Shortcut: WD:AN

Wikidata:Administrators' noticeboard

From Wikidata
Jump to navigation Jump to search

Administrators' noticeboard
This is a noticeboard for matters requiring administrator attention. IRC channel: #wikidataconnect
Translate this header box!
Start a new discussion
On this page, old discussions are archived. An overview of all archives can be found at this page's archive index. The current archive is located at 2024/06.

Requests for deletions

high

~157 open requests for deletions.

Requests for unblock

low

1 open request for unblock.

Report concerning JMagalhães

[edit]

A user has been exhibiting disruptive editing behaviour, focusing on removing aliases and modifying genders. Here are some examples:

We can also see they are removing terms without including them on the alias in several other moments:

This behaviour is problematic for the following reasons:

  • Information Loss: Removing aliases hinders the searchability of entities with different names.
  • Disruption: Edit warring disrupts the collaborative editing process by repeatedly undoing other editors' work.

The core concern is removing information and edits warring, not just modifications. These actions violate established practices and confuse users relying on accurate and consistent information.

 Comment The fact that this account has been persistently pushing for controversial and/or blatantly wrong changes through edit warring, and then complains in this noticeboard posing as the victim is absolutely puzzling. The first half of the links are simply controversial changes being undone. That's it. Stop POV-pushing through edit wars and get consensus for the changes. The second half... I'm just clueless about whatever this editor is trying to prove or claim.

Beyond that, I would like to point out to administrators the systematic disruptive use of edit summaries by this account. Among its last 500 edits, 83 summaries are used to engage in personal attacks to users who simply undid one of his controversial changes or made some sort of edit he didn't like, systematically calling them "vandals" and "trolls". And this is not the first time this user opens blatantly nonsense incident reports. For example, last year he opened an absolutely bizarre and nonsense claim against me claiming, with no proof whatsoever, that I was "doing vandalism driven by sexism". Being an administrator myself for 10 years, I've dealt with many bizarre things, but this is hard to beat. I'm considering open a T&S case against this user because his claims are so outlandish and out of touch with reality that I'm afraid there's underlying issues that should be handled privately. JMagalhães (talk) 18:10, 9 June 2024 (UTC)[reply]

The fact that you attacked me, not defending your actions, says loudly about you. Also, it is not only me that you attack and enter into edit wars. It is not me who is removing information from Wikidata.
It is a clear no-posture when discussing his actions.
Moreover, this "his claims are so outlandish and out of touch with reality" should be enough for a long block. Attacks that no one should receive. Rodrigo Tetsuo Argenton (talk) 22:02, 9 June 2024 (UTC)[reply]

 Comment It has already been explained to this user that when his controversial edits are undone, he should not be pov-pushing through edit wars and should instead engage in consensus building. Yet, not only he keeps pushing unwanted edits and massively using summaries for personal attacks. But, in a bizarre move, he pretends to be some sort of victim and opens incidents reports with completely made-up stuff and claims that are beyond outlandish. Honestly, an administrator intervention would be nice to help this person return to reality. JMagalhães (talk) 10:00, 10 June 2024 (UTC)[reply]

Again, another attack in front of all admins, and nothing is done. Rodrigo Tetsuo Argenton (talk) 20:56, 12 June 2024 (UTC)[reply]
Nothing they have said here appears particularly like an attack, particularly given you have dragged them before this noticeboard. What I see, reviewing items like Wikimedia Commons (Q565), is you making a change, and someone else reverting. You should stop edit warring to re-add your content in and discuss on the talk page. See w:Wikipedia:BOLD, revert, discuss cycle - make a change, it gets reverted, then you discuss. You are being more disruptive here than the other user. Please step back and discuss, preferably on the relevant items' talk pages. -- Ajraddatz (talk) 13:18, 26 June 2024 (UTC)[reply]

Report concerning Olea

[edit]

A few years ago the user @Olea: added a dataset to Wikidata about Digital Guide to the Cultural Heritage of Andalusia (Q5758805). This file, in many cases indicates elements with ambiguous terms in a table, such as house, when in reality what it is talking about in the text is more specific buildings such as a town hall. The source clearly explains that it is a town hall, the term house is very general and not very precise. Furthermore, the source itself indicates that it is a town hall, there is no doubt about it. A few days ago Olea undid here my edition, because he says that the source indicates that, but it is not really a house, nobody lives in it. I have also indicated to him that it was not relevant to put "Contemporary Age", as he has put in all the elements of Digital Guide to the Cultural Heritage of Andalusia (Q5758805), because he is already adding the specific year. It seems that he would like to add more editions because it is not important information. Another error with the source is to interpret that the years that appear in the table of links such this are really the beginning and end of the use of the building, as here has indicated. In this item I deleted the incorrect year because the source indicates it was built in 1929 and he has undone me again, even though I added the referenced year just below. All the elements he added are created like this, and after telling him it refuses to modify them correctly because he says the source says that. The problem is not exclusively this. When he added the dataset, I indicated that he had duplicated hundreds of elements, and I had to merge many of them. He told me he was working on it, but at no time has he responded to my messages about its completion and I have recently found several more, so he is not dedicating himself to solving these problems. He also duplicated property values ​​and I sent him a query with 826 duplicates that I had left there over the years. This problems are added to other datasets that he also added on other topics, use Mix'n'Match and leave the names as indicated by the tool even if they are in capital letters and without more data to identify the elements, other duplicates issues, etc. I ask that the user be told that he cannot add datasets without making prior queries, that he has to review the veracity of the source data before adding them, that he has to review the data after inserting it to erase all those errors from the tools introduce, that he cannot move on to other tasks until he has correctly left the added data and that he answers the messages that are written about it, solving the errors and not hindering the tasks of the rest of the users. Thank you. Vanbasten 23 (talk) 09:55, 17 June 2024 (UTC)[reply]

I have asked Olea to respond here — Martin (MSGJ · talk) 11:59, 17 June 2024 (UTC)[reply]
Thanks @MSGJ, but I don't think I have nothing significant to respond here :-) —Ismael Olea (talk) 18:23, 17 June 2024 (UTC)[reply]
If you have created items containing errors, then it seems reasonable to ask you to fix those up, or to provide a timescale for when this might be done. Are you willing to do this? — Martin (MSGJ · talk) 21:13, 17 June 2024 (UTC)[reply]
@MSGJ It is 2024 and I'm fixing any error I'm finding since that data load in 2020. I'm still updating when I find updates at the source. You can check my edits at 10th of June or before as an example. Other users could confirm I'm actively curating all the Andalusian heritage in Wikidata (more than 24k items) since then. Not just that, some months ago I wrote a retrospective about my experience with this data set I presented in Wikidata Days 2023. So yes, I'm willing to keep and enhance the dataset quality and integrity. —Ismael Olea (talk) 16:28, 19 June 2024 (UTC)[reply]
 Comment I think that in the two answers you have given here you do not provide clear solutions as we are asking. "I'm willing to keep and enhance the dataset quality and integrity" you said it several times, what is going to change now? You say you started in 2020. Shortly after I started writing to you about the matter. After 4 years we continue here. Do you understand the frustration of a user when he does not see solutions to problems, and also does not receive a response to his messages? I had to merge many elements and I told you that there were other problems to solve. You told me you were on it. You made other insertions, you fixed some problems that arose, and you still didn't spend any time on it. I wrote to you again, you were still at it. In June 2022 you proposed other different jobs and I told you again that you were not paying attention to these problems. I have let it be, but now I change a property and you revert it saying that it is fine, when it is not, thereby preventing others from solving your errors. You say "other users could confirm", am I not another user? What do you call "willing to keep and enhance" when you have not yet deleted all the end dates of the instances? That data has been wrong for 4 years and you keep letting it pass. The same as the query I sent you. I dedicate my time so that there are no solutions... Also, I don't think it's appropriate that you give us a link to promote the work you do outside of here, while you are not respecting the rest of the community. I think I am always available to help people, but if there is an intention to do things well. We all make mistakes, but you have to be a little sensitive with the rest. --Vanbasten 23 (talk) 20:13, 19 June 2024 (UTC)[reply]
To be cristal clear:
  • You talk about sensitivity but you always address me in an imperative way. I'm not into this.
  • You don't manage how I spent my time.
  • You describe a partial/subjective view of my work: ok, you have the right to ignore the rest but then don't bring accusations of wrongdoing.
  • It is insulting how you describe the link to a document presented in a Wikimedia community event.
  • Outside Wikimedia I would say you are harassing me, but here I would not make an accusation so serious because I don't have interest spending my time in awful non productive discussions.
In brief, I'm tired how you talk to me. —Ismael Olea (talk) 11:46, 20 June 2024 (UTC)[reply]
@Olea: I'm not going to play into your games so as not to answer the topic that has been opened here. I think you are confusing yourself with someone else, outside Wikimedia I don't know you at all, so I can hardly harass you. I repeat what I have told you other times, if you need me for something here I am, but since you don't answer my messages I have to ask an admin (@MSGJ:) to take action. All the best. --Vanbasten 23 (talk) 18:14, 20 June 2024 (UTC)[reply]

Undelete Kaiman Lee wikidata entry

[edit]

Please undelete the entry for Kaiman Lee on wikidata (Q126734059). More editing needs to be done. Ninakp (talk) 21:15, 19 June 2024 (UTC)[reply]

 Question Have you consulted the notability policy? If yes, could you elaborate a bit how you believe it meets the notability requirements? --KonstantinaG07 (talk) 22:15, 19 June 2024 (UTC)[reply]
courtesy pinging: user:Bovlb as a deleter--Estopedist1 (talk) 06:11, 20 June 2024 (UTC)[reply]
Thanks for the ping. I have undeleted the item. Please add some claims soon to prevent it from getting deleted again. You may like to read User:Bovlb/How to create an item on Wikidata so that it won't get deleted. Bovlb (talk) 15:12, 20 June 2024 (UTC)[reply]

Abuse filter/Revert bot against LTA vandals

[edit]

On Wikidata:Requests for checkuser/Case/LTA vandals we have confirmed some persistent LTA vandals, of which I have yet to find a common pattern. Do you have any ideas for an abuse filter or a revert bot? Pinging some active admins who have edited abuse filters multiple times. @Matěj Suchánek, DannyS712, Ajraddatz, Bovlb: What do you think? --Wüstenspringmaus talk 14:32, 20 June 2024 (UTC)[reply]

There is already more than one abuse filter which deals with this specific LTA (just check Special:Log/block). They have even got their own "callsign", and there is a publicly accessible page somewhere where all information about their activity is collected (I'm not sharing it here for obvious reasons). --Matěj Suchánek (talk) 14:37, 20 June 2024 (UTC)[reply]
From a quick look at the contributions, I see three abuse patterns: text in IMDB id; inappropriate images; inappropriate text in aliases.
The first would be easy to exclude with a filter that required IMDB ids to be correctly formatted. We could do the same thing with most identifiers. This sort of filter might be a net gain for the project as we see a lot of malformed identifiers (usually complete URLs). Historically we have tended to avoid such filters, presumably on the grounds that it's better to get the information in there and fix the format later. Alternatively, we could just keep a closer eye on Wikidata:Database_reports/Constraint_violations/P345#"Format"_violations.
It's harder to write a filter against inappropriate aliases. There's already a lot of variety in what good faith editors consider to be reasonable aliases. We could probably train an AI to do this with some reasonable level of accuracy, but I don't think we could run that as an abuse filter. Still, it would help us to react faster when there's an attack.
Identifying inappropriate images would be quite hard. I'd hope that if an image is appropriate for an entity, then the entity is referenced somewhere in the image's metadata on Commons, but I'm sure there are a lot of valid exceptions. Again, we can't check that in an abuse filter. Bovlb (talk) 15:27, 20 June 2024 (UTC)[reply]

Please revert June 20 edits by User:Jephtah Ogyefo Acquah

[edit]

Jephtah Ogyefo Acquah (talk • contribs • deleted contribs • logs • filter log • block user • block log • SUL (for IP: GUC)) I just spent two months of my life fixing problems with badly merged individual records almost entirely caused by naive users using Magnus Manske's distributed game - duplicate authors tool. I've asked Magnus to fix it and it may be slightly better than it was, but it is still recommending far too many merges of people who are NOT the same person. Anyway, I checked my watchlist and recent changes just now and found the above user has made over 1000 edits this morning that are merges using this tool. If I really have to I'll spend another few weeks fixing these, but I'd prefer this just to be mass reverted if possible. ArthurPSmith (talk) 14:33, 20 June 2024 (UTC)[reply]

Note this is somewhat urgent because Krbot will be moving links to the merged items in the next few days, which makes things much harder to revert. ArthurPSmith (talk) 14:36, 20 June 2024 (UTC)[reply]
I spot-checked a few of this user's merges that were done after your recent warning, and they were all either good merges or the items had too little information to say either way. If there's evidence that the high error rate is continuing despite the warning, we should block them from mainspace until we get some resolution. Bovlb (talk) 15:37, 20 June 2024 (UTC)[reply]
@Bovlb: The error rate has been 30-50% in recent weeks, and I don't think this person is being more careful yet. For example this merge from earlier today is clearly wrong, the names differ significantly. The ones that are "ambiguous" have to be sorted out by looking at the "What links here" relations - these are authors of articles. Considerable care has been made to disambiguate for example authors of high energy physics papers that can have hundreds of authors, some with the same first initial and family name; usually it's based on affiliation. These edits are wholesale throwing out this work. ArthurPSmith (talk) 15:56, 20 June 2024 (UTC)[reply]
Some more examples of clear errors from today: Kunhong Kim vs Kyung-Sup Kim, Heon-Jeong Lee vs Hyang Woon Lee, Fuyi Wang vs Feng Wang, Young Kee Kim vs Young-Min Kim. That last one is especially egregious - Young Kee Kim is the current president of the American Physical Society! ArthurPSmith (talk) 16:04, 20 June 2024 (UTC)[reply]
Unfortunately those specific bad merges were before your warning. I realise that you're seeking mass reversion here, but I was focussing on a narrower issue: whether we should use a partial block to prevent even more problems and encourage communication. Bovlb (talk) 16:08, 20 June 2024 (UTC)[reply]
@Bovlb: A block is not really what I'm asking for. I need to know soon whether a mass revert will happen, because if not I'll be spending my next few days trying to fix as many of these as possible before Krbot hits and complicates things. ArthurPSmith (talk) 19:06, 20 June 2024 (UTC)[reply]
I do understand to errors some of my edits had. It wasn't intentional and I do understand the work loads those errors bring. Offen at times I do edit and realise there isn't anything/button to undo what I have sent already because the moment a click is made it automatically records. Once again sorry for this mistakes. I will be more careful when editing next time. 102.176.66.185 23:03, 20 June 2024 (UTC)[reply]
This addresses the symptoms and not the underlying cause. Unfortunately it seems there is no rollback link for items that were merged _from_, which makes cleanup much harder to do than it should be. Is there any chance at all to add some functionality to this tool or add functionality externally that would aid in any cleanup that has to be done? Infrastruktur (talk) 06:29, 21 June 2024 (UTC)[reply]
ArthurPSmith suggested limiting the use of the tool to confirmed users. Not sure this would help making sure the merges aren't careless. The tool already displays a notification to users urging them to be careful making sure the items are the same, does it not? Maybe there should be an example or two to new users showing what to look for to make sure the items are the same? A graphic or video demonstration would probably be ideal. Time might be better spent on this than on cleaning up after the fact. Infrastruktur (talk) 12:30, 21 June 2024 (UTC)[reply]
Magnus has disabled this tool for now. I don't understand much about the tool itself - it's probably been years since I tried the "distributed game". Treating merges as a "game" for new users seems always likely to be fraught with risk. But if somebody could provide some help for new users on this maybe we can risk it. It certainly does some good in cleaning up duplicates, but right now it's just far to easy to merge things that shouldn't be. ArthurPSmith (talk) 14:56, 21 June 2024 (UTC)[reply]
I've started work on reverting by hand where I can see these are wrong. For the 500 edits from 01:26 to 02:56 there were 215 merges, of which I have now reviewed all and reverted 81. So that's about 38% error rate, roughly in line with how this has gone in the past. The tool just makes it far too easy to cause trouble here. ArthurPSmith (talk) 19:41, 21 June 2024 (UTC)[reply]
FYI this review work is finished, no need for mass reversion. This was a lot of work over a few days but I think I caught it before any bots moved redirected values to the merged ones, in the cases of bad merges. Overall of 702 merges made by this user on June 20, I reverted 325, for a roughly 46% error rate. ArthurPSmith (talk) 03:23, 24 June 2024 (UTC)[reply]

Invisible(?) deletion request

[edit]

Hello, I nominated Translations:Wikidata:WikiProject Roads/United States/New York/Page display title/xh for deletion using {{Delete}}; however, no action has been taken after after almost 2 weeks. The page is not listed in Category:Wikidata:Deletion, but Special:WhatLinksHere/Template:Delete lists it. Is my usage of the template incorrect? Could you delete the page I have nominated for deletion, please? Janhrach (talk) 13:20, 21 June 2024 (UTC)[reply]

✓ Done It might be that the message was only visible in "xh" language? Anyway, deleted. --Lymantria (talk) 15:31, 21 June 2024 (UTC)[reply]
@Janhrach, @Lymantria, @Bovlb: good catch! I suggest to substitute {{Delete}} to nowiki{ {Delete} }nowiki, to keep https://www.wikidata.org/w/index.php?title=Special:WhatLinksHere/Template:Delete&limit=500 clean. Objections? Estopedist1 (talk) 06:33, 24 June 2024 (UTC)[reply]
If you mean the mention of the delete template in Janhrach's request: no objections. --Lymantria (talk) 07:38, 24 June 2024 (UTC)[reply]
@Lymantria: actually I mean e.g. here as well: Talk:Q475182 Estopedist1 (talk) 06:07, 26 June 2024 (UTC)[reply]
@Estopedist1: NP. --Lymantria (talk) 13:46, 26 June 2024 (UTC)[reply]
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Lymantria (talk) 13:46, 26 June 2024 (UTC)[reply]

Protection request for User talk:NKVD UPROOTED

[edit]

Please semi-protect User talk:NKVD UPROOTED. Reason: Excessive vandalism XReport --Leonidlednev (talk) 00:00, 23 June 2024 (UTC)[reply]

✓ Done Full-protected. --Matěj Suchánek (talk) 06:53, 23 June 2024 (UTC)[reply]
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. KonstantinaG07 (talk) 17:10, 26 June 2024 (UTC)[reply]

Report concerning User:000yosaf

[edit]

000yosaf (talk • contribs • deleted contribs • logs • filter log • block user • block log • SUL (for IP: GUC)) — Reasons: vandalism-only account ? ―Mykhal (talk) 16:51, 23 June 2024 (UTC)[reply]

Warned for now, we'll see. --KonstantinaG07 (talk) 17:50, 23 June 2024 (UTC)[reply]

Коьрта агӀо

[edit]

Please redirect to "Wikidata:Main Page". Takhirgeran Umar (talk) 16:55, 23 June 2024 (UTC)[reply]

A redirect (which you've created) already exists, see Wikidata:Коьрта агӀо. We can't redirect from the item-namespace to the main page, e.g. we don't have Hauptseite (German) or Заглавная страница (Russian). --Wüstenspringmaus talk 04:55, 24 June 2024 (UTC)[reply]
@Wüstenspringmaus, When I click on the logo it takes me here, can you fix this? Takhirgeran Umar (talk) 07:37, 24 June 2024 (UTC)[reply]
✓ Done @Takhirgeran Umar: I think I have managed to change the default main page for Chechen. Does this work satisfactory? Infrastruktur (talk) 12:16, 24 June 2024 (UTC)[reply]
@Infrastruktur, Yes, now I get to the Chechen main page. Thank you! Takhirgeran Umar (talk) 16:22, 24 June 2024 (UTC)[reply]
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Estopedist1 (talk) 06:08, 26 June 2024 (UTC)[reply]

Deletion request

[edit]

Q49077449 has the same meaning (in Vietnamese) as Q467056, so I request to delete the first one. Amelia (talk) 03:46, 25 June 2024 (UTC)[reply]

→ ← Merged We merge items, not delete them. Already merged by user:Wüstenspringmaus Estopedist1 (talk) 06:06, 25 June 2024 (UTC)[reply]

Report concerning User:Mas-que-na'

[edit]

Mas-que-na' (talk • contribs • deleted contribs • logs • filter log • block user • block log • SUL (for IP: GUC)) I strongly believe this is a sockpuppet of OppsMetal (indef blocked & globally locked): both speak Spanish, both use #quickstatements, both like to play with changing Wikidata properties a lot, often by adding redundant things. For example, by location (Q56042075) should be used only for metacategories, but this user inserted it in thousands of irrelevant items. I remember well that OppsMetal did the same. I also hope some admin can revert all that. I'm quite sure that yet other sockpuppet did the same things a year ago (can not remember user name), and also got indef blocked, but was not recognized as OppsMetal's sock. --Orijentolog (talk) 09:33, 25 June 2024 (UTC)[reply]

Unfortunately, we cannot make a CU request as this data is only available for 90 days. However, I can also see many similarities between the two accounts. Pinging the blocking administrator of OppsMetal @MisterSynergy: and the locking steward @AmandaNP: What do you think? --Wüstenspringmaus talk 14:21, 25 June 2024 (UTC)[reply]
It is kinda far away in the past for me, thus I cannot tell much for the account in question.
I do remember though that I considered User:OppsMetal to be a follow-up account of User:Hey80'Q, which in turn followed User:&beer&love (all indefinitely blocked by me). They all had plenty of batch-editing experience particularly with the QuickStatements tool; regularly, their jobs were problematic at least, and the willingness to communicate on their user talk pages or to clean up their own mess was pretty much absent. So, if you need more data to compare the user in question to, these connections might be useful. —MisterSynergy (talk) 17:09, 25 June 2024 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Administrators%27_noticeboard&oldid=2189558680"