Wikipedia talk:Wikipedia Signpost/Newsroom/Archive 35

Page contents not supported in other languages.
Source: Wikipedia, the free encyclopedia.
Archive 30 Archive 33 Archive 34 Archive 35 Archive 36 Archive 37 Archive 38

Fourteen hours

I am going to bed, and will be ready to run it then. jp×g 05:11, 21 May 2023 (UTC)

This doesn't seem to have been the case. Similar to last time, it looks possible that once more a computer outage, travel issue, property theft or other mishap might have intervened. @Bri: In case you are around, would you willing to carry out publication again? Regards, HaeB (talk) 01:09, 22 May 2023 (UTC)
Back. jp×g 03:03, 22 May 2023 (UTC)
@JPxG Well, you were back for five minutes, like four hours ago. Is this going to happen now or what? Andreas JN466 06:51, 22 May 2023 (UTC)
Yes. jp×g 06:53, 22 May 2023 (UTC)

Three hours

Three hours left, now ☆ Bri (talk) 17:07, 21 May 2023 (UTC)

Confirming as usual that I should have RR in a publishable state before the deadline. (We had hoped to ramp up our contributor recruitment efforts for this issue, but this fell through for now, as one of our volunteers was not available; so I'm doing the writing by myself again.) Regards, HaeB (talk) 18:09, 21 May 2023 (UTC)

No hours

Done. jp×g 08:38, 22 May 2023 (UTC)

Publication schedule

Adam has set the publication date for the next issue to May 7, which works for me/"Recent Research". But I would like to confirm that it works for our EiC too - JPxG can you confirm?

Zooming out a bit, the current issue was published no less than 10 days after the originally schedule date (April 16). We really should avoid such long delays, for several reasons (e.g. I agree with the reasoning expressed here by Andreas on April 19). Sure, as recorded above, Smallbones' Opinion piece needed some more eyes, and there were some serious travel mishaps on April 20/21 - but these issues did not need to add up to ten days. Next time let's look at alternative options sooner (also considering that this is not the first time that JPxG has had to deal with such serious personal calamities around publication time). Also, JPxG, you have been going above and beyond the EiC duties many times which is much appreciated. But looking at e.g. your last-minute pre-publication edits here, I can't help pointing out that those core duties do *not* involve extensive copyediting of stories that had already been copyedited by other team members (introducing errors no less), or adding an entirely new opinionated story right before publication that was not entirely fact-checked. Don't get me wrong, if you have the time to do extra writing and copyediting beyond the core work of getting the issue out, that's great - but if it means an extra delay of several days, better skip the bonus and focus on the main prize. Regards, HaeB (talk) 04:19, 1 May 2023 (UTC)

I think it's better to have a publication date than to leave it unset for half the prep cycle. I'm quite open to having it changed, I just don't think any of us want to be in a situation where the date gets set two days before publication surprising everyone. It's easy to push a date back, but having at least a rough idea of when to get things done by is important.
FPs. Currently celebrating his 600th FP
!
17:56, 1 May 2023 (UTC)
I agree with HaeB that keeping to deadline is extremely important. In general it
  • Allows everybody to coordinate in the production. If one story is holding everything up (hmm), the other stories often start going bad. Just getting 'em in on time is the easiest method, but when there is no set time - when everybody knows that publication will be late - who's to blame?
  • Readers also should know when the publication comes out. If people are expecting it, there will be more readers.
While there can be good reasons to delay publication, they should be rare.
We do need to have an established backup plan for publishing if the EiC is unavailable. Just a very simple plan to get the discussion going: 2 or 3 editors are appointed to publish if the EiC hasn't made prior arrangements. Whichever of them are available can make the publication decision one day after the original deadline. I assume that @Bri: should be one of the appointed editors or other plans should be made to get a regular publisher.
Just to keep everything really, really simple, let's establish a standard publication schedule, say the 2nd Sunday of the month, and the last Sunday of the month at 4pm Eastern Time. Christmas and sometimes Thanksgiving messes this up, but we can schedule those well ahead of time. Smallbones(smalltalk) 16:41, 2 May 2023 (UTC)
I think I'd somewhat prefer first Sunday of the month, but that's for vague reasons of feeling monthly overviews are more fun if they're as early as possible in the month.
FPs. Currently celebrating his 600th FP
!
17:45, 2 May 2023 (UTC)
So are you saying 1st and 3rd Sunday of the month? Note that there are some months with only 4 Sundays and some with 5 Sundays. (I think it's 8 with 4 and 4 with 5.) Somehow I like the last Sunday of the months and then two weeks later (i.e. the 2nd Sunday of the month). Maybe it's because I'm just used to the last Sunday of the month, but it seems "more regular" to count from the last Sunday. OK, I'll admit it, counting from the 1st Sunday might mess with July 4 (and July 1 for Canadians). Smallbones(smalltalk) 18:35, 2 May 2023 (UTC)
Aye, that'd be my preference, but it's a weak preference.
FPs. Currently celebrating his 600th FP
!
09:59, 5 May 2023 (UTC)
Am in broad agreement. We do need a backup plan. If Bri and you, Smallbones, are willing ...
Moreover, we should make sure the publication instructions are up to date in case the worst ever happens. Andreas JN466 22:02, 2 May 2023 (UTC)
Thanks everyone for weighing in. (And to clarify, I didn't mean to blame Adam for going ahead and updating the deadline template with a provisional publication date for this issue. But I do think that this deadline needs be meaningful and in particular be anticipated to work for whoever will carry out the publication.)
So let's stick with the May 7 date, and if JPxG isn't around by that time, aim to have Bri do the publishing again, presuming that two or three other team members have expressed their agreement that the issue is ready to go out. (And considering the fact that JPxG hasn't yet responded since this thread was started four days ago - but made hundreds of non-Signpost edits elsewhere in the meantime - I would like to mention point 2 here.) Regards, HaeB (talk) 05:32, 5 May 2023 (UTC)
Oh, no offense taken. I'm only doing it because no-one else was. I've been doing it a while, though. Used to post about it, but after a bit...
FPs. Currently celebrating his 600th FP
!
10:02, 5 May 2023 (UTC)
So let's settle on 1st Sunday + 3rd Sunday as the default schedule, as proposed above. For the record: We since published another issue on May 22 (one day after the third Sunday), and I have just updated the deadline template with June 3 (first Sunday) as the publication deadline for the next one. Regards, HaeB (talk) 05:11, 25 May 2023 (UTC)

Sunday

I can be backup publisher but only for a time window on Sunday night, if the issue is ready. After that, I'll unavailable for a couple of days. ☆ Bri (talk) 15:33, 5 May 2023 (UTC)

I can help complete the issue (copyediting, approval if needed) on Sunday if Bri can publish and nobody objects before then. I've got a couple additions to Itm, but won't have another article. Smallbones(smalltalk) 18:35, 5 May 2023 (UTC)
I've kind of accepted that part III of April Fools is delayed by illness.
FPs. Currently celebrating his 600th FP
!
05:47, 6 May 2023 (UTC)

Suggestions subpage

Given that things on the Suggestions subpage keep falling through the cracks, should we get rid of it and invite people to post directly to the Newsroom? Andreas JN466 12:42, 8 May 2023 (UTC)

(For context, right before Andreas' comment I had gone through that page and marked several old suggestions as done, but many others indeed do not seem to have been taken up.)
I agree with the diagnosis but not with the suggested remedy. This talk page here is already busy enough, and I don't think the problem is merely that the regular writers (for N&N and ITM in particular) have so far been unaware of the existence of the suggestions page. I mean, we can and should engage in regular exhortations for them to check and process these suggestions, but ultimately this might also be a bit of a capacity problem. On that note, I think we could do more to encourage the authors of suggestions to already include a write-up themselves, in particular for smaller news items. The edit notice already includes words to that effect (If you write a short description of your suggested item as it would appear in the Signpost, you increase the chances of it making to publication). But the text preload hasn't reflected that so far, it merely said The Signpost should write about.... I have updated it accordingly. Regards, HaeB (talk) 07:11, 26 May 2023 (UTC)

Russian fork

It appears from this on Meta that User:Drbug, the director of Wikimedia Russia, is starting a Wikipedia fork.

User:Drbug, am I getting this right? I've read the Google translation linked above. Do you have any additional comment for Signpost readers? And what is your status with Wikimedia Russia? Best regards, Andreas JN466 21:30, 25 May 2023 (UTC)

Discussion on the mailing list: [1] and following (the Hyperkitty archive is down right now, hence this alternative link) Andreas JN466 18:11, 26 May 2023 (UTC)

I'm out this month

I don't think I'll be able to contribute much this month due to other commitments for this week, and weekend of publication. ☆ Bri (talk) 16:45, 1 June 2023 (UTC)

I'm about the same. If I have a hot story, I'll try to add it, but don't count on me. Smallbones(smalltalk) 18:48, 1 June 2023 (UTC)

In the news

@Smallbones: Just want to mention that I don't mind if you rewrite or remove everything I've added to the 'In the news' section this week. Not quite sure if these texts are what is wanted in this part of the Signpost. Perhaps somewhere else...not sure. I'm not a listed editor anymore. Jonatan Svensson Glad (talk) 17:06, 1 June 2023 (UTC)

@Josve05a: - don't worry about me. I don't think of In the media as owned by anybody in particular. Only that people should feel free to contribute and if I feel too free sometimes, somebody should tell me about it. Smallbones(smalltalk) 18:46, 1 June 2023 (UTC)
Heh, I'll leave "SCOTUS punts on section 230 decision" entirely up to you. Jonatan Svensson Glad (talk) 21:12, 1 June 2023 (UTC)

"Research under a cloud" submission

User:XOR'easter/Research under a cloud was proposed for The Signpost on the submissions page. I don't know if anybody else has seen it there. It's rather unusual and I'm not sure how it fits in in our usual article format. It's a lengthy, footnoted essay. I'm tentatively thinking we publish a summary that links to the original. Thoughts? ☆ Bri (talk) 16:46, 5 May 2023 (UTC)

I've skimmed it twice now. I don't mind the footnotes. It would IMHO benefit from a calmer approach and cutting the length a bit. @XOR'easter: By calmer approach I mean, it looks like you're trying to run the paper's authors off the planet - just assume some good faith, ignore some of the little things, don't go to extremes with the criticism - the reader will believe more of what you say and you won't make any life-long enemies. It's up to XOR'easter to make any changes he wants (except for the headline-that's the editors' job. All suggestions appreciated.), but if he doesn't want to make any changes I'd probably end up flipping a coin or asking @Bri and HaeB: (by email!) whether it should be published. Hope this helps! Smallbones(smalltalk) 19:11, 5 May 2023 (UTC)
I don't think there were any "little things" to ignore: the paper is full of interlocking problems that all compound upon each other. Fumbling basic statistics is a serious issue. So is misrepresenting one's subject matter by selective quotation; I wasn't the one who used the exact phrase "academic misconduct" (that was
talk
) 19:58, 5 May 2023 (UTC)
When I first saw the paper, I thought I would like it. Only after reading did the troubles start to appear. Huh. Reminds me of another paper we've been reporting on a lot of late. Andreas JN466 19:45, 7 May 2023 (UTC)
The topic is obviously in scope for "Recent research". I had already seen this submission earlier and tentatively put XOR'easter down as reviewer in this item's entry on RR's customary todo list (posted here). However, I haven't had a chance yet to read through the full submission (or the paper), will do so later today. In the meantime, on first glance I agree that it could use some editing - for starters, by opening with a more informative summary of the paper's methodology, context and main claims, before going into the details of the criticism. That said (and again, without having examined the specifics here yet), well-argued views are of course in scope - as Smallbones himself is well aware from his many own Signpost opinion pieces, including the one in the last issue that attracted so much discussion above. We do need to make sure they don't distort the facts though, so XOR'easter's desire to back up the review's statements in detail is commendable. The customary format of RR has a "Supplementary references and notes" footnote section for that kind of thing, but let's see if that suffices to make it reader-friendly enough.
I'll see to copy this over to the RR draft later today (Pacific time) and do some edits myself. @XOR'easter: might you be around tomorrow to address feedback and potentially make further edits as required? Regards, HaeB (talk) 20:26, 5 May 2023 (UTC)
I'll be around some tomorrow.
talk
) 22:11, 5 May 2023 (UTC)
Update: I have since added it to the RR draft and made some edits for clarity, brevity and tone - while taking care not to misrepresent the reviewer's views; more eyes still welcome (Smallbones and Bri: see also email). After having fact-checked most - if not all - criticisms in the review, my impression is that its negative assertions seem well-founded factually, even if I myself may not share all the opinions that it expresses.
However, I now have a different concern about the review that may warrant postponing it: It doesn't examine the paper's statistical reasoning per se (only criticizes - justifiedly - various assumptions it makes to enable that reasoning). But having now read the paper myself, there seems to be a serious problem there too, see my detailed notes here: Wikipedia_talk:WikiProject_Women_in_Red#Misinterpretation_of_p-values_in_the_"Primer_index"_comparisons. @XOR'easter: thoughts?
Regards, HaeB (talk) 19:37, 7 May 2023 (UTC)
Might be worth publishing now and leaving that for a sequel? You've got nearly 5,000 words as it is. Andreas JN466 19:48, 7 May 2023 (UTC)
Oh, the review wouldn't need to describe that concern (about the authors' interpretation of their p-values) in the same level of detail as I did here. (And we don't usually feature the same paper twice in RR.)
In any case, you're correct that 5000 words is already quite long. As mentioned, I already tried to reduce length somewhat by removing some text that wasn't directly addressing the paper's claims, but if anyone else sees ways to condense it further without distorting the reviewer's main views, go ahead. We could still add a link to the full version at User:XOR'easter/Research under a cloud. Regards, HaeB (talk) 20:16, 7 May 2023 (UTC)
I could quickly put together a paragraph that summarizes your concerns and slot it in at the end of
talk
) 20:16, 7 May 2023 (UTC)
That should work, yes (and could probably be condensed a bit more still without losing much information). In the last sentence one could also mention that - in fact, the difference in sample medians between kept and deleted BIPOC women was even larger than that for white men, according to figure 2. Regards, HaeB (talk) 20:48, 7 May 2023 (UTC)
I know nobody asked me but I find the vitriol with which this paper has been attacked mildly disturbing... This is beyond the norm and its hard to separate the intensity of the vitriol from the subject matter of the paper. I understand the frustration and the desire to offer effective critique of academic pieces about wikipedia, but why is this bleeding over into personal attacks on the author's credibility and generalized condescension? Horse Eye's Back (talk) 16:12, 1 June 2023 (UTC)
Could you be more specific - which part of Wikipedia:Wikipedia Signpost/2023-05-08/Recent research do you consider to be personal attacks on the author's credibility (the paper actually has three authors btw)?
Since you are commenting here, I assume that your concerns refer to the Signpost's published coverage and not the community discussion in general (where e.g. an AAAS and ACM fellow alleged "academic misconduct" by the authors - a stronger statement that the Signpost piece does not make).
I will say that, generally speaking, in situations like this it is not unreasonable a priori to worry about potential groupthink and defensiveness by the community as a whole. However, 1) the general concerns about biases that this deeply flawed research had appeared to support are actually held by wide parts of the editing community itself, and 2) there have been lots and lots of other academic papers about (in particular) gender gaps and biases on Wikipedia (many of which were likewise covered in "Recent research" and/or discussed elsewhere on-wiki) that did not elicit such a nearly unanimously negative reaction.
Lastly, I will admit that while editing the piece for publication, we didn't expend a lot of effort to make it easy for the reader to distinguish the major flaws identified by the review from smaller issues. (This is also a challenge when reading through the community discussions. E.g. just to take the one aspect of the paper that I examined in detail myself, I would say that the misinterpretation of p-values seriously undermines one of the paper's main conclusions, whereas off-by-one errors do not - even though I agree that they are on some level worth pointing out, and cumulatively might substantiate an impression of sloppy data analysis.) Regards, HaeB (talk) 04:06, 4 June 2023 (UTC)

References

  1. GraphPad Prism
    . Retrieved 2023-05-07.

Fifteen hours

It begins. jp×g 08:25, 2 June 2023 (UTC)

Shall work on FC. If I don't finish it I'll move anything I don't to next week
FPs
.
10:00, 2 June 2023 (UTC)
Sorry about not helping much, Adam. It's been a busy week and I forgot to write the poems. QuicoleJR (talk) 21:34, 4 June 2023 (UTC)
  • Sike! it was 15 hours until it was 24 hours. I think I read the UTC counter wrong. jp×g 22:52, 2 June 2023 (UTC)

I'm going to the beach and won't be back for a day or so. There are only 2 articles close to being ready for publication, although News and notes has a couple of good stories. I'll suggest just postponing publication until a week from tomorrow. Smallbones(smalltalk) 16:23, 3 June 2023 (UTC)

It seems that, right now (technically 25 minutes after the "publish" button was to be pressed) there are only three articles actually ready to be copyedited or published. jp×g 20:25, 3 June 2023 (UTC)

I think delaying publication by a week would be bad, for the reasons discussed repeatedly before (cf. above). And publishing with three substantial sections is fine IMHO; perhaps quite a few readers will actually appreciate being able to read the entire issue for once ;) That said, I just went ahead and moved the date to the usual Sunday (instead of Saturday) publication deadline - looks like I had messed up by one day when setting it earlier, my apologies.

What's more, we also have Wikipedia:Wikipedia Signpost/Next issue/Wikidata, which appears to have been intended for publication back in February but somehow got lost. How about adding it to this issue? Regards, HaeB (talk) 22:41, 3 June 2023 (UTC)

Agree with HaeB. Andreas JN466 22:56, 3 June 2023 (UTC)
I'm going to try to finish FC in time. Or at least get it mostly done and shove the rest to next issue. If you see it unfinished near publication time, just move everything unfinished out to a holding place.
FPs
.
11:02, 4 June 2023 (UTC)
Featured content is ready for copyedit. Sorry it's so late.
FPs
.
16:35, 4 June 2023 (UTC)
  • Two hours and forty-two minutes remaining to publication, if the thing is to be believed. I will be there then, or here now, as the case may be. It looks like we have about four or five articles this issue. jp×g 17:23, 4 June 2023 (UTC)
Thirty seconds. jp×g 19:59, 4 June 2023 (UTC)
I'm all done! It looks ready to me. Smallbones(smalltalk) 20:25, 4 June 2023 (UTC)

Back from the beach just in time!

I've made changes to the last two paragraphs of the Ruviki story in News and notes. They now read

"The move has been welcomed by those media outlets which are still allowed to publish in Russia, such as TASS. Ruviki will not be the first Wikipedia fork to be published in Russia. Runiversalis which had difficulties getting started last fall, now claims to have 1,920,588 entries, but does not seem to be operated according to the usual Wikipedia rules. Adding to the expected confusion is that the new fork's name "Рувики", is pronounced "Ruviki", the same as a common name for the original Russian Wikipedia, RU.wiki."

"Medeyko has since been replaced as director of Wikimedia Russia by Stas Kozlovsky. For related discussion see the Wikimedia-l mailing list thread."

@Jayen466 and JPxG:

I'll also make short changes to the brief Ruviki item at Itm.

Smallbones(smalltalk) 19:57, 4 June 2023 (UTC)

Thanks. I've added a link to the Wikipedia Weekly Facebook discussion as well. Andreas JN466 21:30, 4 June 2023 (UTC)

I just happened to take a look at that ruwiki story too and it still seems problematically unbalanced even after Smallbones' improvements. Per e.g. the Russian Wikinews' coverage or this comment, the purported reasons for the fork as given in the announcement (that the current version of our story very extensively quotes from) are misleading or at least secondary, and the more important context is that "at the moment Russian Wikipedia is on the brink of being banned in Russia. There is a list of Wikipedia articles that are banned in Russia, and if in the past mostly it were a couple of articles related to suicide or illegal drugs, as you can see from the list itself, https://ru.wikipedia.org/wiki/Википедия:Страницы_Википедии,_запрещённые_в_России , it has expanded exponentially since the Russian invasion in Ukraine and now mostly consists of articles related to the Russian invasion. [...] there are forks being created so that the most successful of them gets to be named the Wikipedia's replacement by the Russian government, while the Wikipedia proper gets banned in Russia" (with Runiversalis being another of them; the article Russian Wikipedia currently also describes it as "a pro-Moscow partial fork of the Russian Wikipedia").

I know and respect that Andreas is skeptical of the dominant Western perspective in this geopolitical conflict (recalling e.g. how he vocally defended Runiversalis' take on the Ukraine war against critical coverage in Euronews, in context of a previous ITM). But at the very least we need to give more room to the views of the Russian Wikimedians who objected to the fork and give some hints on why they thought it necessary to remove the fork's founder from his post at Wikimedia Russia and block him indefinitely on Russian Wikipedia. Regards, HaeB (talk) 20:39, 4 June 2023 (UTC)

Ok, I've made my final edit to the New and notes section. A few trouts need to be shared here. @JPxG, Jayen466, HaeB, and Smallbones:
  • JPxG - you are the EiC - you need to take charge here. Then publish. Or decide firmly not to publish. This is why you get paid the big money (from the secret Soros account).
  • Andreas - you missed the obvious story
  • HaeB - go ahead and edit if you think it's needed.
  • Smallbones - quit being such a self-righteous stuffed ass.
Smallbones(smalltalk) 21:41, 4 June 2023 (UTC)
Hey, my time was limited. I was unaware Drbug had been indeffed, hadn't read Bohdan's long comment, etc. I brought the topic here to this page a week ago (three sections up).
Newsrooms are for discussion. So it would be nice if people who had further insight into a story like this said so a bit sooner rather than 5 minutes before publication, or, God forbid, write an article themselves about it. Andreas JN466 21:50, 4 June 2023 (UTC)
Well, I had assumed you would be aware at least of the Russian Wikipedia village pump discussion, given that it was already linked in the May 24 Wikipedia Weekly post (and that you are a regular of that group and "Liked" that post). Bohdan's comment and my own link to Kozlovsky's comment (that Drbug had prepared this fork in secret while serving as director of Wikimedia RU) were also posted there more than a week ago, and don't take that long to read. To miss both the state censorship context and the chapter COI aspect was a major oversight; don't blame this on me for not weighing in earlier. That said, I do appreciate your updates to the story and also just added more myself. IMHO this story is OK to publish now even though I think it still gives too much weight to the less relevant (and possibly actually misleading) parts of the announcement. Regards, HaeB (talk) 22:22, 4 June 2023 (UTC)
I just haven't been on Facebook this past week ... also, I don't think our readers are stupid. While I think the story reads fine as it is now, and am grateful for the additional info, I also would have enjoyed it if it had been left as it was and some of these perspectives had been brought out in community comments on our pages examining the likely background of the good Doctor's statement. Andreas JN466 22:32, 4 June 2023 (UTC)

I did learn something from all of this: if you ping yourself, nothing happens. Smallbones(smalltalk) 04:02, 5 June 2023 (UTC)

Update

It is not interfering with this issue's publishing process at all, since I will be home in a couple days, but it has come to be almost a tradition of mine to post here in a panic whenever something fucked up happens to my laptop, so why stop now? Today I notice that the center pin of the power input jack seems to have completely detached, making it impossible to power on. Oh well. I have a few of those connectors in my workshop, which I can solder on when I get home. What a drag!! jp×g 06:56, 26 June 2023 (UTC)

The standard coaxial connector for low voltage DC, in case this happens to anyone else, is quite cheap (about ten cents) and easy to replace on most motherboards (just make sure to solder it on robustly and deflux afterwards): many "dead" devices can be saved this way in about fifteen minutes of work, most of it waiting for the iron to heat up. jp×g 06:59, 26 June 2023 (UTC)


Twitter

User:HaeB, could you change the pinned tweet of the Twitter account to the current issue, and could you re-tweet [3]? It's nice to see that people appreciated the coverage. --Andreas JN466 15:36, 9 May 2023 (UTC)

Thanks, both done.
Apropos, a note for the record while we're talking about Twitter/social media:
The automatic Twitter --> Mastodon bridge appears to have broken recently. This was not unexpected given Twitter's recent API changes; in fact for @wikiresearch it already stopped working several weeks earlier. However, the operators of the service I have been using for both, moa.party, said earlier that they would try to keep it going. They also appear to have been soliciting donations toward that goal, none of which seem to have been spent yet. So I thought there might be some hope that this would resolve itself. But at this point it looks one should look for a different solution. I already have some options in mind try out next (e.g. involving Buffer or IFTTT), but if anyone has recommendations for a currently working solution, let me know. (Worth noting that the IFTTT-based Twitter --> Facebook bridge still works.)
Generally speaking though one should also be aware that automatic reposting is frowned upon on Mastodon. What's more, as the helpful notes at m:Wikis_World#Content warn Wikimedians, there appears to be a lot of finger-wagging in that community about various other Mastodon-specific norms and practices (I already got gently scolded by Legoktm about violating one of them with last issue's announcement post). In particular, there seems to be a strong community expectation on Mastodon that one must be "engaging", "interacting", "having a conversation" etc. instead of merely posting news. And frankly I don't have the time to do that for @wikisignpost, nor do I think that this should be our role on social media as a news publication. (Checking mentions and doing some boosts/retweets - like in the example Andreas gave above - is OK and I've been doing that ever since I volunteered to post our new issue announcements over two years ago; there hasn't been a lot of mentions anyway.) Ideologically speaking I remain a strong supporter of social network federation based on free, open platforms - that's why I created the Signpost and WikiResearch Mastodon accounts (the latter back in 2017 already). And there seem to be quite a few Wikimedians who enjoy Mastodon as a conversational platform, good for them! But if it becomes too much of a chore to comply with various Mastodon-specific idiosyncracies, then I would also be OK with leaving the Signpost's account there dormant for the time being and focus on social media channels where we can (re)post our new issues and stories more efficiently without platform-specific efforts.
Regards, HaeB (talk) 21:31, 9 May 2023 (UTC)
I agree with your conclusion on Mastodon regarding engagement being what people want, but in reality e.g. @arstechnica has 20k followers and doesn't seem to be replying to anyone and I don't think people care that much. I think the best way to handle it is clearly set expectations in the bio like "This is a primarily announcement-only account, we encourage replies to articles on-wiki". Certainly a low-engagement Signpost account is better than none :)
My gentle scolding was about alt text for accessibility, which is a IMO a much bigger deal on Mastodon and a good practice just in general. A lot of people won't boost posts without alt text, including the @Wikipedia account (hence my message). I hope that's a pretty simple thing to address going forward?
I'll also plug my masto-collab tool, which hopefully makes it easy for other people to participate in writing/approving posts without needing direct account access to help share the workload. Legoktm (talk) 23:31, 9 May 2023 (UTC)
I think the best way to handle it is clearly set expectations in the bio like "This is a primarily announcement-only account, we encourage replies to articles on-wiki" - counterpoint: that's actually much more likely to make the account a target for those who try to "enforce" such expectations, no? Certainly a low-engagement Signpost account is better than none :) - thanks, and to clarify, I am not actually worried about you throwing us off your server ;-) To the contrary, the fact that you are known and trusted as a Wikimedian (and that Wikis.world seems to be well-run, taking good governance seriously with monthly reports and all) made the decision much easier to set the Signpost account there. But due to federation, one also needs to worry about the rest of the fediverse. See e.g. this Columbia Journalism Review article about an entire Mastodon server full of professional journalists (and set up by a NPR journalist) getting defederated (mass-blocked) by hundreds of instances.
I of course agree that alt texts are a good practice, and have already added them a few times myself. But regarding I hope that's a pretty simple thing to address going forward? - is it? E.g. going back to my first question, do you happen to know a currently working crossposting service that preserves alt texts? Again, for me this is a recurring task that I have taken on over two years ago and have so far managed to carry out for every new issue, even at times when I was otherwise entirely off-wiki and not involved with the production of the issue. But I can only justify that if it doesn't become too complicated (it's also not that we seem to be getting tons of readership from any of the three platforms). Or to put it differently: Often the choice might not be between a Mastodon post with or without beautifully crafted alt text that has a 2.37% chance of actually making the life of a visually impaired person a little better for 10 seconds, but between a Signpost announcement with or without image, or a Mastodon post and no Mastodon post.
I'll also plug my masto-collab tool - saw that, looks really interesting and I'm excited to see how this works out for @wikipedia. But I'm not sure how relevant it is for the core task here of posting the announcement for each new Signpost issue. And considering that we are even having trouble processing on-wiki submissions for the Signpost itself (see two sections above), I don't see us operating up a full social media post submission queue anytime soon.
Regards, HaeB (talk) 03:16, 10 May 2023 (UTC)
So... I don't think you or any other Wiki Worlds user should be worrying about the server being defederated, that's for us as moderators to pay attention to and ensure doesn't happen. journa.host was honestly doomed from the start, it was as if 1,000 Citizendium users (ha) joined Wikipedia and started asking how to get their accounts verified (I have a half-written blog post going into detail on this, I'll finish it one day...).
No idea about crossposters, I suspect you could probably try combining nitter's RSS feed (example) with one of the RSS-to-Mastodon scripts? In any case, I would rank your choices as Mastodon post with alt text > post without image > post w/ image but w/o alt text > no post at all. I also think that starting with simple alt text is better than nothing, e.g. for this post the alt text could've just been "Vatican flag", which I hope isn't too onerous.
The reason I brought up masto-collab is that eventually it'll support editing posts, so y'all theoretically could have a workflow in which one person does the main post, and another interested person could add/improve alt text (or the post itself). I don't know if that's more trouble/overhead than its worth... Legoktm (talk) 18:43, 14 May 2023 (UTC)
(Dumping some detailed technical notes and semi-offtopic responses here for lack of a better location; the TL;DR is that I got the Twitter --> Mastodon bridge to work again, with some limitations.)
Thanks for the suggestion about nitter! I actually ended up using that, in combination with feed2toot, and got it to work (also for @WikiResearch). It does have some limitations however compared to the now broken moa.party solution. In particular, it doesn't support images. The nitter RSS feeds contain them though, so it might be possible to implement that by tweaking feed2toot - recommendations welcome. (I already applied a small patch to the feed2toot library anyway to have its toots link back to Twitter instead of nitter.) Also, for some reason the @wikisignpost RSS feed on nitter.nl was missing the last announcement tweet until today, i.e. for over two days after it had been posted. (Same on other nitter instances, which makes me wonder if it is a bug in nitter rather than a random networking glitch.)
This is running as a cronjob my own computer now, and I'll soon move it to an AWS server that I'm maintaining for other purposes anyway, so hopefully that will be a stable, low-maintenance solution. And it looks like the nitter folks expect to be unaffected by the Twitter API changes, so this might be one of the most future-proof free options. Regarding the other options I had mentioned above: For Buffer I got the impression that one would need to have switched to posting everything via their own app/website. And after my comments above, IFTTT announced that their Twitter applets are no longer supported under their free plan, starting this week. I reluctantly signed up for their Pro plan for now since I'm still using that for other purposes - including the Twitter --> Facebook bridge for the Signpost, and for continuously updating a spreadsheet with new Wikiresearch tweets as part of the production process for "Recent research". While this is not very expensive, I'm obviously not super eager to indirectly subsidize Twitter in this way and am looking for alternative solutions there too. (I've tried this script for the spreadsheet with the nitter RSS feed, but haven't quite gotten it to update the sheet consistently, apparently because of this problem.)
I don't think you or any other Wiki Worlds user should be worrying about the server being defederated - I brought up that example from the Columbia Journalism Review just to illustrate in general that on Mastodon is is not sufficient to get along with the local server's community and moderators. journa.host was honestly doomed from the start - I mean, it's actually still up and running, no? On the other hand, the troubles it had were clearly not merely due to missing verification features, but also e.g. due to banning/not banning a user for linking to a New York Times article with the comment "This seemed like careful, thorough reporting" (at least as reported by the New York Times itself). I have a half-written blog post going into detail on this, I'll finish it one day - I would be interested in reading it! Besides the journa.host affair, I've also seen e.g. this Hacker News thread (referring to [4]) which adds to the impression that the fediverse (yes, also its main connected component) contains some rather toxic groups whose attention it is prudent to avoid. Obviously getting a bit off-topic here, but I hope you understand that such things are not entirely theoretical concerns regarding our little news publication's social media presence either. While it might be easy to follow the advice "just avoid <topic X and topic Y> in this community" on a personal account, the Signpost regularly covers contentious or sensitive topics, which often also make it into the highlights that I'm compiling for each issue's social media message. Now, as you might be aware, I recently spent a significant amount of time defending someone else's Signpost writing about an academic publication in such a politically contentious topic area (that I'm personally not particularly interested in) against unjustified criticism of some people heavily invested in one side of that topic. And while I'm OK with doing that within our own community and norms here - it comes with the territory - I'm not eager to have to do similar things in another community whose written and unwritten norms I'm not as familiar with and where I'm less well equipped to deal with what the equivalent of wikilawyering may be there.
the alt text could've just been "Vatican flag", which I hope isn't too onerous - it was not so much about crafting the alt text, but having to re-do the entire post manually on Mastodon because the bridge didn't preserve alt text. Anyway, as mentioned above, the problem is moot for now since the new solution unfortunately doesn't preserve images at all.
Regards, HaeB (talk) 05:41, 26 May 2023 (UTC)
Small technical update (mostly talking to myself here, but it seems worth documenting this somewhere):
Unfortunately the RSS feeds on nitter.nl have broken in recent days (I'm getting errors both for wikisignpost and WikiResearch). So I'm switching the mirroring script to nitter.net instead, where both pages still exist. (However, again with the aforementioned problem that the wikisignpost feed seems to lag; it doesn't yet contain the announcement tweet for the new issue right now, 100 minutes after it was posted; whereas the WikiResearch feed updated with a new tweet within a minute or so). Regards, HaeB (talk) 05:54, 5 June 2023 (UTC)
Another log-like note for the record, apropos finger-wagging on Mastodon: I just happened to notice that the Wikis.world server owners appear to be preparing a new bots policy that my now existing setup would probably violate in several ways. (Fortunately it is not in force yet.) I'll probably try to keep it going nevertheless, perhaps it will turn out similar to Legoktm's optimistic hunch about another potential issue above (I don't think people care that much). But on the other hand they already state explicitly in the draft that (an analogue of) the popular CongressEdits bot likely won't be allowed, so I guess that "usefulness to Wikipedians" won't be a get-out-of-jail-free card. Regards, HaeB (talk) 02:19, 7 June 2023 (UTC)
@HaeB: glad you're watching my edits :) I'll clarify what "bots" means on the page, but in my mind bots are about auto-generated posts rather than how the post was technically posted. Since humans are ultimately writing the posts, I/we wouldn't consider the Signpost a bot account, if that makes sense.
(At some point I will reply to the other stuff you wrote, just haven't had the time to sit down and collect my thoughts!) Legoktm (talk) 03:08, 7 June 2023 (UTC)
Thanks, that sounds like a helpful clarification! (But then again doesn't CongressEdits also just cross-post human-written edits? Anyway, that's up to you folks to sort out; I'm just a bit curious whether this is a reaction to already occurring concrete problems with bots on your server, or an attempt at proactive rulemaking where one tries to anticipate potential problematic behaviors and prohibit them in advance; it seems that the risk of collateral damage/overregulation might be higher in the latter approach.)
PS: Also, just so that you don't feel too much wiki-stalked, the reason I happened to see this yesterday was that I was catching up on Wikimedia-l a bit and followed the link to your first monthly report for @wikipedia. Nice to see this kind of reporting, hope you can keep it up, and looking forward to read your thoughts on those general topics too sometime.) Regards, HaeB (talk) 04:55, 8 June 2023 (UTC)
Currently we don't allow bots (vaguely defined) at all, this is a proposed change to that. We can probably move that discussion to the thread on Meta-Wiki?
Also, I've created MediaWiki:Realme-config.json, so if you add a link to the Signpost wiki page in the profile metadata of the account (see docs), then it should show up as verified. Legoktm (talk) 18:10, 9 June 2023 (UTC)
Currently we don't allow bots (vaguely defined) at all - "at all"? Doesn't https://wikis.world/about say that cute, funny, or useful bots are allowed? Anyway, I'll take a look at that Meta-wiki discussion and see if I have anything useful to contribute.
if you add a link to the Signpost wiki page in the profile metadata of the account (see docs), then it should show up as verified - I did, but it is not yet showing up as verified. Doesn't this still require the linked wiki page to explicitly link back to the Mastodon profile? In that case let's rather use https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/About . I assume that would require you to update the entry at MediaWiki:Realme-config.json?
PS: to log another small technical update here to my June 5 comment above, the nitter.net RSS feed still doesn't contain the tweet from that day (also on several other Nitter instances). Guess it's time to file a bug with the Nitter folks.
Regards, HaeB (talk) 05:13, 11 June 2023 (UTC)
Doesn't this still require the linked wiki page to explicitly link back to the Mastodon profile? no, that restriction only applies to user pages.
Taavi figured out the issue, the main Signpost page just needed to be purged so the <link rel="me"> would show up for logged-out requests too. He did that so if you want to remove and re-add the link from the Mastodon profile it should now work :) Legoktm (talk) 16:10, 14 June 2023 (UTC)
Just did that, and it shows up as verified now. Thank you (and Taavi of course) for sorting it out!
To continue logging updates regarding the bridge: While I haven't gotten around to filing a bug with Nitter yet, the aforementioned tweet is still missing from the nitter RSS feed - but it did catch a subsequent self-retweet of that same tweet. (And I happened to notice that openrss.org also still offers Twitter RSS feeds; theirs doesn't have the same problem, so maybe I'll try that as a plan B in case the Nitter problem persists.) Regards, HaeB (talk) 05:27, 18 June 2023 (UTC)
I really thought this would be my last comment in this thread, but Twitter's management wouldn't have it... The collateral damage of last weekend's (alleged) crackdown on evil data-scraping AI startups killed Nitter too (as well as the aforementioned alternative openrss.org), at least for now. So I spent some time today to adapt the Mastodon bridge to use a different RSS feed that is miraculously still alive. (Also, while there has been much skepticism about whether Twitter's measures will really be temporary as they claimed, it actually seems that anonymous reading is possible again today, at least from some IPs.)
Apropos, we got a nice shout-out for the current issue from Annie [5] (thread with 241K views on Twitter so far). Regards, HaeB (talk) 06:43, 5 July 2023 (UTC)

Good meowning

On schedule. jp×g 17:25, 18 June 2023 (UTC)

A slow news fortnight. Not much seems to have happened. The Germans have a word for that: "Sommerloch" (summer hole). User:HaeB, I added three "Other recent publications" to the Recent Research draft. Andreas JN466 19:38, 18 June 2023 (UTC)
Thanks, much appreciated. Just be aware that as part of the production process for this section, we normally always try to also feature papers on Twitter first and list them in the to-do list Etherpad in case they attract the attention of someone volunteering to cover the paper in a full review, before they get relegated to the "Other recent publications" section (which, as its intro explains, has a bit of a leftovers character).
Regarding Sommerloch, worth noting that (as you had yourself correctly pointed out above recently) there are quite a few items on the suggestions pages which should be suitable. (I had actually envisaged to cover one or two of those myself last time before getting drawn into fixing other things, and may still try later, but perhaps not in time for this issue.)
Regards, HaeB (talk) 20:32, 18 June 2023 (UTC)
Re suggestions – I feared as much but monitoring that page as well is too much for me to take on. :(
Could someone else keep an eye please? <hint>I think traditionally it's an EiC job. </hint> Andreas JN466 21:17, 18 June 2023 (UTC)
Yes. jp×g 02:20, 19 June 2023 (UTC)
  • I note that there are four stories for this issue, which is pretty thin. I will see what can be rustled up from suggestions and submissions, and if there are any drafts that can be prepared for publication. If not, I think it should be delayed until tomorrow, and I can come up with something. jp×g 02:20, 19 June 2023 (UTC)
    I just beefed up ITM and N&N a bit further myself (covering several items from the Suggestions page, but there is still more left there). I think we have enough material to publish now, and would also like to point out that we got positive reader feedback last time for going ahead with a timelier but smaller issue. Regards, HaeB (talk) 03:20, 19 June 2023 (UTC)
    Thanks for the additions – the issue looks a little less impoverished now – and I agree on the timeliness aspect. Andreas JN466 06:34, 19 June 2023 (UTC)
  • Bri did the Traffic Report for me the last time around (after all, just copying the weekly top 10 is easy), but of course I wouldn't be this lucky again. While apologizing for laziness (not helped by how I keep waiting for others to write in the Top 25 Report rather than doing it all myself), to compensate, I'm already starting the next. igordebraga 00:26, 22 June 2023 (UTC)

Watchlist notifications

Usually I get a watchlist notice after publication of The Signpost. It's not appearing for me this time. ☆ Bri (talk) 15:40, 19 June 2023 (UTC)

Looks like it hasn't been updated since May 20, if I'm reading MediaWiki:Watchlist-messages history correctly. ☆ Bri (talk) 15:43, 19 June 2023 (UTC)
MediaWiki_talk:Watchlist-messages#May_Signpost_notice (not sure I agree with all of the arguments there). Regards, HaeB (talk) 01:30, 20 June 2023 (UTC)
Proposal: A three-day watchlist notice for each Signpost issue. This would mean people would see a Signpost WLN for six days per month (or very occasionally more, if we should ever publish on three Sundays in a month). Overall, that would be less than what it was before. Andreas JN466 10:57, 21 June 2023 (UTC)
That sounds basically fine to me, and I think people would be okay with it as well. jp×g 00:23, 22 June 2023 (UTC)
I modified the script at User:JPxG/SPS.js to not leave the watchlist notices automatically, because doing it every two weeks was creating a strange situation: the notices were for one week, so having one every two weeks meant there was a Signpost notice 50% of the time. Some people were getting pissed off, some people were fine with it, but I reckon it is basically fine if we just do a notice once a month -- or else make it for a shorter time. I could not figure out how to make the script alternate notices on different issues, or else I would have done that. jp×g 00:23, 22 June 2023 (UTC)
The script could check the current date and post a request if it is between the 1st and 14th of the month. My suggestion is to have it once a month and advertise both the current issue and the upcoming date of the next issue. I also suggest having a dedicated watchlist template that applies a specific CSS class, so anyone wanting to suppress the message can do so in their personal CSS file. isaacl (talk) 15:56, 26 June 2023 (UTC)
Seriously, are we just going to throw out our work once a month? If we're not watchlist noticing both issues, what's the point in having both issues?
FPs
.
16:10, 26 June 2023 (UTC)
I think having the notice for just three days will miss those who edit in the latter half of the week. Having the notice up 50% of the time risks exacerbating banner blindness, so people will just start ignoring the Signpost notice. Thus I suggest letting people know in advance what the publication schedule is, and advising interested users to subscribe using their favourite method. isaacl (talk) 03:39, 29 June 2023 (UTC)
Perhaps, but I foresee half our issues getting substantially less views this way.
FPs
.
13:27, 1 July 2023 (UTC)
If the notice is only shown for three days every two weeks, that might affect views, too, as might banner blindness. The issues without accompanying watchlist notices so far (the second May issue and the June issues) don't seem to have been drastically affected either way—there was a really big spike for the May News and notes, but that's presumably due to the "Golden parachutes" headline. (Perhaps that also helped garner views from external sites?) My guess is that subscriptions drive views while watchlist notices drive new subscriptions. isaacl (talk) 16:29, 1 July 2023 (UTC)

Next issue

I'll have a Disinformation report this issue, or perhaps a Special report. Maybe even an opinion on a related topic. Unfortunately, I'll likely be out of town after Friday, so may be unavailable after that I'll try to get a finished draft to JPxG on Friday, but would like to keep the story under wraps until Sunday. Smallbones(smalltalk) 13:34, 28 June 2023 (UTC)

@JPxG: In a few minutes I'll post my disinfo report via template. It would be nice to have a couple more stories to go with it. Smallbones(smalltalk) 16:40, 30 June 2023 (UTC)
I'll be doing holiday weekend stuff at least Friday through mid day Sunday so I'll probably miss the opportunity to do major contributions to the issue. ☆ Bri (talk) 16:49, 30 June 2023 (UTC)
Yeah, I'm trying to do the same thing, so there's nobody to blame. The article is posted on one of my user pages and should be ready for ce. I'll get more stories for In the media. News flash Actor lies about her age!!!! Smallbones(smalltalk) 17:47, 30 June 2023 (UTC)
Disinfo report is now at the Newsroom, and I'm satisfied with it. I may copy edit some of the other 3 articles that look close. But basically, I'll be gone until at least Monday. Smallbones(smalltalk) 00:29, 1 July 2023 (UTC)
I am also very short of time this weekend. :(
Nor am I aware right now of any major story that would make a good lead for News and notes.
I was hoping there might be further developments on Meta concerning Gitz6666's controversial global lock, but so far there is nothing. :/ Andreas JN466 22:13, 1 July 2023 (UTC)
The WMUK Open Letter will do as a lead article for News and notes. Will put something together today. Andreas JN466 04:48, 2 July 2023 (UTC)

Just in - a possible obit [Wikipedia_talk:Deceased_Wikipedians#Dthomsen8 here]. I have a COI and will be out-of-town. Smallbones(smalltalk) 00:35, 1 July 2023 (UTC)

Pageviews

As viewership was mentioned a couple sections above, and I haven't compiled these a while, I have actually bothered to make a template for this: Wikipedia:Wikipedia Signpost/Templates/Tables/Issue pageviews. This ought to supersede all of the other templates for issue pageviews, of which there are about three or four, all of which are hopelessly busted and have been for several years. Anyway, here is everything for 2023 to date: jp×g 02:40, 29 June 2023 (UTC)

Thanks! Andreas JN466 22:14, 1 July 2023 (UTC)
Thank you! It would be good to also include the single page view page's page views (right now, the one for the current issue is getting lots of traffic thanks to the tweet by Annie mentioned above). Regards, HaeB (talk) 03:27, 6 July 2023 (UTC)
Big ass tables
Pageviews for 2023-01-01 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Football, football, football! Wikipedia Football Club! 393 502 557 578 589 601 663
Technology report Wikimedia Foundation's Abstract Wikipedia project "at substantial risk of failure" 5188 5616 5836 5981 6033 6062 6166
Serendipity Wikipedia about FIFA World Cup 2022: quick, factual and critical 398 504 537 570 591 613 655
Recent research Graham's Hierarchy of Disagreement in talk page disputes 697 875 981 1060 1088 1104 1169
News and notes Wikimedia Foundation ousts, bans quarter of Arabic Wikipedia admins 7393 8327 8708 8879 8978 9061 9212
Interview ComplexRational's RfA debrief 658 770 804 827 864 901 961
In the media Odd bedfellows, Elon and Jimbo, reliable sources for divorces, and more 1234 1533 1636 1718 1748 1773 1833
From the archives Five, ten, and fifteen years ago 485 601 629 645 655 662 713
Featured content Would you like to swing on a star? 432 537 572 592 601 608 656
Essay Mobile editing 943 1206 1286 1357 1444 1470 1534
CommonsComix #4: The Course of WikiEmpire 455 586 621 634 646 652 703
Arbitration report Arbitration Committee Election 2022 482 586 608 615 632 641 702
Pageviews for 2023-01-16 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report The most viewed articles of 2022 2293 2558 2702 2762 2787 2803 2895
Technology report View it! A new tool for image discovery 735 902 984 1019 1040 1049 1117
Special report Coverage of 2022 bans reveals editors serving long sentences in Saudi Arabia since 2020 14306 14777 15226 15899 15962 16009 16213
Serendipity How I bought part of Wikipedia – for less than $100 1197 1470 1585 1627 1650 1667 1732
Opinion Good old days, in which fifth-symbol-lacking lipograms roam'd our librarious litany 660 777 877 924 942 949 1000
News and notes Revised Code of Conduct Enforcement Guidelines up for vote, WMF counsel departs, generative models under discussion 820 966 1083 1121 1139 1160 1228
In the media Court orders user data in libel case, Saudi Wikipedia in the crosshairs, Larry Sanger at it again 1524 1745 1955 2032 2066 2088 2258
In focus Busting into Grand Central 654 766 830 850 868 885 938
Humour New geologically speedy deletion criteria introduced 740 882 957 991 1003 1016 1078
Gallery What is our responsibility when it comes to images? 662 782 850 875 883 897 952
From the team We heard zoomers liked fortnights: the biweekly Signpost rides again 1109 1434 1609 1663 1678 1692 1764
From the archives Five, ten, and fifteen years ago 473 590 655 681 686 699 760
Featured content Flip your lid 391 477 540 566 577 588 649
Pageviews for 2023-02-04 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
WikiProject report WikiProject Organized Labour 271 356 405 501 515 544 612
Traffic report Films, deaths and ChatGPT 467 650 724 779 792 820 886
Tips and tricks XTools: Data analytics for your list of created articles 479 613 659 692 716 758 828
Special report Legal status of Wikimedia projects "unclear" under potential European legislation 770 980 1044 1087 1106 1130 1216
Section 230 Twenty-six words that created the internet, and the future of an encyclopedia 988 1202 1274 1332 1345 1385 1476
Recent research Wikipedia's "moderate yet systematic" liberal citation bias 2359 4872 4990 5096 5142 5172 5268
Opinion Study examines cultural leanings of Wikimedia projects' visual art coverage 439 599 629 666 678 709 782
Op-Ed Estonian businessman and political donor brings lawsuit against head of national Wikimedia chapter 442 546 595 627 639 669 729
News and notes Foundation update on fundraising, new page patrol, Tides, and Wikipedia blocked in Pakistan 985 1235 1378 1491 1533 1575 1696
In the media Furor over new Wikipedia skin, followup on Saudi bans, and legislative debate 956 1202 1292 1342 1364 1397 1523
From the editor New for the Signpost: Author pages, tag pages, and a decent article search function 770 990 1055 1082 1087 1116 1167
Featured content 20,000 Featureds under the Sea 324 432 477 499 512 538 599
Disinformation report Wikipedia on Santos 2451 2691 2857 2940 3012 3077 3275
Pageviews for 2023-02-20 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Superbowl? Pfft. Give me some Bollywood! Yours sincerely, the world 354 480 576 599 617 661 713
Tips and tricks All about writing at DYK 347 490 557 580 598 630 689
News and notes Terms of Use update, Steward elections, and Wikipedia back in Pakistan 819 1069 1186 1239 1282 1348 1437
In the media Arbitrators open case after article alleges Wikipedia "intentionally distorts" Holocaust coverage 931 1387 1755 1904 1996 2035 2121
Humour The RfA Candidate's Song 368 498 570 590 611 640 734
Gallery Love is in the air 221 326 407 420 433 461 507
From the archives 5, 10, and 15 years ago: Let's (not) delete the Main Page! 379 551 643 672 695 736 790
Featured content Eden, lost. 355 506 590 614 639 684 733
Essay Machine-written articles: a new challenge for Wikipedia 655 990 1121 1178 1213 1253 1341
Disinformation report The "largest con in corporate history"? 17185 18226 18857 19309 19616 19867 20393
Cobwebs Editorial: The loss of the moral high ground 621 839 939 1105 1151 1191 1272
Pageviews for 2023-03-09 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Technology report Second flight of the Soviet space bears: Testing ChatGPT's accuracy 617 865 900 926 974 1014 1085
Recent research "Wikipedia's Intentional Distortion of the Holocaust" in Poland and "self-focus bias" in coverage of global events 2554 3003 3495 3810 3959 4096 4221
News and notes What's going on with the Wikimedia Endowment? 1008 1319 1384 1453 1550 1608 1706
In the media What should Wikipedia do? Publish Russian propaganda? Be less woke? Cover the Holocaust in Poland differently? 1138 1581 1651 1705 1759 1811 1884
From the archives Five, ten, and fifteen years ago 436 628 656 666 708 748 801
Featured content In which over two-thirds of the featured articles section needs to be copied over to WikiProject Military History's newsletter 465 642 668 705 740 782 833
Pageviews for 2023-03-20 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Who died? Who won? Who lost? 463 626 673 720 762 804 866
News and notes Wikimania submissions deadline looms, Russian government after our lucky charms, AI woes nix CNET from RS slate 744 1081 1124 1182 1226 1277 1351
Interview 228/2/1: the inside scoop on Aoidh's RfA 818 1042 1123 1182 1232 1297 1401
In the media Paid editing, plagiarism payouts, proponents of a ploy, and people peeved at perceived preferences 762 990 1045 1105 1149 1200 1267
Featured content Way too many featured articles 484 705 757 829 882 923 986
Eyewitness Three more stories from Ukrainian Wikimedians 426 547 582 613 649 685 754
Pageviews for 2023-04-03 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Recent research Language bias: Wikipedia captures at least the "silhouette of the elephant", unlike ChatGPT 600 868 1086 1151 1208 1237 1299
News and notes Sounding out, a universal code of conduct, and dealing with AI 733 982 1284 1351 1421 1454 1510
In the media Twiddling Wikipedia during an online contest, and other news 563 755 911 948 994 1023 1072
From the editor Some long-overdue retractions 1157 1612 1957 2024 2084 2122 2203
From the archives April Fools' through the ages 463 712 880 926 981 1023 1087
Featured content Hail, poetry! Thou heav'n-born maid 336 469 570 607 656 688 734
Disinformation report Sus socks support suits, seems systemic 640 845 1078 1264 1379 1469 1626
Arbitration report "World War II and the history of Jews in Poland" case is ongoing 572 902 1162 1238 1290 1331 1394
Pageviews for 2023-04-26 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Long live machine, the future supreme 440 613 691 766 813 857 1336
Special report Signpost statistics between years 2005 and 2022 355 486 535 616 648 699 764
Opinion What Jimbo's question revealed about scamming 1040 1354 1472 1581 1623 1695 1778
Op-Ed Wikipedia as an anchor of truth 494 666 716 796 844 883 954
Obituary Remembering David "DGG" Goodman 843 1105 1231 1344 1409 1478 1579
News from the WMF Collective planning with the Wikimedia Foundation 374 521 583 666 704 735 796
News and notes Staff departures at Wikimedia Foundation, Jimbo hands in the bits, and graphs' zeppelin burns 2350 2930 3101 3264 3335 3443 3529
In the media Contested truth claims in Wikipedia 896 1166 1248 1964 2038 2104 2191
Humour The law of hats 386 545 597 647 729 780 845
From the archives April Fools' through the ages, part two 336 494 553 620 660 705 796
Featured content In which we described the featured articles in rhyme again 322 471 519 580 615 641 699
Arbitration report Holocaust in Poland, Jimbo in the hot seat, and a desysopping 840 1151 1257 1357 1397 1441 1509
Pageviews for 2023-05-08 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Special report There Shall Be Seasons Refreshing – Stories from WikiConference India 2023 538 669 730 788 826 858 920
Recent research Gender, race and notability in deletion discussions 1172 1484 1589 1702 1773 1817 1880
News from the WMF Planning together with the Wikimedia Foundation 408 545 588 643 687 718 779
News and notes New legal "deVLOPments" in the EU 1039 1349 1434 1500 1583 1629 1692
In the media Vivek's smelly socks, online safety, and politics 894 1212 1322 1401 1487 1593 1682
Featured content I wrote a poem for each article, I found rhymes for all the lists;
My first featured picture of this year now finally exists!
486 669 727 803 843 878 938
Arbitration report "World War II and the history of Jews in Poland" approaches conclusion 715 953 1005 1065 1091 1124 1179
Pageviews for 2023-05-22 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
WikiProject report Wikipedians Convene for Queering Wikipedia 2023: The First International LGBT+ Wikipedia Conference 516 764 831 897 960 1018 1075
Traffic report Coronation, chatbot, celebs 461 699 748 831 864 897 953
Recent research Create or curate, cooperate or compete? Game theory for Wikipedia editors 2018 2259 2376 2452 2488 2531 2580
News and notes Golden parachutes: Record severance payments at Wikimedia Foundation 51313 52278 52593 52834 53063 53299 53796
In the media History, propaganda and censorship 887 1250 1356 1426 1498 1547 1607
Featured content A very musical week for featured articles 342 539 588 646 684 719 772
Arbitration report Final decision in "World War II and the history of Jews in Poland" 870 1217 1312 1409 1495 1607 1777
Pageviews for 2023-06-05 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Celebs, controversies and a chatbot in the public eye 439 629 707 761 814 848 888
News and notes WMRU director forks new 'pedia, birds flap in top '22 piccy, WMF weighs in on Indian gov's map axe plea 1668 2108 2274 2490 2573 2606 2656
In the media Section 230 stands tall, WP vs. UK bill, Miss Information dissed again 847 1079 1174 1259 1308 1347 1405
Featured content Poetry under pressure 367 526 583 620 662 687 730
Pageviews for 2023-06-19 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Recent research Hoaxers prefer currently-popular topics 733 932 1017 1221 1286 1337 1372
News and notes WMF Terms of Use now in force, new Creative Commons licensing 807 1208 1284 1360 1412 1444 1483
In the media English WP editor glocked after BLP row on Italian 'pedia 2197 2733 3181 3373 3539 3603 3794
Featured content Content, featured 336 471 529 595 652 682 707

Revisionism in Japanese Wikipedia

@HaeB: Further to Yumiko Sato's articles, there is a new-ish little paper on Japanese Wikipedia that I hadn't seen yet: https://wikiworkshop.org/2023/papers/WikiWorkshop2023_paper_22.pdf Would tend to confirm what she has been saying all along. Andreas JN466 17:35, 2 July 2023 (UTC)

Thanks for the heads-up! Generally speaking, always feel free to check the to-do list Etherpad for RR - Wikiworkshop 2023 has already been present there since it took place in May. (As mentioned on the pad, we've been holding out for the video recordings to become available, but on June 28 the organizers announced that those have now been uploaded. They changed some things this year, in particular Wikiworkshop is no longer part of the established WWW conference and all the papers are now only "non-archival" "extended abstracts" - as you indicated, this particular paper too is rather short - rather than full peer-reviewed proceedings publications. So we might end up not covering every single one of them; quite a few are really just research proposals, others - including this one - at least summarize some results.)
Regarding this particular extended abstract, I may take a closer look later (and as always everyone is welcome to offer a review for RR). But on first glance it unfortunately seems that it is largely based on the rather simplistic and questionable approach "that articles with a higher level of controversy [manifested in edit wars] are more likely to be subjected to revisionism." As mentioned in Wikipedia:Wikipedia_Signpost/2022-11-28/Recent_research#Wikimedia_Foundation_builds_"Knowledge_Integrity_Risk_Observatory"_to_enable_communities_to_monitor_at-risk_Wikipedias (reviewing an earlier paper involving one of the same authors that had already made that assumption), there is actually research indicating that the opposite might be true. Or to put it differently, a lot of the concerns about entire projects like Croatian Wikipedia getting hijacked by nationalist viewpoints etc. are about a *lack* of controversy, either because the community is not diverse enough (IIRC a claim that Yumiko Sato made too in her Slate article about jawiki) or because local admins suppress dissent by blocking editors or locking articles.
Regards, HaeB (talk) 03:16, 6 July 2023 (UTC)
Thanks. I agree with you – it is a fairly light paper. On the plus side, they had a Japanese speaker – one of the authors has studied in both Japan and Korea and she did look at some of the surfaced content. As for presence or lack of controversy, don't you think there would have been a period of intense controversy and edit-warring around these topic areas in the Croatian Wikipedia at the time it was captured and dissenting editors were blocked? If so, the Japanese Wikipedia might actually be where the Croatian one was 12 or 15 years ago. (And as you say – things quieting down in the coming months and years might indeed be bad rather than good news.) Andreas JN466 23:42, 6 July 2023 (UTC)

Back from the beach again!

@JPxG and Jayen466: and anybody else around. It looks like there are 5 articles. I'd like to rewrite the Disinfo report lede. Which might take 30 minutes after a 10 minute break. The rest looks like copyediting. JPxG - will you be ready to publish say in 90-180 minutes? Smallbones(smalltalk) 20:29, 2 July 2023 (UTC)

Wikipedia:Wikipedia Signpost/Next issue/Disinformation report could use some copyediting on the lede and a bit in the last 2 paragraphs. Smallbones(smalltalk) 22:36, 2 July 2023 (UTC)
Going over it now, we should be out in about an hour. jp×g 22:53, 2 July 2023 (UTC)
@JPxG: I changed your headline to "Imploded submersible maker foiled trying to sing own praises on Wikipedia?" from "Imploded submarine manufacturer foiled trying to sing own praises on Wikipedia?"
  • 1. "Submarine" is just technically wrong. Since it is dependent on a mother ship, it is a "submersible".
  • 2. "Manufacturer" is very questionable. As I understand it they claimed to have designed it, and have never claimed to have made the hull. And this was the only sub that that could say the they created rather than refurbished. (from The New Yorker article) But in reality the only things they made were the off-the-shelf doodads attached to the hull, and they let the carbon-fiber company pretty much design the hull. "Maker" seems better than "manufacturer".
  • I'm tempted to remove the question mark too. I haven't accused any editor of being the manufacturer. I haven't accused anybody of breaking a law, as far as I know. This story is as solid as it gets.
By the way, feel free to revert me on this - headlines are the E-i-C's call. But when anybody points out the mistake I'll just say "He did it"! Smallbones(smalltalk) 00:50, 3 July 2023 (UTC)
This stuff was all good and I incorporated it into the headline (I thought I had responded here to say this at the time, but I guess I closed the tab without saving). jp×g

And I suggested a blurb at Humour - it had been omitted. "And don't forget the punctuation after the blurb!" Smallbones(smalltalk) 01:00, 3 July 2023 (UTC)

@JPxG: In your finalizing edits to ITM we lost two "In brief" entries (I now have added them to the upcoming issue instead). As I said before here: I can appreciate that sometimes time may run short when pushing a new issue out, but in that kind of situation I would suggest focusing on the basics (including making sure that content isn't messed up by an edit conflict) instead of devoting energy to bonus stuff like stylistic changes or replacing a headline that other team members already came up with a better one. It also looks like the first story could have used some basic fact-checking. (@Smallbones: Apropos, I would appreciate better coordination and cooperation on your part about covering academic research inbetween ITM and RR. Among other things, that could also help avoid such bloopers.)

Regards, HaeB (talk) 00:29, 9 July 2023 (UTC)

@HaeB: I apologize for the blunders. Edit conflicts during publication are frustrating; isn't this why we are supposed to have a deadline? Writing was supposed to be done by 2023-07-01T20:00:00, and publishing to begin at 2023-07-02T20:00:00, but when I got out of bed and clicked "submit" on the version I had prepared, I got an extremely complicated multi-paragraph edit conflict from a bunch of intervening edits during the publication period. In this case, there were still substantial additions being made after the writing deadline, after the editing deadline, and several hours into the period where I am supposed to be pressing the big red button at any minute. I mean, it's better to have improvements made than to not have them made (this obviously made ITM better), so if what we need to do is move the date a couple of days forward I am fine with that. jp×g 12:00, 12 July 2023 (UTC)


Gitz unglock

@HaeB:: Further to Gitz6666's unglocking, you mentioned a couple of weeks ago that in your view Sakretsu's action might not have been consistent with the UCoC enforcement guidelines. Which parts of the enforcement guidelines did you have in mind? Such aspects might be worth mentioning in our reporting in the upcoming issue. --Andreas JN466 16:18, 10 July 2023 (UTC)

User:Kipala has died

As mentioned on the Wikimedia-l mailing list and by User:Risker here, User:Kipala has died. Kipala was a Swahili-speaking German particularly instrumental in developing the Swahili Wikipedia.

Wikimedia Tanzania has posted a tribute to him on Facebook.

Anyone up for writing an obituary? I can quote the Wikimedia Tanzania tribute in News and notes, but a standalone page for him written by someone knowledgeable about his work seems more appropriate. Andreas JN466 10:52, 12 July 2023 (UTC)

@JPxG: I included a section for Kipala in News and notes. I've just noticed that we have a separate obituary section in this issue – you might want to include the write-up for Kipala in that. At any rate, In the media and News and notes (leaving headline and blurb to you) look publishable. Regards, --Andreas JN466 21:22, 16 July 2023 (UTC)

One issue I've never quite been able to get my head around is what we ought to do if there are two of a column in the same issue. Here, for example, we have a few options which all seem kind of dumb:
a) Run one as "Obituary" and one as part of another page like N&N
b) Run them both in the same page as "Obituaries"
c) Name one "/Obituary" and the other "/Obituary 2"
d) Run one this issue and the other one next issue, both as "Obituary"
e) Something hoopty like calling one "Obituary" and the other "Obit"
f) ???
None of these seem particularly good, although I would go with "d" if forced to choose (possibly with a mention in N&N this month, and then a standalone page next month). What do you think? jp×g 21:30, 16 July 2023 (UTC)
Why not b? That seems pretty standard and non confusing.
b
}
21:50, 16 July 2023 (UTC)
I agree that option b is the usual choice for regularly published works with this type of coverage. I feel obituaries will be more appreciated when published in a timely manner. isaacl (talk) 21:57, 16 July 2023 (UTC)
I agree b makes the most sense in this case. As the project along with its contributor base is aging, this is bound to become a more common occurrence ... Andreas JN466 22:39, 16 July 2023 (UTC)

What to do with old drafts/workspaces?

/Next issue

Here I feel that whatever isn't actually considered for the next issue should be moved to /Drafts (All moved to /Drafts

b
} 03:36, 28 July 2023 (UTC))

/Drafts

Some of these are pretty old. If they're no longer considered for publishing, they should be marked with |Final-approval=old or similar. Possibly userfied as well? Those still relevant could be moved to /Next issue

/Newsroom

These mostly seems to be from 2010–2016, reflecting how the Signpost worked around that era. All are marked as old/inactive. I'm not sure we need to do anything with those. Some might make good Cobwebs additions.

b
} 13:35, 17 July 2023 (UTC)

I have been going through all the crap in the omni-index lately. Most of the deprecated stuff I have been delinking and substing (in the case of, say, templates with 3 transclusions), or moving to Wikipedia:Wikipedia Signpost/Boneyard. The old drafts I've been moving to /Drafts/ from wherever they roam, some in very bizarre places indeed. jp×g 08:01, 27 July 2023 (UTC)

Recent Research on Hacker News

@HaeB: [6] --Andreas JN466 15:41, 17 July 2023 (UTC)

Also at the top of Hacker News: The Wikimedia Foundation joins Mastodon and the Fediverse Andreas JN466 21:14, 17 July 2023 (UTC)
Thanks for the ping! And it actually made the front page too ([7][8] - by the way, I tried to re-toot the latter from the @wikisignpost Mastodon account, but unfortunately I didn't find out in time how to do that; seems this is tricky between different fediverse servers. And it appears that on Twitter, the corresponding Hacker News front page bots were killed a while ago). Regards, HaeB (talk) 17:26, 30 July 2023 (UTC)

Mastodon

Wikipedia.social appears to be alive as an official Mastodon instance, see above. Nothing has been posted there yet as far as I can tell. ☆ Bri (talk) 22:16, 17 July 2023 (UTC)

Looks like we don't have this covered yet, I'll try to whip up something for N&N shortly. Regards, HaeB (talk) 17:36, 30 July 2023 (UTC)

Next issue

@JPxG: if you could go through the submissions, there's a few that need your consideration. There's also the couple of /Next issue pieces that have lingered above that we need to decide what to do with.

For the rest of us, Featured content hit a snag. From what I can tell, Adam's father passed away and that understandably snuffed his creative mood, leaving an incomplete piece. We could take it as is, or we could all chip in and write something about the incomplete entries. Or if we don't feel like doing that, we could ask ChatGPT to blurbify some entries [and clearly mark them as such]?

b
} 14:20, 27 July 2023 (UTC)

Attorney doxxing Wikipedians

Possibly useful article illustration to go with the statement "The Mayor further requested that the City Attorney's Office reach out to legal counsel at Wikimedia Foundation to ask that they disclose the identity of the party who uploaded the signature image…" (attribution)

Curious situation here in the draft News and notes. I wondered if this was a legit lawyer letter from a city attorney, and what’s the user Kimlynn69, who twiddled the mayor’s bio, got to do with the situation. The state bar association does list bar number 21004 [9] is the only active member with the last name name Rehberg. The full name is Kimberly Martin Rehberg as the two press items note. The LinkedIn page for an individual named "Kimberly Martin Rehberg" claiming to be the city attorney states undergrad was UNC 1991 [10]. The UNC 1991 commencement document online has one "Kimberly Lynn Martin" [11]. So I don’t want to push this further but maybe the info will help someone writing the piece. ☆ Bri (talk) 06:06, 28 July 2023 (UTC)

[12] ANI, if you haven't seen it. Gråbergs Gråa Sång (talk) 06:17, 28 July 2023 (UTC)
p.s. Just to keep myself out of outing trouble, the operator of the Kimlynn69 account openly claimed to be the attorney. There has been some oversight applied which makes a diff impossible, but you can see it here. ☆ Bri (talk) 06:20, 28 July 2023 (UTC)
I'm not sure any Wikipedian has actually been doxxed in this story at this point, what I've seen so far is the lawyer trying to get their names (conceivably with the intent to doxx, or to send them scary letters), but I may have missed it. Gråbergs Gråa Sång (talk) 06:33, 28 July 2023 (UTC)
Noting also that the story is currently in the WP-articles (and on the talkpages) of all three politicians. Gråbergs Gråa Sång (talk) 08:16, 28 July 2023 (UTC)
@Red-tailed hawk and Smallbones: Just noting that you are both covering the same story in ITM and NAN, so it would make sense for them to link to each other. Andreas JN466 22:49, 28 July 2023 (UTC)

@Red-tailed hawk, Jayen466, and Gråbergs Gråa Sång: I'll leave it to RTH or JN in N&N. It's an interesting enough story, but when I dig down I get the feeling that there's really no harm intended by anybody - just a misunderstanding of what Wikipedia is. I might dig into what the "fight" with the Mayor Pro-Tem was and is there really any possibility of prosecution on criminal charges. Maybe we should suggest strongly that somebody (The Signpost or WMF?) should let people know that if their bio appears on Wiki there are a few thing that they shouldn't do. I think we should remove the signature in our story - nobody seems to want it on Wiki. Take my draft off of ITM and do what you want with it. Thanks! Smallbones(smalltalk) 23:04, 28 July 2023 (UTC)

My opinion is that I think you should keep the signature in the Signpost article based on the newsworthiness of the signature itself, and to show how Streisand effect works. Jonatan Svensson Glad (talk) 23:07, 28 July 2023 (UTC)
What?! How is the signature itself the newsworthy item here? I really think the last thing we want to do to someone firing legal threats at the WMF is to aggravate them more, and this will do exactly that. casualdejekyll 17:22, 29 July 2023 (UTC)
A tempest and a teapot
A teapot that's been through a tempest
I've finished the ITM section and put my recommendations in The Signpost's voice (the editor-in-chief should confirm this paragraph) I'm happy with the section now, but maybe a photo of a teapot or a Durham Bull would help. I'd prefer a teapot. Smallbones(smalltalk) 11:59, 29 July 2023 (UTC)
Personally, I found the two disputes about BLP content a more important issue than the one about the signature. --Andreas JN466 21:51, 29 July 2023 (UTC)
Also now covered in the Carolina Journal. (Added to ITM.) Andreas JN466 10:18, 30 July 2023 (UTC)
I think I've covered it comprehensively in N&N. I like the coverage across both sections. — Red-tailed hawk (nest) 05:50, 31 July 2023 (UTC)

Comparing legal (?) sex scandals

@Igordebraga: The first sentence of Traffic report compares the two sex scandals Phillip Schofield and Huw Edwards. It doesn't seem to be the case that anybody was accused of breaking the law, but it seems to me that saying that Edwards scandal was "similar to" Schofield's that might be interpreted as a BLP violation if one was considered by some to be worse than the other. I couldn't find anything that makes one seem worse than the other, (but what makes a legal relationship "bad" or a blp vio?). I think a 2nd person should check. Smallbones(smalltalk) 00:20, 29 July 2023 (UTC)

BLP applies to all information, good, bad, and anywhere in between. Horse Eye's Back (talk) 14:24, 29 July 2023 (UTC)
Well, can clean it up and I see no issue (even if that was Ollieisanerd writing). igordebraga 15:05, 29 July 2023 (UTC)

RIP Eeglish

Eeglish has passed away; see the talk page and the Wikipedia obituary. Graham87 15:38, 29 July 2023 (UTC)

@
b
}
17:47, 29 July 2023 (UTC)
From the obit page, I think you mean
b
} 17:48, 29 July 2023 (UTC)
@Headbomb: Oh yes, thanks, I meant Eagleash. That username's pronounced oddly by my screen reader. Graham87 04:03, 30 July 2023 (UTC)

Disinfo report

I've started Wikipedia:Wikipedia Signpost/Next issue/Disinformation report with two sections that are ready for copy editing. I could have as many as 5 sections total, but will quit adding in plenty of time to publish. Smallbones(smalltalk) 02:40, 30 July 2023 (UTC)

@JPxG and Bri: The timer tells me publication should have started. I need about an hour to get my disinformation together. It looks like the rest needs a bit of work too. Let me know when you plan to publish! Smallbones(smalltalk) 20:08, 30 July 2023 (UTC)
It will probably be a couple of hours, I'll push the deadline template forwards. jp×g 20:43, 30 July 2023 (UTC)
There's about 3 hours to UTC midnight, after which the script will think it's August.
b
}
21:01, 30 July 2023 (UTC)
It's past midnight UTC btw.
b
}
00:07, 31 July 2023 (UTC)
Surely it would think it was July 31? Unless there is something extremely bizarre going on in the script (which I can't rule out). jp×g 02:29, 31 July 2023 (UTC)
Uh... right. I mixed June and July for some weird reason.
b
}
02:34, 31 July 2023 (UTC)
Well now it'll be August 1st.
b
}
00:03, 1 August 2023 (UTC)

News and notes

Trivial country code issue

Should we spell out what "Wikimedia CH" stands for? ☆ Bri (talk) 21:44, 31 July 2023 (UTC)

(I assume you mean "In the media" instead of "News and notes".)
It is already introduced as "the Swiss Wikimedia chapter" in the preceding sentence, but feel free to make that link explicit there. (IIrc the reason for using that weird country code in the organization's official name was language neutrality.) Regards, HaeB (talk) 21:52, 31 July 2023 (UTC)

User drafts

Here's a list of (mostly) old user pages, all marked as inactive save for

b
} 13:42, 17 July 2023 (UTC)

Yeah, I am making my way through the really old stuff. Some of them are actually very good, although most are nothing (just blank drafts)... hopefully we can be done with Spelunk! 23 soon, as I am starting to get a bit stir crazy from all these damn templates. jp×g 05:04, 21 July 2023 (UTC)
I suggested something for Cobwebs, but there definitely are other pieces that could be dug out (Bri's piece is indeed very decent). I'd double/triple check if those haven't been published before and aren't hard copies of existing articles, and give the authors a heads up/notice to see if they want their piece published as is (with copyediting), or if they want to modernize it and publish in a regular column.
b
}
16:22, 21 July 2023 (UTC)

Pageviews

Note that I have made a template for these which makes it very easy, or as the kids say, "ez clap": {{Wikipedia:Wikipedia Signpost/Templates/Tables/Issue pageviews|2023-07-03}}

Pageviews for 2023-07-03 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Are you afraid of spiders? Arnold? The Idol? ChatGPT? 649 841 918 1005 1043 1084 1125
News and notes Online Safety Bill: Wikimedia Foundation and Wikimedia UK launch open letter 949 1306 1406 1494 1540 1600 1667
In the media Journo proposes mass Wiki dox, sponsored articles on Fandom, Section 230 discussed 979 1261 1341 1451 1501 1538 1598
Humour United Nations dispatches peacekeeping force to Wikipedia policy discussions 583 820 889 957 1002 1041 1074
Featured content Incensed 406 537 589 640 682 713 744
Disinformation report Imploded submersible outfit foiled trying to sing own praises on Wikipedia 1275 1738 1941 2056 2169 2320 2565
Pageviews for 2023-07-17 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report The Idol becomes the Master 364 531 583 657 690 712 742
Tips and tricks What automation can do for you (and your WikiProject) 1072 1213 1320 1461 1540 1585 1655
Recent research Wikipedia-grounded chatbot "outperforms all baselines" on factual accuracy 46363 46818 47047 47232 47336 47426 47539
Obituary David Thomsen (Dthomsen8) and Ingo Koll (Kipala) 496 620 695 1382 1454 1488 1543
News from the WMF ABC for Fundraising: Advancing Banner Collaboration for fundraising campaigns 331 435 470 538 565 587 613
News and notes Big bux hidden beneath wine-dark sea as we wait for the Tides to go out? 827 1175 1300 1376 1417 1445 1487
In the media Tentacles of Emirates plot attempt to ensnare Wikipedia 2216 2561 2817 2949 2989 3022 3072
In focus Are the children of celebrities over-represented in French cinema? 428 576 634 692 726 750 788
Humour New fringe theories to be introduced 580 808 926 1024 1069 1105 1147
Featured content Scrollin', scrollin', scrollin', keep those readers scrollin', got to keep on scrollin', Rawhide! 379 515 583 669 706 723 748
Cobwebs If you're reading this, you're probably on a desktop 647 844 895 983 1014 1035 1068

jp×g 17:50, 21 July 2023 (UTC)

Wow, Tips and Tricks the third most read columns this issue. I did not expect that!
b
}
18:25, 21 July 2023 (UTC)
  • Another unbelievably high spike here for recent research, which as far as I can tell is entirely due to https://news.ycombinator.com/item?id=36757520 submitted by the enigmatic "akolbe" -- who could that be?? ;^) I can't help noticing that our pageviews seem to increase by somewhere around two orders of magnitude whenever they're linked to by a popular website. Much to think about... jp×g 21:32, 21 July 2023 (UTC)

Best, worst, and most ironic drafts

I was looking around, mostly at random, and found Draft:Cat Behaviors. This seems very unlikely to be approved at AfC, or to make it into mainspace, but it is nonetheless cute, and I considered the idea of publishing it. I also looked around in draftspace a little more, which turned up some epic stuff like Draft:Dom is funnier than Sophie and Draft:Colorless-eyed-monkeys. These are kind of funny. We might have a column here. jp×g 20:16, 27 July 2023 (UTC)

If it matters, voting yes on Draft:Cat Behaviors. Speaking as someone with an unusual interest in large rocks, Wikipedia is full of the endearing, quirky interests of our editors, and sometimes they are even notable. ☆ Bri (talk) 21:47, 31 July 2023 (UTC)

Op-ed, July

This seems odd for an op-ed. This is not opinion, this is more of a how-to guide. More suitable for a In focus or Tips and Tricks column, IMO. Both are filled for this issue, but the next one is wide open. This would also give time to adapt this to a more standard Signpost format.

b
} 23:49, 31 July 2023 (UTC)

If we run an op-ed this issue,
b
} 23:51, 31 July 2023 (UTC)
Fine with it being next issue.
FPs
.
00:17, 1 August 2023 (UTC)

Just going to noote I wrote this; if you want it for the Signpost, we can grab it.

FPs
. 01:53, 31 July 2023 (UTC)

Moved to tips and tricks for next issue. jp×g 04:37, 1 August 2023 (UTC)

RSS feed for the Signpost?

User:JPxG, NAN is currently on the Hacker News front page, and someone is wondering how they can get an RSS feed for The Signpost. There's a couple of answers there but you (or others here) might want to add to them. Andreas JN466 18:38, 1 August 2023 (UTC)

Possibly of interest: Wikipedia:Wikipedia_Signpost/Subscribe#RSS_feeds. Regards, HaeB (talk) 20:48, 1 August 2023 (UTC)
PS: Congrats on scoring yet another HN front page hit!
(yes, I'm a bit jealous, considering our recent Twitter/Mastodon/FB engagement rates ;)
Regards, HaeB (talk) 20:52, 1 August 2023 (UTC)

Oppenheimer

I wrote an Op-Ed on Oppenheimer. I can run it in our Milhist newsletter, but I'm offering it to the Signpost first if you guys are interested. Hawkeye7 (discuss) 06:36, 3 August 2023 (UTC)

I should note The Bugle is due to publish sooner. But I'm for this
FPs
.
14:01, 3 August 2023 (UTC)
I'd say go for both. Not sure for op-ed, but definitely somewhere.
b
}
14:10, 3 August 2023 (UTC)
Just call it a special report.
FPs
.
17:38, 3 August 2023 (UTC)
I like it. We can put this in. jp×g 22:25, 3 August 2023 (UTC)

Omni-index for old pages

After finding some bizarre vandalism from June on an old article, that was messing up the Module:Signpost indices, I realized that we probably ought to have some way of tracking edits on old Signpost articles, so I made Wikipedia:Wikipedia_Signpost/Omni-index/Old_pages. RelatedChanges for that is here; this should give you every most-recent edit that's been made to every old article (i.e. prior to the current year), that match the masks of Wikipedia:Wikipedia Signpost/____-__-__*. I can't make it cover more than thirty days, so probably there ought to be an alarm every 29 days to go check this lol. jp×g 22:27, 3 August 2023 (UTC)

This hasn't posted Signpost issues since May. We should fix this!

b
} 09:06, 5 August 2023 (UTC)

This is part of the disruptions caused by Twitter's changes, see this recently archived thread for some context. As mentioned there, this Twitter --> Facebook bridge actually kept still working initially (when the Twitter --> Mastodon bridge had been broken already), but now it is gone too for some reason. Fixing this is on my Signpost to-do list. But honestly not in the top position, also considering the low engagement that these automated posts have been generating (e.g. compared to the manual Facebook announcements in the Wikipedia Weekly Facebook group).
If anybody knows of a good solution for currently working Twitter --> Facebook bridge, let me know! (Or also RSS --> Facebook, using a feed such as ntr.nah.re/wikisignpost/rss which is what I have resorted to recently for the Mastodon bridge, after the previously used RSS feed stopped working earlier this week.) Regards, HaeB (talk) 01:36, 7 August 2023 (UTC)

TIADL

@JPxG:

Smallbones(smalltalk) 23:59, 31 July 2023 (UTC)

Timed Instrumental Activities of Daily Living? I assume this is a clever reference to my stupid ass having the thing out two days late... jp×g 00:25, 1 August 2023 (UTC)
I'm guessing that stands for There Is A DeadLine.
b
}
00:58, 1 August 2023 (UTC)
Have you all thought about publishing content as is ready, then sending out a monthly digest? ScottishFinnishRadish (talk) 01:04, 1 August 2023 (UTC)
It's spelled
TIAD, not TIADL. But yes, why are we already over a day late again? @Bri
: Would you be able to help out with carrying out the publication steps as before in case this doesn't move forward?
I also second the concerns voiced by two other team members about that entirely outdated and (in its present state) extremely reader-confusing "Apocrypha" piece. I would suggest dropping or postponing it, also in the interest of not holding up publication of this issue further. Regards, HaeB (talk) 01:32, 1 August 2023 (UTC)
The issue is out now, but the central page at Wikipedia:Wikipedia Signpost still links to the previous issue's stories, while showing the current issue's headlines. @JPxG: Might this have to do with the recent fiddling and futzing at the "issue" page? (What was the purpose and benefit of that, btw?) Regards, HaeB (talk) 03:56, 1 August 2023 (UTC)
Will fix. Headbomb fixed. jp×g 04:35, 1 August 2023 (UTC)
Fixed here.
b
}
04:23, 1 August 2023 (UTC)
I have no idea what the fuck happened. This is an error with SPS.js having posted the text to Signpost/Issue instead of Signpost/Templates/Issue. If only I had thought to update the script.... but I did! Literally every instance of the pagename was changed in my SPS.js. I just went through and double-checked. There is a specific part where it posts the text to the page, which I updated, and somehow it still did this. This is nuts and makes no sense. jp×g 04:35, 1 August 2023 (UTC)
I'm no expert, but maybe in
b
} 04:41, 1 August 2023 (UTC)
I believe this fixes it. What the absolute hell -- it was concatenated from two separate strings for no apparent reason -- outrageous. Well, I have fixed it now. jp×g 05:08, 1 August 2023 (UTC)
absolute hell, outrageous - really? I find this harsh criticism of Evad37's coding skills (as the author of the script's original version) quite disingenuous. It's an entirely standard programming technique to assemble a URL or file name in this way. In fact the current version of the script still does this in various other places too (search for "data.path + "). Yes, it should have been employed more consistently here, but the upshot is that one can't rely on a string search for the full page name to find all usages, if such a path variable is being used.
Honestly, overlooking this central usage while changing Literally every instance of the pagename was your own error. Of course everybody makes mistakes, we all still love our editor-in-chief and your
WP:STOCKS
time should be kept to the necessary minimum
Regards, HaeB (talk) 00:09, 7 August 2023 (UTC)
@HaeB: You would not believe what a disaster the Signpost namespace was. Twenty years of nobody ever cleaning it out means that we had all kinds of asinine garbage: twenty-year-old redirects and templates with no transclusions or links that weren't tagged as deprecated, templates that had never been used for anything (and duplicated other templates), templates that had two transclusions from 2010 and that was it, and most worryingly, pages like Wikipedia:Wikipedia Signpost/Audio that were just abandoned and seemingly forgotten about (to say nothing of the dozens of random long-forgotten drafts rotting away in invisible sub-sub-directories of deprecated pages). the current omni-index is still pretty big, but is somewhat easier to traverse. One major pain is that templates were more or less never standardized for the Signpost. That means we had stuff at each of these prefixes:
  1. Template:Signpost/________
  2. Template:Signpost-________
  3. Template:Signpost ________
  4. Wikipedia:Signpost/________
  5. Wikipedia:Signpost/Template:________
  6. Wikipedia:Signpost/Template/________
  7. Wikipedia:Signpost/Templates/________
  8. Wikipedia:Wikipedia Signpost/________
  9. Wikipedia:Wikipedia Signpost/Template:________
  10. Wikipedia:Wikipedia Signpost/Template/________
  11. Wikipedia:Wikipedia Signpost/Templates/________
It was quite difficult to get a hold of what did what, where it was, and why. For a while I thought that there had been some kind of method to why various things were at different prefixes, but I eventually had the crushing realization that they did not, and there was just no reason at all why anything was the way it was, and the reason it was such a mess was nobody had bothered to go through and try to straighten it out for twenty years. Anyway, I have been doing my best to fix the stuff where possible: for example there were a bunch of redundant header templates that were used in different years, which I updated all of to use the current header and footer (Wikipedia:Wikipedia Signpost/Templates/Signpost-block-start-v2 and Wikipedia:Wikipedia Signpost/Templates/Signpost-block-end-v2). This way all of the articles will display fine and if, Heaven forfend, we or Internet standards decide to change anything ever for any reason, we don't have to futz around and try to fix eighteen different templates because for one week in 2009 we were trying out a new flavor etc etc etc. Anyway, since I have been doing a large number of AWB runs to fix messed-up stuff in Signpost space, I have been fixing some of the templates that were in stupid places (I am mostly leaving alone the ones with 9,000 transclusions etc because even I don't have editcountitis that bad). jp×g 05:36, 1 August 2023 (UTC)
Ahem.
1. Yes, the Signpost's template and module system is a monster, or perhaps rather a haunted building with many skeletons in many closets. (Recall also that you and I had conversations about aspects of this not long ago.) I'm grateful for any improvements you have made in this area. And I agree that it would have been great if the people who built it in earlier years would have paid more attention to naming consistency and better documentation.
2. However, I don't think it's a big priority to, say, be able to update the design of old articles published back in 2009 to a more modern look with a little less effort. Or to have that "omni-index" look a bit more consistent (especially given that its purpose is described as enabling the monitoring of related changes, rather than being looked at directly). There are much more important technical improvements to be made.
3. And, such benefits are to be weighed against the obviously high risk for disruption when moving such a central template, from a location in which it had existed since at least 2007 (possibly earlier) to a page name that's longer but in your view more aesthetic or consistent. You were clearly aware of this risk: "Gigantic pain-in-the-ass move that breaks everything" [13]. Self-deprecating humor is cool and all, but doesn't change these downsides. To make it concrete: Between you, Headbomb, myself, the reader who pointed out this breakage on your user talk page, and the probably tens or hundreds other readers who encountered this issue, were confused why they were seeing last issue' stories and turned away without alerting us, your error probably wasted several hours of people's lives. Is a nicer page name that is only ever seen by those peeking into the Signpost's technical innards really worth that?
4. And we might not even have seen all the resulting disruption yet. Specifically: It looks possible that this fail at publication time might still have been averted if the redirect resulting from the move had not been deleted the redirect under G6 by an over-eager administrator. @Maile66: Could you explain why you considered this deletion to be "uncontroversial maintenance"? Did you make an effort to ensure that the redirect was not in use, and if yes, did you consider that the moved page is being accessed and changed from scripts? And after JPxG recreated this redirect while scrambling to fix things after publication of the new issue had gone wrong (see above), another over-eager administrator deleted it again, this time with the rationale "R3: Recently created, implausible redirect". Isabelle Belato: Could you explain why you considered the redirect "implausible"? Did you make an effort to understand why JPxG had recreated it, and did you really confirm that the redirect is no more used or accessed anywhere at this point? Again, this moved high-risk template had been in this location since at least 2007 and has a very central function in the Signpost's template and scripting system. It is the primary location where the "publication" of each new issue happens, affecting content in various other places.
All that said, JPxG, if besides the editor-in-chief duties you also have the willingness and required time to clear out some of the aforementioned closets and are not afraid of their skeletons, sure, do it! But that time also includes applying extra diligence when handling high-risk templates, and if you find yourself unable to do that for whatever reason, it might be worth asking for support - we do have a lot of technically skilled people in the community who are adept at dealing with templates and scripts and might be willing to offer a hand, or at least a second pair of eyes.
Regards, HaeB (talk) 00:23, 7 August 2023 (UTC)
HaeB are you the one who pinged me above? Re your question "Could you explain why you considered this deletion to be "uncontroversial maintenance"? - well, not at this point. And that's not normally a phrase I use, so the wording might have been something I clicked on. The best I can tell you, is that based on that delete summary, it must have been listed at Category:Candidates for speedy deletion. First of all, I would not have thought of this myself, and the fact that it has "(G6)" after my deletion, makes it certain it was listed there. About the only time I do delete anything, it's usually from the Candidates for speedy deletion. That shows that I deleted it on July 23. Prior to that, it was a page move by User:JPxG on July 20. After my deletion, it was again deleted on August 1 by User:Isabelle Belato. — Maile (talk) 01:03, 7 August 2023 (UTC)
I tagged the page for deletion myself both times, so if it is anybody's fault, it is mine. jp×g 01:08, 7 August 2023 (UTC)
Thanks for this response, Maile66 and for adding context. (The second deletion was already described above.)
Of course speedy deletions (usually) involve a nominator in addition to the deleting admin, I could have stressed that above. But per the procedure, the admin is supposed to independently double-check the deletion. For mainspace pages that involves inspecting if "What links here" shows remaining references to the page via wikilinks, and by analogy, for this kind of highly unusual deletion involving the move of a high-risk template after over 15 years, due diligence could and should have involved double-checking that the script(s) which had been using it had been updated.
Regards, HaeB (talk) 01:22, 7 August 2023 (UTC)
Well, prior to this, I spent most of a day re-referencing incoming links to the new location; Special:WhatLinksHere/Wikipedia:Wikipedia_Signpost/Issue is completely empty, except for links on this page and a couple places where I have manually made lists of old prefix indices. The only place where the text "Signpost/Issue" appeared on the entire wiki, apart from pages like my own CSD-tagging log, was apparently a single instance in the publishing script, where it was (seemingly for no reason) being constructed by concatenating "/" with "Issue", which is now fixed. As for the namespace in general, it's more of an overall cleanup thing. As far as I know, there's literally no documentation (apart from the incomplete POST/T) about what templates exist, what they do, and the other way around, what tasks are done by templates. So we had, for example, Wikipedia:Signpost/Template:Pageviews, and Wikipedia:Wikipedia Signpost/Templates/Article list maker/Pageviews, and Wikipedia:Wikipedia Signpost/Templates/Signpost edition pageviews. Similarly, all of our serieses are now linked to from (and are subpages of) Wikipedia:Wikipedia Signpost/Series; getting a complete and comprehensive list of all our series can now be done reliably by a single click/transclusion of its PrefixIndex. Previously, they were named weird stuff like Wikipedia:Wikipedia Signpost/Tutorials, Wikipedia:Wikipedia Signpost/Archives/WikiProject report, Wikipedia:Wikipedia Signpost/Templates/flaggedrevisions, Wikipedia:Wikipedia Signpost/Templates/2014 Wikimedia Global Education Program etc -- this meant they were basically inaccessible, not only to readers, but also apparently to us: they weren't really being linked to from anywhere, most of them are never discussed or mentioned, and in the case of the templates, there were sometimes multiple duplicate versions of the same thing because nobody could figure out where to find them. In this, I think that something like fixing file paths (even if extremely tedious for me to do) is still worthwhile, because it prevents the future situation where all sorts of random crap is jumbled into the bare space (i.e. Wikipedia:Wikipedia Signpost/*) -- if the price of doing this is that a cople people get confused checking their watchlists once every twenty years, I am fine with designating myself the official person to get mad at for this iteration, and merely hope that we can do better in the future and avoid such a situation in the first place. jp×g 01:45, 7 August 2023 (UTC)
The added documentation is obviously greatly appreciated (I think I may have said that before), as is the un-orphaning of pages etc. But I still don't see how that necessitates such high-risk renamings and deletions.
The only place where the text "Signpost/Issue" appeared on the entire wiki, apart from pages like my own CSD-tagging log, was apparently a single instance in the publishing script, where it was (seemingly for no reason) being constructed by concatenating "/" with "Issue" - yes, you had said that above already. I'll point to my response there instead of repeating it here.
Anyway, you deserve credit for graciously accepting responsibility here (lots of people have fewer zen skills in such matters). But as you indicate, avoiding this kind of thing in the future is what matters, and one reason why I'm highlighting these matters in some detail now is that this is not the first time we have such issues. HaeB (talk) 02:27, 7 August 2023 (UTC)
Hope everything has worked out to resolve this for you. No excuses, but it helps for people to know what admins have going on. Nobody can know everything. Admin dashboard is just a small sampling of what we have in front of us. And yes, there are admins out there who believe their knowledge and instincts are better than some non-admin. Human beings ... what are we going to do with us? And it's also a given that as much as we do correctly, there are always those situations where an admin action presumed correct by the editor who requested the action, end up being opposed by others after the deed is done. So, we do what we can. And everybody else does what they can. — Maile (talk) 11:59, 7 August 2023 (UTC)

Endowment discussion at User talk:Jimbo Wales#Endowment

FYI. Andreas JN466 10:02, 10 August 2023 (UTC)

Wikimania (August 16-19)

Wikimania always has lots of newsworthy aspects (starting usually with some pre-conference announcements from the WMF Board). I'd encourage anyone so inclined to check the schedule for sessions that may be worth reporting on (remote participation is possible). I'm going to be in Singapore myself, but can't promise too much in this regard (I got drawn into two sessions and aim to keep my participation a bit more leisurely otherwise this time). That said, happy to say hi to anyone else who is there and is involved or interested in the Signpost. At some past Wikimanias we even managed to get a formal Signpost meetup organized, in case anyone is up for trying that. Regards, HaeB (talk) 15:24, 14 August 2023 (UTC)

Thanks for the N&N addition to that effect, and safe journey! Andreas JN466 15:53, 14 August 2023 (UTC)

Out soon

I punted it forward eight hours. There are a bunch of submissions to review and go over... we will be out soon! jp×g 23:45, 13 August 2023 (UTC)

@Adam Cuerden and JPxG: The "once notable" in the ITM headline doesn't quite hit the mark; we're likely to get readers telling us that notability can't be lost (on Wikipedia, at least). As for the main headline and blurb on N&N, up to you, JPxG. (And as HaeB said, I can always link to a subheader.) --Andreas JN466 07:26, 14 August 2023 (UTC)
I have a better headline, perhaps taking a spot of inspiration from that website you like so much. jp×g 07:30, 14 August 2023 (UTC)
This is a joke btw. jp×g 07:36, 14 August 2023 (UTC)
Andreas JN466 15:43, 14 August 2023 (UTC)
  • Soon. jp×g 11:09, 14 August 2023 (UTC)
    @JPxG N&N seemed to be lacking the title in the template, so I copied across what you put in at the top. [14] Andreas JN466 13:48, 15 August 2023 (UTC)

Le views

Pageviews for 2023-08-01 (V)
Subpage Title 7-day 15-day 30-day 60-day 90-day 120-day 180-day
Traffic report Come on Oppie, let's go party 457 668 748 821 868 883 931
Tips and tricks Citation tools for dummies! 631 1158 1259 1372 1440 1497 1580
Opinion Are global bans the last step? 706 899 982 1057 1110 1129 1188
Obituary Donald Cram, Peter McCawley, and Eagleash 460 632 691 807 858 877 934
News and notes City officials attempt to doxx Wikipedians, Ruwiki founder banned, WMF launches Mastodon server 30219 30965 31205 31346 31433 31506 31628
In the media Truth, AI, bull from politicians, and climate change 1057 1318 1443 1605 1671 1723 1796
In focus Journals cited by Wikipedia 1160 1375 1456 1532 1594 1623 1684
Humour Does Wikipedia present neutral perspectives? 561 808 883 961 1011 1036 1099
Featured content Featured Content, 1 to 15 July 252 405 448 497 538 553 577
Disinformation report Hot climate, hot hit, hot money, hot news hot off the presses! 579 874 1169 1342 1476 1590 2001

This will update properly in a day or so. jp×g 18:45, 1 August 2023 (UTC)

Wow, those are some page views for NaN! Jonatan Svensson Glad (talk) 09:04, 2 August 2023 (UTC)
Think it mainpaged on Hacker News
FPs
.
15:10, 2 August 2023 (UTC)
Yeah, the basic way it seems to work is that every issue we get one article slashdotted (well, the modern equivalent, twitter/HN/etc) and it has about 100 times as many views as anything else. Much to think about. jp×g 22:58, 3 August 2023 (UTC)
Actually, Slashdot still exists and can be effective (I once had Elon Musk replying to a Slashdotted Signpost story). However, Slashdot is more demanding in terms of article quality.
I rarely have luck with Reddit. Andreas JN466 23:13, 3 August 2023 (UTC)
I'm honestly surprised the Signpost isn't monitored by more media organization.
b
}
23:13, 3 August 2023 (UTC)
One thing to remember is that headlines are important for attracting outside eyes. In my view, our main headline should always be a concise summary of the main story, which should be top of the page. (Other stories featured further down on the same page can be covered in the blurb.)
Alternatively, the headline should be such that it can be truncated. This worked in this case, e.g.: "City officials attempt to doxx Wikipedians" was a good headline. So was "Wikipedia-grounded chatbot "outperforms all baselines" on factual accuracy" (although people complained that this headline related to the third story on that page rather than the first).
Headlines trying to summarise all the articles on a page, like "Truth, AI, bull from politicians, and climate change", are dead in the water in terms of attracting outside eyes – they are too diffuse, and don't invite clicks. Andreas JN466 14:36, 4 August 2023 (UTC)
To a point. How Wikipedia is covered in the media is also not something that matters a lot to outsiders.
b
}
00:42, 7 August 2023 (UTC)
It varies. HaeB's write-up of the New Yorker story ("Emirate tentacles ... ") got some traction both on HN and on Reddit. Where a press story is intrinsically interesting and/or we can add extra expertise, attracting outside eyes becomes viable. Andreas JN466 00:53, 8 August 2023 (UTC)
Agreed regarding headlines (and, everybody, we should listen to Andreas on this, considering how many additional readers he has brought us in recently with these postings on Hacker News etc). I would say though that something like "Truth, AI, bull from politicians, and climate change" is still more informative - in the sense of appealing to readers who are interested in AI, political misinformation or climate change - than obscure pop culture references or insider jokes, say.
Regarding "Wikipedia-grounded chatbot ...", the problem is our lack of TOCs. That said, section anchors still exist and can be extracted from the page's HTML source as described under Help:Section#Section_linking (or by doing an edit preview with __TOC__ added). I.e. Andreas, you could have posted this link instead on Hacker News: https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2023-07-17/Recent_research#Wikipedia-based_LLM_chatbot_%22outperforms_all_baselines%22_regarding_factual_accuracy .
Regards, HaeB (talk) 01:11, 7 August 2023 (UTC)
That's true. I somehow had it in my head that the HN submission guidelines forbade that, but upon rechecking I find that's not the case. Thanks for mentioning it. Andreas JN466 08:13, 8 August 2023 (UTC)
Now I remember. The problem with that method is that the admins routinely remove the # and everything following it. Andreas JN466 06:40, 16 August 2023 (UTC)
(Outsider thought) VieThis seems odd for an op-ed. This is not opinion, this is more of a how-to guide. More suitable for a In focus or Tips and Tricks column, IMO. Both are filled for this issue, but the next one is wide open. This would also give time to adapt this to a more standard Signpost format.
b
}
23:49, 31 July 2023 (UTC)
If we run an op-ed this issue,
b
} 23:51, 31 July 2023 (UTC)
Fine with it being next issue.
[majestic titan]
01:18, 8 August 2023 (UTC)
I'd actually like the Signpost also to function as a platform for the community to explain Wikimedia and Wikipedia to the public and to inform journalism elsewhere. (Remember our presentation at WikiConference US 2015?) For that to happen it needs to have some visibility outside of Wikipedia. Andreas JN466 06:57, 16 August 2023 (UTC)

Update on Knowledge Equity Fund

The meta:Knowledge Equity Fund recently posted updates.

The Knowledge Equity Fund is a WMF initiative to promote equity. In September 2021 the WMF reported giving US$1.5 million to 6 organizations. As I just mentioned, the WMF just reported on outcomes. In their reporting they also said that one of the grantees previously reported to have received 250k actually received 850k.

For context, getting tabs on WMF funding for Wikimedia community grants is difficult, but United States in 2021 was probably about US$250k, and India was about $100k. The grants to these organizations are substantial as compared to regional Wikimedia community grants.

When the Knowledge Equity Fund awarded grants whatever project proposals or plans the recipients had was private. Each org had a one-paragraph explanation of what they planned to do with the money. Something striking to me was that the plans seemed to have little to do with Wikipedia or the Wikimedia community; in my view this Wikimedia money was going to support other (important, admirable, unrelated) activist issues which had no obvious connection to Wikimedia activities or strategic priorities which the Wikimedia community identified.

Here are my personal reactions at a glance:

  1. These projects are unrelated to Wikimedia community interests
  2. These grants are awarded without Wikimedia community participation and input
  3. This is all weird because it seems out of scope
  4. This money went to causes which do not apparently connect to the Wikimedia platform or the Wikimedia community

Here is the report for the award of US$850k - File:Knowledge Equity Fund (Round 1) - Borealis philanthropy report.pdf. The report says that this org re-granted the money to 28 other projects, but does not identify those projects. The report is dated July 2022, but WMF staff just uploaded and shared it in April 2023. This report either does not mention Wikimedia engagement, or it hardly mentions Wikimedia engagement, or it describes Wikimedia engagement in a weird way that the Wikimedia community would not. The reported impact seems to be that this organization encouraged people to contribute to Wikimedia projects, but they are unable to identify or share any cases where that actually happened. I do not think the WMF has every had a public discussion about what happened with this money.

I appreciate good activist causes wherever they are, but Wikimedia donors expect money to go to Wikimedia projects. I do not think that Wikimedia community grants committees would fund projects which promised these outcomes. I am not sure what is going on here. Bluerasberry (talk) 18:33, 3 August 2023 (UTC)

Thanks. A few months ago it sounded like they had scrapped the project. Now they say, This second round of grants was administered by the Wikimedia Foundation, after all remaining funds for the Equity Fund were transferred back from Tides Advocacy to the Foundation earlier this year..
Back in January, when Maryana Iskaner announced the return of the money (see Signpost coverage), she spoke of Re-centering the Foundation's responsibility in supporting the technology needs of the Wikimedia movement by understanding the needs of our contributor communities, as well as emerging topics like machine learning/artificial intelligence and innovations for new audiences. Words, actions ... Andreas JN466 19:11, 3 August 2023 (UTC)
Hi Bluerasberry, I want to provide clarification on a few of your comments. After our first round of grants, we published a Diff post where we shared learnings from that first round, as well as opportunities for expanded community participation based on those learnings. About Borealis specifically, they received a 250,000 grant from the Knowledge Equity Fund.
You also brought up a point about the relative size of the round 1 equity fund grants. For grantmaking in general, the award size depends on a lot of things like the capacity and grant history of the recipient, among other factors. It's apples to oranges to compare it with other grants given by the Community Resources team, especially given the Equity Fund’s role as a pilot project and that the Equity Fund does not draw from the same pool of resources.The Equity Fund was created in 2020 as a $4.5 million fund, from the Foundation’s operating budget that year. Former and ongoing grants are all given from the initial fund; to date, no additional funding has been added to the Equity Fund and ongoing grant awards do not come out of the Foundation's annual operating budget. NGunasena (WMF) (talk) 23:26, 11 August 2023 (UTC)
I remember now why I thought the project had been scrapped. WMF board member Victoria said on-wiki last November, Equity Grants were an idea of the previous CEO who is no longer with the Foundation so there isn't a chance of them recurring. The Board has done its main job - changed the CEO. I've asked on the mailing list what happened. --Andreas JN466 16:32, 19 August 2023 (UTC)