sentences here encouraging people who disagreed with this decision to join in a discussion, because I am willing to change my mind, but... it turns out that I submitted this article too late for May's publication and, by the time June's Signpost comes around, my plans will already have come to fruition, so... Mwahaha?

Of course, some things were kept off the main page for a reason. In the simplest case, the image is unused, or the article it's connected to is too short. Article improvement or finding a use for them might bring these back into consideration. Similarly, Featured Pictures are as prone to going out of date as any other project, so something like File:WMAP 2010.png, which has been reduced to an infobox decoration, might have been worth main paging ten years ago, but we now have better. Likewise, File:Love or dutyb.jpg has had its scan at the Library of Congress much improved, so can probably wait on a delist-and replace nomination, especially as a nearly identical version of the same restored image has been on the main page before. (This has now happened. As I said, missed the May cutoff date for publication.)

All those, however, could theoretically be put (or, at least, have potentially been put) on POTD in some form. Some types of images have potential harms that may well outweigh any educational value. For example, one featured picture shows a lynching. Not only have such images been used for vandalism, but the specific picture has absolutely no documentation as to where it happened or who the victim was, so the harm of making whole groups of people feel unwelcome is not balanced by any significant educational value whatsoever. Something like this, while we'd probably still want to discuss it going on the main page, feels like a much better way to cover such horrific events sensitively.

It's all a bit of a challenge, but I do promise to do my best.

Oh, right. The gimmick of this article!

In order to destroy Picture of the Day:

• Don't come to me with any issues you might have.
• Don't engage in civil discussion about any problematic pictures of the day so that we can better make decisions about what the cutoff should be for the Main Page, which can, of course, change from what it was in the past.
• Definitely don't go to Wikipedia:Picture of the day and write blurbs for awesome featured pictures that haven't appeared yet on the main page, possibly to celebrate holidays such as Hallowe'en, Christmas, Diwali, or Eid, or possibly for anniversaries of key dates related to the subject, or just because you want that date.

But, most importantly for your goal of dooming Picture of the Day:

• Don't nominate images and vote at Wikipedia:Featured picture candidates to help promote artwork and articles you're passionate about. We promote about fifteen to twenty-five featured pictures a month at present, while putting thirty or thirty-one pictures on the main page almost every month (one month is lazy and leaves early). So by simply not engaging in a fun content process, you can defeat my reign of terror! Rather slowly, but still.

Of course, if you want to actually help make Wikipedia a better place and help Picture of the Day (and make my reign eternal), replace all the "Don't"s with "Do"s.

On June 8, a new page was added to the Wikimedia Foundation's official wiki, bearing the weighty title: "Terrorist and violent extremist content procedures and guidelines". The document, first posted by WMF Tech Law Lead Counsel Charles Roslof, laid out procedures through which the Wikimedia Foundation would accept and respond to a "request for Terrorist Content notice of action" via a "Terrorist Content Sub-Group" of the WMF Trust & Safety team: requests are to be followed by an internal review by the WMF legal team to ensure that they were in fact legally required. However, this review process is not guaranteed to be public, and it is unclear whether even the fact of requests having been accepted will be a matter of public record – the policy says that "the Foundation may be limited by applicable law in disclosing the information about these requests".

Initially, the policy specified that requests would be accepted from "relevant law enforcement authorities in the United States of America (USA), European Union (EU), or a member nation of the EU". However, a subsequent revision on June 10 updated the policy and changed some wording, omitting the specific reference to jurisdiction (as of press time, the policy now refers only to "relevant law enforcement authorities"). It also added a passage clarifying removals that the WMF objected to ("Please note that the Foundation may also be in the course of appealing the Legal Order, but prohibited from reinstating the content in question unless and until it has succeeded in its appeal").

This page's only incoming link is from the site's list of policies, and as of press time it has not been mentioned on the WMF's official news page or Twitter account, making it difficult to tell whether this is a simple formalization of existing practice or a new mechanism entirely.

The policy is fairly short, and does not reference active content removal measures being taken on the WMF's part, instead relating only to the WMF's response to reports from government agencies. It remains to be seen what the ultimate implications of such a policy are. The definition of "terrorism" is notoriously inconsistent – our own perpetual topics of furor on international politics can provide good examples of this – and it is unclear precisely what the interplay will be between this policy and takedown requests from jurisdictions such as, for example, the Russian Roskomnadzor. — J

The Wikimedia Foundation is appealing a ₽5,000,000 (67888.66 USD or 60441.37 Euros) fine issued by a Russian court relating to the decision not to remove information verboten in Russia from several Russian Wikipedia articles. The fine came after the court found that the Wikimedia Foundation operated within the Russian Federation and that the content in question (largely related to the Russo-Ukrainian War) was illegal under Russian law. The Signpost has previously reported that publishers in Russia must only use government-approved facts and terminology when covering military operations.

The appeal was made on June 6; the WMF put out a statement (Russian-language version) outlining their rationale for the appeal on June 13, saying that the decision to fine the Wikimedia Foundation was based on from erroneous claims that the Wikimedia Foundation operated within the territory of the Russian Federation, and that the fine itself violated rights to free expression and access to knowledge. The Wikimedia Foundation also objected to allegations of "disinformation", writing in the statement:

Russian-language Wikipedia is a crucial second draft of history, written by and for Russian speakers around the world who volunteer their time to make reliable, fact-checked information available to all. Blocking access to Wikipedia in Russia would deny more than 145 million people access to this vital information resource. Further, the articles flagged for removal uphold Wikipedia’s standards of neutrality, verifiability, and reliable secondary sources to ensure articles are based in fact. They are well-sourced, including citations to a variety of established news sources. The articles continue to be improved by Wikipedia volunteer editors from all over the world with more sources and up-to-date information.

Russia's telecommunications regulator, Roskomnadzor, previously sought to restrict access to certain articles on the Russian Wikipedia within the Russian Federation, taking umbrage to the characterization of Russia's "special military operation" in Ukraine as an "invasion", "aggression", or "war". On March 31, the regulator threatened the WMF with a ₽4,000,000 fine for publishing what it called "unreliable socially significant materials, as well as other prohibited information".

A list of Wikipedia pages banned in Russia is available on the Russian Wikipedia. As of press time, the list of articles restricted by the Russian Federation has expanded beyond the Russian language articles, and now includes some articles from the English Wikipedia. — M

The WMF's 2020 Form 990, released last month, enables some interesting insight into where the Wikimedia Foundation has been spending its money, especially in light of claims by the Foundation that "a lot" of the money raised through donations is flowing into the Global South (see this issue's In the media section). Firstly, page 1 of the Form 990 shows that the WMF reported:

• Total revenue of $158,987,065 ( up from $120,919,258).
• Total expenses of $111,669,959 ( down from $112,162,122).

According to the Form 990, $92 million of the total expenditure – that is, all but $20 million of it – was spent in the United States. This includes $5.5 million that the Wikimedia Foundation did not actually spend, but added to its own endowment at the Tides Foundation.

As for expenditure in the rest of the world, the Form 990 divides this into expenses for "Program Services" (mainly technical and legal support for Wikipedia and the other Wikimedia websites) and "Grantmaking" (to grow global reach and increase contributor diversity). Of the $20 million spent on "Program Services" and "Grantmaking" outside the United States, most of it – around $15 million (9% of total WMF revenue) – went to Europe and North America outside the US (i.e. Canada and Mexico). This left a little over $5 million – or about 3% of total revenue – for the entire rest of the world. The two main regions in the rest of the world that saw funding for Program Services and Grantmaking in 2020 were Africa and East Asia/Pacific. The regional breakdown was as follows:

World regions (excl. North America and Europe)	Spending (US$)	% of revenue
Sub-Saharan Africa	2.0 million	1.3%
East Asia and the Pacific (Australia, Korea, Taiwan, etc.)	1.3 million	0.8%
South Asia (Afghanistan, Bangladesh, India, etc.)	0.6 million	0.4%
Middle East and North Africa	0.6 million	0.4%
South America	0.5 million	0.3%
Russia and neighbouring states (Armenia, Azerbaijan, etc.)	0.1 million	0.06%
Central America and the Caribbean	0.1 million	0.06%

Total spending in the Global South (understood to comprise the Middle East, North Africa, Sub-Saharan Africa, South Asia, South America, Central America and the Caribbean), therefore, amounted to just $3.8 million. That is 2.4% of total revenue – or 3.4% of global expenditure.

The Form 990 also gives a detailed breakdown for "Grantmaking" alone, without the expenses classed as "Program Services". According to this breakdown, grants given outside the United States totaled $3.5 million, of which $1.2 million went to Europe ($666,875 to organizations and $496,615 to 32 individuals). As for grants given to organizations (page 32–33 of the Form 990) and individuals (page 34) in the Global South, these were mainly focused on Sub-Saharan Africa and South America, with the amounts going to South Asia – home to almost 2 billion people – looking particularly small by comparison:

Global South regions	Grants total (US$)	% of revenue
Sub-Saharan Africa	1,368,343	0.9%
South America	418,934	0.3%
Middle East and North Africa	84,969	0.05
South Asia	78,537	0.05%
Central America and Caribbean	2,925	0.002%

Overall, the "Grantmaking" amounts reported in the Form 990 for the above regions of the Global South totaled $1,953,708, or 1.2% of WMF revenue in 2020 – a very minor part of the WMF budget, especially bearing in mind that the Foundation enjoyed an effective surplus of more than $50 million. It will be interesting to see how these figures will develop in the years to come. See also the Foundation's own 2020–2021 grantmaking report on Meta. – AK

• WMF CPTO selected: The Wikimedia Foundation has selected Selena Deckelmann as its new Chief Product and Technology Officer. Deckelmann comes to the WMF from Mozilla Corporation, where she is currently still serving as the Senior Vice President responsible for Firefox. She will officially join the WMF on August 1, 2022.
• Google employees to join Abstract Wikipedia team: The WMF has signed a deal with Google.org in which seven experienced Google employees (five engineers, a technical PM, and a designer) are to join the Abstract Wikipedia team as "Google Fellows" for six months. The focus of the Fellows will be to support the backend of Wikifunctions, enabling the team to speed up their schedule. – AK
• Wikimedia Enterprise will not hit revenue target this year: Wikimedia Enterprise had achieved 30% of its year revenue goal by the end of the third quarter of the WMF's 2021/2022 financial year (i.e. by the end of March, unchanged from the second quarter). The Advancement department reports: "We have been unable to close additional customers as quickly as we projected due to unanticipated legal and product requirements, and will not hit the revenue target for FY21/22." More can be found on this subject in this month's News from the WMF. – AK
• Fundraising: The WMF had already exceeded its annual revenue goal for 2021/2022 by the end of March 2022, reporting a third-quarter revenue status of $153.6 million (year goal: $150 million) for the Foundation and $13.4 million (year goal: $10 million) for the Wikimedia Endowment. Combined with an underspend, this resulted in net assets increasing by $51.9 million in the first three quarters. The Foundation has been fundraising this month in India, Latin America and South Africa, with banner campaigns in Austria, Belgium, Luxembourg, and Spain to follow next month. The WMF projects that it will once again exceed the previous year's revenue record. – AK
• Endowment update: The WMF has published an update on the status of the Wikimedia Endowment (see the opinion piece in last month's Signpost) saying an Application for Recognition of Exemption under Section 501(c)(3) was filed with the IRS in late 2021. As the IRS is currently experiencing delays, the WMF has no expected date by which processing of the application will be completed. The update also states that bylaws and policies governing the Endowment will be posted on Meta after minutes are approved at the July 2022 Board meeting; however, it says nothing about how much money the Endowment now holds. The last update in this regard was that the Endowment had passed the $100 million mark in June 2021, five years sooner than originally expected. – AK
• New bureaucrats: The Signpost welcomes the English Wikipedia's newest bureaucrat, Lee Vilenski.
• Articles for Improvement: This week's Article for Improvement is Liquor. Please be bold in helping improve this article!
• ARBCOM gives non-binding advice to those wishing to talk about Wikipedia to other Wikipedians while not being on Wikipedia: Basically, just be open and sensible about it.

At its best, Wikipedia offers eloquently written articles, pulled together from a variety of sources. At its worst, Wikipedia offers hastily written abstracts from a single source. And, perhaps just as bad, one-to-one translations from the English Wikipedia to less-active language versions. Why just as bad? Because direct translations fail to address topics that might not be important for an English speaking audience, yet are very important for others. Italian translations of books by James Joyce, for example, are important for Italian readers of Wikipedia – while many English speaking Wikipedians couldn’t care less.

Seven years ago, I uploaded a photograph of the Swiss writer Annemarie Schwarzenbach. Schwarzenbach was a rich, independent and eccentric writer and journalist, who travelled the world as if she was walking through the woods of rural Switzerland, where she was born. She was an ardent photographer too, who took photos in Afghanistan, Belgian Congo, Eritrea, Georgia, India, Iraq, Russia, the United States and 20 other countries.

Photos of her were then quite hard to find: I had to browse through several books before I found a suitable photograph which also had a suitable copyright status. That photo (right), with Schwarzenbach holding a Rolleiflex Standard 621-Camera, looked like a selfie from around 1938 – at first sight.

Schwarzenbach died from a fall on her bicycle in the Swiss Alps. Her mother destroyed most of her letters and diaries. Wikipedia tells us: "A friend took care of her writings and photographs, which were later archived in the Swiss Literary Archives in Bern." That's almost true. The executors of her will, Erika Mann and Anita Forrer, weren't exactly close friends. In the end, Anita Forrer took care of the photographic and literary archives of Schwarzenbach. She treated the archives as a treasure in the library she founded, the Biblioteca Engiadinaisa, and later donated the Schwarzenbach archives to the Swiss Literary Archives. And in that archive I found the supposed selfie again: as a photo made by Anita Forrer in Malans, Switzerland, in 1938.

Anita Forrer, the guardian of the Schwarzenbach archives, was a hell of a woman. Born in 1901 in St. Gallen, she visited a poetry reading by Rainer Maria Rilke when she was 18 years old. Lightning struck and, as Rilke later put it in a letter to her, he became an external reference point to her "in the geometry of the heart to somehow get the measure of the distances and relations in the vast space of feeling". Their correspondence (70 letters over a period of seven years) was published in 1982. Forrer developed, worked in Paris and Luzern, was briefly married and travelled the world. She had an affair with Annemarie Schwarzenbach, and travelled to the United States before WWII. A photograph of Forrer was not hard to find: the UC Berkeley Library's digital collections had a nice photograph, shot by Johnny Florea in 1938.

But there's something odd about the files in the UC Berkeley collections. Three photographs of Anita Forrer are accompanied by a photo card calling "Miss Anita Forrer" a "Swiss woman auto racer". I was not surprised at all. As Forrer was a graphologist, photographer and a spy, she might as well be an auto racer. After all, she had been a driver for the American Red Cross Motor Corps in WWII, so why shouldn't she race cars? But, as one source is no source, I removed that assertion from a Wikipedia article. I hope this piece encourages others to find out if Anita Forrer really was an automobile racer. Any help would be appreciated. The English Wikipedia still lacks an article about Anita Forrer (which will be written soon, of course), but the German and Dutch Wikipedias have made the attempt.