MediaWiki talk:Spam-whitelist

Checked

Checked the links that exist for currency — billinghurst sDrewth 21:44, 1 September 2011 (UTC)

Links Galore

Links Galore is a Google Sheets spreadsheet which is an important list of links to public-domain (or controlled-digital-lending) copies of (mostly) eighteenth-to-twentieth-century scholarly editions of classical and European medieval texts. Some but not all links to this site fall foul of the spam blacklists. The problem seems to affect links which have the query string which is needed to link to a particular page inside the spreadsheet. I've made a request on English Wikipedia that all links to Links Galore should be whitelisted, and I make that request here too. For more details please see request I made on English Wikipedia.

The spam-blacklist error message told me to make this request on MediaWiki:Spam-whitelist, but since the main page is protected, I made the request here on the Talk page instead. I hope that this is the correct procedure: please tell me if I should do something else instead. RW Dutton (talk) 06:10, 18 February 2023 (UTC)

I have found that some links to individual Google Sheets pages do work. However others do not, including the links to individual pages given on the front page of Links Galore. I will still ask that all Links Galore lists be whitelisted, as contributors shouldn't have to experiment with query strings to find links which aren't blacklisted, and many will simply give up instead. RW Dutton (talk) 07:21, 18 February 2023 (UTC)

@RW Dutton: We won't be wontingly adding all links from that spreadsheet into the whitelist, sorry. If you have a specific target page you feel that should be considered, then please identify it, though where would you be looking to have it listed and linked from where?

Can I reflect that building public facing document in that form as a web document is always going to be an issue, and does a disservice to such a neat compilation. That list of resources really would do fantastically well as a wiki construction here, for instance, the ROLLS SERIES at Portal:Rolls Series would be a magical addition. And then they could be built with direct links. — billinghurst sDrewth 01:52, 28 April 2023 (UTC)

Well, by and large all of the links are just ?gid=code_number queries after the main/"front page" URL. The good news is that I've just created some test pages and it seems that all of the links in that format to the internal Links Galore sheets/subpages are presently getting past the spam filter on English Wikisource and on global Wikisource and on English Wikipedia. I'm not sure whether they always were, but if they stay working consistently throughout all pages of the Wikipedias and Wikisources then that's most of the problem solved.

The other remaining issue, then, would be that the links to the inner pages on the front page of Links Galore itself are in a google.com url?q= format where the q= value is the "real" URL encoded. Those links are being blocked, probably very understandably. I assume it would be difficult, and likely not even a good idea, to identify and whitelist the ones which happen to resolve to "real" Links Galore URLs. But that does leave the internal links advertised on the front page of Links Galore blocked by the Wikimedia filter, so it's quite likely that someone else will try using those links inside some Wikipedia or Wikisource and see them bounce off the spam filter, and it's relatively unlikely that that user will find earlier discussions in the archives of old talk:Spam-whitelist pages. To try to ameliorate this, I will turn the Links Galore test page into a permanent Links Galore entry on Wikisource, mainly just to provide the "good" links to the internal pages, in the hope that anyone who is having trouble linking to LG will find it. (I'll also ask the LG author to see if he is able and willing to change the listing of links on the LG front page to use the "real" URLs.)RW Dutton (talk) 15:58, 30 April 2023 (UTC)

@RW Dutton: I still do not understand why we would want to be hosting any of the sublinks on those pages? We don't do redirects, we link directly to works. The sole reason that these sorts of pages and their redirect links end up on the global blacklists is due to their misuse and abuse. Redirecting urls are typically blocked on sight. We don't want them, please use the real target. — billinghurst sDrewth 12:21, 1 May 2023 (UTC)

Edit request

Please add https://t dot co/D8KrMHnFdK from Page:Final Report of the Select Committee to Investigate the January 6th Attack on the United States Capitol.pdf/567 Matr1x-101 (talk) 17:24, 27 April 2023 (UTC)

Similarly, add http://tv<dot>infowars.com/index/display/id/11151 for Page:Final Report of the Select Committee to Investigate the January 6th Attack on the United States Capitol.pdf/593 Matr1x-101 (talk) 18:06, 27 April 2023 (UTC)

Also, add t<dot>co/7JqTWYUgOr%22 for Page:Final Report of the Select Committee to Investigate the January 6th Attack on the United States Capitol.pdf/356 Matr1x-101 (talk) 18:12, 27 April 2023 (UTC)

@Matr1x-101: I have whitelisted the infowars, the other two links do not work anyway, so I suggest to wrap them in nowiki tags like <nowiki>https://t.co/D8KrMHnFdK</nowiki> which will produce https://t.co/D8KrMHnFdK. --Jan Kameníček (talk) 19:15, 27 April 2023 (UTC)

@Jan.Kamenicek: Thanks, but please whitelist the third one (the one on p356). It is in a search string in a larger link, so whether it works doesn't really matter. Matr1x-101 (talk) 22:13, 27 April 2023 (UTC)

@Matr1x-101: I have tried the longer link and it seems it does not work either. --Jan Kameníček (talk) 22:39, 27 April 2023 (UTC)

@Matr1x-101: please wrap any requested link for consideration in <nowiki></nowiki>. To also note that as a waiting measure that you can do that on the page on which you are working, as the blacklist just stops active urls. I will also note that I truly have a preference that the actual target url beyond the redirect can be added in the form [https://(deep target of url) https://(blacklisted redirect)]. The whole reason for the blacklist is the abuse of these redirects, and there is zero requirement for need to list the redirect as the active url. — billinghurst sDrewth 22:55, 27 April 2023 (UTC)

The https://t.co/D8KrMHnFdK link redirects to a dead link, however, as can be seen at https://web.archive.org/web/20201220090138/https://t.co/D8KrMHnFdK it once redirected to https://www.washingtonexaminer.com/news/wh-adviser-navarro-releases-report-election-fraud-swing-victory-to-trump so it makes sense to link there and use the original redirector link as a nowiki label only, like: https://t.co/D8KrMHnFdK —Uzume (talk) 07:26, 28 April 2023 (UTC)