• Always interesting to hear from Adam, one of the most consistent contributors to The Signpost. His research and restoration work on old portraits is particularly impressive.--Ipigott (talk) 06:11, 24 October 2023 (UTC)[reply]

• I once again thank you for your incredibly inspiring work, Adam! Oltrepier (talk) 08:11, 24 October 2023 (UTC)[reply]
• Since I'm a follower of
  WP:OPERA I've seen the tons of your work for years. Isn't it time to write an article on how you do it and what tools you use? - kosboot (talk) 17:24, 24 October 2023 (UTC)[reply
  ]

  @

  FPs. 17:32, 24 October 2023 (UTC)[reply

  ]

• What a great tip Adam, thank you! Innisfree987 (talk) 22:03, 10 November 2023 (UTC)[reply]

Humour: New citation template introduced for divine revelations, drug use, and really thinking about it (1,078 bytes · 💬)

← Back to 2023-10-23

Discuss this story

• Documentation: The invention, creation or use of this template necessitates an altered state -- GreenC 18:57, 23 October 2023 (UTC)[reply]
• Speaking of God, I see that Elon Musk has offered to donate 1 billion dollars to Wikipedia if we change our name to "Dickipedia." Now that's an offer I would have trouble in refusing. "Dickipedia" doesn't sound so bad, does it, for our collective endeavors to bring some sense into this disintegrating world? (Ok, guys, I'm joking -- but I'm tempted to call Musk's bluff. Send us a check, Elon, and we'll open an RFC to discuss the subject.) Smallchief (talk) 13:33, 27 October 2023 (UTC)[reply]

• But only after Musk's check clears the bank. Cullen328 (talk) 23:38, 27 October 2023 (UTC)[reply]

In the media: Thirst traps, the fastest loading sites on the web, and the original collaborative writing (7,657 bytes · 💬)

← Back to 2023-10-23

Discuss this story

"Slate" celebrates encyclopedic selfies

• @Smallbones and HaeB: I think the wiki-link for Slate is wrong... Still, shout-out to @LittleT889: and the legend Annie Rauwerda! Oltrepier (talk) 08:05, 24 October 2023 (UTC)[reply]

  Thanks for the head-up; it looks like Ca already fixed it shortly before your comment. Regards, HaeB (talk) 03:53, 25 October 2023 (UTC)[reply]

Wikipedia is the second-fastest website in the US

• so... Wikipedia is spending money elsewhere, which some believe is unnecessary, and disbanding teams that do essential work? Why am I not surprised? —usernamekiran (talk) 11:14, 23 October 2023 (UTC)[reply]
  • I'm sorry, but I can't figure out which section in this article you are referring to. Smallbones_(smalltalk) 13:55, 23 October 2023 (UTC)[reply]

    @Smallbones This success is certainly in part due to the longtime work of the Wikimedia Foundation's recently disbanded Performance team... Counterfeit Purses (talk) 16:12, 23 October 2023 (UTC)[reply]

• OK, I got it. With no disrespect to the performance team (which appears to have been fairly small and effective), I took the main reason for the speed of Wikipedia to be the simple design, as emphasized in the "TechNewsWorld" article. Smallbones_(smalltalk) 17:44, 23 October 2023 (UTC)[reply]

AI finding references

This item could have benefited from a bit more context, e.g. the fact that the paper was already published last year in preprint form and received media attention back then. We covered it in both "In the media" ("Facebook experiments with Wikipedia fact-checking") and "Recent research" ( "Facebook/Meta research on "'Improving Wikipedia Verifiability with AI'") at the time, and the current story doesn't really offer any new information about this research project. That said, we might still run a fuller review in "Recent research" now that the published version of the paper is out. Regards, HaeB (talk) 19:36, 23 October 2023 (UTC)[reply]

Yes @HaeB: - technically this is well above my pay grade ($0). And a short paragraph couldn't possibly cover it as well as an article in Recent research, it would be great for this. Smallbones_(smalltalk) 20:10, 23 October 2023 (UTC)[reply]

To follow up on a remark by Piotrus (moving here, as a more suitable location):

unlike most coverage (and research), this seems actually useful. Underlying research is here: https://www.nature.com/articles/s42256-023-00726-1 The research say the code to reproduce the study is somewhere here: https://github.com/facebookresearch/side . Can anyone convert it it into a usable tool, assuming this has not been done already?

I agree this could be super interesting. Two things to be aware of though:

• The models alone appear to require two terabytes of disk space, so the server requirements are not quite trivial (e.g. it's not clear whether one could make this a
  Toolforge tool). That said, perhaps the Foundation's new "Lift Wing"
  machine learning infrastructure could be open to hosting such projects?
• In July 2022, the lead author stated that Note, Side is a POC [proof of concept] that shows the technology is there. To build a production system there is still lots to do. :) and the code repository hasn't seen new commits since then.

Regards, HaeB (talk) 21:03, 23 October 2023 (UTC)[reply]

Empirically we found that only using the first sentence in front of the claim and also adding the Wikipedia article’s title to the query did yield the best BM25 results. I know from my experiments that their approach, undoubtedly the best of what they tried, doesn't isolate the entire claim being sourced more than half the time. Associating a reference with the article text it is intended to support is very difficult even after examining the source in full. Sandizer (talk) 16:32, 25 October 2023 (UTC)[reply]

Interesting! Yes, I have been wondering what the state of the art is regarding this kind of

entailment problem. This paper highlighted in the July issue of "Recent research" (see also talk page there) seemed to have encountered more difficulties than the authors of the Facebook/Meta paper. (By the way, in case you are interested, we would still like to run a fuller review of the paper in "Recent research", which could touch on the various issues mentioned above.) Regards, HaeB (talk) 03:42, 27 October 2023 (UTC)[reply

]

See https://twitter.com/EdKrassen/status/1716103049863663962 and The Hill. Elon Mush offers $1 Billion if Wikipedia will change it's name to "Dickypedia" for 1 year. @JPxG: I got dibs on this story for the next issue! Smallbones_(smalltalk) 20:12, 23 October 2023 (UTC)[reply]

Weirder things

happened. ☆ Bri (talk) 20:39, 23 October 2023 (UTC)[reply

]

And don't forget

Joe, Montana Or the fact that somebody actually bought Twitter for $XX billion. Smallbones_(smalltalk) 20:56, 23 October 2023 (UTC)[reply

]

Don't we have

WP:POST/TIPS

and/or the Newsroom talk page for this kind of remark?

Anyway, thanks in advance for your sacrifice ;) Some background on how what triggered this one-side schoolyard row: [1]. And the Guardian is making some hay of it too [2]. Regards, HaeB (talk) 20:50, 23 October 2023 (UTC)[reply]

News and notes: Where have all the administrators gone? (31,068 bytes · 💬)

← Back to 2023-10-23

Discuss this story

Record low number of active administrators

The net -40 admins this year refers to all administrators, not active ones, which is partially a function of changes to inactivity desysop practice. Aside from the weird one-off drop from 8 February (498) to 9 Marchish (462), depending on how exactly you assess it the drop in active admins during 2023 is somewhere around 10-15 (eg. there are 449 as of last update). CMD (talk) 07:17, 23 October 2023 (UTC)[reply]

• Re: Record low number of active administrators – Long and short of it: it's not that being an admin is unattractive. For me personally, having the tools would mean more areas where I could help out and not having to bother an admin to edit a fully-protected page. While I've had a few run-ins with admins, the vast majority of my relationships with them have been highly educational, leading me to honor them, all of them, at every opportunity for their help and for the trust they have earned before and during their tenure as admins. So maybe it's the way an editor becomes an admin that is not so attractive? RfA with all its shortcomings over the years has given us many really good and trusted mop wielders; however, we all know how stressful it can be for some editors. Is there a better way? Perhaps when an editor meets certain requirements, longevity, number of edits and such, they become eligible to attend a formal admin instruction system. This school of sorts could show them the basics to help them decide where they want to work and such. I've always thought that becoming an admin should be much more educational and far less the rigors of a confrontational gauntlet. We could learn something from various other training programs and adapt them to what is inarguably an original and unique situation. But that's just me; I could be wrong.
  ed. ^{put'er there} 11:03, 23 October 2023 (UTC)[reply
  ]
• this article is going to generate a whole lot of discussion about administrators, and as usual, nothing will be done. ltbdl (talk) 13:13, 23 October 2023 (UTC)[reply]

  +1 unfortunately. Clyde _{[trout needed]} 03:19, 31 October 2023 (UTC)[reply]

• I'm not really active on enwiki, but are we sure that a low number of active admins is a bad thing? Like are there problems in areas of enwiki that are caused by a lack of admin participation? I feel like those problems should be mentioned when covering this "issue" if that's the case... Lectrician1 (talk) 14:20, 23 October 2023 (UTC)[reply]
  • The answer is pretty plainly yes, we need administrators. As a single data point, as I write this, there are 75 cases in the
    trust- and cooperation-based environment. ☆ Bri (talk) 17:53, 23 October 2023 (UTC)[reply
    ]

    But most admins aren't checkusers, right? That's a similar problem, but still a separate problem. -- asilvering (talk) 14:07, 24 October 2023 (UTC)[reply]

    Checkusers are needed for the part of the backlog that is pending checkuser, not the others. They probably need more clerks. I asked if they are short handed, but didn't ever get a reply. Clerks can be administrators but are not necessarily; of the current roster, about half are admins. If you are a non-admin clerk, you can't see deleted contributions so there are some technical limits to the amount of investigation that can be done. Clerking is frequently listed as a qualification of people seeking adminship. ☆ Bri (talk) 15:41, 24 October 2023 (UTC)[reply]

    Right, so again, the issue there is not a lack of admins. -- asilvering (talk) 16:29, 24 October 2023 (UTC)[reply]

    That's not my takeaway from the link to RfA discussions. The folks who bring it up are often saying to the community "I need to be an admin to be effective as a SPI investigator". And most of the time that is a good reason to be made an administrator according to the RfA outcomes. Quoting the nomination for Ivanvector, We need more SPI clerks, and we desperately need more who are administrators. I think that says it plainly enough. Oh, that was in 2016 by the way, when there were about 100 more active admins than today. ☆ Bri (talk) 17:29, 24 October 2023 (UTC)[reply]

  • added later See what Liz says below that seems to corroborate my observations. ☆ Bri (talk) 22:20, 23 October 2023 (UTC)[reply]
• I've long been in favor of an academy of some kind, as well as a rehabilitative academy/program where trust may be re-acquired after someone has had their mop removed. I still believe in the gauntlet of RfA. I believe the core value one brings to the position is trust. I'm biased towards a demonstration of commitment though time served. The entire process runs on AGF and willingness. I must remark, not very many folks at
  WP:ORCP any time recently. Of Archive 16, about 5 are non-serious queries (Carolina Heart, Ekdalian, L1amw90, Idontknowwhattouseasmyusername300, Seawolf35), five are serious inquiries where the OP received significant feedback (Epa101, Adolphus79, AirshipJungleman29, Mattdaviesfsic, Taking Out The Trash), one (Partofthemachine) is a deliberately disruptive self-nom, two (Curbon7, Ganesha811) are clear agreements and one (Ingenuity) is now an admin. BusterD (talk) 14:35, 23 October 2023 (UTC)[reply
  ]
• I notice several admins amongst several others, retiring this year. I wonder if its the start of a trend, lots of the original folk retiring and leaving. A lot of these folk have been here since the beginning. Nothing will change until we get rid the circus that is RFA. We need a much more professional system that doesn't puts people through the mill. Wikipedia lost a good editor this year who was massively productive as an NPP coordinator. He was really put throught the mill, beyond belief and then disappeared immediately afterwards. An application system and that is what it is, which is so destructive needs to go. If its a trend, it might accelerate, bringing the WMF in here in 3-4 years to appoint admins. We will have no choice. They are not going to let their product go under, due to lack of progess our part. scope_creep^Talk 16:28, 23 October 2023 (UTC)[reply]
• I can certainly understand the admin issue. It is an intersection of several problems. One is that administrators wield tremendous power compared to us ordinary yokels and have far greater prestige, by which I include the "soft power" of their status even when their "tools" are not used. Anyone who has ever ventured into ANI can attest to that. They have to be selected with great care because it is a lifetime position, very much like a federal judge in the U.S. Yet at the same time, the process of becoming an administrator is protracted and can be cruel. Longtime users avoid burnout by staying out of fights if at all possible, and finding a little niche in which they may be of use. This is a far better way of using one's limited spare time than becoming an administrator. I know of no way of resolving this problem, and it is a problem. Coretheapple (talk) 16:54, 23 October 2023 (UTC)[reply]

  There are some niches for administrators to avoid burnout as well, the thing is of course anybody unhapppy even with a single administrative action can just require a full report on everything an administrator has ever done, which in the best case scenario leads to a huge time loss. Ymblanter (talk) 19:14, 23 October 2023 (UTC)[reply]

  That "full report" you mention is news to me. First I've ever heard of such a thing. Is that something new? Coretheapple (talk) 16:37, 24 October 2023 (UTC)[reply]

  Not really. It is just a literal application of

  WP:ADMINACCT - in theory, an admin must be able to explain every single edit, and sometimes users insist on explaining every single edit in a certain episode. (Basically saying, that if the explanation is not forthcoming immediately, the matter gets referred to Arbcom). Ymblanter (talk) 19:43, 24 October 2023 (UTC)[reply

  ]

  Oh come on. Users, good or bad, haven't an icicles' chance in hell against admins in the drama boards unless they have a huge phalanx of supporters to back them up, especially at AN and ANI. I've seen abusive admins taunt proles about going to ANI. They know the deck is stacked. Sure you can ask for a "full report," and you will then be tasked with pounding sand. Coretheapple (talk) 20:55, 24 October 2023 (UTC)[reply]

  Take a look at Wikipedia:Arbitration/Requests/Case/RHaworth#Previous attempts at resolution to see how hard it is to desysop an admin who wasn't actually malicious or abusive, just a nice guy (I've met him in real life, he is) who was sub-optimal over years of issues. Ritchie333 ^{(talk) (cont)} 10:26, 26 October 2023 (UTC)[reply]

  I'm not familiar with the admin or arbcom case in question, but the record of that arbitration paints a different picture than you are presenting (e.g. he or she "regularly performed deletions that do not comply with the deletion policy" among other things), and it substantiates my point, that it takes the Wikipedia equivalent of World War II to dislodge bad admins from their posts. Calling such positions the "mop" as some do, and saying it is "no big deal" is delusional, as if it was such a triviality, admins that bad would be desysopped right and left. Coretheapple (talk) 14:22, 26 October 2023 (UTC)[reply]

  @Coretheapple I read Ritchie's comment as a supplement to your own, not a refutation of it. -- asilvering (talk) 20:30, 26 October 2023 (UTC)[reply]

  You know, you're right? I'm striking out my comment. Sorry Ritchie. Coretheapple (talk) 20:43, 26 October 2023 (UTC) Pinging Ritchie333 to be sure he sees my mea maxima culpa. Coretheapple (talk) 20:52, 26 October 2023 (UTC)[reply]

• Re "Record low number of active administrators", I think folks are just busier in real life since 2020, myself included, often caretaking for relatives with
  long Covid. I would love to be an admin again, and perhaps will run again in 2024. Bearian (talk) 18:03, 23 October 2023 (UTC)[reply
  ]
• It's not just a low number of RFAs with passing candidates that is an issue, we have areas that need admin care that are seeing fewer and fewer admins participating in them. This is a volunteer job so that is their perogative to work when and where their interest takes them, but I've seen a change just over the past few years. When I started closing AFDs nearly two years ago, I could probably name a dozen admins who would go over to close AFD discussions. Now, you look at a daily log page and you'll see the names of 3 or 4 (or sometimes only 2) admins on any given day closing discussions (along with helpful NAC closers). And it's not just deletion discussions but noticeboards as well. I'm not sure if it is burnout or getting called to Deletion Review to be scrutinized or just a loss of interest and moving on to other activities. But it's much easier to handle the load with a larger group of admins each doing a little bit. Of course, it could be that there is an overabundance of deletion discussions now compared to past years but someone would have to run the numbers to see what past averages were.
  It reminds me of when I was a graduate student working as an adjunct professor. We were all anxious because we were getting extremely low pay for teachig and they were not tenure-track jobs. But the regular professors were concerned because with fewer non-adjunct professors, there were fewer professors to sit on committees, fewer advisers to students, fewer bodies to handle the non-classroom duties that are part of their role as tenured professors that go largely unacknowledged. So, while we need more new recruits, through RFAs, we also have to keep existing admins interested in helping out where and when they can. Liz ^{Read! Talk!} 21:32, 23 October 2023 (UTC)[reply]
• A not-insignificant cause can probably be attributed to RFA. I'm not entirely saying it should be a cakewalk to become an administrator, especially given the sheer amount of things that can be done as a sysop that, for necessary reasons, not everyone has access to review and criticize about. One notable example should be the ability to delete pages. These are then entirely hidden from everyone except other administrators, so it's not like the permission should be handed out to anyone who tosses their hat in the ring. But perhaps something that is a problem is the immense and overwhelming pressure at RFA. I'm fairly sure anyone who's looked at an RFA oppose section can, more often than not, see commentary that can range from mildly distasteful to outright incivility, and this is perfectly fine in an RFA, while it would result in a warning or even a block in other places. It isn't necessarily a bad thing that the bar is a little higher for what counts as incivility at a forum to decide whether someone is trustworthy enough to wield some of the most powerful tools on the site, but the bar seems way too high for what is actually acceptable before an admin removes it for breaking
  WP:CIVIL. EggRoll97 ^(talk) 03:33, 24 October 2023 (UTC)[reply
  ]
  • I don't like the idea of setting the bar for incivility higher on RFA than elsewhere; it would just make RFAs the new ANI. If someone has an issue with a candidate that would cause them to oppose their candidacy, they should be able to express it without the use of incivility. Civility is
    notability, and discussions on talk pages are generally centered around those. RFA has no guiding Polaris, so people are essentially free to support or oppose based on whatever they choose. You can't help but feel rather disheartened when people just try a little too hard to find (idiotic) reasons to oppose. While coming up with such terrible arguments would at best get you laughed off a talk page, such a circus directed at the candidate can dissuade them from running again, if their impression of RFA is a place where editors go to turn their brains off. —k6ka 🍁 (Talk · Contributions) 01:06, 26 October 2023 (UTC)[reply
    ]
  • The civility rules are not uniformly enforced on Wikipedia to begin with, and in fact one of the perennial complaints one hears is that certain admins flout the civility rules themselves, while expecting those same rules to be enforced strictly if they are subject to criticism. If admins required re-confirmation every few years, that issue would be largely alleviated. Indeed, the fact that these are permanent positions, and that abusive admins cannot be readily removed, is one of the reasons RfAs are so demanding in the first place. Coretheapple (talk) 14:44, 29 October 2023 (UTC)[reply]
• While the fall in admin numbers is concerning, I think there are other metrics we should look at alongside those statistics. From what I can tell, we've managed to keep (most of) the admin backlogs down to a reasonable level fairly consistently over the last few years, even with fewer of us around to do it. Noticeboards like UAA, AIV etc. are usually cleared fairly quickly, people get a quick response at AN and AN/I, AfDs are often closed or relisted promptly 1 week after they open... so while the numbers might be dwindling, the "mopping and stopping" seems largely unhindered by having a smaller team doing it.
  I think the advent of various filters, bots, helper scripts and other tools have taken a fair chunk of the admin strain away so while we may be fewer in number, we're still a fairly efficient bunch on the whole. Perhaps another change is that a lot of situations that previously were new and needed in-depth discussion now have clear precedents and policies for us to fall back on, enabling us to act quicker and more decisively in a lot of cases. WaggersTALK 09:49, 24 October 2023 (UTC)[reply]
• I'm with the others skeptical that lower admin numbers per se are a problem, though obviously it isn't an encouraging trend. Liz's comment about service work in grad school strikes me as illustrative here: I think the problem is a lack of committed, capable volunteers in various parts of the project. In my nearly three years of AfD participation, I have only very rarely noticed a concerning backlog in closes - usually none at all, and never more than a couple of days, and then only for difficult cases. My enduring impression of AfD is that it desperately needs more contributors, not more closers. I am an AfC reviewer, and we need many more AfC reviewers (also not admins) to keep on top of that backlog. SPI is cited above as another area, and again that's one that doesn't need admins as such, but rather clerks and checkusers. The only area I can think of that needs more admin help is DYK (and I remember leeky's first run at RFA and the absolute deluge of ageism she received for doing so). A common RFA oppose is "no need for the tools". Frankly, it doesn't look like Wikipedia is desperate for people to use those tools, nor does the community tend to support people who offer to use them. What Wikipedia does need is sensible people working in various areas of need. We need work on mobilizing more people to do these back-end maintenance jobs. -- asilvering (talk) 16:53, 24 October 2023 (UTC)[reply]

  I have noticed this as well. One of the areas most starved for attention is

  Copyright cleanup investigations. Most of the work there doesn't need admin rights, just somebody prepared to go through and do a bunch of what I'd consider to be boring and repetitive work, with an added bonus of being yelled at if you make a mistake. To be honest, it's one of those things you won't make a serious dent in unless you paid someone to do it, but the whole stigma of "paid editors" (as opposed to simply crap paid editors with a massive conflict of interest) seems to shut down this avenue of thinking. Ritchie333 ^{(talk) (cont)} 10:23, 26 October 2023 (UTC)[reply

  ]

  @

  WP:AFC in Chinese and Arabic. The fluently multilingual volunteers we have have more interesting and less burnout-inducing things to do with their time. For the copyright cleanup, I assume that will be backlogged to hell until someone sues the WMF over it, at which point we may indeed get a "paid editor" working in that space. -- asilvering (talk) 20:39, 26 October 2023 (UTC)[reply

  ]

  @

  WP:RfD instead of AfD, as of the time of writing, the backlog goes all the way to September. And that's not just a one-off - RfD is pretty much always backlogged a ridiculous amount. RfD is hardly the most important thing that admins could be doing, but that's exactly what makes it such a great barometer of administrator backlogs. It is the routine, boring running of the wiki that is backlogged, not necessarily the high-profile stuff that admins (and regular users) want to do. --NYKevin 17:29, 31 October 2023 (UTC)[reply

  ]

• In my own humble, but experienced opinion, there have been two factors at work with the number of admins. One contributed to an unreasonable inflation of numbers: it's obvious to anyone who has edited Wikipedia for more than a few weeks that since anyone can edit Wikipedia, there's no way to prove one is an "important" member of the community -- as opposed to a dabbler who is sorta involved -- which led to a lot of people wanting to be an admin. However, when the Foundation evolved from a paper existence to a real one, those needing to prove they're an "important" member can get involved in the functionary aspect of things, thus reducing the attraction of being an admin. (By "functionary" I mean those volunteers who occupy themselves in the numerous committees or the Wikimedia chapters. Not to say all, or even most, people who are involved in those are simply looking to add a line to their resumes. However, it is a fact most committees will have one or more members who are there to simply prove they are doing important work -- the more important the committee, the more likely it will have deadwood.)
  The second factor effecting the number of admins is a negative one: most of what an admin does is not all that exciting. Spending some time doing it with little acknowledgement -- & more often negative than positive or constructive -- results with admins losing enthusiasm for the task. I'm not pointing fingers here -- I'm definitely as bad if not worse than anyone else about not providing positive reinforcement -- but for the Wikipedia community to survive, we need to provide more support for each other. -- llywrch (talk) 16:59, 24 October 2023 (UTC)[reply]
• Update: Daily update, 444 active admins - Rick Bot, latest. Another record. ☆ Bri (talk) 17:37, 24 October 2023 (UTC)[reply]

  Working on it, give me a minute... jp×g 01:38, 26 October 2023 (UTC)[reply]

• tl;dr: Wikipedia does not have enough administrators → People complains about it → People all agree that RfA is the main root cause for it → Make an RfC on how to fix RfA → People complain that doing so would lower the administrator standard → These changes are not enacted → Wikipedia does not have enough administrators → ... CactiStaccingCrane (talk) 13:07, 26 October 2023 (UTC)[reply]

  @CactiStaccingCrane: You ever look at the instructions for nominating: Wikipedia:Requests for adminship/Nominate It's by far the most discouraging page on here. There's an essay somewhere too that lists all the things you have to do including make your signature custom but boring, which pops to my memory because I just looked JPxG's RfA and one of the top questions is Are you open to changing your signature....Rjjⁱⁱⁱ (talk) 21:43, 28 October 2023 (UTC)[reply]

  Time to fix it then. CactiStaccingCrane (talk) 04:07, 29 October 2023 (UTC)[reply]

• I just got a notification of a new RfA discussion, meaning as of date there are 3 candidates, all of whom have high numbers of supporters. Seems like this edition of Signpost is doing something. GeraldWL 04:08, 28 October 2023 (UTC)[reply]

  Ha ha ha ha, this page actually helped a lot. Kinda like when three busses follow the same path... The 🏎 Corvette 🏍 ZR1^{(The Garage)} 23:29, 29 October 2023 (UTC)[reply]

And bang goes another admin: Wikipedia:Arbitration/Requests/Case#Lourdes. Oh well. Gråbergs Gråa Sång (talk) 09:50, 2 November 2023 (UTC)[reply]

Permalink for the reader of the far future. Rotideypoc41352 (talk · contribs) 17:43, 2 November 2023 (UTC)[reply]

Knowledge Equity Fund

• "Data for Black Lives which we connected with AfroCrowd and Black Lunch Table... Through these connections, we have seen positive synergies within the movement at large". That's very nice. Now, can someone translate this from corporation speak into plain language? We have spend five six(!) digits on that particular KEF grant, presumably. And what does it mean for Wikimedia? All I can find is more happy marketing corpo speak at [3] that "They will be receiving a one year grant of $100,000, which they will use to launch a Movement Scientists Fellowship. This Fellowship will match racial justice leaders with machine learning research engineers to develop data-based machine learning applications to drive change in the areas of climate, genetics, and economic justice. They will also launch a new series of educational programs, such as free and open oral histories that promote data literacy." I applaud what those NGOs are doing - but I don't see why we should be funding them?? (Bonus points for anyone who can point me to where that particular grant proposal actually exists, I couldn't even find it through Google or meta search). --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 10:55, 23 October 2023 (UTC)[reply]

News from Diff: Sawtpedia: Giving a Voice to Wikipedia Using QR Codes (16,284 bytes · 💬)

← Back to 2023-10-23

Discuss this story

Sawtpedia is more than just a tool; it's a leap forward in making Wikipedia accessible to a broader audience. - is it possible to quantify this claim a bit more? How often is Sawtpedia used per day currently, and how do the usage numbers per topic compare to the pageviews of the Wikipedia article(s) about the same topic, say?

check this article: 7 reasons why museums should use Sawtpedia! - there seems to be a link missing here? The story could also have used some copyediting (e.g. there's a word missing here: QRpedia, the precursor Sawtpedia).

Regards, HaeB (talk) 19:12, 23 October 2023 (UTC)[reply]

Thank you @HaeB for your feedback ! This should be possible in the next version of the tool as we would like to include a dashbaord that displays the number of scan, some statistics on the languages used and other useful information . This will be very helpful to identify for instance which recordings or articles are missing.

Re the missing article "7 reasons why museums should use Sawtpedia!" the article will be available on diff soon and the link will be added. Apologies for this inconvenience. Yamen (talk) 00:50, 30 October 2023 (UTC)[reply]

Thank you for this clarification. It's good to hear that it is planned to obtain data that will enable quantifying the actual impact of Sawtpedia. Until then, I would suggest to refrain from claims that may turn out to be exaggerated. Regards, HaeB (talk) 21:12, 4 November 2023 (UTC)[reply]

How is a QR code "more accessible" than an ordinary clickable link to an .ogg file???? Contrary to the delusions of some technophiles, not everybody on this planet has a smartphone which they use for QR codes. --Orange Mike | Talk 21:19, 23 October 2023 (UTC)[reply]

All this is about people walking around in the real world - not sitting in a chair with a laptop, so we're not really comparing apples to oranges. Now if you are carrying your laptop around with you and you can type in the Wiki article title, then clicking a link would be easier. But you have to admit, if you have a phone (or a laptop) that can scan the photo, this will take you straight to the audio. I would guess (with a bit of reprograming) the QRcode could be replaced with a regular link for desktops and then the audio could be called up easier in that context. And maybe (with a bit more reprograming) you could hear a complete reading of any Wiki article - in the most recent version. That would be super cool, but it is getting away from what they are actually doing now, which is good enough for me (so far). Clicking a link might be very useful for vision-impaired folks, but designing it for the way those folks actually would use it would probably be way more important. So now I'm really getting ahead of things, but I think QRcodes are a good step along the path. Smallbones_(smalltalk) 00:56, 24 October 2023 (UTC)[reply]

Thank you @Orangemike for the question and @Smallbones for the answer ! I will add that a QR code can be very helpful for instance for museums to provide visitors with a better experience by scanning the QR codes and listening to a description of the displayed collection especially that with the one QR Code displayed we can have access potentially to all the languages supported by Wikipedia no need to have a QR Code by language. Yamen (talk) 00:55, 30 October 2023 (UTC)[reply]

Thank you so much for presenting this project, all of the overall goals you set definitely resonate with me! Obviously, this tool still needs improvements, and I have a few doubts about the availability of ready-made audio files (since not many users seem to contribute regularly anymore), but still, I can see huge potential in here! Oltrepier (talk) 08:37, 24 October 2023 (UTC)[reply]

Thank you @Oltrepier for your feedback! Yes definitely the tool needs improvements especially to support the iOS devices which have an issue reading ogg files, also to be able to support local languages/dialects that are not installed on the phone and to provide more statistics. Regarding the availability of ready-made audio files, actually the tool can be a good reason to motivate people to record more articles so their voices can be heard (it can be a way to foster a sense of community engagement for example within a city or a museum community). Worth noting also that if the audio file is not available then a Text To Speech will be triggered and the user will be able to listen to the article and with the improvement of the TTS technology we can think in the future to automatically generate audio files and upload them to commons (we need just to sort out the copyright issue). Yamen (talk) 01:02, 30 October 2023 (UTC)[reply]

Nice news Yamen Bousrih et al ... a clever idea and thanks for the acknowledgement. Now, I wonder if you/we have thought about upgrading the existing QRpedia code so that codes in Australia, Ukraine, Germany, Wales etc suddenly start reading stuff out aloud. There were codes on Tutu's, Mandela's house, US heroes graves and on Gandhi's house. Obviously we wouldnt change the plaques but the code... more ideas available if we open the code box. Victuallers (talk) 09:11, 24 October 2023 (UTC)[reply]

Thank you @Victuallers for your feedback! actually the tool is not meant to replace QRpedia as providing the possibility to read the article is definitely a good thing. Nevertheless what we can think is when scanning the exsiting QRpedia Code, we can ask the user to choose between reading or listening to the article. This of course won't require to change the plaques but we will need to do a small adjustment to the QRpedia code. I will try to contact Roger Bamkin to explore this idea with him. thank you very much for this suggestion !! Yamen (talk) 01:07, 30 October 2023 (UTC)[reply]

{{ping|Yamen}} I am Roger Bamkin. You have contacted him - see my talk page?. Your interesting work is not seen as a threat but as an opportunity. Victuallers (talk) 08:33, 30 October 2023 (UTC)[reply]

Thank you @Victuallers and apologies for not making the link with your username :) . I have sent an email and will be happy to have a call to further explore the idea of offering the option to read or to listen to the article when scanning an exisiting QRpedia code. Yamen (talk) 23:01, 4 November 2023 (UTC)[reply]

QR codes will also be available soon for each article. See phab:T242467 and currently in testing at beta: https://en.wikipedia.beta.wmflabs.org/wiki/Barack_Obama available in the tools menu as "Download QR code". —TheDJ (talk • contribs) 12:11, 24 October 2023 (UTC)[reply]

I immediately imagined a Wikipedia triple sticker on a sign in front of a tourist attraction with QR targets for (mobile) text, audio, and video versions. Or a whole cluster of QR for languages. No, not good. Probably better to have one QR leading to a paragraph or just to choices starting with language. Audio versions would be robot voices unless the article attracts so much interest that we make a good narration, and video would I think be rare but eventually who knows? Jim.henderson (talk) 20:55, 27 October 2023 (UTC)[reply]

@Jim.henderson: I think you've got the idea here! There's a lot that you can do with this type of QR code. I'll make a few suggestions, though, of things I think you've missed, e.g. there would be little need for a slew of language QR codes to be posted. The program recognizes your preferred language on your phone and should read the introduction of the article in whatever language version you prefer. In case there isn't an article in that language version, I believe it defaults to English.

I'm interested in how much people like or dislike the mechanical voice. I'm pretty impressed. I'll rate it 95/100 for pronunciation, 90/100 for general naturalism, but only 75/100 for reading tempo (there are occasional long pauses in the wrong places). You can listen to the article introduction for the Medina simply by scanning the QR code at the top of the article. It's a bit short, so I've put the code for a much longer introduction (for the Liberty Bell) down here. Well maybe it's not just scan and listen, but almost. With my phone it takes 4 steps:

1. turn on my phone

2. turn on my phone's camera and point it at the QR code

3. (the QR code appears amazingly quickly at the point) click the link

4. (generating the voice can take some time) press the little triangle symbol to hear the voice.

That's it (Yes, i was trying to be as detailed as possible)

By the way, the mechanical voice is required to read out the most recent article version every time. Having pre-recorded voice files would get in the way of this (in the current version - I'm writing about things here that could probably be done with minor changes in the code).

Your idea about linking to a menu paragraph could turn out great! Maybe just 4 items: 1) current article text; 2) current article audio; 3) pre-recorded audio; 4) video.

For on-site use (e.g. in front of a building) I don't think a lot of people would need to use video. Why not just look at the building in front of you?!

BTW, I love video on-Wikipedia. You might enjoy some of my NRHP efforts, e.g. my best such video is on the right. They are hard to make, and hard to edit (I couldn't make this now). Some other such efforts are: A Walk up Main Street, Adamstown, PA (2 minutes), File:A Walk down Main Street, Delta PA.webm, and File:Wellsville, PA.webm. On second thought, maybe posting a QR code to video would be useful at the start of these walks. As always, Smallbones_(smalltalk) 19:04, 29 October 2023 (UTC)[reply]

Thank you @Jim.henderson for the idea of the video option and I think this is definitely a very good idea to add as I believe in the importance of video content especially for new generation who prefer watching rather than reading.

@Smallbones thank you again for your valuable feedback and I like very much your idea of having a menu paragraph with 4 choices. I'm on the same page as you regarding the importance of having videos on Wikipedia and I like the videos that you shared. I tried few years ago on the French Wikiepdia to replace the main photo in the infobox by a video as I think it will be more useful for user to get quick information before reading all the article. You can check this example here: https://fr.wikipedia.org/wiki/Utilisateur:Yamen/Brouillon/Amphith%C3%A9%C3%A2tre_d%27El_Jem

I'm still waitng for a better version of the tool Video Wiki to be able to produce easily more videos. Yamen (talk) 01:16, 30 October 2023 (UTC)[reply]

Something very important to share about the QR codes: using QR codes requires the installation of a plaque or a sign having the QR code to be scanned which is not possible sometimes without the cooperation of the owner/manager of the building or the GLAM institution. To bypass this obstacle we are trying to have another version of Sawtpedia based on an augmented reality and IA which means that you don't need to scan a QR code anymore but just to open the Camera in front of a monument or a museum collection and the tool will identify which Wikipedia article corresponds to it. I will keep you posted about this tool :) Yamen (talk) 01:22, 30 October 2023 (UTC)[reply]

So, I see a few ways to get into the guide.

First we talked about the signpost method. A sign in front of the castle or other item has a QR code. The user knows what a QR code is, scans it, and enters the system. Variant version, a paper map has the QR code. In both of these, no Wikipedia app is necessary but the sticker on the signpost (or paper) is.

Second, the points of interest app version. The phone is running an app that knows its location, either by AI recognition or more likely by GPS. App sounds a gong and user looks to see what's here. No sticker, but the user must have installed and started the app.

Third, the robot tour guide. The app knows its location, and suggests a route. "Walk ahead 30m and turn left towards the statue" same as an automobile navigating app, but when you're there, it asks whether you want to know about the place you have reached. In younger years I was sometimes a tour guide here in New York, and I'm sure the roboguide won't be as smart as I was, but not every tourist was lucky enough to have me. No sticker; app yes.

Fourth, the WikiShootMe model. Being a Wikiphotographer, I often open that page when I'm walking in a place not close to home. It shows a map with my location and various green and red dots for places that have, or don't have, a Wikiphoto. I snap and upload a picture and the red dot becomes green. Based on that idea, use an app that shows all the nearby Wikipedia articles so you can tap on the item before or after going there. Again no sticker, and again yes it needs the user to install the app. Jim.henderson (talk) 01:46, 1 November 2023 (UTC)[reply]

Thank you @Jim.henderson for the idea of the map. This is an amazing idea and can be very useful in historic cities like the old towns listed as World Heritage e.g. the medina of Tunis in Tunisia, median of Fez in Morocco or Stone Town in Zanzibar, etc. These cities can be described as an open air museum with several building/monuments documented on Wikipedia. Displaying a map of these buildings/monuments can be used as an audio guide! I really like this and we will try to implement it asap under this url: map.sawtpedia.wiki

The map can't be used in closed spaces like museums and in this case an app can be built using AI to recognize the displayed items (in case there is no QR code displayed).

Thanks Yamen (talk) 23:10, 4 November 2023 (UTC)[reply]

Interesting idea with QR, there are a few comments:
1. The Svg format is not as often used as some others (JPEG), as far as I understand. This may be a problem for some readers.
2. It is not clear whether this will work with “suspicious” IP (such as URL-shortener). Any reader can get “suspicious” IP. See also Wikipedia:Blocking IP addresses#Problems and solutions [4][5].
3. This is only for the desktop version, as far as I understand (less than 50% page views).
4. It is not clear how users will pay attention to it. More over, if it is promoted, it will mislead users of the mobile site (see 3.)--Proeksad (talk) 11:21, 30 October 2023 (UTC)[reply]

Traffic report: The calm and the storm (0 bytes · 💬)

Wikipedia talk:Wikipedia Signpost/2023-10-23/Traffic report