Wikipedia:Village pump (miscellaneous)

 Policy Technical Proposals Idea lab WMF Miscellaneous 

• Table of contents
• First discussion
• End of page
• New post

The miscellaneous section of the village pump is used to post messages that do not fit into any other category. Please post on the policy, technical, or proposals sections when appropriate, or at the help desk for assistance. For general knowledge questions, please use the reference desk.

Discussions are automatically archived after remaining inactive for a week.

« Archives, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78

Banners and changes at the Wikimedia Foundation

I've been the CEO of the Wikimedia Foundation for nearly 11 months now. I am posting here as a follow up to the Request for Comment to change fundraising banners.

I agree that it is time to make changes at the Wikimedia Foundation, including more direct community input into fundraising messaging. We have taken the guidance provided by the close of the RfC to change banners on the English Wikipedia campaign as early as Tuesday. The fundraising team welcomes your help and ideas on the specifics.

The task at hand in responding to the guidance provided by the RfC is that Wikipedia's existence is dependent on donations. Donated funds are used primarily to support Wikipedia. I think what we heard is that while this may be true, how we say it matters. We need banners that better recognize the real stake our communities have in how we communicate to our donors.

In the next few months, the fundraising team will work more closely with local communities to guide future campaigns. The Foundation will measure the financial results of using new banners in this year's English campaign, and we will share this information when the campaign is completed.

I will briefly address a few other areas of concern that were raised about the future direction of the Wikimedia Foundation, and commit to writing again in January after we finish this campaign. I believe some things at the Foundation can in fact be different, because they already are:

• I agree that the Foundation has grown very rapidly over the past years and that the budget should be stabilized. In my first six months, I did two things to act on this: first, I informed the Board that we would dramatically reduce hiring until we were sure that we were getting the maximum benefit from the resources we already had. Second, I started reversing the trend of prior growth by setting this year's budget to account mainly for year-on-year costs like inflation.
• I brought in a Chief Product & Technology Officer, @SDeckelmann-WMF:, who has experience supporting online communities and collaborating with technical contributors. She has been on the job for 17 weeks and has already directly responded to community concerns on New Pages Patrol and Wikimedia Commons. In March 2023, Selena will be ready to host forums on- and off-wiki with what she thinks are needed improvements to the Foundation's processes, including technical support and collaborative product development. We collectively have to respond to decades of growing technical debt, poor processes for maintaining software, and staying relevant in a world where technology keeps going faster.
• I have already started frank conversations with the Board and Foundation staff about what roles the Foundation should grow (like support for technology), and what activities we should hand over to others or stop altogether. Looking ahead, the size of our budget should be driven by what the Foundation should be doing, and can actually do well. I think the 2030 movement strategy provided guidance (and motivated much of our historic growth), but was short on the specifics. I await the Movement Charter for further clarity, but believe we'll need to make some decisions sooner.

None of these things may happen as quickly as those of you who have been very frustrated for many years would like. I think we are heading more in the right direction, and I am sure you will tell me if we aren't.

I will write again in January with more information. In the meantime, you can reach me on my talk page or by email. MIskander-WMF (talk) 23:28, 25 November 2022 (UTC) (I am getting on a long flight shortly but will check this thread afterwards.)

@MIskander-WMF: Can you clarify Second, I started reversing the trend of prior growth by setting this year's budget to account mainly for year-on-year costs like inflation? The year or year budget increase is over 20% (after an even larger 29% increase the year before), which is more than twice inflation; can you provide a breakdown of where this extra money is flowing to?

Regarding the movement, the current format deters most editors from participating; most editors don't leave their home wiki to participate on meta, and of those that do even less will register and regularly check an unrelated forum. Would you consider acting to encourage a broader range of participation, such as shuttering the forum and returning the conversation to meta, as well as looking at technical changes to allow editors to participate in discussions on meta without leaving their home wiki? BilledMammal (talk) 23:48, 25 November 2022 (UTC)

Sure, @BilledMammal: thanks for the question. Yes, the budget did grow 20%, which is down from 31% year-on-year growth in the previous year - this was my main point about stabilising the budget following a period of more rapid growth. As I will detail in a moment, this takes into account some increased costs carried over from the prior year, as well as a return to travel and community convenings which were reduced during the pandemic. The Foundation's actual expenses for FY2022 were just under $146M, and our current budget was projected to be $175M in FY2023 (although this remains uncertain pending fundraising outcomes). The $29M difference is broken down into 4 broad categories: about half of that is personnel costs that includes full-year salaries for people hired in the prior year and cost-of-living and other adjustments across 40+ countries; the other half is from increases in grants, travel and events, and depreciation which includes developed software and our newest data centre in Marseilles. This was offset by some in-kind expenses and reductions in donor processing costs. I am curious to hear more about what you think are the highest priorities for the budget (besides infrastructure like servers which I don't dispute). I noted above that for me, this has to start with what roles the Foundation is expected to do, and can do well. MIskander-WMF (talk) 13:48, 28 November 2022 (UTC)

I agree that the movement strategy forum may be better placed on-wiki, since that is the software that Wikipedians are used to. –Novem Linguae (talk) 06:15, 26 November 2022 (UTC)

Hi @BilledMammal: and @Novem Linguae: - I will be honest that I am still learning about all the different channels folks use to communicate and for what purposes. I understand that some forums are there to support multilingual participation, but I am happy to keep learning more. MIskander-WMF (talk) 13:48, 28 November 2022 (UTC)

@MIskander-WMF: I don't know why the forum is in use, but I know that it decreases participation. If the goal is to encourage multilingual participation I would encourage users to contribute to discussions in their own language, and use Google Translate to understand comments made in languages they can't read. BilledMammal (talk) 00:25, 29 November 2022 (UTC)

I am curious to hear more about what you think are the highest priorities for the budget (besides infrastructure like servers which I don't dispute). I believe the highest priorities are technical support. For example, the WMF should be able to respond to 400+ editors asking for NPP to be upgraded with "Yes, we can" and "When can we have a meeting to discuss the exact requirements of the NPP team?" rather than struggling to find the resources. BilledMammal (talk) 00:25, 29 November 2022 (UTC)

I second that question by BilledMammal. For reference, the relevant board meeting minutes state: FY22-23 is not anticipated to be a year of rapid growth. The Foundation anticipates 17% growth to a budget of $175 million with moderate growth in terms of staffing. Next year, the fundraising team will be increasing targets in each of their major streams, with a particular focus in Major Gifts. (Expenses in 2021–2022 were just under $146 million, so $175 million does in fact represent a 20% increase).

I have an additional question. Last year, an INSEAD fellow published an article containing the following passage: The vast majority of Wikimedia’s value to ordinary people – the website we know and use – costs the firm about 30 percent of their $112.5 million operating budget ($33.75 million) to maintain according to Lisa Seitz Gruwell, Chief Advancement Officer at Wikimedia. Would you please provide a comment on this that my colleagues could quote in the WP:Signpost? Thanks, Andreas JN466 08:41, 26 November 2022 (UTC)

@MIskander-WMF: Thank you so much for posting here Maryana, it really is a refreshing change. While it is obvious you cannot be expected to micromanage the WMF, this comes as a welcome respite from a lack of direct engagement by people at the top with the volunteers since around 2014. This new trend in dialogue is also evidenced by the video meetings I and the NPP team have had very recently with Selena. While the role of the CEO is extremely important as an ambassador for the movement, most of us feel reassured when its leaders know about what needs to be done on the factory floor to keep the volunteers happy and retain their enthusiasm. I am aware that my comments do not always come across as particularly friendly towards the Foundation, but over the years I and others have had to strive very hard on occasions to impress upon some its engineering departments that they can sometimes help by rethinking their practice and policies - and it worked. Thank you again. Kudpung กุดผึ้ง (talk) 01:09, 26 November 2022 (UTC)

+1 to what @Kudpung said for both you and Selena. Best, Barkeep49 (talk) 17:36, 26 November 2022 (UTC)

Thank you @Kudpung and @Barkeep49. Most of the real work happens on the proverbial "factory floor," so it's where I also need to spend a lot of my time. As I said in my note above, we have some really tough stuff to figure out on the product/tech side. No matter how hard SDeckelmann-WMF works, she isn't a magician (at least I don't think she is!), so I hope we can keep rebuilding these channels of communication to get more collective brainpower solving problems together. MIskander-WMF (talk) 13:52, 28 November 2022 (UTC)

• Thanks for your statement. While I have your attention, can a staff member please summarize how us volunteers can best participate in 1) movement strategy and 2) annual plan? It seems that these two processes direct what WMF staff work on and where our donation dollars go so they are very important. But I have no idea how to participate. One question I left on a meta talk page has gone unanswered. If we are given real, effective channels to discuss and vote on these issues, I think that would go a long way towards reducing tensions. If the existing channels are poor or non-existent then the processes should be overhauled. Thanks. –Novem Linguae (talk) 06:14, 26 November 2022 (UTC)

  "If we are given real, effective channels to discuss and vote on these issues" History has proven again and again btw that this isn't true. There are plenty of channels both past and present and almost no one ever shows up. Ppl only show up to complain when things go wrong (cause they want to write articles, not built communities). Due to low participation, none of the channels ever become effective. The most effective are the affiliates, but those seem to be hated even more so by the subset of wiki editors complaining this time round. —TheDJ (talk • contribs) 10:11, 28 November 2022 (UTC)

  Hi @Novem Linguae - As someone with relatively new eyes, I can see that the Foundation has grown its communication channels for how to engage in these processes but we clearly aren't winning with everyone, which I am not sure will ever be possible. Though I know invites like this can easily be missed, I did experiment with a few new things in the Annual Plan last year on-wiki and saw higher levels of volunteer response – we had about 17,000 pageviews of the plan and more discussion on Meta-wiki (4x increase from last year). I also agree with @TheDJ that affiliates can serve as a more helpful channel for some communities. I am trying to figure out what changes for the coming year would be productive, and achieve the intended goal. What would you find most helpful? (Can't promise I will be able to do it, but would appreciate knowing.) MIskander-WMF (talk) 13:57, 28 November 2022 (UTC)

  Hey @MIskander-WMF. Thanks for replying. For better communication with enwiki specifically, I'd recommend 1) that WMF continues to post invites to annual plan and movement strategy stuff at WP:VPWMF, 2) that WMF tries to make all its posts at WP:VPWMF (instead of at this board and others), and 3) for the most important stuff, you can request a watchlist notice by making a post at MediaWiki talk:Watchlist-messages. Maybe do this once or twice a year for the most important items. I have WP:VPWMF watchlisted now, so that would get people like me involved in these conversations. Another issue is that it sounds like movement strategy uses a third party forum, and annual plan uses video meetings. This is not software that Wikipedians may be most fluent in. Wikipedians are most comfortable using wiki talk pages to communicate, so that leads to 4) provide wiki talk pages and invest some time in onwiki discussion of both these issues (movement strategy, annual plan). The idea here is to reduce the # of steps it takes for a community member to participate in these processes, which reduces bureaucracy and reduces how much of an "insider" you have to be in order to know the steps to participate. Hope this helps. –Novem Linguae (talk) 04:54, 29 November 2022 (UTC)

Thank you for this very welcome statement. I understand that an organisation as large as the WMF has become can't turn instantly, but it gives hope for movement in the right direction. I had planned to stop editing Wikipedia on 29 November when the banners appear, possibly returning in January, but I'll carry on while we see whether your good words fall on the right ears. Certes (talk) 20:08, 26 November 2022 (UTC)

@MIskander-WMF: am I correct in my reading of extrapolation from this report that your 2020—2021 current salary is at least US$400,000? If so, how would you explain to our small donors why your salary is worth 200,000 recommended donations ($2, in the banners going out this year)? And how would you explain to the volunteers who work for free what value you bring to the Wikimedia community that is worth $400,000? — Bilorv (talk) 10:36, 27 November 2022 (UTC).

(Factual correction per Jayen466—my comment was based on both misremembering and then misreading The Signpost. Original comment here.) — Bilorv (talk) 17:54, 27 November 2022 (UTC)

That's an easy one to answer: she's the CEO of one of the world's largest nonprofits, in charge of hundreds of employees and hundreds of millions of donated dollars. Good luck finding anyone on earth competent and willing to do that job for less than $400k/yr. That's a lower-than-average CEO salary for an organization of that size. Levivich (talk) 16:45, 27 November 2022 (UTC)

The WMF is smaller than a typical UK university, and VC salaries in the £300k+ range are widely seen as excessive. —Kusma (talk) 17:36, 27 November 2022 (UTC)

Why would you compare a US nonprofit to a UK university? Compare it to other US nonprofits, or better yet, the WMF is a tech company, so compare it to other tech companies. Levivich (talk) 17:37, 27 November 2022 (UTC)

Honestly, I think anyone who is complaining about executive salaries for non-profit companies should possibly take a look at themselves and their own jobs and ask if anyone could do their own job for significantly less than they're making, not just within their own home country but also around the world. It's never a pleasant conversation whenever someone's livelihood (paying for their bills, supporting their families, putting their kids through college, etc..) is scrutinized, and I feel a bit more compassion is needed here. 🌈WaltCip-(talk) 13:34, 29 November 2022 (UTC)

A rant about WMF salaries, for my volunteer colleagues:

• What kind of software engineers do we want the WMF to hire to work on MediaWiki? Below average, average, above average? Well, according to Indeed.com, the average base salary for a software engineer in the US is $122k [1]. At Apple, $174k. At Meta, $165k. With benefits, that'd be over $200k in total compensation.
• Software engineering managers, of course, get more: $141k base salary is average; at Apple, $219k, at Meta $245k [2], and with benefits, you're now approaching $300k.
• It's a no-brainer that the Chief Technology Officer at a tech company, who supervises all the managers (and probably some vice presidents as well), would be at or above $300k [3]. Base salary for Apple's CTO is $378k according to Salary.com [4]. WMF's CTO made $319k in 2020 per its Form 990, page 7, and that's total compensation, not just base salary; their base salary is surely below $300k.
• The other officer-level executives (Chief Financial Officer, General Counsel) would of course be on par at $300k total compensation.
• So how much should the Chief Executive Officer, who oversees the other officers, make? $400k total compensation is totally reasonable; in fact, it's low.
• All the executives are taking "pay cuts" to work at the WMF -- they'd all make more doing the same jobs elsewhere.

There are lots of problems with WMF spending but executive salaries aren't one of them. Lowering them would only be shooting ourselves in the foot because we wouldn't be able to attract top personnel, who would go work elsewhere for more money. Levivich (talk) 17:36, 27 November 2022 (UTC)

The WMF very much pretends not to be a San Francisco tech company, but a movement empowering people all over the world. If it needs the money to pay outrageous San Francisco tech salaries, it should say so on the banners it shows to people on far lower salaries all over the world. —Kusma (talk) 17:45, 27 November 2022 (UTC)

Kusma, these aren't "outrageous San Francisco tech salaries". These are average base pay in the United States--the whole country. San Francisco tech salaries are much higher. Levivich (talk) 17:50, 27 November 2022 (UTC)

It is news to me that the WMF is a tech company. (Usually, tech companies have more functional mobile apps). —Kusma (talk) 17:56, 27 November 2022 (UTC)

That's a cute rhetorical line but in all seriousness, you do agree that the WMF's primary function is maintenance and development of technology, right? That's what makes it a "tech company". I don't think it's news to you. Levivich (talk) 18:01, 27 November 2022 (UTC)

The WMF has important tech roles such developing MediaWiki. Personally, I'd like it to become more of a tech company, simply by terminating some peripheral non-tech activities. However, what proportion of its income is spent on tech? For $200+ million a year, you could pay 1000+ tech salaries and buy some very impressive server hardware. I don't think that happens. Certes (talk) 18:15, 27 November 2022 (UTC)

Believe it or not but I happen to know a large group of people willing to donate their labour for 100% less than market rate—some world-leading experts, many with PhDs, professional educators among them—because they believe in the mission of making human knowledge accessible to all. Many of those people actually pay in order to volunteer (buying books, newspaper subscriptions etc.), and are unemployed or employed in jobs with salaries less than they could potentially get.

It is one thing for employees to receive living wages and fair working conditions: I've never expected a WMF employee to work outside Monday-Friday, 9 a.m.-5 p.m., although some seem to. It is another to pay $400,000 of small donors' money to a single person. I don't care what other companies do. The inequalities of capitalism are not news to me. Though I don't disclose my job/entire education status for safety reasons, let me assure you that with my qualifications if I wanted to work at Meta, I could be working at Meta as a software engineer; instead I chose a course far less well-compensated because of my values. — Bilorv (talk) 17:49, 27 November 2022 (UTC)

I flat out don't believe you that you know anyone who is qualified to be the WMF CEO and who would do that job for free. It's a more-than-full-time job, nobody would do that for free, and if they did, I wouldn't trust them. A CEO who isn't afraid to be fired is dangerous. Levivich (talk) 17:52, 27 November 2022 (UTC)

It's strange to disbelieve me over something I didn't say and don't believe, with which I'll let this be my last reply. You are better than this comment Levivich and I don't mind disagreeing with you but I'm sorry that you are not interested in hearing my perspective. — Bilorv (talk) 18:00, 27 November 2022 (UTC)

"100% less than market rate" == free, no? "I happen to know a large group of people willing to donate their labour for 100% less than market rate" == you know people who would work for free. Do you know anyone who would serve as WMF CEO for free? I understood that is what you were saying? Levivich (talk) 18:02, 27 November 2022 (UTC)

As you're asking a clarifying question, I will respond to this to say: read it again. The people I'm referring to in the first paragraph are you, me, and the Wikimedia volunteer community, and I am not talking about WMF employees in that paragraph. — Bilorv (talk) 18:08, 27 November 2022 (UTC)

My apologies for misunderstanding, then. Since your post about knowing people who would work for free was in reply to my post about WMF salaries, which in turn was in reply to your post about the WMF CEO's salary, I thought we were talking about WMF salaries, specifically the CEO's. Otherwise, what's the relevance that there are people who would do some work for free? Of course we all know people willing to work for free -- you and I are among them -- but that's not relevant to the question of WMF CEO and executive salaries. Nobody is going to do those jobs for free, and the people doing them are already taking a discount on their compensation. Levivich (talk) 18:11, 27 November 2022 (UTC)

Like you Bilorv I choose to take some of my compensation in belief for the organization I work for rather than maximizing my compensation based on my skills at a different organization. Like both of us so do many Wikimedia employees. Software engineers do take a discount to work at the Wikimedia Foundation. Most employees, and all senior employees, take a discount over their for profit rate to work at the Wikimedia Foundation. We want those people to be professional and to act accordingly so of course we need to pay them a market rate for their non-profit work. We want them to be professional so that they respond in a professional manner when they're told - as many people in mission based organizations or roles (i.e. teachers) - that they should not value their labor appropriately and that they should do their work for less or, in this case, for free. We want those people to be in the range of what they could get at other organizations so we get skilled employees doing skilled work rather than people whose skills are comenserate with low compensation or people who can afford to gift their labor for less than a fair rate because they have outside support. The former would lead to a poorly run org, the latter would lead to a workforce that represented the Wikimedia movement even less than what we have today. Best, Barkeep49 (talk) 18:02, 27 November 2022 (UTC)

The WMF should minimise its US-based staff. Even rates in Europe are considerably lower; rates in Asia even more so. Andreas JN466 18:58, 27 November 2022 (UTC)

The WMF has been becoming less US-centric in its staff composition for much of the last decade. Seddon ^talk 07:33, 29 November 2022 (UTC)

I'd love it if this very general statement could be made more meaningful by backing it up with specifics. Andreas JN466 13:48, 29 November 2022 (UTC)

I do think Bilorv's question is highly relevant in the context of WMF fundraising in India or South Africa or even Latin America, especially fundraising that implies Wikipedia might go offline or start charging a subscription.

For reference, about one billion people in India have less than $2 a day to live on. Almost half a billion people have less than $1.25. [5] So if you are asking for 150 Rupees, what you are asking for is the mathematical equivalent of asking a programmer on a $200,000 salary for $550 on a fundraising banner – more actually, because that programmer spends most of their money on luxury items (one day without such spending makes hardly a difference), whereas a poor person in the developing world spends most of their money on bare necessities for survival.

And you need about 250,000 people in India donating that 150 Rupees just to pay the salary of the WMF CEO.

Fundraising messages in such countries should be toned right down so that only the people who truly can afford it (yes, there are billionaires and rich professionals in India) feel called upon to donate. And even then you are actually pulling money out of India and into the US that could be more profitably spent on more urgent charitable causes in India itself. Andreas JN466 18:35, 27 November 2022 (UTC)

Yes, "a poor person in the developing world spends most of their money on bare necessities for survival", and so is pretty unlikely to be reached by wp appeals at all, most especially on en:wp. Johnbod (talk) 19:41, 27 November 2022 (UTC)

@Johnbod: According to the Business Standard, "India had 1.2 billion mobile subscribers in 2021, of which about 750 million are smartphone users." I recall that the WMF made special efforts to support the lower-spec operating system used by many mobile phone users in India. Andreas JN466 21:16, 27 November 2022 (UTC)

It seems a suspiciously high figure given the whole population is supposed to be only 1.375 Bn... Anyway hard to reconcile with your first statement: "about one billion people in India have less than $2 a day to live on. Almost half a billion people have less than $1.25". Lots of Indian statistics have to be treated with caution. Johnbod (talk) 23:44, 27 November 2022 (UTC)

That's not really the issue though, is it? Mobile phone use is pretty ubiquitous in India and Africa; it's one item even people very poor by Western standards try really hard to afford.

Our article claims over 600 million smartphones in India in 2021, up from 440 million in 2020, based on a source also cited by Reuters.

The WMF constantly speaks of its efforts to reach the most disadvantaged in Asia and Africa. I would assume these efforts are having some degree of success. Andreas JN466 00:06, 28 November 2022 (UTC)

Chesterton's fence, how does it work? The Blade of the Northern Lights (話して下さい) 05:29, 28 November 2022 (UTC)

Mobile phone usage is. Smart phone usage, not so much. Seddon ^talk 07:34, 29 November 2022 (UTC)

750 million smartphones is a lot. Even if you mistrust the article linked above: the number of smartphones in India went up from 440 million in 2020 to over 600 million in 2021 according to Newzoo, a source deemed reliable by Reuters and the New York Times. It's a fairly reasonable assumption that the number has risen to about 750 million today and will continue to rise. Andreas JN466 14:00, 29 November 2022 (UTC)

You're saying a 25% increase in mobile smartphones in a year is a reasonable assumption? I don't. I'm not saying 750 million is wrong only that if you have 600 million as your 2021 basis it's not reasonable to suggest 25% more people have them today. Best, Barkeep49 (talk) 15:36, 29 November 2022 (UTC)

Against the background of a 36% increase in mobile smartphones in the year before, and a 28% increase in the year before that, it doesn't strike me as unreasonable to assume there might have been a 25% increase this year.

The rate of increase has actually been rising.

For reference, smartphone penetration in India rose from 25% in 2019 to 32% in 2020 (that is the 28% increase) and then to 43.5% in 2021 (the 36% increase). See List of countries by smartphone penetration.

Look, the WMF is specifically targeting this demographic: "I think in the US, we’re primarily familiar with Apple and with Android, Google, those types of products. But in India and in other kinds of emerging markets in the world, there’s different operating systems, more simple operating systems that we’ve built an app for. So, we formed a partnership with KaiOS, and we built a KaiOS app to get that out to users in these places. So, these are much, much less expensive phones and devices that they’re using. And so, we have to build an app for that market. And those are the markets we’re trying to reach." (Interview with Lisa Seitz-Gruwell on Wikimedia fundraising.)

The fundraising messages shown to these users should bear the economic disparities between the prospective donors and the donation recipients in mind. Regards, Andreas JN466 17:13, 29 November 2022 (UTC)

Hi Andreas,

We do not run fundraising banners in the KaiOS app.

Best, JBrungs (WMF) (talk) 10:11, 30 November 2022 (UTC)

@JBrungs (WMF): Thanks, that is very good to know. Does that mean the only mobile users in India who see fundraising banners are those that have an Android or Apple phone, or are there apps for any other operating systems that show them?

And are the fundraising banners limited to English Wikipedia, or are they also shown in other Indian language versions (Hindi, Marathi, etc.)? Regards, Andreas JN466 13:49, 30 November 2022 (UTC)

@JBrungs (WMF): Also, I understand KaiOS includes a browser (note screenshot showing Wikipedia). So, what is the situation of KaiOS users who don't download the app, but visit the mobile online version of Wikipedia? Would they see the banners? Regards, --Andreas JN466 19:02, 30 November 2022 (UTC)

During the last campaign we ran a test on the Android app in India. This was the first time we did any in app banners in India. We only run banners on English Wikipedia in English not on any local language versions. I am finding out about the KaiOS browser but this might take me a few days. JBrungs (WMF) (talk) 12:43, 1 December 2022 (UTC)

Thanks, Julia, I very much appreciate your efforts. What may also be relevant here is the proportion of KaiOS users taking the trouble to download the WMF app versus those simply using their built-in browser to visit Wikipedia. I would imagine most casual users will do the latter.

Restricting banners to English Wikipedia means that most people confronted with a banner will probably be above India's 90th percentile in terms of education and income, simply based on the number of English speakers in India. This is good.

However, it doesn't mean these users are wealthy by US standards. According to Lant Pritchett, The “statistical” rich in India are very poor by rich country standards. This 90th percentile household [he shows a picture] has a measured consumption per person per day of $8.47 (in purchasing power adjusted dollars), which is higher than the highest poverty line the World Bank reports of P$5.50. But the US “guideline” poverty line for a family of 4 in 2019 was $17.63 [per person] per day. The “rich” Indian household would have to have income twice as high to not be poor in the USA. Sources: [6][7] Andreas JN466 14:40, 1 December 2022 (UTC)

Thank you for this message and clear communication, Maryana, it's good to see! A question: how do you think the WMF should balance the needs of readers vs the needs of editors? Both groups are crucially important to Wikipedia's long-term mission, but while we editors are generally skilled at making noise and alerting the WMF to our wishes, readers (who vastly outnumber us) are almost entirely silent. The recent discussion over the adoption of Vector 2022 was a fine case in point. We editors are a select group of hobbyists whose hobby is Wikipedia; readers might include almost everyone in the world with an internet connection. In the end, though, all editors started as readers and without a healthy reader base and a solid pathway from reading to editing, the editing base will eventually become endangered. How does the WMF plan to determine the needs of readers? How does the WMF view this issue? —Ganesha811 (talk) 02:11, 29 November 2022 (UTC)

@Ganesha811 Thank you for helping us all remember that readers also have a stake in these discussions. The Foundation's product team has had this as a focus area, and in particular how to help them deepen their engagement with the wikis, wherever they are on their journey - whether it’s reading more frequently, or if they are ready, to try to start editing. The newcomer experience built by the Growth team is one example of a project that engages readers when they are farther along in their journey, encouraging them to try out editing with guidance and suggestions along the way (this seems to have increased the engagement of new editors). I could use more perspectives from you and others about how the Foundation should be balancing the needs of existing contributors with the needs of readers. It’s an important question. MIskander-WMF (talk) 17:47, 1 December 2022 (UTC)

Thanks for the answer. In general, I would say that it would help editors if the WMF was as transparent as possible about how they assess what readers need, and ask for input on surveys and questionnaires. The more information shared, the easier it will be for editors to trust that the WMF is not just using readers as an excuse to make changes the WMF would have wanted anyway. That's not something I believe, but I've seen that sentiment expressed here. I would also say that the newcomer experience is helpful. There has been some talk on the fundraising banners page about possibly launching an "editing banners" campaign to encourage readers to edit - that's another idea worth exploring. —Ganesha811 (talk) 01:09, 5 December 2022 (UTC)

I thought the WMF was going to abide by the outcome of the RfC. The latest banner once again asks people to donate money to "preserve" Wikipedia. :(

(The wording goes: "Wikipedia is a place to learn, free from bias or agenda. Together, let's preserve this special space. If Wikipedia has given you £2 worth of knowledge this year, please support the technology that makes our projects possible ...") Andreas JN466 00:09, 6 December 2022 (UTC)

I think you may be over-interpreting here, Andreas. The WMF *does* play a key role in preserving Wikipedia; if it disappeared tomorrow, so would Wikipedia. I generally agree with your views on the fundraising campaigns, but you might not have a lot of success with this argument. The word is not a form of crisis language or implying that Wikipedia is in danger. —Ganesha811 (talk) 00:14, 6 December 2022 (UTC)

@Ganesha811, see Wikipedia:Fundraising/2022_banners#"Together, let's preserve this special space" (current banner wording). The WMF say thay have removed the sentence. Andreas JN466 14:42, 6 December 2022 (UTC)

Fair enough! I still think the word is innocuous, but I'm glad they are being responsive to editor opinion in any case. —Ganesha811 (talk) 16:43, 6 December 2022 (UTC)

Paywalled 1970s biochemical articles used widely as references

Tracked in Phabricator
Task T198456

When looking at a recent AfD on an article about a company, I saw an unusual use of a 1975 biochemistry journal article as a reference. Looking further, I could see it also used as a reference for articles on a document destruction firm, on athletic records, to support a summary of later writings by Cornell West, etc. and could also see similar use of other journal articles.

I wrote a program to generate tables summarising their uses across articles (at User:AllyD/BiochemReferences). I was hoping to identify a pattern, but, while this does show them being used to reference all manner of topics (from entertainer's biographies, to town patron saints, to geo-political disputes), and shows they are being added to more articles recently, I am not really seeing why these references are appearing. (I had expected to find them to be "Potemkin village" references to bolster new articles, but that doesn't seem to be the pattern.)

I am not sure if this is an appropriate forum, and what can and should be done about this, but felt it could be useful to bring this to others' attention. I have removed some of the more infeasible uses but while the likelihood of some others being relevant is vanishingly-small, have left them in place becauss of being unable to confirm or deny the relevance of the paywalled articles. AllyD (talk) 10:23, 26 November 2022 (UTC)

I notice they both have unusually low PMIDs of 1 and 3. Maybe that has something to do with it? PMID 3 has the same metadata as appears in the references, but a totally unrelated abstract. the wub "?!" 10:34, 26 November 2022 (UTC)

Perhaps someone has used the reference generator and put a 1 or 3 in the pmid box by accident. Graeme Bartlett (talk) 10:46, 26 November 2022 (UTC)

Yes I think that must be it. PMID 2 shows up in a bunch of unexpected places too [8]. the wub "?!" 10:51, 26 November 2022 (UTC)

Thanks for pointing that out. I had been bogged down in the typical human pattern search thing of looking for intent rather than accident. It could be worth a bit of validation in reference generation tools, for example insisting that a single-digit PMID be input as "01" etc.? AllyD (talk) 11:56, 26 November 2022 (UTC)

I can offer a similar example of irrelevant/inappropriate IDs. While editing an article, I researched several databases to collect references on the subject and then proceeded to drill down to the individual references. One of the database aggregators was EBSCOhost, which supplies its own identifier known as an "accession number" to quickly retrieve the particular document. However, the identifier is not present in all EBSCOhost records. It seems that when an established identifier (such as PMID) is already assigned to a document, it is used by EBSCOhost as a defacto accession number. Unfortunately, there doesn't seem to be any routine to avoid an EBSCOhost-assigned accession number being the same as a PMID-assigned number. If one adds the (duplicate) identifier to a reference (as an EBSCOhost id), the wrong document may pop up. One example from EBSCOhost is below. When querying the identifier 14297630 the result is 2 entries:

• "The Kenotic Convict: A Divertissement on Contemporary Contemplative Spirituality in its Social Context." Merton Annual. Nov2003, Vol. 16, p152-171. 20p. EBSCOhost accession number 14297630
• "SENSITIVITY OF VIRUSES TO BILE SALTS AS TESTED IN TISSUE CULTURE SYSTEMS." The Indian journal of medical research [Indian J Med Res] 1965 Apr; Vol. 53, pp. 304-8. PMID 14297630 (defacto EBSCOhost accession number)

Emphasis added for clarity. 50.75.226.250 (talk) 19:00, 26 November 2022 (UTC)

Yep, this is mostly coming from VisualEditor users. I suspect people either are trying to reuse existing reference, or think they need to manually assign a number to each reference. In any case, all you need to do is open in a page in VisualEditor, click "Cite", then type in "1" and out comes a citation to "Formate assay in body fluids: application in methanol poisoning". Edit filter 979 (hist · log) tracks these. Perhaps it should tag or warn, also. It's very easy to mistakenly assume bad faith when someone is doing this; if you don't know about this VisualEditor "feature", it looks like https://xkcd.com/906/. Suffusion of Yellow (talk) 21:48, 26 November 2022 (UTC)

• Question… does this mean the visual editor is generating inaccurate citations? Blueboar (talk) 22:27, 26 November 2022 (UTC)

  Yes and no. It's not spitting out random citations when used correctly, AFAIK. But it's a very easy mistake to make. Suppose you are new to Wikipedia and want to reuse citation number 2. The natural thing is click "cite" and type in the number "2". The correct process is to click "cite", then click "re-use", then type in "2" but some people miss the middle step. Suffusion of Yellow (talk) 22:39, 26 November 2022 (UTC)

Agree with comment above. I never use VE, but I doubt the tool is solely to blame. The editor is supposed to check the generated citation before it is committed. If it looks foreign to the article's subject matter, why should one agree to insert it? Unless the source has been examined manually, in detail (as it should happen for every citation) and passes requirements for context and relevance, despite the strange title. 172.254.222.178 (talk) 22:51, 26 November 2022 (UTC)

Or, never ever ever use numbers as ref names. Use words. Use phrases. Fix VE so that it doesn't offer people the option to use numbers as ref names. DS (talk) 03:20, 28 November 2022 (UTC)

• Thanks to all for the above. It is interesting that the problematic edits are being captured already in an Edit Filter; it has long been the case in IT that Audit Lists are useful but a poor cousin to prevention at point-of-entry: in this case, while the faulty Biochem refs could be erased, we can't retrieve the original editor's intended ref.
• I am not familiar with VE so tried it on a dummy article today. This confirmed all that is said above about the vulnerability of the single input field in VE's Add a citation / Automatic tab to defaulting to these Biochem articles, and also that it does present the retrieval to the editor for approval - but clearly that isn't enough in practice. (I have placed some screenshots here.)
• In terms of the overall remit of the Visual Editor, these show a further issue: Someone using VE and opting to reuse ":1" might have reasonable expectations that the existing reference appearing as [1] will be reused, but it won't: it will be the one which has been allocated :1 as a ref name (as per my screenshot where the BBC ref [1] is not used but the Journal article is).
• In my opinion, we need to request an enhancement to safeguard against these situations: I am not sure where is the best place to workshop this and raise a consensus request for change? Based on everything above, I am inclined to favour DragonflySixtyseven's suggestion that pure numeric ref names should be replaced by alphanumeric. AllyD (talk) 08:33, 28 November 2022 (UTC)

There seem to be some good user behaviour observations here, that someone should capture in phabricator, if you want them to at least be remembered beyond the date range of this page and have any hope of them ever being fixed. It also sounds like a perfect little small achievable project for the community tech wishlist, so note it down, so you can propose it during the wishlist as well. —TheDJ (talk • contribs) 10:37, 28 November 2022 (UTC)

I've been holding off writing a phabricator task until there's a better idea of what we actually want. Blocking the automatic reference generator from working on PMIDs below a certain threshold? A design change to make the "Re-use" tab more obvious?

While VE's lack of properly named references (phab:T92432 / phab:T245199) is certainly a real problem that should be fixed, I doubt it's relevant to the situation here. Without looking at the wikitext users won't even see the autogenerated :0 :1 style names. What seems more likely is naive users are just taking the number they see in the and plugging that into the automatic reference generator. the wub "?!" 10:46, 28 November 2022 (UTC)

Well one idea might be to pop up a 'did you mean?' whenever a number is entered matching a reference number/name in the current version. It could give you a button to 'autocorrect' to re-using an existing reference. —TheDJ (talk • contribs) 11:12, 28 November 2022 (UTC)

@The wub, I'd love to have you file a Phab ticket (tag it with citoid, which should add the visualeditor tag automatically when you submit the ticket), but I suggest that you file the ticket about the problem, and leave the solution out of it. It might be the kind of thing that needs to be addressed from several different perspectives, and having a clear problem statement without suggesting solutions could help the team evaluate the problem without any preconceived notions. Whatamidoing (WMF) (talk) 01:50, 30 November 2022 (UTC)

@Whatamidoing (WMF): Isn't this just phab:T198456? Which was tagged as "low" priority over four years ago. Suffusion of Yellow (talk) 01:58, 30 November 2022 (UTC)

Thanks for identifying that "Users entering small numbers into automatic citations, expecting it to re-use an existing citation" change which was proposed in 2018 to alleviate this problem. I have added a comment there to support the change. To my mind, preventing use of a single-digit number at VE's Add a citation | Automatic input box would not materially detriment users (any who really do want to reference one of these 1975 Biochemical articles can enter it manually) and would prevent further inadvertent corruption of article references. AllyD (talk) 12:33, 30 November 2022 (UTC)

I've attached a naïve first take on this as a patch (864858) on that phab ticket. It just makes it so that pressing “generate” after you’ve entered any number less than 1000 in the auto input will switch to the reuse panel and fill in its search field with that number. The editor would then be responsible for clicking the item in the search results that corresponds to the one they meant. DLynch (WMF) (talk) 23:06, 5 December 2022 (UTC)

See WP:Edit filter noticeboard#Set filter 979 to warn? Suffusion of Yellow (talk) 23:49, 29 November 2022 (UTC)

• Following from the above both I and EdJohnston have each been editing some of the erroneous biochem references. When I started doing this with the hope that by looking at the revision prior to the biochem reference being introduced - in particular looking at its references [1] and [:1] - I might discern and be able to replace with the editor's actual intentions. In practice I didn't succeed, and ended up just excising the biochem refs. The current situation is one of Inattentional blindness (the gorilla experiment): after keying 1 the biochemical article info does preview at the bottom of the screen, but it is so far outside expectation (e.g. if the editor is working on an article about an entertainer) as to not be noticed. Both Pcoombe and I have recently added comments to phab T198456, which I do feel should be progressed to staunch this flow of dodgy refs. AllyD (talk) 10:56, 5 December 2022 (UTC)

  The match would be to whatever was in the revision when the person added the wrong ref. Whatamidoing (WMF) (talk) 19:15, 5 December 2022 (UTC)

Join the Movement Charter Regional Conversation Hours

You can find this message translated into additional languages on Meta-wiki.

More languages • Please help translate to other languages.

Hi all,

As most of you are aware, the Movement Charter Drafting Committee (MCDC) is currently collecting community feedback about three draft sections of the Movement Charter: Preamble, Values & Principles, and Roles & Responsibilities (intentions statement).

How can you participate and share your feedback?

The MCDC is looking forward to receiving all types of feedback in different languages from the community members across the Movement and Affiliates. You can participate in the following ways:

• Attend the community conversation hours with MCDC members. Details about the regional community conversation hours are published here
• Fill out a survey (optional and anonymous)
• Share your thoughts and feedback on the Meta talk page
• Share your thoughts and feedback on the MS Forum:
  • Preamble
  • Values & Principles
  • Roles & Responsibilities (statement of intent)
• Send an email to: movementcharterwikimedia.org if you have other feedback to the MCDC.

Community consultation hour for the United States and Canada will take place on Monday, 5 December 2022 on Zoom. You can check out more times here. The conversations will not be recorded, except for the section where participants are invited to share what they discussed in the breakout rooms. We will take notes and produce a summary report afterwards.

If you want to learn more about the Movement Charter, its goals, why it matters and how it impacts your community, please watch the recording of the “Ask Me Anything about Movement Charter” sessions which took place earlier in November 2022.

Thank you for your participation.

On behalf of the Movement Charter Drafting Committee,

MNadzikiewicz (WMF) (talk) 13:52, 1 December 2022 (UTC)

• Hey all, just wanted to follow up about this conversation which begins in just under 9 hours (at 9 PM EST/6 PM PST). Although it's regionally schedule/titled "USA & Canada", all English Wikipedians (and English speakers, for that matter) are welcome regardless of location. Here is the Zoom link: https://wikimedia.zoom.us/j/88407930105?pwd=VVpBaFpKRThtd29mVzk5RCtOa1FaUT09 . Hope to see some of you there! Xeno (WMF) (talk) 17:04, 5 December 2022 (UTC)

  I don't have time to join the zoom but I posted my thoughts on meta. Levivich (talk) 19:12, 5 December 2022 (UTC)

  I appreciate you providing your thoughts in that format! There's also a survey available. Xeno (WMF) (talk) 19:20, 5 December 2022 (UTC)

  Thanks -- as it turns out, that page is how I found the place on meta to post a public comment. So, to say something positive: it was easy to find a place to complain voice my opinion on the subject, and I appreciate that, as well as the reminder about the zoom call, which reminded me to complain voice my opinion on the subject. :-) Levivich (talk) 19:24, 5 December 2022 (UTC)

  Ha =) Just under an hour to go for this conversation. Of course anyone else who isn't able to join should feel free to leave comments on Meta-Wiki also! Xeno (WMF) (talk) 01:02, 6 December 2022 (UTC)

Sound logo vote + asking for permission for a banner?

Hi everyone,

Voting in the Wikimedia sound logo contest has started. Crowds, pages turning, drums, chimes, vocals, and the sound of keyboards typing. Wikimedia is alive with sound, music, and everything in between. From December 6 to 19, 2022, please play a part and help identify the Sound of All Human Knowledge. Voting is open until 19 December, 23:59 UTC. Learn more on Diff.

On a related note, we would love to activate banners for the voting component of the sound logo project to ensure Wikipedians are notified. Would this be ok? We will be running banners on Commons and Meta-Wiki with fewer impressions than the submission phase. This has already been approved through the appropriate community means. Would extending the "vote now" banner to English Wikipedia be acceptable as well? Thank you. MPourzaki (WMF) (talk) 16:32, 6 December 2022 (UTC)

@MPourzaki (WMF) who do you want to target with such a banner (asusming this is a CN Banner). Would a watchlist banner suffice? — xaosflux ^Talk 16:36, 6 December 2022 (UTC)

Hi @Xaosflux, yes this would be for CNBs so online editors are also notified of the sound logo vote happening. We will be running them on Meta and Commons already starting December 8. I didn't know about watchlist banners. MPourzaki (WMF) (talk) 17:41, 6 December 2022 (UTC)

@MPourzaki (WMF) it's probably fine, can you give us a link to one for example? — xaosflux ^Talk 17:54, 6 December 2022 (UTC)

Hey @Xaosflux, here is the example. MPourzaki (WMF) (talk) 20:23, 6 December 2022 (UTC)

Thank you. — xaosflux ^Talk 21:21, 6 December 2022 (UTC)

• The banner campaign (for logged in users) sampled above looks fine to me, anyone have an objection? — xaosflux ^Talk 21:21, 6 December 2022 (UTC)

  Watchlist message draft and request here. 🐶 EpicPupper ^{(he/him | talk)} 05:08, 7 December 2022 (UTC)

  Noted there, if this is going to be a CN, we don't also need a WLN. CN will attract more editors. Looks like this is configured for logged in users only. — xaosflux ^Talk 14:36, 7 December 2022 (UTC)

  @Xaosflux: The banner impressions are limited to 1 because there's a lot of ongoing discussion involving the WMF and banners, so I'm not actually sure that a banner would attract more editors. I believe our preference would be to use both. Ed Erhart (WMF) (talk) 17:43, 7 December 2022 (UTC)

  @Ed Erhart (WMF) well, as noted at meta:CentralNotice/Request/Wikimedia Sound Logo contest 2#Central Notice admin comments - apparently this was launched to ~500 projects, including here already. So doesn't seem like there is anything much for us to do. I don't think it is much of an issue, but this was a very bungled communication compared to what was deployed. — xaosflux ^Talk 00:41, 8 December 2022 (UTC)

  @Xaosflux: Mmmm. Thankfully that's been fixed. I think you can see from that Meta discussion that our intentions changed starting on 2 December, which is what led to the impression limits given by me above and (eventually) this village pump post. Unfortunately, there was a miscommunication along the way. Ed Erhart (WMF) (talk) 05:07, 8 December 2022 (UTC)

  Thank you. Assuming the community has no issues, I can add enwiki back on there when ready. Let's see if anyone has anything to say. — xaosflux ^Talk 11:10, 8 December 2022 (UTC)

• Advertised this to WP:VPR, holding to see if there is feedback. This sounds reasonable on its face, will default to "OK" unless there are some objections below. — xaosflux ^Talk 11:15, 8 December 2022 (UTC)

  I think this should be advertised, and I wouldn't mind both CN and WLN. If I understand what this is, this will become a replacement for the logo if pictures can't be shown. As such choosing the one that everyone likes the best is important. Dreamy Jazz ^{talk to me | my contributions} 12:17, 8 December 2022 (UTC)

  No objection to CN, but don't think we need both if the CN doesn't fall under one of the 3/4 categories most editors turn off. If it does (which may well be the case), I'm neutral on also having the WLN. Nosebagbear (talk) 13:24, 8 December 2022 (UTC)

  @Nosebagbear the banner type is 'special', so prob not turned off. The impression diet is very low, so it should only appear 1 or 2 times per device, over 2 weeks. — xaosflux ^Talk 13:32, 8 December 2022 (UTC)

  Wonderful, then yes, I'd back the +CN, -WLN position Nosebagbear (talk) 14:28, 8 December 2022 (UTC)

---

• I've enabled the CN here as there haven't been any objections to it. — xaosflux ^Talk 10:30, 9 December 2022 (UTC)

ISA Tool Workshop - Dec 12th

Greetings.

We released a few weeks ago a new version of the ISA tool on the Test Server. The new version propose semi-automatic tagging of images, which seeks to assist community in identifying and labelling depicts statements for Commons files.

Now we are looking for feedback with regards to

• the look, feel, and usability of the tool with the machine vision / metadata-to-concept enhancements in place.
• and most importantly, the quality of the recommandations provided.

This is why we are organizing an online workshop on Monday 12th of December, 16h-17h30 UTC+1 (on zoom). We will be very happy if you find the time to join us.

More information and sign-up : https://commons.wikimedia.org/wiki/Commons:ISA_Tool/Image_to_Concept#ISA_Workshop_-_Dec_12th

Thank you in advanceǃ

Anthere (talk) 09:05, 8 December 2022 (UTC)