Wikipedia:Link rot/URL change requests/Archives/2023/August

Facebook

facebook.com/photo.php urls no longer work, you can check every link here and it will not work, please fix these links Notrealname1234 (talk) 21:28, 25 July 2023 (UTC)

While I can confirm some of the URLs are broken most seem to be working. In the first 10 results that appeared for me 4 are broken, with the remaining 6 loading as expected. There's no difference in behaviour when I'm logged in versus logged out of Facebook, except that two of the errors are replaced with login screens. Sideswipe9th (talk) 21:34, 25 July 2023 (UTC)

i am logged in to facebook, it just shows me a error "this content is not available" Notrealname1234 (talk) 21:43, 25 July 2023 (UTC)

Could you give some example URLs that are broken for you? From the first ten results, this one linked on ENGERERcycrus' user talk page is broken, and this one linked on File:The Decemberists - The King Is Dead.jpg is fine. Sideswipe9th (talk) 21:45, 25 July 2023 (UTC)

The links are in 2,403 pages. Some probably work and some don't, the page needs to be scraped for keywords because the header's return 200 (soft-404s). I might run into bot blockers when at volume, Facebook.-- GreenC 21:50, 25 July 2023 (UTC)

From the small sample I checked, now up to the first twenty results, around 70ish percent were still live. I suspect what's happened here is just that the original files have either been deleted or made private on Facebook by the original uploader. There doesn't seem to be a URL schema change, and some of these like the broken one I linked above were added ten or more years ago. Unless there's some A/B testing on Facebook's end, where Notrealname1234 is being directed to a different version of the site which might in the future become the live version, I don't think there's really anything to do here bot wise. Sideswipe9th (talk) 21:56, 25 July 2023 (UTC)

At the 13th link, for the brazilian version (still the same link) of facebook, it says this: "Este conteúdo não está disponível no momento" @GreenC @Sideswipe9th Notrealname1234 (talk) 00:39, 26 July 2023 (UTC)

and the 13th link is this https://www.facebook.com/photo.php?fbid=195469590484313&set=a.180687601962512.39754.180681685296437&type=1 Notrealname1234 (talk) 00:43, 26 July 2023 (UTC)

This link https://www.facebook.com/photo.php?fbid=461539374654&set=a.56890239654.63538.5229624654 you provided does work. Notrealname1234 (talk) 00:44, 26 July 2023 (UTC)

If you check the user talk page where that URL is linked, it seems it was added in May 2011. I suspect what has happened here is that the original uploader of the image has deleted it sometime in the last twelve years. Sideswipe9th (talk) 03:46, 26 July 2023 (UTC)

Still, most links do not work and needs to be fixed. Notrealname1234 (talk) 20:27, 27 July 2023 (UTC)

I think if a page scrape finds "The link may be broken, or the page may have been removed. Check to see if the link you're trying to open is correct" it should be converted to an archive URL. Example. There might be other variations of a dead-link landing page. Any help identifying other versions appreciated. -- GreenC 01:57, 29 July 2023 (UTC)

there is also another dead link landing page here: https://www.facebook.com/photo.php?fbid=1986142294321&id=1266183626&notif_t=photo_comment&refid=0#!/pages/I-LOVE-MOCHUDI-CENTRE-CHIEFS/83747398194/ Notrealname1234 (talk) 22:25, 31 July 2023 (UTC)

Thank you. -- GreenC 01:50, 1 August 2023 (UTC)

I've looked through the first 100 links in the search, and the "This page isn't available" and "Sorry, this content isn't available at this time" seem to be the only variations I've come across. The error message is localised though, the link that Notrealname1234 provided here appears in English not Brazilian for me. I think it might be IP geolocation based, as the localisation I get whenever I'm not logged in is always English. Sideswipe9th (talk) 01:58, 1 August 2023 (UTC)

Thanks for checking. -- GreenC 03:33, 1 August 2023 (UTC)

Sideswipe9th & Notrealname1234: When logged into Facebook a URL takes me to one page. When not logged in it goes to another. Example: https://www.facebook.com/photo.php?v=1405623292982594&set=vb.108930319155814&type=2&permPage=1IAAF .. it works when not logged in, does not work when logged in. The bot runs in a 'not logged in' mode, and initial testing shows all links work, that otherwise don't work when logged in. I don't know what to make of it other than Facebook is a complicated service whose behavior is subjective to the viewer. Is anyone seeing something different? -- GreenC 14:19, 1 August 2023 (UTC)

The URL that Notrealname1234 provided in this comment doesn't appear to work when logged in or out. Of the first ten links from the search above: links 2-4, 6, 9-10 are not broken; links 1 and 5 provide a "This page isn't available" message when logged in, and a login screen when logged out; links 7 and 8 provide a "Sorry, this content isn't available at this time" when logged in or out. I can't seem to find any way to make the broken links work. Sadly I don't have time right now to test any deeper than this.

The link you provided though, provides a "This page isn't available" message when logged in, and redirects to an entirely different URL (https://www.facebook.com/WorldAthletics/videos/1405623292982594/ ) when logged out. Not sure what the reason for this change in behaviour is though, other than as you say Facebook being a complicated service with subjective behaviour. Sideswipe9th (talk) 15:18, 1 August 2023 (UTC)

Sideswipe9th & Notrealname1234: Sorry, I am abandoning this project. There is too much weird stuff happening. Links work in a browser (when not logged in) but don't work when checking via bot. Every single link I check in the first 100 articles has the same result. This is probably bot protection by Facebook to prevent mass scraping of content. This is what I was afraid of, Facebook is more sophisticated than my tools, which is not surprising! -- GreenC 19:00, 1 August 2023 (UTC)

It's fine. GreenC Notrealname1234 (talk) 19:02, 1 August 2023 (UTC)

Mediastatements.wa.gov.au

As far as I can tell, every single one of the hundreds of links to this archive of Western Australian government media statements going back 30 years or so is now dead (random example which is how I found this: "Health centre renamed in honour of Busselton doctor"). Please get a bot to add archive links to all URLs in this domain. Thanks! Graham87 11:23, 30 July 2023 (UTC)

User:Graham87, it looks like the example page is a soft 404. My bot WP:WAYBACKMEDIC can deal with these after some checking to see where they redirect to and configuring the bot to treat those landing pages as dead. It will check each URL because it's possible some are still working. Based on the results of this discovery, we can then decide to mark the entire domain as dead in the IABot database, for purposes of fixing dead links in 100s of other wikis. -- GreenC 14:30, 30 July 2023 (UTC)

@GreenC: Hmmm, it looks like *some* of them are working, but from an unscientific check, only the ones for 2023 (which I hadn't checked before. They're grouped by the name of the state premier at the time then the year and month so this link search will catch those from early 2017 to Mid-2023 when Mark McGowan was premier. https://www.mediastatements.wa.gov.au/Pages/McGowan/2023/01/King-Neptune-statue-given-heritage-recognition-within-Sun-City-Precinct-.aspx works fine but https://www.mediastatements.wa.gov.au/Pages/McGowan/2022/01/Mandurah-Line-now-open-following-successful-20-day-Shutdown.aspx and https://www.mediastatements.wa.gov.au/Pages/McGowan/2022/12/Premier-unveils-new-team-with-a-focus-on-renewal-and-experience.aspx don't work. (Probably needless to say but my first example was from 1993). Graham87 15:02, 30 July 2023 (UTC)

I should have noted this before but I also started a discussion about this at Wikipedia talk:WikiProject Western Australia#Fair warning: incoming watchlist onslaught. There are some 2022 media statements that were transferred over but as noted there, the URL pattern is not *quite* amenable to a pattern-based change. As I said there the earliest I could get to was https://www.wa.gov.au/government/media-statements/McGowan-Labor-Government/Support-for-young-offenders-to-turn-their-lives-around-20220731. Graham87 15:13, 30 July 2023 (UTC)

User:Graham87: Ugh it's probably one of those sites with a really short expire time which means it would need constant checking for dead status, but since they are soft 404s the standard tools don't work. It might be better to mark the entire domain as permanent dead in IABot. Unless there was a way to find them at www.wa.gov.au/government/media-statements but I think converting to archive URL via IABot is going to be best solution long term in particular for all wikis not only enwiki. -- GreenC 15:17, 30 July 2023 (UTC)

I can see statements back to 1991 here, and as far as I can tell, all have been migrated to the new site. I don't know what it's like for you on a screen reader, but on my screen, there is a box on the right where I can select which "Administration" which allows me to select different premiers. Steelkamp (talk) 15:20, 30 July 2023 (UTC)

Oh wow I completely missed that box (which does work but it's fiddly ... but I never really got to know the old site and always navigated it using Google searches ... which can't be done now). For example, this one from 2008 still works. Hmmm ... Graham87 15:33, 30 July 2023 (UTC)

FWIW I ran IABot on the domain, after setting it to status "permadead" ie. treat all instances as a dead link. It added archives for most of them. Some are actually still alive but they'll be dead soon enough. For the parallel version at www.wa.gov.au if/when someone figures out how to map between the sites I can update the links. -- GreenC 13:42, 1 August 2023 (UTC)

ARA News - aranews.(net|org)

Before it had news, now aranews.net has Indonesian blog-styled advertisements while aranews.org redirects to an app. Count Count (talk) 11:21, 1 August 2023 (UTC)

aranews[.](net|org|com) has 282 pages. -- GreenC 13:32, 1 August 2023 (UTC)

Done. Also updated IABot database, it will propagate to other wikis. -- GreenC 16:17, 1 August 2023 (UTC)

Deleting the query

Looking to have something that delete the queries out of google books. So basically https://books.google.com/text&q=abc&moretext becomes https://books.google.com/text/test&moretext. Note, I'd also like to do the same with dq= . (I think a query can only have one). I've managed to figure out how to do this with Autowikibrowser, but it seems like a *very* large thing to do, so are there bots that can do it? (I'm using AWB to do the ones related to Fraternities and Sororities)Naraht (talk) 21:40, 3 August 2023 (UTC)

I don't have a good-enough understanding of Google Books links to know if removing queries is a good idea (in any, some or all cases). However someone who might be able to help is User:AManWithNoPlan who maintains WP:CITATIONBOT which already does Google Books link maintenance. -- GreenC 21:49, 3 August 2023 (UTC)

Often, the reference should only have the book ID, and no query or page numbers. Secondly, even when the query is present, the URL is wrong, since much of GB is javasccript driven, so the "click me to link" correct URL is often not used, and people use the browser URL. Also, usually the search or page needs removed since only one of them is really what is intended (either: all instances of the word "jones" or explicity page 23, but I used "jones" to find it) - and that requires human intervention and thought. Another problem is that the stuff after the hash # is often the final location, and much of the URL is the path taken to get there. https://github.com/ms609/citation-bot/blob/master/expandFns.php function normalize_google_books() deals with the vq, dq, q both before and after the # sign. It also has to deal with the evil & within quotes. Other oddities, like if article_id is set, then you need #v=onepage to be set. Other than that one oddball, the code trims URLs down so that they have only one search, one book ID, and one page. AManWithNoPlan (talk) 12:42, 4 August 2023 (UTC)

Thank, AMWNP. User:Naraht, it looks like GB URLs are complex and error-prone if the wrong things are removed/changed. I would suggest rely on Citation bot, it can be run for selected articles. -- GreenC 16:48, 4 August 2023 (UTC)

OK. Let me take a look at running Citation Bot on the articles in the WikiProject.Naraht (talk) 16:57, 4 August 2023 (UTC)

drdo.org

Another WP:JUDI case. I have not seen it listed there. Count Count (talk) 16:47, 4 August 2023 (UTC)

Done. Or will be. Thanks, Count Count! -- GreenC 16:50, 4 August 2023 (UTC)

User generated genealogy site

I look to start removing links to gov.genealogy.net, a user generated website, befoee realising it has several thousand uses. Could a bit replace all such uses with {{citation needed}}? -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 09:33, 6 August 2023 (UTC)

ActivelyDisinterested: are there discussions about removing it, like at RSN or somewhere? I want to link to something in the edit summary. Otherwise it looks like no problem for the bot. -- GreenC 12:58, 6 August 2023 (UTC)

Sorry I thought there was more, but checking the archives shows I was mistaken. I'll start a new discussion and come back to this later. -- LCU ActivelyDisinterested ∆transmissions∆ °co-ords° 13:17, 6 August 2023 (UTC)

In case you want to delete them all, there are 1,934 pages with gov.genealogy.net and 2,116 with *.genealogy.net -- GreenC 00:16, 7 August 2023 (UTC)

The RSN discussion closed with the outcome of waiting until November, when it will be discussed again. -- GreenC 21:59, 8 August 2023 (UTC)

Canadian Indigenous profile links

Links to http://fnp-ppn.aandc-aadnc.gc.ca say there's no such site; but if you replace "aandc-aadnc" with "aadnc-aandc" it works. To avoid changing archive links I would appreciate if someone would change "url=http://fnp-ppn.aandc-aadnc.gc.ca" to "url=http://fnp-ppn.aadnc-aandc.gc.ca" (548 pages) and also changing "[http://fnp-ppn.aandc-aadnc.gc.ca" to "http://fnp-ppn.aadnc-aandc.gc.ca" (632 pages). See my last few edits for examples.

(Background info AFAIK: These are pages provided by the Canadian federal government about indigenous nations/bands/groups. Depending on which province the headquarters of the department is in, the French or English acronym comes first; maybe the headquarters moved. Now the department is no longer called AANDC anyway; it's CIRNAC and most of its web pages are now at rcaanc-cirnac.gc.ca but this one for example https://www.rcaanc-cirnac.gc.ca/eng/1100100013791/1535470872302 links to fnp-ppn.aadnc-aandc.gc.ca to provide these pages.) ☺Coppertwig (talk) 22:52, 11 August 2023 (UTC)

Coppertwig: Done: Special:Contributions/GreenC_bot. -- GreenC 04:52, 12 August 2023 (UTC)

Wow! Thanks!! ☺Coppertwig (talk) 13:52, 12 August 2023 (UTC)

LOL! GreenC, I see you fixed a significant typo in my request. I thought I had read it over carefully. Anyway, you knew what I meant. Thanks again! ☺Coppertwig (talk) 14:49, 12 August 2023 (UTC)

You are welcome. I found this change to be unusually easy to make mistakes with, I made a few during the programming process. -- GreenC 15:21, 12 August 2023 (UTC)

Coppertwig, any ideas what new URLs there might be for these?

• https://pse5-esd5.ainc-inac.gc.ca/fnp/Main/Search/FNMain.aspx?BAND_NUMBER=564&lang=eng (391)

-- GreenC 15:21, 12 August 2023 (UTC)

Maybe simply replace the domain https://pse5-esd5.ainc-inac.gc.ca --> http://fnp-ppn.aadnc-aandc.gc.ca ? -- GreenC 15:28, 12 August 2023 (UTC)

Yes I think this is the case, the site moved: https://web.archive.org/web/20170810144106/http://pse5-esd5.ainc-inac.gc.ca/fnp/Main/Search/FNMain.aspx?BAND_NUMBER=564&lang=eng .. I'll update these 391 pages -- GreenC 16:14, 12 August 2023 (UTC)

Looks right. ☺Coppertwig (talk) 17:25, 12 August 2023 (UTC)

They're all converted (except 1 dead link). I also standardized the metadata so it's |website=Crown–Indigenous Relations and Northern Affairs Canada and |publisher=Government of Canada, and converted everything to https. I could also work on the title field with web scraping but I think this is enough for now. Government links are a mess, each new administration might redo the last admin, and not be so polite to maintain redirects. -- GreenC 20:45, 12 August 2023 (UTC)

Thanks!! I did some limited regex searches looking for BAND_NUMBER and didn't find any other old url's for the same thing, though my searches wouldn't have found everything. You've done way more than I asked for; great work. ☺Coppertwig (talk) 21:46, 12 August 2023 (UTC)

smmercury.com and asianrehub.com

Both serve online gambling spam now (WP:JUDI). There are only three links for asianrehub but it should be marked permadead in IABot in any case. --Count Count (talk) 19:16, 23 August 2023 (UTC)

Added. Thank you! -- GreenC 04:00, 26 August 2023 (UTC)

UOL (Natelinha)

natelinha.ne10.uol.com.br is a dead link, if you replace the link with natelinha.uol.com.br, it will work, there are over 91 pages with this link. (see http://natelinha.ne10.uol.com.br/noticias/2012/08/31/a-grande-familia-e-confirmada-na-grade-da-globo-para-2013-145754.php, it does not work, but if you remove the ".ne10" like this http://natelinha.uol.com.br/noticias/2012/08/31/a-grande-familia-e-confirmada-na-grade-da-globo-para-2013-145754.php, it will work.) Notrealname1234 (talk) 00:38, 26 August 2023 (UTC)

Notrealname1234. Done. Example. In 7 pages an archive URL was added example. -- GreenC 04:49, 26 August 2023 (UTC)

thisisjersey.com

Apologies if this has already been done, but thisisjersey.com: Linksearch en (insource) - meta - de - fr - simple - wikt:en - wikt:fr • Spamcheck • MER-C X-wiki • gs • Reports: Links on en - COIBot - COIBot-Local • Discussions: tracked - advanced - RSN • COIBot-Link, Local, & XWiki Reports - Wikipedia: en - fr - de • Google: search • meta • Domain: domaintools • AboutUs.com a formerly reliable source, has been hijacked and is now a gambling site. The former articles are in the Wayback Machine, so I believe a bot can rescue them? — Trey Maturin™ 11:43, 28 August 2023 (UTC)

User:Trey Maturin: Added to WP:JUDI .. this is part of a larger problem and will get processed in batches with other domains. If you find any more, please add them to JUDI, or post here, there are likely others, thank you! -- GreenC 15:32, 29 August 2023 (UTC)

Ooh, that's useful to know! Bookmarked. Thank you, GreenC! — Trey Maturin™ 16:31, 29 August 2023 (UTC)