This page is for requesting modifications to URLs, such as markingdead or changing to a new domain. Some bots are designed to fixlink rot; they can be notified here. These bots includeInternetArchiveBot andWaybackMedic. This page can be monitored by bot operators from other language wikis since URL changes are universally applicable.
SlideWiki's website has been usurped by a gambling/ scam site. Likely due to a domain expiration being taken advantage of, but the legitimacy of that speculation is currently unknown.DrCzyżew (talk)23:48, 6 July 2025 (UTC)[reply]
Site has bot blocking technology that I successfully breached, with future application for other sites (building up a library of methods to circumvent bot blockers). Also testing a new AI-driven method for soft-404 detection to reduce manual checking. And a new AI method for findingruled mapped redirects. --GreenC19:21, 14 July 2025 (UTC)[reply]
Enwiki
Pass 1 (00001-05000): Checked 5,000 pages andedited 2,303 pages. Moved 2,332 links to a new URL: 647normal redirects, 1,612ruled mapped redirects, 73ghost mapped redirects, Resolved 300soft-404s. Removed 1{{dead link}}. Added 126{{dead link}}. Switched 26|url-status=dead to live. Switched 86|url-status=live to dead. Added 605 archive URLs (547 Wayback).
Pass 2 (05001-10000): Checked 5,000 pages andedited 2,184 pages. Moved 2,232 links to a new URL: 422normal redirects, 1,742ruled mapped redirects, 68ghost mapped redirects, Resolved 297soft-404s. Removed 1{{dead link}}. Added 85{{dead link}}. Switched 21|url-status=dead to live. Switched 71|url-status=live to dead. Added 514 archive URLs (476 Wayback).
Pass 3 (10001-20000): Checked 10,000 pages andedited 4,510 pages. Moved 4,469 links to a new URL: 802normal redirects, 3,550ruled mapped redirects, 117ghost mapped redirects, Resolved 726soft-404s. Removed 4{{dead link}}. Added 131{{dead link}}. Switched 50|url-status=dead to live. Switched 156|url-status=live to dead. Added 1,175 archive URLs (1,065 Wayback).
Pass 4 (20001-33500): Checked 13,500 pages andedited 6,104 pages. Moved 6,438 links to a new URL: 1,343normal redirects, 4,952ruled mapped redirects, 143ghost mapped redirects, Resolved 912soft-404s. Removed 1{{dead link}}. Added 164{{dead link}}. Switched 80|url-status=dead to live. Switched 211|url-status=live to dead. Added 1,505 archive URLs (1,374 Wayback).
Pass 5 (33500-47048): Checked 13,565 pages andedited 5,971 pages. Moved 5,862 links to a new URL: 1,082normal redirects, 4,635ruled mapped redirects, 145ghost mapped redirects, Resolved 937soft-404s. Removed 4{{dead link}}. Added 175{{dead link}}. Switched 56|url-status=dead to live. Switched 240|url-status=live to dead. Added 1,573 archive URLs (1,463 Wayback).
IABot DB
Checked over 500,000 unique URLs and updated about 32,000
It appears ctvnews.ca changed the format of their urls at some point. They were of the formatarea.ctvnews.ca and are now ctvnews.ca/area/. A couple of examples, inAugust 8 the url "https://toronto.ctvnews.ca/former-premier-of-ontario-william-davis-dead-at-92-1.5539011" will auto redirect you to "https://www.ctvnews.ca/toronto/article/former-premier-of-ontario-william-davis-dead-at-92/". However inMay 1 the url "https://atlantic.ctvnews.ca/it-s-got-to-make-some-kind-of-change-boycott-of-loblaws-owned-stores-begins-1.6869575" results in a 404, even though the article still exists at "https://www.ctvnews.ca/atlantic/article/its-got-to-make-some-kind-of-change-boycott-of-loblaws-owned-stores-begins/". The redirect fails because it tries to redirect to "/it-s-" while the article is at "/its-". I'm not sure how many areas there are but they include; Toronto, Atlantic, Montreal, Edmonton, Regina, and bc. There could be others. --LCUActivelyDisinterested«@» °∆t°11:56, 17 July 2025 (UTC)[reply]
Domain required a lot of special rules, it is now able to convert about 70% to live links. The other 30% are dead links mostly, but also some edge cases the rules can't catch, maybe 10% of that 30%. --GreenC21:05, 26 July 2025 (UTC)[reply]
WikiWiXis was an archiving platform which has gone defunct. The domain registration expired at the end of April this year. It was used on4,291 articles on enwiki (not as bad as the39,851 on frwiki). Can anything be done to switch to an alternative archive? In at least one case I've seen WikiWiX used to archive a web.archive.org link.Cabayi (talk)14:46, 21 July 2025 (UTC)[reply]
I'm glad you brought this up. On French wiki it is actually ingrained through a custom patch to MediaWiki that adds an "[archive]" link automatically next to every external URL (egfr:Aubelin_Jolicoeur in the Sources section). It started as a small non-profit that had external funding. Recently they lost their sponsor and the owner took it over as a personal project. He then got hit badly by DoS (or AI scrapers) and have had trouble keeping it up. The main tech guy left, making it a one-man operation. The site reliability is terrible going up and down. Attempts by the French community to escape this trap have been unsuccessful because the owner has a bunch of supporters typical of small wiki politics. Nobody seems willing or able to get rid of WikiWix, so long as the owner keeps making promises and telling positive stories.
On English Wikipedia, I made previous attempts to convert WikiWix links to Wayback Machine. There are probably new ones added since then. The links also exist in the IABot database - I also unwound most of those where possible but they still exist propagating through 300+ wikis via IABot. I recall it was a difficult operation, but I also wrote a lot of code for it, so maybe that code can still be applied.
It's unclear the site is actually defunct. Until the owner literally says so, the site has a tendency to keep popping back up in hobbled form. I would like nothing better than definitive proof of defunct. --GreenC16:14, 21 July 2025 (UTC)[reply]
FWIW the registrar still propagates, the domain is in the root servers.dig +trace wikiwix.com. Not sure why who.is information says expired. TheICANN tool can't connect to the registrar, and the registrar's whois.ovh.com doesn't work. Odd setup. --GreenC18:00, 28 July 2025 (UTC)[reply]
Hi, Many football-related Wikipedia pages currently link to archived versions of my website (https://www.statto.com/), which was offline for a while. It’s now fully restored, and the original content is live again. Since I can’t use a bot or edit all the pages manually, I was suggested that you can help update those Webarchve links back to the original URLs.Please let me know what’s possible or how best to proceed.Ggvanncaa (talk)07:59, 30 July 2025 (UTC)[reply]
The article archives for theAnandTech website have been removed, and links to AnandTech or its article archives are now being redirected to the AnandTech forums front page. Links to AnandTech articles on Wikipedia need to be modified to go to an archive site.Jesse Viviano (talk)06:05, 5 August 2025 (UTC)[reply]
This now redirects to a website for a New Jersey state representative, for doubtlessly inscrutable reasons, so link templates should be updated accordingly. --Slowking Man (talk)16:03, 5 August 2025 (UTC)[reply]
Not done only because there was nothing to do. It was already done. Of the 1,461 citations, 1,459 have the same URL. --GreenC23:53, 6 August 2025 (UTC)[reply]
Musicline.de used to host the Germany music charts. Although these chart positions can be found at offiziellecharts.de, they can't be converted over because they have an numerical id in the URL. Some pages already have archives. Not sure ifThe Fixer (song) can be fixed because it's in a wikitable.~1400. Thanks!MrLinkinPark333 (talk)02:32, 8 August 2025 (UTC)[reply]
Batch 1: Checked 4,000 pages andedited 3,428 pages. Moved 9,023 links to a new URL: 7,463normal redirects, 1,523ruled mapped redirects, 37ghost mapped redirects, Resolved 2,187soft-404s. Removed 25{{dead link}}. Added 209{{dead link}}. Switched 456|url-status=dead to live. Switched 223|url-status=live to dead. Added 2,000 archive URLs (1,912 Wayback).
Batch 2: Checked 5,748 pages andedited 4,997 pages. Moved 13,138 links to a new URL: 10,913normal redirects, 2,137ruled mapped redirects, 88ghost mapped redirects, Resolved 3,326soft-404s. Removed 6{{dead link}}. Added 235{{dead link}}. Switched 474|url-status=dead to live. Switched 421|url-status=live to dead. Added 2,530 archive URLs (2,476 Wayback).
IABot DB
Checked 15,000 URLs and updated about 2,867
Notes
The soft-404s in this case were difficult because while the page content was displaying a "home" page, the URL itself didn't redirect, so it required foreknowledge of HTML keywords to know when the page landed on a soft-404. There were about 3 dozen different "home" page variations to be discovered. For this reason I had to redo Batch 1 a couple times before it was ready for upload.
The article archives for the 1UP.com website as seen in1Up Network have been removed, and links to 1UP.com or its article archives are now being redirected to the IGN front page. Links to 1UP.com articles on Wikipedia need to be modified to go to an archive site.Jesse Viviano (talk)09:29, 10 August 2025 (UTC)[reply]
Urls ending with /number/ can be fixed by removing the number and ending slash of the URL.This is nowhere forAnne Hathaway. I do not know how to extract these URLS as the number id's are at the end of the URL. Ive only found that oneandthis forBruce Willis so far from the 10k overalllinks. May only be a handful to convert. In any case, thank you!MrLinkinPark333 (talk)03:25, 13 August 2025 (UTC)[reply]
It's OK, I'm here, might as well check them. It didn't take long because most return 200 or simply convert to https. As for filtering, I don't know how with insource, need to load all links anyway. Also, it's usually best to check the entire domain because it creates more data for soft-404 algorithms to learn from, and there are usually unknown-unknowns to be discovered. In this case I discoveredthis is nowhere, added the rule and repaired a couple hundred. I won't do IABot because there isn't much that can be done, IABot doesn't support URL moves which is mostly what this is (other than 52 archives in 10,000+ URLs). --GreenC02:51, 26 August 2025 (UTC)[reply]
This website used to host South African music charts. Now, it's been taken over by a gambling site. Some of the articles already have archived links.I would like to request this site to be added to JUDI. Of the48 pages, there's a few lists with many URLs to this site. Thanks!MrLinkinPark333 (talk)20:51, 16 August 2025 (UTC)[reply]
Following a URL move from nzhistory.net.nz → nzhistory.govt.nz, the former has now been taken over by a gambling website.Could a bot change ".net" to ".govt" in all the outdated refs? Currently looks like~1,160 pages. Thanks,Nil🥝03:49, 18 August 2025 (UTC)[reply]
Chinapost.com.tw was the URL before the publication got bought by Now News and moved to Chinapost.nownews.com. The newspaper no longer exists, making both the old and new urls broken.
There are two urls because this is the second time the project site has moved:1. cis.k.hosei.ac.jp/~F-rep/ — dead link2. hyperfun.org — spammer (link to casino on homepage, no impersonation)The second link appears to have been usurped a year or two ago.
I've manually changed a few links (regular and archive) on the pagesHyperFun andFunction Representation to a site which matches the archive.
Idolator was part of SpinMedia before it got sold in 2016. The website waslast updated in 2022. If any of these links on Wikipedia are live, I request archive copies of them in case the website fully shuts down. Otherwise, archived copies of broken links are appreciated. ~3700 articles. Thanks!MrLinkinPark333 (talk)02:55, 21 August 2025 (UTC)[reply]
WaybackMachine has no archives available:example. The site requested not to be archived. That leaves archive.today, or any that are still live. The rest will be{{dead link}}. --GreenC00:55, 31 August 2025 (UTC)[reply]
I'm not sure why anything would specifically ask to not have its pages be archived, but for the record,this January 2025 piece is the most recent thing I could find from them after a few 2024 articles. It's unclear what happened with activity there when there's nothing to be found for 2023.SNUGGUMS (talk /edits)20:38, 31 August 2025 (UTC)[reply]
I checked each link if it is live, before adding an archive URL or{{dead link}}. Probably 5% to 10% of domains request to be removed. The problem is when a company goes defunct, they never lift the block, it has to stay until someone requests to remove it, which never happens, effectively banishing them forever. Except from sites like archive.today which generally don't do archive blocks, but for that reason makes them more vulnerable to legal action and potential takedown. --GreenC21:42, 31 August 2025 (UTC)[reply]
The idea of getting sued never crossed my mind. If archive.today and other things used to preserve Idolator's entries are all taken down for any reason, then that might render some or all pieces from the site unusable. I hope it doesn't come down to that.SNUGGUMS (talk /edits)21:57, 31 August 2025 (UTC)[reply]
Found in a GAN spotcheck and redirects tothis, which sure as hell looks like a usurpation from what I've seen. See also Wikipedia's redirectMediaBistro, which says the parent company liquidated in 2015.Departure– (talk)03:35, 21 August 2025 (UTC)[reply]
The thing is that links that once led to content that at one point verified information now lead to a site that doesn't. I apologize if this is the wrong venue to bring this, but I was hoping this was the place to report link changes of this type.Departure– (talk)22:03, 30 August 2025 (UTC)[reply]
You are in the right place. I see now what is happening. I'll process the domain. If any are legit live, it will keep them live otherwise the rest look like soft-404s ie. they look like working pages, but redirect to a home page. Those will be converted to archive URLs. --GreenC22:14, 30 August 2025 (UTC)[reply]
Departure–, the data was showing a connection toAdWeek. I asked Google "has mediabistro.com been bought by adweek.com" it replied:
Acquisition Date: Prometheus Global Media acquired Mediabistro in May 2014 for $8 million. Acquiring Company: The parent company of Adweek, The Hollywood Reporter, and Billboard, Prometheus Global Media, bought Mediabistro. Integration: The acquired Mediabistro sites, including blogs and job boards, were merged into the Adweek Blog Network, with their web addresses now starting with Adweek.com. Result: The Mediabistro brand was absorbed by its parent company, and its content became part of Adweek's offerings
Thank you @Departure– for bringing this up and @GreenC for tackling this. Mediabistro was indeed absorbed. It ran the blogs TVNewser (covering US national TV news) and TVSpy (covering US local TV news), so I'm not surprised that there was a link inGary England, a local TV meteorologist.Sammi Brie (she/her · t ·c)22:27, 31 August 2025 (UTC)[reply]
groklaw.net has been usurped, some content has been removed, other content has apparently had crypto spam added.https://en.wikipedia.org/wiki/Groklaw#Later_history "As of 18 August 2025 the site points to a crypto gambling site."
tiebreakertimes.com is no longer active and was moved to tiebreakertimes.com.ph. It seems that all of the news articles fall under the new URL.MarcusAbacus (talk)12:00, 30 August 2025 (UTC)[reply]
Many of these redirect to new URLs at bbc.com. However, some will stay at the current URL likethis one. I cannot predict which ones will redirect. As this is a huge site, I have to break it up into multiple requests.
BBC has a very large number of pages on enwiki, probably one of the largest domains across all Wikipedia language sites. Will finish smaller requests first,WP:JUDI is backlogged, then return to look at BBC. It's a good idea to break it up. --GreenC21:57, 30 August 2025 (UTC)[reply]
Preliminary data suggests this is a well maintained and stable site with few surprises. The number and type of changes are relatively minor, like http/https, or adding a "www", that sort of thing, based on existing redirects. The most exotic change is removing ".amp" from URLs added by mobile users. In 500 pages there were 0 dead links. It should go fast. This is nice, I run it and go do something else while it cranks through 10s of thousands of pages making small changes. Hope they don't block the bot is the concern. --GreenC17:12, 3 September 2025 (UTC)[reply]
Anyone able to assist here? By not having "https" the links to the old site do not redirect to the new site, so user will see a "Page not found" message. Or, as an editor for over 10 years, can I make the changes myself? (I just need a few instructions).DivermanAU (talk)19:16, 8 September 2025 (UTC)[reply]
You asked "If any of these don't redirect, please let me know": 108 archives, 54 live to dead, and 16 dead link templates. --GreenC04:32, 10 September 2025 (UTC)[reply]
ClumsyOwlet: I would normally say this is impossible. There is no way to mapthis tothat. However there is logic: given the last field "probation-act-for-ducie-after-donation-made" is common to both URLs, make a Google search of the site independent.ie for this common string. It correctly returns a match for the new URL. However, I can't automate Google searches without being blocked. But I can run Google Gemini (AI) queries, and ask it to run Google searches. This loophole works. It's not free, but Google seems OK with it so long a there is payment involved. Whose paying? My boss,The Internet Archive. I did some cost analysis, if I run the query 2,000 times it will cost $3.67 US total. I think we can afford it to repair all these URLs is cheap. This would be a new AI approach never done before. --GreenC05:08, 10 September 2025 (UTC)[reply]
AI is not working it is hallucinating too much. I came up with a different solution using"Ruled mapped inferred redirects" (last section) - basically it searches the WaybackMachine index for the common string. It misses some because the URLs are not in the WaybackMachine. I am out of tricks to find those, they will be converted to archive URLs. --GreenC20:50, 10 September 2025 (UTC)[reply]
Enwiki
Batch 1 (0001-0200): Checked 200 pages andedited 188 pages. Moved 161 links to a new URL: 161ruled mapped inferred redirects, Switched 31|url-status=dead to live. Switched 7|url-status=live to dead. Added 104 archive URLs (73 Wayback).
Batch 2 (0201-1508): Checked 1,308 pages andedited 1,227 pages. Moved 988 links to a new URL: 988ruled mapped inferred redirects, Resolved 1soft-404s. Added 2{{dead link}}. Switched 192|url-status=dead to live. Switched 75|url-status=live to dead. Added 701 archive URLs (518 Wayback).
Thank you for finding so many URL replacements for bbc.co.uk. There are 11k left, but not all of them will need fixing:
URLs that end in .shtml tend to be working, with no changes needed. These pages will primary say that BBCarchived the page. However, I found a broken link atChordate.~5k
URLs that are not sport, news, or shtml tend to be working or redirect like thisone.~6k
The main things I see are either changing HTTP to HTTPS or archive fixes. As some of these links already have archived links in the article, this should hopefully be resolved quickly. Thank you again!MrLinkinPark333 (talk)22:22, 5 September 2025 (UTC)[reply]
Using a different method of searching (SQL query), for #1 it returns over 29,000 pages, and for #2 is 134,000 pages. I think CirrusSearch can't accurately search in this case because if there is /news anywhere in the page it will not be reported. For example a page has two URLs - one with /news and the other not - it will skip the entire page since it contains /news. SQL shows every URL, you can filter and see which pages contain a URL pattern. --GreenC15:26, 12 September 2025 (UTC)[reply]
Better, but same issue:-insource:"bbc.co.uk/news/" means if this string appears anywhere on the page, don't list the page, even though there might be other URLs on the page that should be included. According to SQL, the number of pages containing a BBC url is about 150,000. There might be some /news or /sport in that 150,000 but those pages also contain other BBC links. It excludes pages that only contain /news or /sport. Since there is no real difference in how the URLs are processed, I suggest we consider the 150k as the primary set, then break down into smaller batches. It could be a lot of batches. If it runs as well as last time, very large batches are possible then it won't be many. I can start slow with small batches to see what problems come up. --GreenC16:41, 12 September 2025 (UTC)[reply]
Hmm. I'm not sure which batch to focus on. A lot of them look to be working with no issues. I've found some with various issues:
Maybe you misunderstood what I wrote. There is no sense separating based on URL path, because every BBC urls needs to be processed. /teach, /sound/ etc.. all of them need to be checked: *.bbc.co.uk/* --GreenC20:11, 12 September 2025 (UTC)[reply]
It's alright so far over 80% of the pages have a change. The results are similar in quality but more in quantity than /news and /sport .. The work is on the computer. It's actually more work to do separate projects because it requires creating a new project, updating configurations, downloading a list of target articles. By keeping it under the same project I only need to start a new batch ("Batch 1", "Batch 2" etc) which is fairly easy. If the projects require different configurations they need separate, but this project it's looking all the same. --GreenC00:02, 13 September 2025 (UTC)[reply]
If you could extract a list of sections to go through, that'd be great. I'd only need the section names, not the URLs. I don't think all of them will need checking. I can then check, and post batches in later requests. I'll just leave the 4 above here, so you can work on other requests. --MrLinkinPark333 (talk)18:32, 12 September 2025 (UTC)[reply]
@GreenC: This seems to be changing lots of news.bbc.co.uk links to https when the https actually redirects back to http. Examples:[10][11][12][13] In addition it's changing between news.bbc.co.uk/1/* and news.bbc.co.uk/2/* which seem to randomly redirect to each other. According tometa tags the former is theUKFS_URL and the latter is theIFS_URL whichapparently stand for "UK facing site" and "international facing site", but something is seemingly misconfigured as they now redirect randomly from the same IP.EvenTwist41 (talk)02:17, 17 September 2025 (UTC)[reply]
There are two issues isolated to news.bbc.co.uk : A) https redirects to http B) /1/ redirects to /2/ and other way randomly .. there are also two piles of links: X) URLs already modified listed below. Y) URLs yet to be modified.
I think for A+X, it is best to leave them alone, it causes no harm, and maybe one day they will properly support https anyway. For A-Y, there is no compelling reason to switch to https. For B-X, this is harmless best left alone. For B-Y, same, best left alone not make any more changes.
End result: do nothing, except add code to skip processing news.bbc.co.uk going forward, at least when they only change is of type A or B. --GreenC20:19, 19 September 2025 (UTC)[reply]
Just saw this on a page I frequently edit. What did your bot do? It changed it from http to https. What was that for? Are you saying that whoever copied the url is wrong? There was nothing wrong with it in the first place!RandomEditorofWiki (talk)14:01, 27 September 2025 (UTC)[reply]
An HTTP era domain found inRudolph G. Wilson that is now usurped; I assume it most likelywas a high-school newspaper forGranite City, Illinois, but now it seems the type of site that'd go on the spam blacklist, with the Chinese text and the markedly not-high-school-friendly content of the site.Departure– (talk)20:28, 6 September 2025 (UTC)[reply]
For what it's worth, I don't know if it's cited in any other articles and I'm going to bring the article I found it on to AFD momentarily, but it doesn't hurt to check.Departure– (talk)20:34, 6 September 2025 (UTC)[reply]
MrLinkinPark333: Apparently I already permadead'd many of the .biz links in May 2021 (example). This time through I found another ~200 archive.today links missed. Possibly I wasn't checking for archive.today in 2021. It's interesting that so many more links in 2025 needed updating:1,142{{dead link}}. Switched 1,031|url-status=live to dead. Added 1,056 archive URLs (854 Wayback). Maybe these links have since died but were active in 2021, maybe my methods in 2021 were inaccurate, or maybe IABot was unable to parse/fix them on-wiki. Anyway, things continue to move in the right direction. --GreenC02:19, 21 September 2025 (UTC)[reply]
OnIranian handicrafts, http://www.yjc.news/fa/news/6228037 is now at https://www.yjc.ir/fa/news/6228037/توتن-قایق-یکی-از-صنایع-دستی-سیستان-دریاچه-هامون-چشم-انتظار-حیات-دوباره-آن-است (I just changed .news to .ir in the original link and it turned into the correct link).
OnThe Accused Escaped, https://www.yjc.news/fa/news/6397790/%D8%B1%D8%A7%D8%B2-%D8%A7%D8%B5%D8%BA%D8%B1-%D9%81%D8%B1%D9%87%D8%A7%D8%AF%DB%8C-%D9%BE%D8%B3-%D8%A7%D8%B2-%D8%B3%D8%A7%D9%84-%D9%87%D8%A7-%D9%81%D8%A7%D8%B4-%D8%B4%D8%AF-%D9%81%DB%8C%D9%84%D9%85 is now at https://www.yjc.ir/fa/news/6397790/راز-اصغر-فرهادی-پس-از-سال%E2%80%8Cها-فاش-شد-فیلم (Just changed .news to .ir. Also works if you take the "https://www.yjc.news/fa/news/6397790" part of the original link and change .news to .ir.).
OnHeshmatollah Falahatpishe, https://www.yjc.news/en/news/38193 is now at https://www.yjc.ir/en/news/38193/iran-to-claim-compensation-from-us-for-chemical-weapons-victims-mp (Same thing for English).
On2022 Hormozgan earthquakes, https://www.yjc.news/fa/amp/news/8162021 is now at https://www.yjc.ir/fa/news/8162021/زلزله-های-پی-در-پی-در-غرب-هرمزگان-زلزله-۵۲-ریشتری-چارک-را-لرزاند-فیلم-و-تصاویر (Removed /amp and changed .news to .ir).
OnList of Esteghlal F.C. managers, https://www.yjc.news/00U4Iq is now at https://www.yjc.ir/fa/news/7166384/وریا-غفوری-سرمربی-موقت-استقلال (URL Shortening. .news to .ir works.)
This will be multi-step. Because the domain name has changed fromhttp://www.yjc.news/fa/news/6228037 -->https://www.yjc.ir/fa/news/6228037 I can do a domain move on existing URLs. It will be configured so any that can't be moved will be set as|url-status=dead and archive URL added (or dead link tag). After that is complete, I will add the old domain toWP:JUDI, so those remaining links in the old domain get the usurpation treatment, as part of a future JUDI batch run. That should cover both moving the domain where possible, and the usurpation where a move was not possible. Unfortunately I can't easily do both at the same time as moving and usurpation are different types of processes. --GreenC02:31, 21 September 2025 (UTC)[reply]
Enwiki
Checked 136 pages andedited 135 pages. Moved 153 links to a new URL: 153ruled mapped redirects, Resolved 71soft-404s. Removed 1{{dead link}}. Switched 8|url-status=dead to live. Added 4 archive URLs (4 Wayback).
IABot DB
Set domainpermadead (IABot does not support URL moves)
"Due to a high number of AI bots scrawling our website we've had to take the decision to ask visitors to please log-in or register before browsing this website. We apologize for the inconvenience and hope we will find a better solution in the future."
Since I don't have a login, it requires a "blind move" ie. switch the URL without verifying. Blind moves are risky, there are usually some links that don't work, but since there are only about 120 pages containing coa links, it is a better option than nothing. If it breaks things let me know I can try to repair. --GreenC00:48, 22 September 2025 (UTC)[reply]
The old TED video URL format was "http://www.ted.com/talks/talk_name_here.html", which now return 404. The current TED video URL format is: "https://www.ted.com/talks/talk_name_here" with the trailing ".html" removed (and HTTPS). A quick search suggests there could be about 1,000 affected links.UnlikelyEvent (talk)07:02, 15 September 2025 (UTC)[reply]
I can't seem to find a search to give you a rough idea of how many there are, including various fixes such as archive refs, but I think there'll be a few hundreds.
Is there a way of excluding a domain name? There are hundreds (probably thousands) of citations using villagevoice.com sourcing Scientology topics. Although VV copied the text of their articles from a prior website style, the new copies are shit (images and formatting gone). The latest run was on articleList of Scientologists which screwed up. By marking newer VV URLs as live, we are losing formatting and images of the Wayback Machine archived articles... well, not exactly losing, but deferring to the newer shit copy. Can you just make villagevoice.com an exclusion of GreenC Bot? There's no documentation atUser:GreenC bot to explain what this bot does or how to exclude it, etc., and there are about 800 articles in the Wikiproject:Scientology; likely every one of them linking to a VV article. Fixing this by hand is not an option. ▶ I am Grorp ◀08:13, 10 October 2025 (UTC)[reply]
User:Grorp this was a job request made on a noticeboard initiated another editor to process the entire domain. The link to this location was in the edit summarySpecial:Diff/1314954569/1316049965 ..User:MrLinkinPark333 I have put the other half of the domain on hold until it is clear how to proceed because right now there is no longer consensus. I have not looked too closely at the issue raised by Grop yet. Also Grop, you are looking for documentation about the bot, that is also in the edit summary. --GreenC16:52, 10 October 2025 (UTC)[reply]
@GreenC: That link (i.e., this thread) doesn't have documentation about whatGreenC bot is doing... it's just some request about villagevoice.com. I would have expected documentation atUser:GreenC bot that explains what the bot is intended to do -- keeping in mind that I (and most editors) were not part of any request to invent the bot. Are you saying GreenC botonly does villagevoice.com? Even if so, I would expect some sort of documentation (or pointer to it) on its page.I'm not sure what other article types cite to villagevoice.com articles. Keeping in mind that the previous editor of the paper,Tony Ortega, has been covering the Scientology topic for a very long time. As editor and reporter, there are many articles on the topic, and many are cited withinWP:WikiProject Scientology-tagged articles, which were duly copied into the Wayback Machine.The Wayback Machine copies are much more desirable than the 'new' VV URLs, which I swear were created by an automated process when they updated their website to a new coding format. ▶ I am Grorp ◀20:22, 10 October 2025 (UTC)[reply]
The bot is actually WaybackMedic, indicated in the edit summary. The account that runs the bot is GreenC bot, which also runs many others bots with various names. That's how bots work on Wikipedia: 1 user account, many bots attached to it. Otherwise I'd need 10 different accounts for each bot program which is not practical. What WaybackMedic does is address link rot issues - it's been in development for over 10 years and has thousands of features due to the complexity of link rot. When you ask "what is it doing", for this page it is for user's to request modifications to URLs so that dead links become live again, or add an archive URLSpecial:Diff/1315452952/1316054727.
Back to VV: Comparearchive vs.live. The formatting doesn't look too bad, but I agree the image link is broken. Sometimes Wayback Machine copies are superior to live pages, this might one such case. --GreenC00:16, 11 October 2025 (UTC)[reply]
GreenC: Village Voice redesigned/reformatted/reprogrammed their website in 2017. Anything prior to then had its text copied from the original article -- minus images, formatting, bolding, etc. Of course a Wayback archive of the new design looks just like the new (current) design, like in your example above which is a 2017 article. But take older articles and you'll see the difference.Here is one example of a 2008 article (used in 11 Wikipedia articles):current article,archived original. Hardly the same. If the Wayback Machine's archived copy is better than VV's archive copy, I don't see why we need to primarily point to VV's archived copy, especially not to mark the newer url as live even if there is already a validolder url-archive listed.Now here's the rub... inthe edit which alerted me to GreenC bot's failures, the very first change it made is blatantly incorrect. The bot took thisJune 20, 2011 article and changed it to thisJune 28, 2011 article. They are not even the same article. I don't know what the bot does to attempt to find the same article, but this one was indeed a failure, and an editor unfamiliar with looking up old edits and correcting the bot's error might simply check the June 28 article and realize it doesn'tverify the text which precedes it, and might delete the "unverifiable" content from the article.Second example: The second change in that edit illustrates the loss of images:Wayback archive versuscurrent VV version; especially noticeable is the loss of the last image which is referred to in the text of the article.I don't know how bots get programmed, but maybe for villagevoice.com it can (a) skip over articles that are pre-2017, and (b) skip over any citations that already contains an archive-url parameter. A better solution for villagevoice.com/blog urls is to just find the Wayback Machine archived copy. I just don't understand what problem GreenC bot is trying to solve that isn't better solved byUser:InternetArchiveBot. ▶ I am Grorp ◀10:50, 11 October 2025 (UTC)[reply]
Seems like it was an error on the site. Usingthis redirect instead points to the right article. For some reason, had to remove the _t at the end, even though it was already there in the url per your revision. If IA copies are needed instead, then that works for me.MrLinkinPark333 (talk)15:57, 11 October 2025 (UTC)[reply]
This has become more complicated, the details were not understood before and kindly brought to attention by Grorp. The first thing is to stop more changes, the bot was halted two days ago. The second is to undo the controversial changes made in Batch 1, namely "Moved 2,043 links to a new URL" and "Switched 400 |url-status=dead to live". There are logs so it should be possible with programming work. It might be possible to convert to archive URLs in the same pass, or possibly two passes (first revert, second archive). --GreenC21:45, 11 October 2025 (UTC)[reply]
@PARAKANYAA: After a discussion (see above) GreenC decided to reverse their earlier edits on villagevoice.com citations. You'll notice that these new edits are the reverse of ones the GreenCBot did earlier. ▶ I am Grorp ◀12:52, 22 October 2025 (UTC)[reply]
Yeah unfortunately this is a complex site with a lot of dead links that were not being repaired by IABot. I was left with a bad choice, or a very bad choice. So I went in the direction of the bad choice: adding archives in some cases where the link is still live. The alternative, the very bad choice, was tonot add archives when links are dead. This later category is much larger than the former, so I went with the former because it did the least amount of harm for the most good, relative to the other option. Blame the website owners for having such a messed up website. --GreenC15:18, 22 October 2025 (UTC)[reply]
Perhaps a slightly odd request, but I'd figured I'd bring this one to attention of users on this board. The template was deprecated and now isWP:TFDHd, but needs its uses changed to{{Cite POWO}}. There is not an automatic translation of URLs in the one that can be used with the other apparently. This board has the right background to do something about that perhaps.Izno (talk)23:39, 6 October 2025 (UTC)[reply]
Izno: There is no map between old and new URLs they are using different schemes. The best I can do is convert the ~1,000 instances of{{WCSP}} to{{cite web}} and add archive URLs (or{{dead link}}). --GreenC04:48, 18 October 2025 (UTC)[reply]
Izno, I think the template is now gone from main, file and template space. There are two other template pages transcluding it somehow, but I can't figure out where. --GreenC17:33, 19 October 2025 (UTC)[reply]
Thedomainhostname in theFQDN needs to be removed to create redirects to the new URL. Example: Changingthis tothat redirectshere. These are in two categories:
MrLinkinPark333: Almost 5,000 links from dead to live (3,956 + 788). The 1,000 dead links are unfortunate, nothing to be done, but only about 1 in 6 of the total. This process moving dead links live is a particularly good idea when WaybackMachine excludes the domain. Good catch and an important domain. --GreenC01:18, 20 October 2025 (UTC)[reply]
The entire website blockmrecords.org redirects. It is in31 pages. I tried to find a way to move the existing URLs to the new domain without success. I'll add archive URLs. --GreenC01:38, 20 October 2025 (UTC)[reply]
These Mexican government agencies will likely be dissolved this month, and I'm not sure what will happen to references and other materials used within.Sammi Brie (she/her · t ·c)19:00, 10 October 2025 (UTC)[reply]
It took place on October 17 for the former and presumably a similar time for the latter. Sites are still up for now (but with the replacing agency's logo). Unclear whether they will use it or another page for their own business. I suspect the domain rpc.ift.org.mx (full of PDFs containing broadcasting technical information) will be retained intact at some other domain at some point.Sammi Brie (she/her · t ·c)06:32, 21 October 2025 (UTC)[reply]
If it's a new agency at the same domain.. what does this mean we should do in terms of archiving URLs? Options are do nothing. Or treat all URLs as dead and add archive URLs. --GreenC16:06, 21 October 2025 (UTC)[reply]
I understand the need to update the URLs to reflect the updated location, but am not sure of the changes made inthis edit where two URLs (one valid/live; the other incorrect/dead) were changed to SKIPDEADURL. —Archer1234 (t·c)12:43, 22 October 2025 (UTC)[reply]
User:The C of E, there is no obvious way to automate replacement, for examplethis does not gohere. I will add archive URLs via bot. Even better, there are so few links, recommend if you can manually search for the new URLs and replace them. --GreenC18:54, 22 October 2025 (UTC)[reply]
That works. They have bot blocking that I can not surmount ("Are you human?" from CloudFlare), I am unable to verify the new link works, so I made a "blind move" ie. simple search-replace without verification. I did manually verify 8 pages so I assume the other 4 are OK. The problem is some "/local-news/" redirect to "/lifestyle/" which is fine, except it might be a problem later with the Wayback Machine archives, which might not be able to follow the redirects. Ideally if you can go through and fix those redirects it would be best long term, I can't do it automatically because of the bot blocking. --GreenC16:18, 24 October 2025 (UTC)[reply]
Old links for The Post and Courier unfortunately are not redirecting to their new URLs. For example,this is nowhere. Unless ghost redirects are found, I think archives will be needed. This is because the new URLs are not easily converted. Some of these links are already archived.450. Thank you!MrLinkinPark333 (talk)23:19, 23 October 2025 (UTC)[reply]
I can't find examples where changing the domain name alone makes sense, except for the home page:https://nla.gov.au -->https://library.gov.au .. everything else is either the old Trove, Catalog and Pandora/Webarchive links .. or the links at the new domain have new paths, it's basically an entirely new website. Mapping the old links to new, without redirects, may or may not be possible, and it would take serious investigative work. Assuming the new website content is even the same as the old. It's like they abandoned the old site and started a new one. --GreenC17:07, 24 October 2025 (UTC)[reply]
I found a match inHelena Blavatsky:this [dead link] is now availablehere. InRobert Louis Stevenson,this matcheshere. The matches are imperfect (eg. "german-colonies-in-the-pacific" vs. "german-colonies-pacific") it would require AI or fuzzy matching. Unfortunately there are only37 pages. There is also "nla.gov.au/research-guides/" (19 pages) availablehere. I'm hesitant because of the low count and bespoke coding. Will think about how it might be done. --GreenC02:22, 25 October 2025 (UTC)[reply]
Done perWP:JUDI batch #28.GreenC21:53, 5 September 2025 (UTC)[reply]
Not done only because there was nothing to do. It was already done. Of the 1,461 citations, 1,459 have the same URL. --GreenC23:53, 6 August 2025 (UTC)[reply]
On hold per above --GreenC05:11, 17 September 2025 (UTC)[reply]