Wikipedia talk:Wikipedia Signpost/Newsroom/Archive 36
 | This is an archive of past discussions about Wikipedia:Wikipedia Signpost. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
| Archive 30 | ← | Archive 34 | Archive 35 | Archive 36 | Archive 37 | Archive 38 | → | Archive 40 |
Opinion vs op-ed
Okay, so here is the query:
SELECT page_namespace AS ns, page_title AS title FROM page WHERE ( page_namespace = "4" AND page_title LIKE "Wikipedia_Signpost%Opinion" ) OR ( page_namespace = "4" AND page_title LIKE "Wikipedia_Signpost%/Op%d" ) ORDER BY ns ASC, page_title ASC;
Here is the table of every article we've run that is titled "Opinion", "Op-ed" or "Op-Ed" (the latter of which we seem to have started doing in 2019 and is inconsistent with old articles, but whatever).
So now here's my deal. What is the difference? I have put some considerable effort into figuring out what the deal is... I cannot really find any sources online that differentiate between an "op-ed" and an "opinion piece". Our own article says that an op-ed is a type of opinion piece that "expresses the opinion of an author or entity with no affiliation with the publication's editorial board".
For the Signpost, though: the most recent /Op-Ed is from 2023-04-26, written by Lambiam. Meanwhile, 2023-08-15/Opinion is by Josve05a, 2023-08-01/Opinion is by Lemonaka, and 2023-04-26/Opinion is by Smallbones (the only one written by a listed staff member). Even going back a few years: 2018-04-26/Opinion (Guideline for Organization Notability revised, by Jytdog) seems to be the same basic tier of thing as 2018-04-26/Op-ed (World War II Myth-making and Wikipedia, by K.e.coffman). There seems to be no real distinction between the two, so I think that we should stop having different categories for them. If the issue is just that publishing two opinion pieces in the same issue requires us to name one of them Opinion and the other Op-Ed at random, well, this is dumb and I guess we should fix it, but not by having two separate synonym names for the same department. jp×g 21:37, 19 August 2023 (UTC)
- Seems like a reasonable enough approach to me. Clovermoss🍀 (talk) 22:19, 19 August 2023 (UTC)
- Yeah, I was confused as well when submitting the latest piece. Jonatan Svensson Glad (talk) 22:25, 19 August 2023 (UTC)
Boom goes the 2023-08-15
Wikipedia talk:Wikipedia Signpost/Single/2023-08-15
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Traffic report | Come on in, and pull yourself up a chair | 393 | 580 | 648 | 705 | 738 | 750 | 772 |
| Tips and tricks | How to find images for your articles, check their copyright, upload them, and restore them | 589 | 857 | 922 | 972 | 1087 | 1140 | 1223 |
| Special report | Thirteen years later, why are most administrators still from 2005? | 21472 | 21929 | 22177 | 22355 | 22436 | 22490 | 22623 |
| Serendipity | Why I stopped taking photographs almost altogether | 27550 | 27893 | 28027 | 28126 | 28194 | 28232 | 28286 |
| Opinion | Copyright trolls, or the last beautiful free souls on this planet? | 683 | 860 | 931 | 1009 | 1056 | 1071 | 1117 |
| News and notes | Dude, Where's My Donations? Wikimedia Foundation announces another million in grants for non-Wikimedia-related projects | 14440 | 15296 | 15653 | 16049 | 16280 | 16391 | 16557 |
| In the media | An accusation of bias from Brazil, a lawsuit from Portugal, plagiarism from Florida | 2529 | 2819 | 2934 | 3000 | 3049 | 3073 | 3121 |
| In focus | 2023 Good Article Nomination drive is underway: get your barnstars here! | 347 | 492 | 540 | 585 | 606 | 620 | 663 |
| Humour | Arbitration Committee to accept case against Right Honorable Frimbley Cantingham, 15th Viscount Bellington-upon-Porkshire | 529 | 749 | 834 | 898 | 938 | 953 | 1003 |
| Featured content | Barbenheimer confirmed | 529 | 723 | 811 | 879 | 911 | 933 | 978 |
| Cobwebs | Getting serious about writing | 370 | 502 | 555 | 603 | 622 | 631 | 663 |
It will take a day or so for these to be anything (they are all -1 right now because the pageviews aren't available yet). jp×g 19:46, 15 August 2023 (UTC)
- Did not think my opinion piece would be published without quite a bit of CE, but perhaps not :) Jonatan Svensson Glad (talk) 20:26, 15 August 2023 (UTC)
- Kind of disappointed my tips and tricks has average readership, but there ye go. Probably not hook-y enough. Adam Cuerden (talk)Has about 8.5% of all FPs. 05:26, 21 August 2023 (UTC)
- By the way, I've added links to the single talk page of the current issue on the Newsroom, so it should be easier to get to. jp×g 02:08, 16 August 2023 (UTC)
19-16 Recent research
As usual, we are preparing this regular survey on recent academic research about Wikipedia, doubling as the Wikimedia Research Newsletter (now in its thirteenth year). Help is welcome to review or summarize the many interesting items listed here, as are suggestions of other new research papers that haven't been covered yet. Regards, HaeB (talk) 13:11, 24 August 2023 (UTC)
Frontend that works properly
I guess I have to go buy a VPS because Toolforge's ingress doesn't support external domain names. Boy oh boy!!! I am holding off on getting pissed about this until I see how much it costs. jp×g 03:53, 27 August 2023 (UTC)
Unusual ban
BrownHairedGirl, an editor since 2006 with 2.9 million edits (#2 on the list of Wikipedians by number of edits) was banned a few hours ago by Arbcom. ☆ Bri (talk) 04:13, 27 August 2023 (UTC)
- This should be an arb report, there was a big-ass case about this (iirc "SMALLCAT dispute"). jp×g 19:19, 27 August 2023 (UTC)
- I know her. Does a lot of theatre work. Nice person. Weird she's banned Adam Cuerden (talk)Has about 8.5% of all FPs. 19:41, 27 August 2023 (UTC)
19-16 In the media
Another politics thing? Oh well. Last issue, I did my best to cover things in as balanced a way as I could, and we still got cooked in the comments with what seems to have been fairly justified criticism (or, at the least, criticism nobody had a snappy retort to). I am probably going to do the same this issue. For the record, I have no issue with getting yelled at for publishing stuff that actually serves the public interest or is useful for us to have published, but I think we have finite room to piss people off, and should avoid doing it unless necessary. jp×g 01:02, 28 August 2023 (UTC)
19-16 News and notes
Test of new section-transclusion links from newsroom. jp×g 06:02, 28 August 2023 (UTC)
In the media - UPE sockfarm
Concerning the item I added recently to In the media. It apparently is showcasing the Zavia345 sockfarm. The connection is pretty obvious based on the interviewee's name and one of the socks. I don't know if we ought to add this to the piece. Additionally, I suspect that "Z News Service" itself is a PR outlet hoping to ride on the coattails of and/or be mistaken for Zee News. ☆ Bri (talk) 23:58, 29 August 2023 (UTC)
19-16 Arbitration report
This is extremely unlikely to be ready by publication time, so I think it should be left for next issue. jp×g 20:59, 30 August 2023 (UTC)
Boom goes the 08-31
Wikipedia talk:Wikipedia Signpost/Single/2023-08-31
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Traffic report | Raise your drinking glass, here's to yesterday | 410 | 574 | 645 | 706 | 732 | 751 | 780 |
| Recent research | The five barriers that impede "stitching" collaboration between Commons and Wikipedia | 781 | 977 | 1046 | 1160 | 1185 | 1202 | 1235 |
| News and notes | You like RecentChanges? | 1051 | 1273 | 1360 | 1460 | 1497 | 1520 | 1557 |
| In the media | Taking it sleazy | 912 | 1154 | 1267 | 1381 | 1416 | 1439 | 1487 |
| Humour | The Dehumourification Plan | 663 | 845 | 927 | 1013 | 1057 | 1090 | 1149 |
| From the editor | Beta version of signpost.news now online | 873 | 1173 | 1294 | 1363 | 1383 | 1407 | 1443 |
| Draftspace | Bad Jokes and Other Draftspace Novelties | 506 | 679 | 744 | 814 | 847 | 870 | 912 |
Won't be filled out till I get back. Whoever posts the new issue on le epic social media, go ahead and give the signpost.news urls a shot -- they'll preview way better and I expect they will get Moar Clix (maybe we could do half with .news and half with enwp? and wee what happens?) jp×g 00:55, 31 August 2023 (UTC)
- (note that these are going to be artificially low for the next four days since I'm not going to be online to run the script) jp×g 03:57, 1 September 2023 (UTC)
Fixable redlink in Humour
Can the redlink at the end of Wikipedia:Wikipedia Signpost/2023-08-31/Draftspace be fixed? It uses tag references which I don't know how to fix. ☆ Bri (talk) 15:41, 31 August 2023 (UTC)
- Got it. Whoopsies!!! jp×g 18:03, 31 August 2023 (UTC)
Fixes
I have completed the most recent set of things that needed to get done, and so my next efforts in organization and technology are yet to be determined. @HaeB: I recall that you had a list somewhere of stuff that should be addressed sooner rather than later: what do you think would be the best course of action here? jp×g 17:46, 6 September 2023 (UTC)
Wikipedia search-by-vibes through millions of pages offline
https://www.leebutterman.com/2023/06/01/offline-realtime-embedding-search.html
https://news.ycombinator.com/item?id=37355487
Would appreciate if someone more technically-minded than me could have a look at this. Is it any good? Worth writing about? Andreas JN466 10:41, 2 September 2023 (UTC)
- @JPxG, Josve05a: The upcoming News and notes is already quite substantial, so if we want to run something like the text below, I'd be in favor of making it a Technology report and giving it a standalone page.
- If so, I think it would be worth including the two links (leebutterman.com and the Hacker News discussion) for reference. Andreas JN466 13:49, 7 September 2023 (UTC)
- Could be a thing for News and notes or Technology report perhaps. Seems interesting enough!
"A groundbreaking browser-based search engine for Wikipedia is making waves. This innovative tool allows users to search the vast Wikipedia database in real-time, entirely offline on mobile devices. Developed by Lee Butterman, this search engine leverages sentence embeddings to provide rapid search results.
What sets this search engine apart is its remarkable speed and efficiency. It can deliver search results every 10 milliseconds on a mobile device, with search results updating as the database is scanned sequentially. The engine's performance metrics are equally impressive, with initial results appearing in just 21 milliseconds, and 70% of the final results available in 116 milliseconds.
To achieve this level of efficiency, the search engine employs product quantization to compress embeddings and ONNX models for computation. The database, which includes two million Wikipedia pages and their titles, is compact, taking up only about 100MB. In addition, Arrow is used instead of JSON for optimal memory and disk storage.
The search engine's unique approach has garnered both praise and constructive criticism. Users have found it to be a valuable tool, with its ability to search in real-time offline environments being particularly noteworthy. However, some have noted that search results may not always align with their expectations.
Despite any potential room for improvement, this browser-based Wikipedia search engine represents an exciting development in the world of information retrieval. Its ability to deliver rapid, offline search results opens up new possibilities for accessing knowledge on the go."
— Jonatan Svensson Glad (talk) 11:16, 2 September 2023 (UTC)
AI and Arbitration Report
Are we sure that we should be using AI to summarize ArbCom cases, which are some of the most controversial parts of Wikipedia? QuicoleJR (talk) 15:47, 7 September 2023 (UTC)
- Pinging JPxG who wrote the prompts for the Arbitration Report in question. QuicoleJR (talk) 15:54, 7 September 2023 (UTC)
- I wrote the Arbitration Reports for all of 2020 (I think) and a few since then; these LLM summaries of the on-wiki discussions look reasonable to my eye. However, I'm not going to read the Wikipediocracy board, so no telling how accurate that one is. ☆ Bri (talk) 16:05, 7 September 2023 (UTC)
- Well, if you say that it's fine, it's probably fine. QuicoleJR (talk) 16:13, 7 September 2023 (UTC)
- I'd only go so far as to say that the summary looks reasonable in terms of summarizing the surface level of the discussion. Even with an accurate summary, we are missing out on the kind of analysis and context that I think a really good Arbitration Report would provide. Does it set a new standard (we're not supposed to say "precedent" for arbitration), or reflect a new community consensus? The AI can't speak to that, but that's probably more important than the immediate decision for a few editors involved in a specific case.
- One possible positive is this. We are using an AI to summarize a complex discussion. In my own ARs I'd do this by picking out some quotations that I thought illustrated interesting parts of the conversation. But that hazards introducing my own bias. The machine-mediated summary maybe levels out that bias and tends to present more of the conversation in its conclusion. But I don't really know yet if a) that is the case and b) if it is actually better for the reader. Perhaps having a biased reviewer point out something they as a human found interesting is of value after all. ☆ Bri (talk) 16:55, 7 September 2023 (UTC)
- Well, if you say that it's fine, it's probably fine. QuicoleJR (talk) 16:13, 7 September 2023 (UTC)
- Writers are fact-checked whether they are animal, vegetable or mineral. We will have to see how this mineral has done -- it might have woofed completely for all I know. Obviously the material is sensitive and requires verification either way. jp×g 00:35, 8 September 2023 (UTC)
- First fact check: The machine generated summary left out the reminder to XfD editors in both the proposed decision and the final decision. I'll highlight that but probably won't have a chance to fix it. ☆ Bri (talk) 01:16, 8 September 2023 (UTC)
WikiCup 2023
it's out jp×g 04:35, 8 September 2023 (UTC)
AI-generated "empty blather" is just fine, apparently
This is starting to grow beyond a mere mention at News and notes. Maybe other editors want to take a look and chime in, either in the piece I started, or in the ongoing discussion at ANI. ☆ Bri (talk) 17:46, 4 September 2023 (UTC)
- As a non-native, Ive sometimes used AI to reformat some responses Ive written in order to be more concise and clear. Wasn’t aware that using a tool to help writing in foreign languages was forbidden, even in cases where care is given to ensure the produces text is good… Jonatan Svensson Glad (talk) 21:29, 4 September 2023 (UTC)
- The responses on the ANI thread are interesting. The problem isn't the language models per se. Judicious use, that actually facilitates discussion, is probably just fine. My concern here is that in the case of the guy paying to create mayhem, it's being weaponized as a way to randomize the community. ☆ Bri (talk) 23:42, 4 September 2023 (UTC)
- Well, I would love to write a fiery op-ed about how it was imperative for us to get off our asses and approve WP:LLMP, but I am the editor-in-chief, and I wrote WP:LLMP, so it doesn't seem prudent. jp×g 00:27, 5 September 2023 (UTC)
- For reference, I just asked ChatGPT to help me write an apology in response to the accusation that started the ANI thread:
- Me
- Please write an effective apology to the following accusation made on Wikipedia: "Your main contribution to Wikipedia is to get an article on Parvati Kurakula, a non-notable writer as decided by community consensus following a discussion that I started at Wikipedia:Articles for deletion/Parvati kurakula. You have been suspected of conflict of interest multiple times, however you seem to ignore this every time you are asked about this. However, the main issue here is the use of ChatGPT. The Kurakula article was written in a promotional nature, mainly due the use of AI - this was noted by DreamRimmer in the discussion. After you reached out to me on my talk page, I have tried to communicate with you, to no avail. You have confessed on the linked discussions that you are using ChatGPT to facilitate more effective communication. You may not know an adequate level of English to proactively create useful contributions and communicate effectively, given that my efforts to explain notability and redirect you to Hindi Wikipedia have been a waste of time."
- ChatGPT
- Subject: Apology and Clarification Regarding My Wikipedia Contributions
- Dear [Accuser's Name],
- I hope this message finds you well. I want to address the concerns you have raised regarding my contributions to Wikipedia, particularly my involvement in the creation and maintenance of the article on Parvati Kurakula. I am writing to apologize for any misunderstandings or frustrations that may have arisen as a result of my actions.
- Firstly, I want to acknowledge your dedication to maintaining the quality and notability of Wikipedia articles. Your commitment to the platform is evident, and I respect your efforts to ensure that Wikipedia remains a valuable resource for all users.
- Regarding the Parvati Kurakula article, I understand that there was a discussion on Wikipedia:Articles for deletion, which ultimately resulted in the consensus that Ms. Kurakula's notability was not established according to Wikipedia's guidelines. I accept this decision, and I apologize for any inconvenience my previous contributions may have caused in this regard.
- I must clarify that my use of ChatGPT was not intended to create promotional or non-notable content. It was a genuine attempt to improve my ability to communicate effectively on Wikipedia, especially when English is not my first language. I fully appreciate the importance of maintaining a neutral and encyclopedic tone in all Wikipedia articles, and I did not intend to deviate from these standards.
- I also want to apologize for any perceived disregard of your concerns regarding conflict of interest. I understand that such suspicions can compromise the integrity of Wikipedia, and I assure you that I have always taken such allegations seriously. Moving forward, I will be more diligent in addressing any concerns regarding conflict of interest promptly and transparently.
- Furthermore, I am open to any constructive feedback and guidance you can provide to help me improve my contributions on Wikipedia. Your expertise and dedication to the platform are valuable resources, and I am eager to learn from your experience.
- In conclusion, I want to express my sincere apology for any shortcomings in my past contributions and any frustration they may have caused. I remain committed to making positive and valuable contributions to Wikipedia, and I hope we can work together to ensure that the platform continues to thrive.
- Thank you for your time and understanding. I look forward to a more collaborative and productive engagement on Wikipedia.
- Sincerely,
- [Your Wikipedia Username] Andreas JN466 08:49, 7 September 2023 (UTC)
- Jeesh, compare to [1] Here's an example:
I must clarify that my use of ChatGPT was not intended to create promotional or non-notable content. It was a genuine attempt to improve my ability to communicate effectively on Wikipedia, especially when English is not my first language. I fully appreciate the importance of maintaining a neutral and encyclopedic tone in all Wikipedia articles, and I did not intend to deviate from these standards.
— ChatGPT with Andreas' promptI understand your concerns about my use of ChatGPT as a communication tool. I want to clarify that my intention in utilizing this tool is to facilitate more effective communication and content creation. I find it helpful in generating coherent responses. However, I wholeheartedly respect your recommendation and understand the importance of ensuring that my contributions align with Wikipedia standards, regardless of the communication method used. I am dedicated to working on this aspect and will strive to communicate more effectively.
— COI editor's response at ANI- A takeaway to me from both of these incidents is how fragile our editor-centric, labor-intensive processes are if they are confronted by a bunch of bad-faith editors using technology to cheaply generate superficially communicative text we have to wade through. It's a real attack on the immune system of the WP organism. Is this worth an op-ed? ☆ Bri (talk) 13:56, 7 September 2023 (UTC)
- To be honest, editors have been manually generating empty text for a long time, swamping conversations. Trying to deal with this has always been challenging, given that the community tries to be encouraging to even those whose communication skills need improvement, in part as this can coincide with giving under-represented views a better chance at being heard. Also, sometimes what one person considers to be off-topic musings is what another person sees as valuable insight into the topic's issues. isaacl (talk) 17:20, 7 September 2023 (UTC)
- Our current WP:Wall of text problem, though significant, is somewhat self-limiting in the sense that is usually caused by zealots with time to kill for no compensation, and only occasionally by compensated individuals. But now, with the LLM tool, anyone who can push a button can generate arbitrarily large amounts of it, for tangible reward, and it doesn't even require knowledge of English as we are finding out. Rather than "communication skills [that] need improvement" you are conversing with a machine that can't hear you. An LLM fed orders to generate garbage won't be improved by engaging in conversation. It's an utter waste of time. This is about as far as you can get from a good-faith editor or even someone who can be led to future positive contributions. And it's not fair for the remaining good-faith editors who have to live in such an environment. It's just pollution of the cognitive-intellectual commons, and at some point, people will have enough of it. ☆ Bri (talk) 01:12, 8 September 2023 (UTC)
- Sure, it's an existing problem that can increase in magnitude. To deal with it, the community will have to shut down off-topic, verbose discussion more quickly. This will make it less welcoming to good-faith editors who aren't good communicators. The advent of better text generating programs may hasten this, however it's a shift that has been gradually happening already. isaacl (talk) 01:35, 8 September 2023 (UTC)
- @Bri, JPxG, HaeB: Note that there is an ongoing research project on this topic: m:Research:Implications_of_ChatGPT_for_knowledge_integrity_on_Wikipedia
- An interesting aspect HaeB mentioned on the mailing list yesterday is that some chatbots do cite their sources.
- Having played around with them a bit, Bing cited sources quite consistently (in a way that could actually help someone write a sourced article if they know what they are doing) and Google's Bard (which seems more reliant on Wikipedia) indicated the parts of its responses that were copied from Wikipedia.
- Note that Bing's sources were hit and miss. Citing them blindly will lead to errors. Andreas JN466 07:09, 8 September 2023 (UTC)
- Our current WP:Wall of text problem, though significant, is somewhat self-limiting in the sense that is usually caused by zealots with time to kill for no compensation, and only occasionally by compensated individuals. But now, with the LLM tool, anyone who can push a button can generate arbitrarily large amounts of it, for tangible reward, and it doesn't even require knowledge of English as we are finding out. Rather than "communication skills [that] need improvement" you are conversing with a machine that can't hear you. An LLM fed orders to generate garbage won't be improved by engaging in conversation. It's an utter waste of time. This is about as far as you can get from a good-faith editor or even someone who can be led to future positive contributions. And it's not fair for the remaining good-faith editors who have to live in such an environment. It's just pollution of the cognitive-intellectual commons, and at some point, people will have enough of it. ☆ Bri (talk) 01:12, 8 September 2023 (UTC)
- To be honest, editors have been manually generating empty text for a long time, swamping conversations. Trying to deal with this has always been challenging, given that the community tries to be encouraging to even those whose communication skills need improvement, in part as this can coincide with giving under-represented views a better chance at being heard. Also, sometimes what one person considers to be off-topic musings is what another person sees as valuable insight into the topic's issues. isaacl (talk) 17:20, 7 September 2023 (UTC)
- For reference, I just asked ChatGPT to help me write an apology in response to the accusation that started the ANI thread:
Better grants reporting from WMF
This report just came about 7 August and I am seeing it now. According to meta:Special:WhatLinksHere/Community_Resources/Reports/Funding_Report_2022-2023 no one there is no evidence that anyone has yet discussed it, so it is still hot news. What follows are just due-diligence notes for reporting. For the purpose of the Signpost I think it is enough to say that this report came out and that anyone can ask questions. Nothing below is a news-worthy scoop, but only possible context.
The Wikimedia community is increasingly interested in understanding Wikimedia Foundation spending. This report is better than previous. WMF responded requests for clarification on previous reports, and I think they would take constructive criticism to improve particular things on this one. I do not have full thoughts on this but here is what more I would want:
- There is a great budget breakdown of grants here, but all awarded grants have Meta pages and categories. It should be possible to link to categories and see actual grants. As the number of grants in each category is typically about 20, such lists of grants should be comprehensible to humans.
- Beyond categorizing in Meta, undoubtedly the future of grant management is making Wikidata items for each grant. If we did that, then grants would be endlessly queryable and set the standard to all foundations for grant transparency. Perhaps the Wikimedia community should apply for a grant to do that conversion, so that community can set the data model.
- We still have no standard for determining grant "success". This is extremely difficult, but that conversation is not currently in community forums. The WMF probably has internal standards and internal data but I think it is on the community to determine this.
- Down at the bottom are "other grant programs". There is email and more already scrutinizing the Knowledge Equity Fund. Similarly underreported are 3 others on that list. I asked about them at meta:Talk:Campaigns/Organizer Lab, meta:Talk:Education, and meta:Talk:Wikimedia Foundation Community Programs team. The convention that is bent here is that community grants are conventionally documented for community to see and understand, but these are not. I advocate for the Wikimedia community's right to know everything done in the name of the Wikimedia community, so I asked for info.
- WMF is in my opinion still heavy handed in deciding what it wants community to do. For example, WMF greatly favors conferences. I do not think the community ever agreed or consented to establishing conferences as a primary outreach strategy, and typically it is the major funded program in a region. I am not opposed to conferences; I just question what it means that the WMF chose conferences as the default grantmaking strategy, and when community would choose to do otherwise with the same money given the choice. There are other WMF decisions that bound what they will and will not fund, and there is one set of global rules which may not apply equally everywhere.
- Here is what has improved
- There are clearly bottom line numbers now, like total grants awarded per category. Numbers were scant before.
- WMF seems to be increasing grant allocations.
- We now get reports annually. I am grateful. See 2020, 2021, 2022
- The precedent for financial transparency with questions and answers is being normalized
Bluerasberry (talk) 15:12, 6 September 2023 (UTC)
- Thanks, Bluerasberry, included in N&N. Andreas JN466 12:07, 9 September 2023 (UTC)
Misdirected "save Tater Tot" canvassers
I’m not sure how to frame this, but apparently some off-wiki canvasser posted the wrong address for the clueless those unfamiliar with WP procedures to canvas. See Wikipedia talk:Articles for deletion/Tater Tot (cat) (yes, talk space) with over 35 "votes". ☆ Bri (talk) 04:33, 13 September 2023 (UTC)
- Now 45 editors canvassed to the talkpage according to stats, all anon or new, single-edit accounts. ☆ Bri (talk) 15:40, 13 September 2023 (UTC)
- The discussion provides some perspective on how the rest of the world sees WP regulars as process-bound gatekeepers. I'd recommend everyone read "Gatekeeping A Dead Kitten, Are We?" Possibly worth a mention in the issue? ☆ Bri (talk) 17:24, 13 September 2023 (UTC)
- I'm a little reminded of Wikipedia:Why is BFDI not allowed on Wikipedia?. Gråbergs Gråa Sång (talk) 20:00, 13 September 2023 (UTC)
Tomorrow morning
I will get the thing out then. jp×g 10:37, 15 September 2023 (UTC)
Wikipedia talk:Wikipedia Signpost/Single/2023-09-16
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Traffic report | Some of it's magic, some of it's tragic | 419 | 581 | 689 | 746 | 780 | 805 | 833 |
| Serendipity | Yea, though I walk through the valley of the shadow of death, I will fear no paywall, for thou, Wikipedia Library, art with me | 652 | 879 | 1218 | 1346 | 1436 | 1504 | 1690 |
| Obituary | Nosebagbear | 805 | 1039 | 1206 | 1283 | 1321 | 1369 | 1438 |
| News and notes | Wikimedia power sharing – just an advisory role for the volunteer community? | 1493 | 2255 | 2550 | 3091 | 3236 | 3359 | 3468 |
| In the media | "Just flirting", going Dutch and Shapps for the defence? | 634 | 835 | 959 | 1040 | 1070 | 1102 | 1141 |
| Featured content | Catching up | 380 | 508 | 601 | 654 | 674 | 701 | 741 |
| Concept | Strange portal opened by CERN researchers brings Wikipedia articles from "other worlds" | 630 | 875 | 1039 | 1104 | 1130 | 1164 | 1199 |
jp×g 04:43, 16 September 2023 (UTC)
Is tone of RfA changing?
Wikipedia:Requests for adminship/Hey man im josh: editor with barely three years' experience, 100+ unopposed votes in favor of adminship, virtually no questions for the candidate. This is different.
We have had 7 successful candidacies so far in 2023, which – if it does not change by end of year – puts us at a tie with the worst-ever year for RfA, which was 2021.
By the way, I'm keeping an eye on the number of active administrators -- we started the year at 500, it took a big dive in mid February for no reason that I can see, and lately has steady been around 460. ☆ Bri (talk) 16:42, 19 September 2023 (UTC)
Automating the updating of the deadline template
Now that we have settled into the first+third Sunday schedule, it would be great to always update the publication deadline template right after publication to the default next date (it can always be overridden of course). I seem to recall that this used to be routine, but it hasn't been for the last half year or so - I have done this manually for the last couple of issues, and before that Bri and Adam jumped in a couple of times [2]. But it should really be part of the publication process.
@JPxG: or other fluent scripters: Could this be integrated into the publication script? (Same for cleaning up the Newsroom page, where we seem to have completely fallen off the wagon for the last couple of issues).
Regards, HaeB (talk) 07:08, 9 July 2023 (UTC)
- Un-archiving this prematurely archived thread (although it should probably have been posted it at /Technical instead, will post a note there now).
- We've fallen off the wagon again with this after the publication of the most recent issue - Adam had to jump in again after several days, setting the next issue's date to August 13. And the newsroom page still hasn't been reset either. (This isn't a mere formality - some may recall how it caused great grief last year when a story accidentally wasn't included in a new issue because of this.)
- Clearly a busy editor-in-chief could benefit from automation here. And it might also help with sticking to the previously agreed upon first+third Sunday schedule (or was it a deliberate decision to add a third issue for July?).
- Do other folks have thoughts on how one might go about implementing this in the publication script?
- Regards, HaeB (talk) 02:10, 7 August 2023 (UTC)
- I tried something like this a few issues ago, which was to add {{Wikipedia:Wikipedia Signpost/Templates/Single talk}} to the talk for the single-issue page, i.e. Wikipedia talk:Wikipedia Signpost/Single/2023-07-17. It seemed pretty straightforward, but for some reason I could not get it to work when I ran the script. I think that the whole damn thing is probably long overdue for a look-see and some documentation -- it was originally written by Evad37, who is not maintaining it (and isn't very active on the project in general). This is definitely one of the things that should be in the queue for whosoever ends up being the person who does this. jp×g 02:37, 7 August 2023 (UTC)
- I am not at all a JavaScript expert, but perhaps this newly added function fails because it still needs an entry in
taskDescriptions? Just a wild guess. Regards, HaeB (talk) 05:36, 7 August 2023 (UTC)
- I am not at all a JavaScript expert, but perhaps this newly added function fails because it still needs an entry in
- I tried something like this a few issues ago, which was to add {{Wikipedia:Wikipedia Signpost/Templates/Single talk}} to the talk for the single-issue page, i.e. Wikipedia talk:Wikipedia Signpost/Single/2023-07-17. It seemed pretty straightforward, but for some reason I could not get it to work when I ran the script. I think that the whole damn thing is probably long overdue for a look-see and some documentation -- it was originally written by Evad37, who is not maintaining it (and isn't very active on the project in general). This is definitely one of the things that should be in the queue for whosoever ends up being the person who does this. jp×g 02:37, 7 August 2023 (UTC)
- In the meantime, seeing that the template had again not been updated several days after publication, I have done it myself again, picking Sunday Aug 27 for the next issue.
- As detailed above, this problem screams for a technical solution, especially now that we have agreed on a default publication schedule (first and third Sunday). But in the meantime, we could also consider devolving this particular EiC responsibility (setting the next issue's publication date) to a new role in the team - basically, have someone take formal responsibility for what Adam, Bri and I have been doing for the last half year or so. Regards, HaeB (talk) 03:01, 19 August 2023 (UTC)
- I think that since the dates are regular and predetermined, it doesn't really matter who resets the template; I will try to add it to the script, but I have been somewhat occupied lately, and there are several things in the script that also need attention (albeit things that have a lot of commonalities, so the time for all of them will likely overlap a lot). jp×g 11:00, 28 August 2023 (UTC)
- OK, so @Bri and Adam Cuerden: would one of you be willing to take on this chore until it is automated? (i.e. updating the deadline template right after publication of each issue, by default to the first and third Sunday of the month)?
- In the meantime, I just jumped in myself again, setting the next issue's date to October 1 (two weeks and one day after the publication of the most recent one).
- Regards, HaeB (talk) 20:16, 19 September 2023 (UTC)
- Sure. Adam Cuerden (talk)Has about 8.5% of all FPs. 20:20, 19 September 2023 (UTC)
- I think that since the dates are regular and predetermined, it doesn't really matter who resets the template; I will try to add it to the script, but I have been somewhat occupied lately, and there are several things in the script that also need attention (albeit things that have a lot of commonalities, so the time for all of them will likely overlap a lot). jp×g 11:00, 28 August 2023 (UTC)
Labeled section transclusion for newsroom
I swear I did this a few months ago, and demonstrated it here, but both of the templates show no such edits. What??? Anyway, I guess I will do it again.
The Newsroom tasks in the template are now set up to automatically transclude sections from this talk page. Thus, the section above this one ("In the media") is transcluded to the Newsroom under the "In the media" task. This seems like a pretty obvious no-brainer fix, which is why I did it some months ago, or I guess hallucinated doing it some months ago -- from now on we can just have all the discussion live here (and have some record of it persist in history) rather than just clearing it off the Newsroom every time we publish. There is an added benefit of not having to clunkily reset the Newsroom (or fail to) each issue, which was a pain in the ass. jp×g 01:14, 28 August 2023 (UTC)
- The task items now all have links to create talk page sections here if there aren't sections already. Christ, what a disaster -- that took an hour and a half. jp×g 02:41, 28 August 2023 (UTC)
- The related discussion is at Wikipedia talk:Wikipedia Signpost/Newsroom/Archive 34 § Newsroom formatting reorganization. isaacl (talk) 04:59, 28 August 2023 (UTC)
- Ah, yeah: the sections do need to be distinguished. I think what I'll do for this is automatically prefix them with the volume/issue, i.e. this issue's N&N section would be
19-17 News and notes
-- very easy to do this with the way the template is formatted. I did have to spend some hours rassling with the template to get it as far as it is right now, though, and I have some other crucial shizz that needs to get done and we need to publish, so it will have to be a "coming soon" type beat. jp×g 05:35, 28 August 2023 (UTC)- I think the problem is that no-one added it to the reset template. Adam Cuerden (talk)Has about 8.5% of all FPs. 05:48, 28 August 2023 (UTC)
- Well, now it is part of Wikipedia:Wikipedia_Signpost/Newsroom/Task so it will be there regardless. God, this tech stack is so cursed. Twenty years of lasagna piled up with Express on NodeJS on Ubuntu running API queries to MediaWiki on PHP loading JSON out of Lua via a Python script -- and get this -- it doesn't even use normal MediaWiki extensions, it uses bizarre idiosyncratic English Wikipedia custom templates-within-templates. Fuck me. Hopefully when I have a fatal aneurysm from an error message I have at least documented this stuff enough for the next person. jp×g 06:05, 28 August 2023 (UTC)
- Sorry no. The problem here was not
PHP loading JSON
etc., but once again your own code. As per the discussion that Isaacl kindly linked above, back in February you had overlooked a pretty obvious problem with your solution as implemented it then, and when I immediately pointed that out here on this talk page (politely framed as a question), you didn't respond at all. After these issues didn't get solved and your solution caused confusion and inconsistencies otherwise too, Adam reverted to the previous setup in early April, transparently posting about it on this talk page - which you apparently didn't see either, hence now theWhat???
etc. I think what I'll do for this is automatically prefix them with the volume/issue, i.e. this issue's N&N section would be 19-17 News and notes
- OK, appreciate that after that aforementioned obvious problem has now again been raised to you, you have come up with another idea. But I already can see various problems with that too, starting with the very cryptic naming - I think very few of us on the team are permanently aware of which volume and issue number the Signpost is at. I.e. your naming scheme is likely to create constant friction and confusion about finding the right section to comment at. (I would recommend using the upcoming issue's date instead; you can pull that from the deadline template too.) In any case, it seems you already managed to mess this up with a new bug, considering that the test section you created below actually has the volume and issue for the recently published issue (19-16), not the upcoming one as you propose above (19-17).- I also noted that in your reset of the Newsroom page, you forgot to handle the existing discussion section about the "Recent research" section (e.g. by converting it to the new format); it's now still in the page's source code and will probably remain there until someone else stumbles over it in the future. Same for the instructions about the previous method that you propose to abandon, i.e. resetting the newsroom page manually after the publication of each issue - these are actually still present in the HTML comments of the current revision.
- The upshot here is that your sloppy work on such tech issues has again and again been causing significant confusion and lost time for others on the team. Relatedly, have you had a chance yet to look at User talk:JPxG/SPS.js#Footer bug (about bugs in your version of the publication script that caused various issues last month which several users had to fix post-publication?) Also, did this by any chance help to fix another issue with the script that appears to be blocking an imho more important tech improvement?
- I'm sorry to have become more direct about such issues in recent weeks. But as mentioned above, some less direct attempts to bring them up weren't quite successful earlier. What's more, honestly, your own constant expletive-laden rants about the Signpost's existing tech stack are also not very respectful toward the people who built it (which, for clarity, mostly does not include myself). And while I think you have a point to some extent (especially with regard to the lack of documentation and clear structures), their code might often had a lower error density than certain current efforts.
- Regards, HaeB (talk) 10:02, 28 August 2023 (UTC)
- I don't blame anybody for it being confusing, since nobody who did any of it was on any sort of payroll -- so a task carried out in addition to whatever personal and professional obligations they had to begin with. Indeed, the existence of something like the module and the tag search system is nothing short of miraculous. But there is nonetheless a very large amount of technical debt, which I suspect is the prime explanation for stuff like the tag browser being partially written in 2015, not fully implemented, and not mentioned much after that. There was an entire system of Python scripts that were supposed to automate tag aliasing, ten years ago, and I am not quite sure what happened to them. The stack is not cursed for lack of effort or diligence, it's a result of being developed as a patchwork of individual components by dozens of volunteers over the course of nearly twenty years. Currently, it is 3:30 in the morning, since I have been applying several years of Ubuntu upgrades to a webserver that will allow me to run Wegweiser, so I apologize if I am not very awake.
- With respect to pulling from the next issue date, this isn't feasible because that changes sometimes -- and auto-incrementing the date would necessitate extremely complicated handling of December issues (i.e. is the next issue V and I+1, or V+1 and 1). I have already spent several hours today writing code to handle the talk page sections, so if you want to come up with some way to algorithmically parse the next volume/issue number in a way that meets your standards, it would be appreciated. jp×g 10:49, 28 August 2023 (UTC)
- I take your word regarding the incomplete implementation of that tag browser back in 2015. But that's honestly a feature of secondary or tertiary importance compared to the process of publishing each new issue and organizing the discussions about each upcoming story (the two things negatively affected by the aforementioned recent technical problems with your own changes).
pulling from the next issue date [...] isn't feasible because that changes sometimes
- I was aware of that when I wrote the above, but one could envisage various technical solutions to that. You are right though that implementing them is work, like your suggestion that I shouldcome up with some way to algorithmically parse the next volume/issue number
. But that's all something that one should think through *before* implementing a change that causes the need for such fixes. Given that above a mere 36 minutes had passed between someone pointing you (again) to the quite obvious problem with your initial idea, and you responding withAh, yeah [...] I think what I'll do [...]
, I think it's fair to assume that you had not given your second idea much thought either. In other words, here it was again yourself who created significant tech debt and inferior usability, not theTwenty years of lasagna piled up
by past Signpost technical contributors. Don't be surprised (What???
etc.) if your solution gets reverted again at some point in the future. Regards, HaeB (talk) 00:51, 28 September 2023 (UTC)
- Sorry no. The problem here was not
- Well, now it is part of Wikipedia:Wikipedia_Signpost/Newsroom/Task so it will be there regardless. God, this tech stack is so cursed. Twenty years of lasagna piled up with Express on NodeJS on Ubuntu running API queries to MediaWiki on PHP loading JSON out of Lua via a Python script -- and get this -- it doesn't even use normal MediaWiki extensions, it uses bizarre idiosyncratic English Wikipedia custom templates-within-templates. Fuck me. Hopefully when I have a fatal aneurysm from an error message I have at least documented this stuff enough for the next person. jp×g 06:05, 28 August 2023 (UTC)
- I think the problem is that no-one added it to the reset template. Adam Cuerden (talk)Has about 8.5% of all FPs. 05:48, 28 August 2023 (UTC)
- Ah, yeah: the sections do need to be distinguished. I think what I'll do for this is automatically prefix them with the volume/issue, i.e. this issue's N&N section would be
Knowledge Equity Fund community call
Scheduled for Oct 6; for further details see Wikipedia:Village_pump_(idea_lab)#Knowledge_Equity_Fund_community_call. Andreas JN466 19:06, 26 September 2023 (UTC)
- @Jayen466 We had some issues with not receiving confirmation emails to the Equityfund@ email, but resolved the issue last week and folks have stopped receiving bouncebacks. Since I saw its mentioned in the Signpost article coming up, I wanted to confirm that people who are interested can email that alias without issue. NGunasena (WMF) (talk) 19:27, 2 October 2023 (UTC)
- Thanks for letting me know, NGunasena (WMF). I'll remove that sentence. Andreas JN466 20:29, 2 October 2023 (UTC)
19-18 Recent research
As usual, we are preparing this regular survey on recent academic research about Wikipedia, doubling as the Wikimedia Research Newsletter (now in its thirteenth year). Help is welcome to review or summarize the many interesting items listed here, as are suggestions of other new research papers that haven't been covered yet. Regards, HaeB (talk) 00:29, 28 September 2023 (UTC)
- Working on this, should have something publishable up in less than two hours from now. (@JPxG: what's the current best guess for the expected time of publication?) Regards, HaeB (talk) 20:01, 1 October 2023 (UTC)
I will be home in a couple hours. jp×g 01:00, 2 October 2023 (UTC)
- Mountain was a little higher than I expected. I am kind of tired; I figure I can publish a lame issue now or a good one in the morning. jp×g 05:41, 2 October 2023 (UTC)
- Thanks! I got diverted from this yesterday, but will have it ready before the new deadline today. Regards, HaeB (talk) 14:41, 2 October 2023 (UTC)
- @JPxG: Can we slowly get this issue out the door? There's some interesting discussion happening at User_talk:Jimbo_Wales#Endowment_financial_information ... if we wait much longer, we'll end up behind the curve, or having to edit N&N.
- @HaeB: We're quoting the following sentence in the Flagged Revisions story:
In analyzing its side effects, we found, contrary to expectations and most of our hypotheses, little evidence that the system neither raises transaction costs sufficiently to inhibit participation by the community as a whole, nor measurably improves the quality of contributions.
Is anyone going to understand that sentence? - From the other quotes you've included I suspect what they meant to say was
In analyzing its side effects, we found, contrary to expectations and most of our hypotheses, that the system doesn't raise transaction costs sufficiently to inhibit participation by the community as a whole, yet measurably improves the quality of contributions.
- But if that is what they meant, they spectacularly failed to say it. Do we need this quote? Andreas JN466 19:06, 3 October 2023 (UTC)
- Double negative singly negated? Here is my parsing:
- In analyzing its side effects, we found, contrary to expectations and most of our hypotheses, little evidence that the system neither raises transaction costs sufficiently to inhibit participation by the community as a whole, nor measurably improves the quality of contributions.
- becomes
- In analyzing its side effects, we found, contrary to expectations and most of our hypotheses,
littleevidence that the systemneitherraises transaction costs sufficiently to inhibit participation by the community as a whole,nor measurably improves the quality of contributions.
- In analyzing its side effects, we found, contrary to expectations and most of our hypotheses,
- jp×g 20:10, 3 October 2023 (UTC)
- Well yes, the authors messed up that wording a bit, but from the context (i.e. the findings in detail further below) the meaning is clear. I had noticed that too but I guess I was a bit too hesitant to go sic! sic! on them.
- And yes, a concise summary of the headline results is surely needed. I have reworded the quote a bit to fix that. (Since absence of evidence is not evidence of absence, these statements can't simply be reduced to a double negation that we could eliminate.) Regards, HaeB (talk) 20:39, 3 October 2023 (UTC)
- Double negative singly negated? Here is my parsing:
FC title, issue 19
Was the dangling comma [3] intentional? ☆ Bri (talk) 20:33, 3 October 2023 (UTC)
- Yes; posting habit from the old country. I forget if it is cool or uncool here and now. We can keep or remove it. jp×g 21:52, 3 October 2023 (UTC)
ChatGPT plug-in
I am sure we can do better than the average news piece (including the cited source) and actually link the mediawiki or other relevant page about the plug in for those who want to download it or discuss it or such? Piotr Konieczny aka Prokonsul Piotrus| reply here 10:11, 2 October 2023 (UTC)
- @Piotrus Thanks. I had meant to but had to rummage around for the link first.
- As far as I am aware, you can't download it. It is only available as an experimental beta feature to ChatGPT Plus subscribers. I've now added a link to the most relevant page on Meta-Wiki I'm aware of to ITM, along with a link to the July edition of Recent Research, where User:HaeB covered the plugin briefly. Andreas JN466 15:36, 2 October 2023 (UTC)
- @Jayen466 Thanks. I do wonder about ethics of having a non-free plugin - developed by WMF and effectively sold by another company to its subscribers? Piotr Konieczny aka Prokonsul Piotrus| reply here 06:07, 3 October 2023 (UTC)
- ChatGPT Plus costs $20/month. That does put the plugin out of reach for most people that the WMF wants to reach.
- I've added a mention of the price. I agree the plugin and the wider topic area deserve more substantial coverage than we have given it so far. @JPxG, HaeB, and Bri: Thoughts? Andreas JN466 08:44, 3 October 2023 (UTC)
- Yes, that's a good point (which I also mentioned in our earlier coverage). Some months ago, the WMF was offering to sign up community members for a free testing version of the plug-in (apparently OpenAI has a mechanism for that); but I'm not sure whether that offer still exists.
- In any case, besides the cost, there are also the more general concerns about building products exclusively for proprietary ecosystem. But, without dismissing these (I agree there is a slippery slope that way), this is not a new discussion - e.g. the WMF has long invested in building an iOS Wikipedia app, which you can basically only use after paying Apple lots of money (and also providing them with your identity, via your credit card), and very much forms part of a walled-garden ecosystem. In such situations, the argument has always been that there are a lot of people using these proprietary platforms which we wouldn't reach otherwise, i.e. that increasing the availability of free knowledge directly in this way is worth the downside of indirectly supporting something that goes against our values (in that case, the closed iOS/iPhone ecosystem). So it's a trade-off.
- Relevant questions to ask in such situations are whether that trade-off is worth it, e..g how much donor money/WMF staff time is spent to reach how many of those walled-garden readers. Also, whether at least the WMF's own code is open source, which it is in case of the plug-in. (By the way, that's also what enabled me to write a bit about how it actually works in that previous Signpost story, as the published code was more informative that the WMF's public descriptions.)
- In any case, yes, we should cover this topic more; I've been envisaging to do so based on the metrics that the WMF has been publishing (so far only in preliminary form and without being very transparent on how some of them are defined). Regards, HaeB (talk) 11:55, 3 October 2023 (UTC)
- Interesting paper about ChatGPT in Recent Research. Anecdotally, I had exactly the same experience today. My wife told me about a press story mentioning government shutdowns in the US. She knew more about what government shutdowns in the U.S. are like and what their effects are, but I still thought it might be worthwhile to look it up. So I asked ChatGPT and got the following well-formulated paragraphs that repeated all the points my better-informed wife had mentioned and added a few more details.
- Government shutdowns in the United States occur when Congress fails to pass a budget or a continuing resolution to fund the government. This can happen due to disagreements between political parties or issues related to specific policies. The shutdown results in the temporary suspension of non-essential government services and the furlough of federal employees.
- During a government shutdown, essential services such as national security, law enforcement, and emergency activities continue to operate. However, many government functions, agencies, and national parks may be affected. Federal employees who are deemed non-essential may be furloughed, meaning they are placed on temporary unpaid leave.
- The economic impact can be significant, affecting various sectors, including tourism, government contractors, and the overall economy. Additionally, government shutdowns can lead to a loss of public confidence in the government's ability to function effectively.
- It's worth noting that my information is up to September 2021, and I recommend checking the latest sources for updates on this topic.
- I then compared that to the lead of Government shutdowns in the United States:
- In the United States, government shutdowns occur when funding legislation required to finance the federal government is not enacted before the next fiscal year begins. In a shutdown, the federal government curtails agency activities and services, ceases non-essential operations, furloughs non-essential workers, and retains only essential employees in departments that protect human life or property. Shutdowns can also disrupt state, territorial, and local levels of government.
- Funding gaps began to lead to shutdowns in 1980, when Attorney General Benjamin Civiletti issued a legal opinion requiring it. This opinion was not consistently adhered to through the 1980s, but since 1990 all funding gaps lasting longer than a few hours have led to a shutdown. As of September 2023, ten funding shutdowns have led to federal employees being furloughed.
- The most significant shutdowns have included the 21-day shutdown of 1995–1996, during the Bill Clinton administration, over opposition to major spending cuts; the 16-day shutdown in 2013, during the Barack Obama administration, caused by a dispute over implementation of the Affordable Care Act (ACA); and the longest, the 35-day shutdown of 2018–2019, during the Donald Trump administration, caused by a dispute over funding an expansion of barriers on the U.S.–Mexico border.
- Shutdowns disrupt government services and programs; they close national parks and institutions. They reduce government revenue because fees are lost while at least some furloughed employees receive back pay. They reduce economic growth. During the 2013 shutdown, Standard & Poor's, the financial ratings agency, said on October 16 that the shutdown had "to date taken $24 billion out of the economy", and "shaved at least 0.6 percent off annualized fourth-quarter 2013 GDP growth".
- I found ChatGPT's readability vastly superior to the somewhat tortured Wikipedia prose. Andreas JN466 15:30, 3 October 2023 (UTC)
- Thanks for the ping. I find that the human-written version has a more complete history and coverage of the legal basis for the shutdowns. Interesting that both leave out a sore point for many: military members must continue to serve without pay during the shutdown. Unlike civilians, they (at least enlisted under contract) don't have the option of looking for another employer who pays them on time. I do wish we had a graphic as clear and concise as this one in the Washington Post. So far, the LLMs don't do graphical summaries automatically (do they?). ☆ Bri (talk) 16:38, 3 October 2023 (UTC)
- Interesting paper about ChatGPT in Recent Research. Anecdotally, I had exactly the same experience today. My wife told me about a press story mentioning government shutdowns in the US. She knew more about what government shutdowns in the U.S. are like and what their effects are, but I still thought it might be worthwhile to look it up. So I asked ChatGPT and got the following well-formulated paragraphs that repeated all the points my better-informed wife had mentioned and added a few more details.
- @Jayen466 Thanks. I do wonder about ethics of having a non-free plugin - developed by WMF and effectively sold by another company to its subscribers? Piotr Konieczny aka Prokonsul Piotrus| reply here 06:07, 3 October 2023 (UTC)
- Jumping in here to say:
- Any Wikimedian who is not a ChatGPT Plus user but is interested in getting access to the Wikipedia plugin can do so! Just let me know the email address associated with your ChatGPT account (you can email me at mpinchuk
wikimedia.org). - Plugins are an experimental beta feature of ChatGPT, not yet a stable product. This beta plugin framework is providing us with an easy, low-cost, low-risk way to learn about how/whether we can deliver knowledge to users via an AI assistant before we consider making any future (much more substantial) investments in this space (e.g., training our own models, adapting/hosting 3rd party models, etc. – all of which would require significant resourcing and discussions with the movement).
- Some early learnings are here: meta:Future_Audiences/Experiments:_conversational/generative_AI#Experiments_with_off-platform_AI:_Wikipedia_ChatGPT_plugin
- I've been sharing updates at the monthly open call on Future Audiences: meta:Future_Audiences/Community_discussions
- I'd be delighted to talk more with anyone who's interested in writing a more in-depth Signpost piece on this or anything else in Future Audiences :) Don't be a stranger! Maryana (WMF) (talk) 21:18, 3 October 2023 (UTC)
- ETA: just changed my sig to avoid any confusion. I realize newer editors might not remember me, the oldbie non-CEO Maryana ;) Maryana Pinchuk (WMF) (talk) 21:22, 3 October 2023 (UTC)
- Thanks, @Maryana. I for one remember sitting next to you at a table in a restaurant in London when Sue Gardner took a bunch of us for dinner, ca. 2012 ... Andreas JN466 21:49, 3 October 2023 (UTC)
- Maryana, your input here is very appreciated! Piotr Konieczny aka Prokonsul Piotrus| reply here 00:54, 4 October 2023 (UTC)
- Thanks, @Maryana. I for one remember sitting next to you at a table in a restaurant in London when Sue Gardner took a bunch of us for dinner, ca. 2012 ...
- Any Wikimedian who is not a ChatGPT Plus user but is interested in getting access to the Wikipedia plugin can do so! Just let me know the email address associated with your ChatGPT account (you can email me at mpinchuk
This issue
Wikipedia talk:Wikipedia Signpost/Single/2023-10-03
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Traffic report | There shall be no slaves in the land of lands, it's a Bollywood jam | 377 | 512 | 627 | 716 | 758 | 792 | 826 |
| Recent research | Readers prefer ChatGPT over Wikipedia; concerns about limiting "anyone can edit" principle "may be overstated" | 1718 | 2076 | 2336 | 2450 | 2513 | 2567 | 2620 |
| Poetry | "The Sight" | 352 | 454 | 558 | 594 | 625 | 649 | 683 |
| News and notes | Wikimedia Endowment financial statement published | 829 | 1117 | 1366 | 1418 | 1468 | 1511 | 1560 |
| In the media | History is written by whoever can harness the most editors | 743 | 988 | 1223 | 1301 | 1340 | 1381 | 1445 |
| Featured content | By your logic, | 423 | 584 | 717 | 785 | 832 | 870 | 924 |
| Concept | Wikipedia policies from other worlds: WP:NOANTLERS | 468 | 641 | 774 | 834 | 867 | 895 | 931 |
jp×g 18:50, 5 October 2023 (UTC)
19-19 Humour
This is a little thing I threw together. I don't know if it is publishable. If someone wants to finish it, go ahead. jp×g 18:54, 22 October 2023 (UTC)
19-19 Gallery
This is in good shape and ready for a copyedit. jp×g 18:54, 22 October 2023 (UTC)
19-19 Featured content
@Adam Cuerden: For "polar bear" it says Pictured: The subject of the article in a snowstorm:
. But the picture is of the RAAF officer. Is this a... wait, no, durr, most people have black-on-white display skins. Okay. Do we need stuff for the FLs? jp×g 18:52, 22 October 2023 (UTC)
- If someone has time. I don't think it's worth holding up publication. Adam Cuerden (talk)Has about 8.5% of all FPs. 19:05, 22 October 2023 (UTC)
- I've swapped in a blank white square for those not using white backgrounds. Adam Cuerden (talk)Has about 8.5% of all FPs. 19:17, 22 October 2023 (UTC)
19-19 Traffic report
I cannot think of a good reason why a nonstandard flag should represent Israel in the traffic report. ☆ Bri (talk) 05:43, 22 October 2023 (UTC)
- @Bri and Igordebraga: Yes, we need to be very careful about this, especially since the non-standard flag was uploaded less than 2 weeks ago and may be very controversial. Also I have no idea what the added text says (but I'm pretty sure it's in Hebrew). Let's suppose, just for this discussion, ... (removed by Smallbones). Either of those would be incredibly controversial and would look like The Signpost is taking a very strong stand on one side or another. We don't want to do that, especially not knowing what side we are taking. In fact even that sentence starting "Let's suppose, just for this discussion ..." is probably too controversial. Anybody should feel free to remove it here. I'll replace the non-standard flag with the standard one in about 10 seconds (unless anybody objects). Smallbones(smalltalk) 12:53, 22 October 2023 (UTC)
- OK I've replaced the flag (and removed my overly bold sentence above). BTW the section heading in Traffic report "With their tanks and their bombs, and their bombs and their guns (October 8 to 14)" is perfect IMHO. It says something meaningful without taking sides - and yes they are still fighting. Smallbones(smalltalk) 13:04, 22 October 2023 (UTC)
- @JPxG: Traffic report looks ok to me now. I always have trouble copy editing it, so consider my efforts just a walk-through, just meant to avoid any major problems. The flag thing above was by far the main thing to consider - and I think the EiC should at least take a look. The blurb I added makes sense to me - but it would, wouldn't it? I added a sentence to Dick Butkus, roughly "He was considered the most feared tackler of all time," based on something in Dick Butkus. I should say that I once almost met Butkus (I was about 5 yards away from him for 10-15 minutes in a very controlled environment) about 1979. He could barely walk because of his knee injuries. Smallbones(smalltalk) 14:00, 22 October 2023 (UTC)
- OK I've replaced the flag (and removed my overly bold sentence above). BTW the section heading in Traffic report "With their tanks and their bombs, and their bombs and their guns (October 8 to 14)" is perfect IMHO. It says something meaningful without taking sides - and yes they are still fighting. Smallbones(smalltalk) 13:04, 22 October 2023 (UTC)
- My friend, a poster known as "train", says that this is Psalm 60, last line, verse 14: בֵּאלֹקים נַעֲשֶׂה-חָיִל וְהוּא, יָבוּס צָרֵינוּ ("Through God we shall do valiantly; for He it is that will tread down our adversaries"). It would seem somewhat better in this case to just use an unadorned flag. jp×g 00:50, 23 October 2023 (UTC)
Nature: AI tidies up Wikipedia’s references — and boosts reliability
I expect this is already on our radar, but I wanted to mention that unlike most coverage (and research), this seems actually useful. Underlying research is here: https://www.nature.com/articles/s42256-023-00726-1 The research say the code to reproduce the study is somewhere here: https://github.com/facebookresearch/side . Can anyone convert it it into a usable tool, assuming this has not been done already? Ping User:Headbomb who may be interested in this. Piotr Konieczny aka Prokonsul Piotrus| reply here 03:23, 20 October 2023 (UTC)
- 1. Thanks for the heads-up; as I may have previously mentioned, if it was covered in WikiResearch Twitter/Mastodon feed, it can indeed be assumed that it's on said radar.
- 2. This paper was already published last year in preprint form and received media attention back then. We covered it in both "In the media" ("Facebook experiments with Wikipedia fact-checking") and "Recent research" ("Facebook/Meta research on 'Improving Wikipedia Verifiability with AI'") at the time. (The story that's now in our new issue unfortunately omits that context and doesn't really offer any new information about this research project.) That said, we might still run a fuller review in "Recent research" now that the published version of the paper is out (are you interested?).
- 3. Please use the Suggestions page instead next time, that helps to avoid duplicate efforts.
- 4.
Can anyone convert it it into a usable tool
- yes, good point, that could be very interesting. However... (moving to Wikipedia talk:Wikipedia Signpost/2023-10-23/In the media#AI finding references as a more suitable location). - Regards, HaeB (talk) 21:04, 23 October 2023 (UTC)
19-19 Arbitration report
This one is a real walloper. I am going to need to spend a couple more hours on it. jp×g 03:51, 23 October 2023 (UTC)
- I am on the fence about this one and I am leaning towards next week.
- The core of it is that I used Claude to write the rough draft. This is something that was very well-received a year ago; there were two long articles in 2022-08-01, including an arbitration report, written with GPT-3. At that time, there had been virtually no usage of language models in journalism, and as such most people treated it as at worst an eccentricity, and at best an innovation. Now I don't think we have the same benefit of the doubt: there have been several high-profile instances of publications shooting themselves in the feet with this sort of thing. Maybe fairly and maybe unfairly; the CNET brouhaha did actually involve language models, albeit applied in a very aggressive way and with very little oversight. The story about the wildly inappropriate MSN article, on the other hand, had nothing "AI" about it at all: they just ran a different outlet's article through an automated thesaurus script (for an obituary no less).
- Of course, I know that this has got no bearing on Signpost matters: all of the controversies about papers running "AI-written articles" are, at root, controversies about papers running poorly written articles without editorial oversight. If that weren't the case, there would be nothing to point to. Maybe this is a tautology, but I don't know of any incidents in which somebody got in trouble for publishing something that was written by a language model and adequately fact-checked prior to being sent out. There is not any kind of poisonous substance in these models' output that makes them incapable of saying true things.
- But I don't think the zeitgeist really cares about the difference between a RNN, a LSTM and a transformer: people seem to be quite willing to hate something (or at least refuse to read it) if it was brought to you by the letters "A" and "I". The only thing I can think of to counteract that is some kind of personal guarantee that I'll pay anyone $10 who finds a factual error in the article, for which I would have to do some quite extensive fact checking beforehand lest I risk draining my wallet completely. It is getting rather late, and I'm not confident that I can go through this to a thorough enough extent to feel comfortable publishing it in this issue, nor am I confident that delaying the issue further would be worth having an extra article (which we can just as easily run next issue). I apologize for putting this article off again, but I think I will be fully ready to take the $10-per-error challenge with it next issue.
- @Vaticidalprophet, Bri, and Adam Cuerden: jp×g 06:27, 23 October 2023 (UTC)
- (edit conflict) I'm not sure if you saw my comment here. Having an AI program write about a highly contentious arbitration case, and sending the result to hundreds of people, does not strike me as a very good idea. Adding: intense face-checking, if this is to run, is of course a good idea, but in that case it might be no more time-consuming to write the article from scratch. Newyorkbrad (talk) 06:31, 23 October 2023 (UTC)
- Well, that is what I mean: I think the record is pretty good on this. Specifically, for writing Signpost arbitration reports with the aid of transformer networks. I did the same thing twice a year ago, one of them being the deletion report where I told it to use florid profanity and the other being the arbitration report where I told it to imitate Oliver Wendell Holmes. These had only one person fact-checking them (me) and neither drew any criticism for errors. For the arbitration report, the only amendments in the year since have been that I typoed "was been made" instead of "was made", and accidentally wrote
{{noping|Black}} Kiteinstead of{{noping|Black Kite}}. Alas. jp×g 06:56, 23 October 2023 (UTC)- Thanks for the reply. I missed that prior arbitration report, or I would have commented then. I can understand the appeal of experimenting with these programs, but I don't think the arbitration report—which deals with the most bitter and protracted disputes among our editors—is a sensible place to do it. (And I'm afraid that in the prior report I see no signs of Holmes's writing style, but that is a different discussion.) Regards, Newyorkbrad (talk) 12:35, 23 October 2023 (UTC)
- Well, that is what I mean: I think the record is pretty good on this. Specifically, for writing Signpost arbitration reports with the aid of transformer networks. I did the same thing twice a year ago, one of them being the deletion report where I told it to use florid profanity and the other being the arbitration report where I told it to imitate Oliver Wendell Holmes. These had only one person fact-checking them (me) and neither drew any criticism for errors. For the arbitration report, the only amendments in the year since have been that I typoed "was been made" instead of "was made", and accidentally wrote
- (edit conflict) I'm not sure if you saw my comment here. Having an AI program write about a highly contentious arbitration case, and sending the result to hundreds of people, does not strike me as a very good idea. Adding: intense face-checking, if this is to run, is of course a good idea, but in that case it might be no more time-consuming to write the article from scratch. Newyorkbrad (talk) 06:31, 23 October 2023 (UTC)
@Newyorkbrad: Just for your awareness, we discussed this specific issue in Newsroom archive 36. I opined that I'm satisfied with the quality of this particular summary, with the addition of a few editorial snippets of my own. ☆ Bri (talk) 15:39, 23 October 2023 (UTC)
- Thanks for the link. I hadn't seen that discussion. For the record, I still think the arbitration report, particularly one discussing a case that resulted in indefinite bans, is far from the best place to be experimenting with AI. Regards, Newyorkbrad (talk) 14:05, 24 October 2023 (UTC)
@JPxG: I'm done with News from Diff. The intent is to copy [4] as closely as possible. I may add a bit more to the editor's intro and might need to reformat the pix.
I think this tool has a lot of possibilities if rejiggered a bit, e.g. easy sound files for reading every Wikipedia article. But that's beyond this article. Smallbones(smalltalk) 05:29, 22 October 2023 (UTC)
- I'm done with the News from Diff article. The only thing original here is putting the generated QR code in the editor's intro at the top of the article. I'm off to work in the garden of earthly delights for most of the rest of the day, but might check back once or twice before publishing deadline. Smallbones(smalltalk) 17:35, 22 October 2023 (UTC)
- By the way, I don't know what the editorial process is on Diff these days, but this example indicates that we can't rely on it (e.g. the story appears to be missing a link, and offers various other copyediting opportunities, see here). Regards, HaeB (talk) 21:07, 23 October 2023 (UTC)
- Indeed. My general procedure on republished articles from Diff is to edit with a very light hand -- stuff like "teh" and curly quotes I will fix, but since the whole point of the column is that we're republishing it from somewhere else, it feels kind of awkward to actually copyedit it. Sometimes this results in us publishing something that's a little strange in places. Ought we to just say to hell with it, and edit them as we see fit? jp×g 02:15, 27 October 2023 (UTC)
Missed obituaries
Did we miss one for User:BeenAroundAWhile? How many have we missed? Is this something we can fix? jp×g 00:52, 26 October 2023 (UTC)
- We tried to have one, but obittuaries are hard without someone who knew them to write it. Adam Cuerden (talk)Has about 8.6% of all FPs. 21:54, 29 October 2023 (UTC)
I requested to interview WMF representatives for Signpost
In the Wikimedia-l mailing list the Wikimedia Foundation board of trustees offered to make appointments with community for conversation. I requested a conversation between any WMF representative and me, recorded, to be published in The Signpost. See my request here.
- meta:Wikimedia_Foundation_Community_Affairs_Committee/Talking:_2024/Interested#Username:bluerasberry
If they accept I will come back to the newsroom with some interview questions to discuss here in advance.
I propose to talk about money. Briefly: my general philosophy is that Wikipedia is unique among new media platforms for being community governed through democracy and other community-empowering processes. I feel that when donors give money to Wikipedia, then they do this because they believe in Wikipedia's user community and our governance process. The questions that I would explore with a WMF representative would compare and contrast WMF and community goals for spending Wikimedia donations. Bluerasberry (talk) 16:55, 23 October 2023 (UTC)
New board selection method
Thanks, Lane. Speaking of the board, one thing we seem to have missed covering in the current issue is the new board selection method: https://meta.wikimedia.org/wiki/Talk:Wikimedia_Foundation_elections/2024
Bad because the feedback period only runs until 29 October.
TL;DR: 2024 will once again have the affiliates make a shortlist of candidates for the community to vote on. The shortlist will be longer than last time – up to 12 (or even 15) rather than 6 candidates.
The thinking behind this might also be worth touching upon in your interview, if they take you up on the idea. --Andreas JN466 20:13, 26 October 2023 (UTC)
- Democracy in the Wikimedia Movement should be its own conversation topic. Unfortunately the subject is not well developed, and I have looked and failed to find Wikimedia Foundation attempts to invest in this subject. The new election model comes from the newly established or empowered meta:Wikimedia Foundation elections committee. My own personal wish is that somehow, either through WMF investment or its own fundraising, that this committee could gain independence of WMF pressure. I served two terms on an earlier iteration of this committee, and I felt that WMF staff were bold, aggressive, and even threatening in demanding election practices which I felt were unethical. I hope someday that I find an open forum to discuss these things. The best, cheapest, and easiest solution which I would recommend is transferring WMF oversight of elections to any third party, such as an established university institute for democracy or elections, or any similar neutral expert organization.
- It is not plausible that the Wikimedia community can organize a response to the new structure on this time frame. I do not think that any neutral party would grant the election concessions which the WMF routinely requests. The point of the election is for the Wikimedia Community to self-organize and offer candidates who can convey Wikimedia community positions in board governance. I would be happy to talk elections with anyone who would talk to me on recorded video, but otherwise, cannot fit the subject into the separate discussion of WMF/community relations with regard to general budget reporting. Bluerasberry (talk) 18:06, 27 October 2023 (UTC)
I commented. Not many comments because the call for comment period is hardly announced. For a process which decides the direction of ~US$600 million dollars in a typical electees term, this whole process is very quiet as compared to the Wikimedia Foundation outreach norm. Compare the much more developed voting campaign around meta:Communications/Sound Logo, which included democratic process review, outreach, comment recruitment, and the rest. Bluerasberry (talk) 18:07, 30 October 2023 (UTC)
It's happening again
- https://www.theguardian.com/commentisfree/2023/oct/23/why-is-elon-musk-attacking-wikipedia-because-its-very-existence-offends-him
- https://www.ndtv.com/world-news/elon-musk-willing-to-offer-1-billion-to-wikipedia-if-it-changes-name-to-4505917
- https://thehill.com/policy/technology/4270360-elon-musk-offers-1m-to-wikipedia-if-theyll-change-their-name/
- https://www.pcmag.com/news/elon-musk-starts-another-feud-this-time-with-wikipedia
- https://www.livemint.com/news/world/elon-musk-offers-1-billion-to-wikipedia-but-there-is-a-condition-11698043965167.html
- https://www.yahoo.com/entertainment/elon-musk-says-ll-wikipedia-222531368.html
- https://www.vice.com/en/article/7kxd8a/elon-musk-still-doesnt-understand-how-wikipedia-works
jp×g 23:41, 23 October 2023 (UTC)
- This Vice article is really bizarre. It seems like primarily an attempt to epicly own Elon Musk, and then incidentally there is some stuff about Wikipedia thrown in as window dressing. For example, I just found out today that the WMF spends all of its money responsibly, everyone agrees on this fact, and there's no controversy over it. 🤔 jp×g 23:46, 23 October 2023 (UTC)
- https://twitter.com/elonmusk/status/1350203005971398659 Sandizer (talk) 16:50, 25 October 2023 (UTC)
- Honestly, if Musk agrees to divide the money between all the editors who agree to this, I'm pretty sure we could finally get an RfC to pass unanimously. AryKun (talk) 06:56, 31 October 2023 (UTC)
Next seven days
I may be of decreased availability. jp×g 06:01, 26 October 2023 (UTC)
- What up!!!!!! jp×g🗯️ 20:11, 2 November 2023 (UTC)
- OK, congratulations on finding a spare broom. But don't let it go to your head. It's not really a demotion since you still get to keep the EiC position. BTW, please don't run for arbcom anytime soon. The last time (and only time) this happened the result was not pretty, neither for The Signpost, nor for arbcom, and especially not for the EiC. (Don't mind me, just feeling a bit grumpy today!) So really, congrats! Smallbones(smalltalk) 20:38, 2 November 2023 (UTC)
Lourdes / Wifione
I posted on the suggestions page already, but I am mentioning it again here. Should we cover the Lourdes issue? QuicoleJR (talk) 14:49, 2 November 2023 (UTC)
- user:Lourdes admitted that they are user:Wifione and has been indefinitely blocked.
- Wikipedia:Arbitration/Requests/Case#Statement_by_Lourdes
- Previous
- Seems like a story. Bluerasberry (talk) 14:58, 2 November 2023 (UTC)
- You mean a banned paid editor-slash-admin who has been operating a sockpuppet admin account since 2018? Yeah, I'd say newsworthy. ☆ Bri (talk) 16:07, 2 November 2023 (UTC)
- I started an arbitration report. Other editors are invited to contribute. I don't know whether to keep it short and sweet, or go into the implications of a wolf-in-the-fold who has been here since 2006 or earlier, and apparently takes glee in subverting our system built on trust. The former case resulting in the siteban was bad enough that it actually resulted in off-wiki press coverage, and now we potentially top that by not being able to recognize and root them out a second time. IMO, a WP:LTA who is able to repeatedly gain administrator privileges, and both creates bent content and attempts to undermine core community-based processes like RfA – then, after being discovered, expresses the intent to disorganize and disrupt our entire system – is as close to a worst-case scenario as I can imagine. Maybe that's better left for a separate op-ed or the like. ☆ Bri (talk) 17:18, 2 November 2023 (UTC)
- @Bri and JPxG: yeah, you got a story - run with it, take it as far as it goes. I won't be joining in, unless I stumble across something really juicy, in which case I'll email it to you. But what could be more juicy than what you got? The only thing I can think of is if Eostrix/Icewhiz and/or Wiki PR/Status Labs/(new name) was somehow involved - which I doubt - but you never can rule something like that out until you check! (I'll try, but your time may be better spent elsewhere).
- I will be putting a short paragraph or two on Lourdes/Wifione in an opinion piece I'm writing. It started being about Musk, but will likely include 3-5 other recent oddities. It a pretty simple piece about the sky falling. My advice about Lourdes would be just to include all the details about them (not about arbcom talking heads). Readers will want to ponder some details, so feed them what they want. They won't all fit into a nice neat storyline though unless you are really lucky. Smallbones(smalltalk) 18:40, 2 November 2023 (UTC)
- Should this be part of the writeup in light of your "all the details about them" guidance?
Editing habits, interests, and contributions to sister Wikimedia projects suggest the editor is in India.
It seems relevant to their connection to the org that was at the center of the Wifione case, and their former username matches a well known person from that country except for one character (a connection which they obliquely disclaimed in their last message), but beyond that, I dont' know. ☆ Bri (talk) 21:23, 2 November 2023 (UTC)- @Bri: I don't know which former username you have in mind, but the "celebrity" that Lourdes referred to in the RfAr post is clearly not the person you mention. I'd prefer not to identify the celebrity, because her name was misused by a troll without her knowledge and there is no reason to drag an innocent non-Wikipedian into this mess. Regards, Newyorkbrad (talk) 03:13, 3 November 2023 (UTC)
- The earliest username associated with the Wifione case is Mrinal Pandey, which was renamed to Empengent. Empengent/Mrinal was linked to Wifione at arbcom. But you probably know this?? ☆ Bri (talk) 16:46, 3 November 2023 (UTC)
- @Bri: If I knew it once I'd forgotten it. I wasn't on the Committee the year of the Wifione case. Regards, Newyorkbrad (talk) 23:08, 3 November 2023 (UTC)
- The earliest username associated with the Wifione case is Mrinal Pandey, which was renamed to Empengent. Empengent/Mrinal was linked to Wifione at arbcom. But you probably know this?? ☆ Bri (talk) 16:46, 3 November 2023 (UTC)
- @Bri: I don't know which former username you have in mind, but the "celebrity" that Lourdes referred to in the RfAr post is clearly not the person you mention. I'd prefer not to identify the celebrity, because her name was misused by a troll without her knowledge and there is no reason to drag an innocent non-Wikipedian into this mess. Regards, Newyorkbrad (talk) 03:13, 3 November 2023 (UTC)
- Should this be part of the writeup in light of your "all the details about them" guidance?
Sure (but I did say "ponder some details" as well). My guess is that you know a lot more than 98% of Wikipedians about this story and you're probably thinking "everybody knows that already" simply because it's so obvious to you. But please think about the new Wikipedian who has only been here a month who needs this story to get themselves up to speed. And everybody could use a review. The only thing I'll suggest is that for the "well known person" think about whether Arbcom could do something silly - like accusing you of outing somebody. In that case a quote from Wifione or Lourdes on-wiki might be the way to go. Also the long fight against 1 named Indian about the "Institute", that resulted in the arb case. (wasn't there an earlier one?) But really, they (W1 & L) have been around forever, just summarize some of the more interesting or bizarre things they've said or done. Hope this helps. Smallbones(smalltalk) 22:50, 2 November 2023 (UTC)
Permalink to arb request at Special:Permalink/1183169230. jp×g🗯️ 00:25, 3 November 2023 (UTC)
Maybe something useful: Wikipedia:Administrators'_noticeboard#2,000+_admin_actions_in_violation_of_WP:BAN. Gråbergs Gråa Sång (talk) 08:49, 5 November 2023 (UTC)
Forward a day?
What do we think about doing so? I was busy most of today rather than editing happily. jp×g🗯️ 03:17, 5 November 2023 (UTC)
- As usual, my vote would be in favor of trying to keep the originally envisaged publication time.
- For what it's worth, I spent some time reviewing, fact-checking and enlarging ITM and N&N earlier today, and think both should be almost good to go already, without requiring much additional work.
- Regards, HaeB (talk) 04:49, 5 November 2023 (UTC)
Note: I fixed a very strange bug in the deadline template that was causing it to say the writing deadline was the 2nd instead of the 4th at the top (it was passing through a "Nov" rather than a "11" which I guess it thought meant "2"?) jp×g🗯️ 05:40, 5 November 2023 (UTC)
- It's that pesky autoconvert from Roman numerals to Arabic (jk) ☆ Bri (talk) 17:36, 5 November 2023 (UTC)
WikiCup report
Is one of the upcoming editions going to have one of those "WikiCup reports"? Just wondering :) BeanieFan11 (talk) 18:37, 19 October 2023 (UTC)
- @BeanieFan11: - can you write it up? Smallbones(smalltalk) 00:29, 20 October 2023 (UTC)
- @Smallbones: I mean, I might be a little biased as I'm one of the finalists, but I could write it depending on the publication date (I probably wouldn't be able to if its published by November 1/2, but after that date I'd probably have the time to). BeanieFan11 (talk) 00:44, 20 October 2023 (UTC)
- @Smallbones: It seems to say that the next edition plans on being published by November 5 – if I were to write a WikiCup report on November 3, would that be alright? (I'm actually a little excited at the thought of writing something for Wikipedia's newspaper ) BeanieFan11 (talk) 13:50, 26 October 2023 (UTC)
- @BeanieFan11 and JPxG: I'd think if it was written by the 2nd or 3rd it would work, but the editor-in-chief, JPxG, should let you know. Smallbones(smalltalk) 16:58, 26 October 2023 (UTC)
- @JPxG: Would it be alright if I were to write a WikiCup report on November 2 or 3? BeanieFan11 (talk) 23:30, 31 October 2023 (UTC)
- Absolutely, that'd be beautiful. I want to run something about the WikiCup, it's a vital part of the content side of the project. Exposure is good. jp×g🗯️ 01:46, 1 November 2023 (UTC)
- @JPxG: I've written up a comprehensive report on the WikiCup (including notes about every person to score) – it could probably use some copyediting as I have never written a newspaper article before – also note that I won the contest, and I included some of my personal thoughts (which I'm fine being removed if you want), so you might want to check it for bias, etc. BeanieFan11 (talk) 20:10, 3 November 2023 (UTC)
- @JPxG: I see the publication will be soon – do you think my write-up of the WikiCup is alright? Should I make any changes, etc.? BeanieFan11 (talk) 15:09, 4 November 2023 (UTC)
- It looks good to me. Queued for this issue. jp×g🗯️ 23:30, 5 November 2023 (UTC)
- @JPxG: I see the publication will be soon – do you think my write-up of the WikiCup is alright? Should I make any changes, etc.? BeanieFan11 (talk) 15:09, 4 November 2023 (UTC)
- @JPxG: I've written up a comprehensive report on the WikiCup (including notes about every person to score) – it could probably use some copyediting as I have never written a newspaper article before – also note that I won the contest, and I included some of my personal thoughts (which I'm fine being removed if you want), so you might want to check it for bias, etc. BeanieFan11 (talk) 20:10, 3 November 2023 (UTC)
- Absolutely, that'd be beautiful. I want to run something about the WikiCup, it's a vital part of the content side of the project. Exposure is good. jp×g🗯️ 01:46, 1 November 2023 (UTC)
- @JPxG: Would it be alright if I were to write a WikiCup report on November 2 or 3? BeanieFan11 (talk) 23:30, 31 October 2023 (UTC)
- @BeanieFan11 and JPxG: I'd think if it was written by the 2nd or 3rd it would work, but the editor-in-chief, JPxG, should let you know. Smallbones(smalltalk) 16:58, 26 October 2023 (UTC)
- @Smallbones: It seems to say that the next edition plans on being published by November 5 – if I were to write a WikiCup report on November 3, would that be alright? (I'm actually a little excited at the thought of writing something for Wikipedia's newspaper
- @Smallbones: I mean, I might be a little biased as I'm one of the finalists, but I could write it depending on the publication date (I probably wouldn't be able to if its published by November 1/2, but after that date I'd probably have the time to). BeanieFan11 (talk) 00:44, 20 October 2023 (UTC)
19-20 Recent research
As usual, we are preparing this regular survey on recent academic research about Wikipedia, doubling as the Wikimedia Research Newsletter (now in its thirteenth year). Help is welcome to review or summarize the many interesting items listed here, as are suggestions of other new research papers that haven't been covered yet. Regards, HaeB (talk) 21:46, 29 October 2023 (UTC)
- As usual, I'll work to make sure this is in publishable form by the publication deadline (whether or not it gets moved, cf. below). Regards, HaeB (talk) 04:45, 5 November 2023 (UTC)
- Running a bit late with this one, sorry, hope to have it publishable in an hour from how. (But from Wikipedia:Wikipedia Signpost/Newsroom#Article status and below, it seems that that is unlikely to hold up publication, as several other sections haven't yet progressed to the "Ready for copyedit" stage either, and we just received a large and potentially controversial new opinion submission last minute.) Regards, HaeB (talk) 20:00, 5 November 2023 (UTC)
- @HaeB: How are we doing on it? jp×g🗯️ 23:56, 5 November 2023 (UTC)
- Getting there (I saw the deadline update). Publishable in half an hour from now is my last word. Regards, HaeB (talk) 00:01, 6 November 2023 (UTC)
- @HaeB: How are we doing on it? jp×g🗯️ 23:56, 5 November 2023 (UTC)
- Running a bit late with this one, sorry, hope to have it publishable in an hour from how. (But from Wikipedia:Wikipedia Signpost/Newsroom#Article status and below, it seems that that is unlikely to hold up publication, as several other sections haven't yet progressed to the "Ready for copyedit" stage either, and we just received a large and potentially controversial new opinion submission last minute.) Regards, HaeB (talk) 20:00, 5 November 2023 (UTC)
Do you need help copyediting?
I guess it depends on how close you are to publication, If too close it might only get in the way. I'll try 2 articles
- first Traffic report
- second Featured content
@JPxG: ping me if you want me to get out of the way.
Smallbones(smalltalk) 00:10, 6 November 2023 (UTC)
- Yeah, that would be helpful, I am finishing up the Lourdes thing and we should be ready to go fairly soon. I think HaeB is going to have his thing ready in a few minutes and then it'll be poggers. jp×g🗯️ 00:12, 6 November 2023 (UTC)
- @JPxG: - sure I'll ce Featured content. You should ok the Gaza war stuff in Traffic and come up with a head line. Smallbones(smalltalk) 00:39, 6 November 2023 (UTC)
- Confirming that RR should be publishable now (I also did some copyediting myself, but obviously a second pair of eyes would be good). I'll still add illustrations from the featured paper considering we don't seem to be at the brink of publishing yet, but then I can also help out with copyediting the remaining stories. Regards, HaeB (talk) 00:43, 6 November 2023 (UTC)
19-20 In the media
For the most part I like this, but there is one thing I am not so sure about. Mostly the Elon Musk thing – we have this featured very prominently but I'm not really clear on what he did. As far as I'm aware, it boils down to:
- He alleges that we are run by Freakin' Libs. This is a pretty common complaint, as is the converse (i.e. that we're carrying water for the Freakin' Cons). It's at least somewhat borne out by research. Of course, so is the converse. At any rate, I do not personally think that having 52% of our references be to the Daily Lib Times and 48% of our references be to the Daily Con Post (or the other way around) is a big deal that seriously endangers the credibility of Wikipedia as a resource, but I don't think it is a prima facie crazy-go-nuts thing for someone to be mad about. It can be incorrect without being insane.
- He disagrees with some of the stuff in his article (namely, that it's "cow emoji poop emoji"). A gigantic proportion of famous people complain about their Wikipedia articles all the time, so I don't think this is really noteworthy.
- He claims that the WMF spends money in strange ways that seem unrelated to sustaining the project. This is something we've reported on extensively; in fact, our 2017 publication of WP:CANCER is among the highest-readership articles in Signpost history.
- He makes a goofy suggestion that we change our name to a dick joke and in return he donates a billion dollars to the WMF. I don't think he is serious about this, but either way, who cares.
- He made some throwaway comment that we were losers.
- He got into some kind of slapfight with Jimbo on Xwitter recently.
Maybe I am an idiot who's incapable of reading between the lines, and in actuality this is all some kind of veiled reference to a mercenary detachment barreling towards San Francisco. But if not, this really just doesn't seem like that much of a threat to me. Some rich guy thinks we suck. Whoopty doo. It seems somewhat silly to react to this as though he's actively working to destroy us; rather than an impassioned defense of our autonomy, it feels kind of corncobbish and pointless. My preference would honestly be that we run something else as the lead story and run this lower, if at all (it could just as easily be a brief). jp×g🗯️ 17:10, 5 November 2023 (UTC)
- @Smallbones and HaeB: who did the writeup. What do you think of this? jp×g🗯️ 17:13, 5 November 2023 (UTC)
- I'd be fine with making "Does Wikipedia's Gaza coverage show an anti-Israel bias?" the top story. I think the Musky thing got #1 just because it was the longest writeup, at the time. ☆ Bri (talk) 17:39, 5 November 2023 (UTC)
- I think your observations are by and large correct. Then again, as we discussed before, a main purpose of ITM is summarizing the media's coverage of an issue, and it's fair to say that quite a few media outlets were less sanguine about this one. And we do allow Signpost writers (including, often, yourself) some leeway on how to frame things, as journalism standards tend to do. So I was reluctant to come down too hard on Smallbones' writeup while editing and augmenting it, and I do think it could be published in its current state. That said, I would still support in particular toning down the "we are under attack from the world's richest man" alarmism and spending a bit more time on finding better headlines and illustrations for the piece. (Then again, the Signpost currently frequently features humor attempts that may only work well for a fraction of the audience.)
- Regards, HaeB (talk) 17:39, 5 November 2023 (UTC)
- EC with Bri and HaeB @JPxG, HaeB, and Bri: There are two things in the In the media Musk section that are newsworthy. 1) the richest person in the world effectively calls all Wikipedia editors "dicks" (I frankly think this is a new landmark in our relation to the rest of the web and deserves a second article (see below), 2) a flotilla of reliable sources comes to our defense, or criticizes Musk (or both) about the incident. That's definitely something that needs to be prominently noted in In the media. BTW I started the section and never had any doubt that it should be right at the top. I've been doing this for about 4 years now, and I've never seen anything that more deserves to be at the top of Itm. HaeB added to and polished my start. I think he did a great job, but if he wants to remove what he added (without removing his copyediting) that's up to him. There's one other thing that's unlikely, but we might as well talk about it once. It only happened to me 3-4 times during my tenure as EiC. If somebody important has contacted you and said that this material is unfortunate or against the rules, etc. - well, they are allowed to say that, and you should consider what they are trying to say, and a soft answer turns away wrath. But you don't want to get the reputation of censoring your reporters at the Signpost at the behest of bigshots. Once you give in, the requests will start coming in fast and furious. (next topic will follow soon) Smallbones(smalltalk) 17:40, 5 November 2023 (UTC)
- "censoring" "at the behest of bigshots"? Are you implying that Musk has contacted JPxG with demands to spike your story? Regards, HaeB (talk) 17:51, 5 November 2023 (UTC)
- Not really, it's more like some admin who thinks he has the authority to dictate what we write, or maybe even an arb, or ... Smallbones(smalltalk) 17:55, 5 November 2023 (UTC)
- Thankfully, as of the first of this month, I need only be governed by truth, justice and the Wikipedian way, so we can report confidently on stuff even if it pisses people off. My concern here is just that, well, it seems more to me like this is some kind of personal beef between Elon Musk and various online newspapers -- his main complaint about us is that we include them as reliable sources -- the only thing he did to us was vaguely call us losers on the Internet, something which any hundreds of random people do on any given day, so why does he get the privilege of special treatment for it? jp×g🗯️ 18:27, 5 November 2023 (UTC)
the only thing he did to us was vaguely call us losers on the Internet
- to be precise, that "losers" comment was from a separate incident which we covered a month ago (Wikipedia:Wikipedia_Signpost/2023-10-03/In_the_media#"Losers_just_got_a_lot_of_time_on_their_hands"). The "Dickypedia" joke that the current story centers one is a new insult. But this just goes further to show that this incident is by no means as unusual or important as some here seem to assume - Musk is trolling on Twitter, film at 11.- (Also, generally speaking we should clarify that admins - congrats! - are still bound to BLP and other Wikipedia policies. Then again, not every accusation of BLP violations leveled against the Signpost has turned out to be correct in the past, and it seems that Smallbones has some well-earned scar tissue in that regard, but that still doesn't justify overly paranoid assumptions on what prompted JPxG's remarks above.)
- Regards, HaeB (talk) 19:08, 5 November 2023 (UTC)
- I'm sorry if this was unclear but I wasn't accusing JPxG of anything improper, just suggesting that he might run into this type of thing and how I would and have handled it. And of course admins need to enforce WP:BLP, we all do, especially JPxG as EiC. I'm perplexed why this calculated insult from Musk isn't considered big news - in fact it is considered big news by multiple mainstream media (i.e. reliable sources) we should note what the RS say of course, especially in ITM. The ITM section is not so much about Wikipedia or Musk as about the super strong reaction by RSs. The headline might go. I'll go check ITM again, and maybe some folks should check the open letter to Elon Musk mentioned below. Smallbones(smalltalk) 19:24, 5 November 2023 (UTC)
- Thankfully, as of the first of this month, I need only be governed by truth, justice and the Wikipedian way, so we can report confidently on stuff even if it pisses people off. My concern here is just that, well, it seems more to me like this is some kind of personal beef between Elon Musk and various online newspapers -- his main complaint about us is that we include them as reliable sources -- the only thing he did to us was vaguely call us losers on the Internet, something which any hundreds of random people do on any given day, so why does he get the privilege of special treatment for it? jp×g🗯️ 18:27, 5 November 2023 (UTC)
- Not really, it's more like some admin who thinks he has the authority to dictate what we write, or maybe even an arb, or ... Smallbones(smalltalk) 17:55, 5 November 2023 (UTC)
- "censoring" "at the behest of bigshots"? Are you implying that Musk has contacted JPxG with demands to spike your story? Regards, HaeB (talk) 17:51, 5 November 2023 (UTC)
- EC with Bri and HaeB @JPxG, HaeB, and Bri: There are two things in the In the media Musk section that are newsworthy. 1) the richest person in the world effectively calls all Wikipedia editors "dicks" (I frankly think this is a new landmark in our relation to the rest of the web and deserves a second article (see below), 2) a flotilla of reliable sources comes to our defense, or criticizes Musk (or both) about the incident. That's definitely something that needs to be prominently noted in In the media. BTW I started the section and never had any doubt that it should be right at the top. I've been doing this for about 4 years now, and I've never seen anything that more deserves to be at the top of Itm. HaeB added to and polished my start. I think he did a great job, but if he wants to remove what he added (without removing his copyediting) that's up to him. There's one other thing that's unlikely, but we might as well talk about it once. It only happened to me 3-4 times during my tenure as EiC. If somebody important has contacted you and said that this material is unfortunate or against the rules, etc. - well, they are allowed to say that, and you should consider what they are trying to say, and a soft answer turns away wrath. But you don't want to get the reputation of censoring your reporters at the Signpost at the behest of bigshots. Once you give in, the requests will start coming in fast and furious. (next topic will follow soon) Smallbones(smalltalk) 17:40, 5 November 2023 (UTC)
Well, I've made changes to the article suggested by this conversation. I changed the headline to "Media reacts to "Dickypedia"" , but that might offend some other folks. More later Smallbones(smalltalk) 20:06, 5 November 2023 (UTC)
- I think the topic needs a second article to fully cover it. I've just started a much longer article at the rubric "An open letter to Elon Musk" and I should have the basic formatting and CE done within an hour from now. Feel free to copy edit it, but after 10 minutes I'll be working on it full time. And please do not change my opinion - it is prominently labeled as opinion. Smallbones(smalltalk)
- I'm done on both ITM and the Open letter. I'll check back in a couple of hours. Smallbones(smalltalk) 21:12, 5 November 2023 (UTC)
- @Smallbones: Moving this to /Opinion, that is correct, or should it be /Op-ed? jp×g🗯️ 23:37, 5 November 2023 (UTC)
- @JPxG: - I guess opinion is better. That would mean changing the headline to "Open letter to Elon Musk" and I think the blurb to "You should learn some of our rules" - I'll do that, but you can do the move. Also, I'll change the ref in the PS to the endowment which lists $2,000,000 +, but doesn't give a date. And I chopped off the last PA sentence. Thanks for your help! Smallbones(smalltalk) 23:51, 5 November 2023 (UTC)
- I think the topic needs a second article to fully cover it. I've just started a much longer article at the rubric "An open letter to Elon Musk" and I should have the basic formatting and CE done within an hour from now. Feel free to copy edit it, but after 10 minutes I'll be working on it full time. And please do not change my opinion - it is prominently labeled as opinion. Smallbones(smalltalk)
- Issue is ready to go. @Smallbones: Let me know what you think of the ITM. jp×g🗯️ 02:58, 6 November 2023 (UTC)
- This still needs cross-links between ITM and Smallbones' opinion piece, assuming we go with both. Regards, HaeB (talk) 03:22, 6 November 2023 (UTC)
- @JPxG: You are the EiC so it's your call. It's not bad enough that I would ask that my name be taken off of it. On the other hand the story is about Musk, so why are we being coy about using his name? A key sin in journalism is "burying the lede" but his goes beyond that to "burying the subject". And as you know, I disagree about most of the criticism of WMF finances. The only news released since the tweets is that so many RSs said that Musk was off-base on this. So, I guess my advice is publish and get it over with. Smallbones(smalltalk) 03:32, 6 November 2023 (UTC)
Another for the pile
Wikipedia pays more attention to events in richer countries, say data scientists. Contra: "Wikipedia would love to pay more attention to events in poorer countries but can't because the damn news refuses to write anything about them, say Wikipedians"? jp×g🗯️ 21:13, 14 November 2023 (UTC)
Pageviews, 2023-11-06
Wikipedia talk:Wikipedia Signpost/Single/2023-11-06
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Wikidata | Evaluating qualitative systemic bias in large article sets on Wikipedia | 577 | 691 | 732 | 774 | 802 | 819 | 863 |
| WikiCup report | The WikiCup 2023 | 351 | 482 | 510 | 544 | 584 | 638 | 712 |
| Traffic report | Cricket jumpscare | 476 | 633 | 680 | 721 | 758 | 772 | 832 |
| Recent research | How English Wikipedia drove out fringe editors over two decades | 12788 | 13224 | 13417 | 13534 | 13592 | 13629 | 13725 |
| Opinion | An open letter to Elon Musk | 1264 | 1623 | 1713 | 1781 | 1836 | 1862 | 1925 |
| News from Wiki Ed | Equity lists on Wikipedia | 434 | 566 | 594 | 632 | 668 | 686 | 730 |
| News and notes | Board candidacy process posted, editors protest WMF privacy measure, sweet meetups | 815 | 980 | 1017 | 1086 | 1128 | 1153 | 1213 |
| In the media | UK gov bigwig accused of ripping off WP articles for book, Wikipedians accused of being dicks by a rich man | 1114 | 1378 | 1440 | 1520 | 1565 | 1608 | 1703 |
| Featured content | Like putting a golf course in a historic site. | 465 | 623 | 678 | 723 | 753 | 765 | 807 |
| Arbitration report | Admin bewilderingly unmasks self as sockpuppet of other admin who was extremely banned in 2015 | 53006 | 54419 | 54996 | 55482 | 55738 | 55879 | 56196 |
Not sure why these stats are showing up for some articles and not others; some kind of weird 24-hour rollover thing. Unsurprisingly, the ones that Andreas posted to HN look like they have about a hundred bajillion times more readers than the others. Much to think about! jp×g🗯️ 06:09, 7 November 2023 (UTC)
- Numbers aren't everything. What kinds of people does the Signpost want to reach? Ed [talk] [OMT] 06:12, 7 November 2023 (UTC)
- The Hacker News community is a pretty good match for Wikipedia demographically. Similar interests. Andreas JN466 14:42, 7 November 2023 (UTC)
- To be clear, Andreas, I wasn't trying to implicitly criticize you with that post! (Apologies if it came off that way.) It was just an observation about valuing pageviews/popularity vs. intentionally reaching an audience. Ed [talk] [OMT] 15:35, 7 November 2023 (UTC)
- Thanks, Ed. Best, Andreas JN466 22:16, 8 November 2023 (UTC)
- To be clear, Andreas, I wasn't trying to implicitly criticize you with that post! (Apologies if it came off that way.) It was just an observation about valuing pageviews/popularity vs. intentionally reaching an audience. Ed [talk] [OMT] 15:35, 7 November 2023 (UTC)
- Ed raises a valid point. Then again numbers are also not nothing, and I do think we should continue to quantify the impact of our work here.
- Apropos, Ed, since there are a few parallels (and differences, of course) and it's always nice to talk shop, perhaps you would like to share a bit on how you folks are handling this nowadays in the Wikimedia Foundation's Communications department: How does the department (or your team) define the audiences that you want to reach with your communications, and what metrics are you using to measure the impact and success of that work?
- Regards, HaeB (talk) 08:35, 9 November 2023 (UTC)
- The Hacker News community is a pretty good match for Wikipedia demographically. Similar interests. Andreas JN466 14:42, 7 November 2023 (UTC)
- https://twitter.com/depthsofwiki/status/1721932638556991971 Expecting about eight bajillion from this, as is the normal situ. jp×g🗯️ 20:01, 7 November 2023 (UTC)
- Alternative link to the full thread, for those not logged into a Twitter/X account: https://threadreaderapp.com/thread/1721932638556991971.html
- Regards, HaeB (talk) 21:50, 7 November 2023 (UTC)
- Hell of a question, and worth thinking about in more detail. Personally, I am not totally certain. I think that, as far as English Wikipedia power users, we are probably at or close to saturation (lots of getting-started guides, help pages, welcome messages and the like mention the Signpost). I think we could call this our core audience, and the people who we care the most about informing (or at least not misinforming). Hence all the stuff about the WikiCup, the traffic reports, featured content, arbitration and deletion reports, et cetera; I think this is primarily interesting for en.wp editors and there's no problem with it lacking "broad appeal". On the other hand, I think there is a much wider base of people who would be interested in learning about Wikipedia, and Wikimedia projects, across the wider WWW: a website or a project or a movement needs to continually bring in new participants, and Wikipedia is famously opaque to outsiders. I think that a news outlet that provides accessible, interesting stuff that teaches you about how things work in a way that isn't obtuse or boring serves a role in getting people comfortable and excited and familiar with the "whole deal of it".
- Even if it doesn't, and they stay readers forever, I'd like to think that reading intelligent and accurate stuff about stuff that happens on Wikipedia would help people to be more intelligent and informed consumers of news and of media and of our own articles. For example, how often do you see some noob on a talk page saying that they're going to "stop donating to Wikipedia" if we don't XYZ? Or hear someone talk about Wikipedia needing to "hire more editors" or whatever? I think the wider public just lacks a bunch of key information about how we work, and part of that is that there isn't a lot of news coverage that explains it in a way that is easy or fun to read. Maybe the most poignant examples are the regular kerfluffues where someone (whether it's a rando online or a famous pundit) accuses us of being biased, and refers to some random editor who said a dumb politics thing as a "mod" who's sopping up your hard-earned donation money to et cetera. Or there's some big circus about a "censorship campaign"... referring to one person making a goofy AfD which doesn't even close as "delete". And so on and so on.
- Anyway, this is kind of a long and disjointed post, but I really do think that if we're able to get a bunch of people to read what we write, we do a service both to ourselves, to the broader editoriat, and to the project, although of course this is always contingent on writing things that are fair, accurate, true and good. jp×g🗯️ 23:25, 9 November 2023 (UTC)
Popular Signpost articles in 2023
I was curious what articles published this year have been popular. Here are the data from the pageviews widget used above:
2023 Signpost articles with >20,000 pageviews so far Issue and item Title Pageviews 2023-02-20 Disinformation report The "largest con in corporate history"? 20393 2023-05-22 News and notes Golden parachutes: Record severance payments at Wikimedia Foundation 53796 2023-07-17 Recent research Wikipedia-grounded chatbot "outperforms all baselines" on factual accuracy 47453 2023-08-01 News and notes City officials attempt to doxx Wikipedians, Ruwiki founder banned, WMF launches Mastodon server 31496 2023-08-15 Special report Thirteen years later, why are most administrators still from 2005? 22455 2023-08-15 Serendipity Why I stopped taking photographs almost altogether 28205 2023-11-06 Arbitration report Admin bewilderingly unmasks self as sockpuppet of other admin who was extremely banned in 2015 54589
It's an interesting mix, and I'm pleased to see a diversity of authors and article types represented, including research. ☆ Bri (talk) 22:40, 8 November 2023 (UTC)
- These are interesting, thanks for posting! — Frostly (talk) 06:14, 16 November 2023 (UTC)
- I think Recent Research is a bit of an unsung hero here. It's the Signpost's most consistently high-quality segment. Andreas JN466 14:40, 16 November 2023 (UTC)
More
| Subpage | Title | 7-day | 15-day | 30-day | 60-day | 90-day | 120-day | 180-day |
|---|---|---|---|---|---|---|---|---|
| Wikimania | Wikimania 2024 scholarships | 343 | 471 | 547 | 590 | 614 | 627 | 664 |
| Traffic report | If it bleeds, it leads | 539 | 743 | 794 | 846 | 868 | 886 | 951 |
| Recent research | Canceling disputes as the real function of ArbCom | 996 | 1263 | 1332 | 1405 | 1435 | 1469 | 1533 |
| News and notes | Update on Wikimedia's financial health | 1168 | 1516 | 1877 | 2020 | 2072 | 2154 | 2320 |
| In the media | Propaganda and photos, lunatics and a lunar backup | 1108 | 1535 | 1632 | 1735 | 1803 | 1822 | 1882 |
19-22 In the media
@HaeB: Hello, I've just incorporated my "In brief" bit into the current draft for next issue, as you suggested.
Actually, I was thinking about doing another one: I've found an article by an Italian sports magazine that explicitly cites the article for the Peruvian Clásico (an association football derby between Alianza Lima and Universitario) and encourages readers to take a look at it. That's because the latest, title-deciding match between the two teams took place just a few days ago and ended with some... interesting post-game antics, let's put it that way.
I fear this would be a little more off-topic than the "In brief" bits you usually publish, though. What do you think about it? Oltrepier (talk) 17:33, 11 November 2023 (UTC)
- @Bri: Thank you for taking a look at my entry! I don't know if it really deserves to be a lead story, though, since the original article I used as a source is pretty brief... Oltrepier (talk) 11:21, 12 November 2023 (UTC)
Odd heading for item
There's a heading under In brief that says "Israel and Palestine" but the text that follows says nothing about that, rather, it's about translation of Stephen Harrison's Slate article into Spanish. ☆ Bri (talk) 19:20, 16 November 2023 (UTC)
News and notes section about the Foundation’s audit report
Happy to see the Foundation’s audit report being covered in the Signpost. Question for you all: would you consider swapping out the image for something more neutral and less charged? The article points to our reserves, so maybe this image would add more context? Or even just a generic image could work fine.
A couple of additional notes for the editors’ consideration:
- 1. “The Foundation took $180 million in total support and revenue, vs. total expenses of $169 million.” It might be helpful to contextualize these numbers a bit more for the reader. We could add a sentence like “Both the revenue and expenses were within 5% of the $175M budget set in the Annual Plan. Revenue was exceeded primarily due to unrealized investment income, which is determined by fluctuations of investment values that can’t generally be predicted.”
- 2. “Its net assets grew by $16 million over the year prior, to $255 million at the end of the year.” I think context would also be valuable here, to understand what the $255 million in assets represents. Could we add something along the lines of “This is in line with the Foundation’s working capital reserve target of 12-18 months of operating expenses - which is standard practice for non-profits of the Foundation’s size.”
Let me know if you want to discuss any of this further. Thanks again for helping bring the audit report to a wider audience. –JBaldwin (WMF) (talk) 23:13, 16 November 2023 (UTC)
- I didn't pick the picture, but as for the text, I am rather against having you micromanage our coverage and insert corporate spin. (The 12 to 18 month reserve line in particular is trotted out every year ... it is misleading if you don't at the same time tell people by how many tens of millions you've increased your planned expenses.) Up to the EiC, of course, but my answer would be a "No". A "Hell, no", in fact.
- If you want your version in the Signpost it would be best to submit a text for the "News from the WMF" slot. Regards, Andreas JN466 02:22, 17 November 2023 (UTC)
Question for you all: would you consider swapping out the image for something more neutral and less charged? - Is this request about the image a joke? Sometimes humor does not translate well and I am having trouble understanding the request. Bluerasberry (talk) 16:06, 17 November 2023 (UTC)
- The quark image was a snarky placeholder that I substituted for the original. See Special:Permalink/1185359699 for what was there when Foundation personnel requested a change. ☆ Bri (talk) 16:50, 17 November 2023 (UTC)
- I've now put in a bar chart showing WMF finance data from 2003-23. Andreas JN466 19:57, 17 November 2023 (UTC)
- The quark image was a snarky placeholder that I substituted for the original. See Special:Permalink/1185359699 for what was there when Foundation personnel requested a change. ☆ Bri (talk) 16:50, 17 November 2023 (UTC)
19-21 Essay
@Theleekycauldron: Looks short. I'm anxiously awaiting the 2nd half! Is this a more general problem? Is there a good solution for the specific problem? Smallbones(smalltalk) 00:21, 13 November 2023 (UTC)
- not nearly ready for copyedit yet – it's gonna get a lot longer – but glad you like the pitch so far! theleekycauldron (talk • she/her) 00:32, 13 November 2023 (UTC)
- Considering that this hasn't progressed further since the above conversation and that this issue's publication deadline has passed, this piece should probably be dropped or postponed. Regards, HaeB (talk) 02:19, 20 November 2023 (UTC)
19-21 Recent research
As usual, we are preparing this regular survey on recent academic research about Wikipedia, doubling as the Wikimedia Research Newsletter (now in its thirteenth year). Help is welcome to review or summarize the many interesting items listed here, as are suggestions of other new research papers that haven't been covered yet. Regards, HaeB (talk) 21:13, 16 November 2023 (UTC)
- @JPxG: Should be publishable now if need be, although I may still be adding a bit unless I hear the publication engines revving up. Regards, HaeB (talk) 20:04, 19 November 2023 (UTC)
- My piece is rather long but I have so much more I'd like to contribute regarding analysis of what the author's conclusion means to us. I'd like to say that a) it was reinforced by another paper [5] that called the arbcom decisions amoral or "purely procedural ethics". More of this still resonates with me, especially whether Arbcom's form and function are a deliberate design from the beginning: that although it's described by outsiders as a "Supreme Court" of sorts for us [6], its function to quash editor conflict at the cost of everything else is "just how it is". ☆ Bri (talk) 21:29, 19 November 2023 (UTC)
- I do think there is room for another sentence or two; we also often include "see also"s for related research publications (ideally linking our own previous coverage). Feel free to add in case there is time (looks like we have a significant publication delay). Regards, HaeB (talk) 05:29, 20 November 2023 (UTC)
Can anybody figure out if Law & Social Inquiry that has the research I reviewed is the same as Law and Social Inquiry? It's very confusing, American Bar Association might be involved or might not. The website says "Published by Cambridge University Press on behalf of the American Bar Foundation".Never mind, I determined the ISSNs are the same. ☆ Bri (talk) 00:55, 20 November 2023 (UTC)