Jump to content

Wikinews:Water cooler/technical/archives/2025/May

From Wikinews, the free news source you can write!

Issue with archival

Hi @Leaderboard How was this possible? Could Leaderbot archive (including protect from editing) everything 24 hours after publication, automatically, please? Gryllida (talk) 00:38, 23 April 2025 (UTC)[reply]

@Gryllida The bot hasn't been running automatically due to a couple of bugs reported at my talk page (which caused incorrect templates to be applied). I'll need to fix them first. Leaderboard (talk) 04:15, 23 April 2025 (UTC)[reply]
Hi Leaderboard, thank you for replying. Which bugs are these? May I fix these bugs for you? Over a week when there is increasing number of foolish users is a bit concerning. Archiving articles is possibly a decent way to reduce load of vandalism fighting. Gryllida (talk) 04:22, 23 April 2025 (UTC)[reply]
@Gryllida See User_talk:Leaderboard#Archiving. To summarise: the problem is that the bot currently uses the date/time of the last edit in determining whether to, for instance, protect a page. As shown there, such an approach doesn't always work, which can result in cases such as incorrect templates applied, and more importantly, will result in the bot not protecting pages or otherwise taking action when it should. The fix is not difficult (the bot will need to figure out when the page has been published) - just annoying. I'll try to find some time to do this. Leaderboard (talk) 05:24, 23 April 2025 (UTC)[reply]
Hi @Leaderboard
As there was increased vandalism of published articles recently, I made a script, and a tracking category for it (in article edit it links to script source). This category contains articles which I archived without checking anything else: I didn't check pending changes to them, and I didn't add 'archiveurl=' for sources; some of them are double archived because the Published category addition date for them was hijacked by something and they were in the 'February to April' range despite having been published earlier. I encourage you, or anyone else interested, to process and empty this category as these articles are checked for whatever else needs to be done other than protecting them.
If nothing is done in near future, I'll write a script which adds 'archiveurl=' to all sources within the articles in that category.
Have a good day. Gryllida (talk) 13:32, 28 April 2025 (UTC)[reply]
Got it - but it may take some time for me to get to this. Leaderboard (talk) 14:36, 28 April 2025 (UTC)[reply]
@Gryllida Can you clarify more about the "archiveurl=" part? My script doesn't do that right now, and I'm wondering if I need to take that into account. Leaderboard (talk) 16:52, 28 April 2025 (UTC)[reply]
Hi @Leaderboard
1) Apparently I overdid it, should have done over 7 days old. Now {{Main headlines}} is practically empty. I was sure it is 1 day old. :-(
2) Example:
  • {{source
|url = https://www.bbc.co.uk/news/world-asia-58223231
|title = Afghanistan conflict: Taliban push into Kabul as Ghani flees
|author =
|pub = BBC News Online
|date = August 15, 2021
|archiveurl = https://web.archive.org/web/20210815165730/https://www.bbc.com/news/world-asia-58223231
}}
I'm not sure why this archiveurl supposed included, it seems to show only in source and there is no user-facing output of it. I think someone mentioned it would be good to run our own instance and archive all sources. Was it @SVTCobra?
Hope it helps.
Regards, -- Gryllida (talk) 00:42, 29 April 2025 (UTC)[reply]
My feeling is that archiveurl is not important right now and that this can be done separately - is that right? Leaderboard (talk) 04:26, 29 April 2025 (UTC)[reply]
Yup can be done later with some risk that some sources will be gone by the time this is started. But yes this can be done later. Gryllida (talk) 05:13, 29 April 2025 (UTC)[reply]
So how does this work usually? Do you manually get the web.archive link for that page? Leaderboard (talk) 06:10, 29 April 2025 (UTC)[reply]
Yes someone used to do that manually a long time ago. Gryllida (talk) 06:12, 29 April 2025 (UTC)[reply]
OK got it. Meanwhile I think I've fixed the bugs in my script, but it seems like it doesn't need to do anything for a while. Leaderboard (talk) 10:05, 29 April 2025 (UTC)[reply]
Thanks. May I please have ssh access as co-maintainer? I will, in case of needed changes, keep backups and notes on-wiki. I can program for wiki in python, node, perl, and javascript. Gryllida (talk) 11:57, 29 April 2025 (UTC)[reply]
See https://github.com/Leader-board/WMF-projects/tree/master/enwikinews Leaderboard (talk) 12:42, 29 April 2025 (UTC)[reply]
Thanks a lot Gryllida (talk) 11:40, 4 May 2025 (UTC)[reply]

┌──────────────────────────────────────────────┘
ArchiveURL can help recover lost source links. I am not aware of an automated process for this. User:InternetArchiveBot was proposed to assist, but its current status is unclear (see discussion here).

We possibly include ArchiveURL to support Wiki-wide developments, similar to this recent tech news update:

The Editing Team is working on making it easier to populate citations from archive.org using the Citoid tool, the auto-filled citation generator. They are asking communities to add two parameters preemptively, archiveUrl and archiveDate, within the TemplateData for each citation template using Citoid.

I am not aware of us using Citoid in relation to {{Source}}. —Michael.C.Wright (Talk/Reviewer) 14:46, 5 May 2025 (UTC)[reply]

We can't actually use Citoid with the Source template, as it is designed to automatically generate inline references, which are not used on here. Although we can use its API for similar functionality, to auto generate sources, like what this script does; But Citoid doesn't automatically generate archive URLs. Asked42 (talk) 18:10, 5 May 2025 (UTC)[reply]

Review template is not replyable

When an article was reviewed, a new section is added at it talk/collaboration page. It does not have a signature. As a result, it does not have a clickable "reply" link for it. What would be a nice way to resolve this? Gryllida (talk) 10:24, 4 May 2025 (UTC)[reply]

Add the signature code to the reviewing template? Leaderboard (talk) 11:25, 4 May 2025 (UTC)[reply]
If it works it would be very good, could you please add it? Gryllida (talk) 11:40, 4 May 2025 (UTC)[reply]
To add a signature after the review template, we can modify the variable reviewText at line 260 of easyPeerReview Gadget with the following code:
var reviewText = "<!-- " + Bawolff.review_i18n.review_comment + " --> \{\{" + Bawolff.review_i18n.review_peer_reviewed + "|" + Bawolff.review_i18n.review_revid + "=" + r.EasyReview_revid + "|" + Bawolff.review_i18n.review_copyright + "=" + r.EasyReview_copyright_comment + "|" + Bawolff.review_i18n.review_newsworthy + "=" + r.EasyReview_news_comment + "|" + Bawolff.review_i18n.review_verifiable + "=" + r.EasyReview_sources_comment + "|" + Bawolff.review_i18n.review_npov + "=" + r.EasyReview_npov_comment + "|" + Bawolff.review_i18n.review_style + "=" + r.EasyReview_sg_comment + "|" + Bawolff.review_i18n.review_reviewer + "=" + mw.config.get('wgUserName') + "|" + Bawolff.review_i18n.review_comments_field + "=" + r.EasyReview_comment + "|" + Bawolff.review_i18n.review_time + "=~~" + "~~" + "~}} \n--" + "~".repeat(4);
-- Asked42 (talk) 14:50, 4 May 2025 (UTC)[reply]

We will be enabling the new Charts extension on your wiki soon!

(Apologies for posting in English)

Hi all! We have good news to share regarding the ongoing problem with graphs and charts affecting all wikis that use them.

As you probably know, the old Graph extension was disabled in 2023 due to security reasons. We’ve worked in these two years to find a solution that could replace the old extension, and provide a safer and better solution to users who wanted to showcase graphs and charts in their articles. We therefore developed the Charts extension, which will be replacing the old Graph extension and potentially also the EasyTimeline extension.

After successfully deploying the extension on Italian, Swedish, and Hebrew Wikipedia, as well as on MediaWiki.org, as part of a pilot phase, we are now happy to announce that we are moving forward with the next phase of deployment, which will also include your wiki.

The deployment will happen in batches, and will start from May 6. Please, consult our page on MediaWiki.org to discover when the new Charts extension will be deployed on your wiki. You can also consult the documentation about the extension on MediaWiki.org.

If you have questions, need clarifications, or just want to express your opinion about it, please refer to the project’s talk page on Mediawiki.org, or ping me directly under this thread. If you encounter issues using Charts once it gets enabled on your wiki, please report it on the talk page or at Phabricator.

Thank you in advance! -- User:Sannita (WMF) (talk) 15:07, 6 May 2025 (UTC)[reply]

Seeking consensus to install mw:Help:Extension:Newsletter

There are three purposes: to announce published articles, to announce new drafts, and to announce new requests for revision, to interested wiki users. The announcements can be delivered via Echo, and are visible also when logged into a sister wiki. This extension provides an opportunity to do this automatically, and users subscribe or unsubscribe themselves on special page. To get started, consensus is needed here: do we want to add this extension on this wiki? Thanks. Gryllida (talk) 12:42, 9 May 2025 (UTC)[reply]

Discussion (and ping any users who you think could be interested here)

{{flag}}

Yes FranklySiva (talk) 16:02, 10 May 2025 (UTC)[reply]
https://phabricator.wikimedia.org/T394022 added. Gryllida (talk) 12:28, 13 May 2025 (UTC)[reply]

Votes

Script to sort sources

In https://en.wikinews.org/w/index.php?title=India_and_Pakistan_agree_to_a_ceasefire_after_days_of_mutual_attacks&oldid=4857433#Sources the fresh sources should go first. I'd like to write a script to sort them. Have no clue how to even start. For a given section, how can I retrieve a list of templates? For a retrieved list of templates, how do I get list of their keys and values which were specified as parameters? cc @Leaderboard @Asked42 @Polygnotus Thanks Gryllida (talk) 20:04, 14 May 2025 (UTC)[reply]

Fresh sources as in? The more recent, the newer? Leaderboard (talk) 06:29, 15 May 2025 (UTC)[reply]
source May 15, then May 14, then May 13, etc. the 'today' dated go first. Gryllida (talk) 11:31, 15 May 2025 (UTC)[reply]

New software request

Hello,

I'd like to request that someone writes a script which opens a dialog with a list of category members of 'Revision requested' category, for each member it goes through the article edit history and finds where a new section titled starting with 'RR:' was added and adds column with (i) stuff that is in section title after 'RR:' wikilinked to that section (ii) date of addition of that section (iii) nick of who added it (iv) link to article page.

I may be able to code something like this but I haven't started yet, so you're welcome to start now and/or share some pointers if you like. Thanks heaps Gryllida (talk) 20:49, 14 May 2025 (UTC)[reply]

Template requests

I will update this section with more template requests over time:

1. Excerpt (from Wikipedia): allows for essentially transcluding content from one article to another. Perhaps it could be possible to even transclude Wikipedia content to a Wikinews article? GreekApple123 (talk) 17:52, 15 May 2025 (UTC)[reply]

2. Would something like w:Template:Category redirect be appropriate here? This way, something like Category:Politics can continue to be empty, but at least exist as a redirect to the currently-used Category:Politics and conflicts.

3. Something like the series of w:Template:R with possibilities templates could be helpful as a way of creating redirects in the moment, but making it clear they can be developed further later. GreekApple123 (talk) 18:09, 15 May 2025 (UTC)[reply]

It is needed to understand that there is a scarcity of content writers and reviewers. Your assistance with that editing of articles is very welcome.
Wikipedia articles are written differently. Verb tense and inline citations and they do not do "attribution" as consistently and they do not follow "inverted pyramid". I would not recommend to transclude. They are very different contents.
I have no position on category redirects. It is less effort to just use existing categories. HotCat aurocompletes them anyway.
I have not seen a need in creating redirects. They consume time.
Hope it helps. Gryllida (talk) 20:54, 15 May 2025 (UTC)[reply]
There may be a use for transcluding a paragraph or two from Wikipedia (I'm guessing transclusion may not run afoul of copyright license incompatibility?), or at least transcluding between articles on Wikinews. GreekApple123 (talk) 21:45, 15 May 2025 (UTC)[reply]
From Wikipedia I dare you to find examples. Usually information is not included into article unless it is verifiable in sources, unless it is forever ever kkowledge like Paris capital of France. So if you are writing article about new Pope, you cannot put five pages from Wikipedia into a background section. It has to be sourced from something more specific than Wikipedia. Gryllida (talk) 22:03, 15 May 2025 (UTC)[reply]
From Wikipedia I dare you to find examples. Well, in the final section of the page I am editing about the India-Pakistan ceasefire, there is background offered on the Kashmir conflict. I think the first paragraph and one of the images from the Wikipedia article on that would offer a more comprehensive overview essentially. GreekApple123 (talk) 23:17, 15 May 2025 (UTC)[reply]
Wikipedia articles are written differently [...] inline citations Just FYI, Excerpt allows for transcluding content while excluding all inline citations and references. GreekApple123 (talk) 04:00, 16 May 2025 (UTC)[reply]
That does not solve content. For example, verb tense. Gryllida (talk) 07:29, 16 May 2025 (UTC)[reply]
For article bits transclusion, I guess would need to be subst'ed as cannot edit a part of article after archival. Dunno value of it. With such as wars, the context keeps changing a bit too. Gryllida (talk) 21:49, 15 May 2025 (UTC)[reply]
@GreekApple123, regarding transcluding content between articles on Wikinews, it's technically possible—usually done by transcluding specific sections. However, since our articles rarely use subsections, this may require structural adjustments.
As for transcluding content from Wikipedia, we legally cannot do so. Our copyright licenses are incompatible: Wikinews uses CC BY 2.5, while Wikipedia uses CC BY-SA 4.0, which includes a ShareAlike requirement we do not support. The same applies in reverse—Wikipedia cannot reuse our content either.
Regarding category redirects, I think that’s a solid consensus compromise and provides a useful tracking mechanism. One concern to keep in mind: many pages here are edit-protected to admin-only. In my view, this often causes more issues than it prevents, especially given how few active admins we have to handle tasks like archiving, deletion, and counter-vandalism. As part of any category overhaul, I suggest we also reconsider our protection practices—possibly lowering some protections to allow editing by reviewers, or even auto-confirmed users, to ease the workload and improve flexibility. —Michael.C.Wright (Talk/Reviewer) 21:50, 15 May 2025 (UTC)[reply]
Changing categories of archived articles is a bit of a problem. I will post here if I find a working solution. I presume content should not be changed. Gryllida (talk) 22:06, 15 May 2025 (UTC)[reply]
regarding transcluding content between articles on Wikinews, it's technically possible—usually done by transcluding specific sections. The Excerpt template allows for transclusion by paragraph (e.g. paragraphs 3 through 5), and so since archived articles don't change, it should be possible to excerpt from them in a stable manner.
The same applies in reverse—Wikipedia cannot reuse our content either. This seems to be somewhat contradicted by the WN:For Wikipedians page: ... content from Wikinews may be copied word-for-word to Wikipedia if proper copyright attribution is provided GreekApple123 (talk) 23:08, 15 May 2025 (UTC)[reply]

4. Extending the functionality of {{W}} so that if there is no link found either at Wikinews or at Wikipedia, then the link text instead shows up as regular text. GreekApple123 (talk) 17:31, 16 May 2025 (UTC)[reply]

In my common.js there is a script that color codes this. Please check. If you cannot find it, please ping me, I will look at it when I am at desktop. Gryllida (talk) 10:32, 17 May 2025 (UTC)[reply]

5. Porting over various templates related to conversion, such as converting between measurements, currencies, indicating language translation and transcription more explicitly, etc. GreekApple123 (talk) 15:33, 23 May 2025 (UTC)[reply]

Timezone converting is so painful. :-( It would be nice to have a code. Gryllida (talk) 12:00, 28 May 2025 (UTC)[reply]

Talk pages indexed by search engines

Today I discovered that our article talk or collaboration pages are indexed by search engines. This is problematic, as these pages may contain discussions about unsupported statements that were never included in the published article, yet are visible in search results. Just as Wikipedia prevents indexing of BLP talk pages[1], I believe we should have NOINDEX applied to of all of our talk pages, given our role as a news site. We'll remain transparent by having it all publicly available, just not (hopefully) indexed by search engines.

The solution seems to involve namespace control[2], but I’m unsure how we would implement this or who to approach to begin the discussion. Suggestions? —Michael.C.Wright (Talk/Reviewer) 22:02, 27 May 2025 (UTC)[reply]

I don't know if this is a good idea - is it a bad idea for people to see unsupported statements that are clearly part of a discussion? Leaderboard (talk) 07:22, 28 May 2025 (UTC)[reply]
I saw today that unpublished stories are also indexed. How did this work before? Gryllida (talk) 11:58, 28 May 2025 (UTC)[reply]
I did a search using DuckDuckGo for "Maksim Lepushenka" and it still returns our article (good) plus the article's talk page (bad).
Gryllida, which unpublished stories showed up in a search and what was the search? I just searched for the headline of one of our developing articles "Sound Talent Group employees among casualties in private plane crash at San Diego, California" and did not get our article or its talk page in the search results.
It works through Robots.txt. See also this similar and recent request. {{develop}}, {{tasks}}, {{abandoned}} and {{delete}} all add NOINDEX to individual pages.
I don't believe we've ever explicitly prevented the indexing of our article talk pages.
I do not see the talk page listed in our Sitemap, which is good. That means a crawler probably went from the article page to the talk page through the link.
It could be that we add Disallow: /wiki/Talk: to MediaWiki:Robots.txt. —Michael.C.Wright (Talk/Reviewer) 12:46, 28 May 2025 (UTC)[reply]
It was the union rally story, which is developing. Gryllida (talk) 20:45, 28 May 2025 (UTC)[reply]
It looks like Robots.txt is not being applied, even when we use the magic word __NOINDEX__. Do local admins have access to Wiki settings discussed in this mediawiki help page: mw:Help:Controlling search engine indexing to help troubleshoot? If this isn't something local admins can help with, should it be a phabricator request? Thanks in advance. —Michael.C.Wright (Talk/Reviewer) 23:37, 28 May 2025 (UTC)[reply]

┌─────────────┘
Leaderboard said ...is it a bad idea for people to see unsupported statements that are clearly part of a discussion? NOINDEX won't hide them or prevent anyone from finding them. It should just stop search engines from indexing them and then presenting them in search results. For published articles, it's fine if the article is indexed and the talk page is not. Readers can still follow the "Collaboration" link and see the discussion behind the article. NOINDEX should also prevent AI/LLMs from scraping talk pages and article drafts...should. —Michael.C.Wright (Talk/Reviewer) 23:42, 28 May 2025 (UTC)[reply]

But is it a bad thing to allow search engines from seeing them? Leaderboard (talk) 05:52, 29 May 2025 (UTC)[reply]
Yes, I believe so. Wikinews should index the content we deem as our primary "product," the published news articles. These are the polished, peer-reviewed stories we intend the public to read first. Conversely, Talk pages and collaboration spaces represent the “behind-the-scenes” process of how we produced those stories, including discussions that may reference unsupported or draft ideas. By preventing these pages from being indexed, we signal a clear distinction: the article is the authoritative, intended presentation for the public, while the Talk page is supplementary for those who choose to explore the process. Transparency remains, but we control the first impression that search engines deliver. —Michael.C.Wright (Talk/Reviewer) 13:27, 29 May 2025 (UTC)[reply]
The other way to think about this is SEO optimisation. It may instead be better to put some sort of template on the talk page so that newbies are not confused as to where the article is? Leaderboard (talk) 19:05, 29 May 2025 (UTC)[reply]
What would you recommend for verbiage? I can put that together easily enough. —Michael.C.Wright (Talk/Reviewer) Michael.C.Wright (Talk/Reviewer) 20:49, 29 May 2025 (UTC)[reply]
It is also currently possible to search for unpublished (developing) articles in Wikinews search. Should that be disabled/restricted too? GreekApple123 (talk) 15:38, 29 May 2025 (UTC)[reply]
I don't believe so. We want contributors to easily find stories already started so that we avoid duplicating efforts. It's also an "internal" thing versus an external search engine, where people are more likely searching for news and other 'official' information about a topic.
I assume someone who is using our Wikinews search is more likely to be aware of the difference between a talk page and a published article. Someone doing a Web search for a person's name, for example, may not even know what a Wiki talk page is. —Michael.C.Wright (Talk/Reviewer) 17:03, 29 May 2025 (UTC)[reply]

Draft assistant

https://gryllida.neocities.org/news/writer/v3/

cc @GreekApple123 @Md Mobashir Hossain @BigKrow @Wikiwide

Feedback welcome

Thanks Gryllida (talk, live chat) 20:54, 29 May 2025 (UTC)[reply]

@Gryllida, seems complicated for me.... Sorry... BigKrow (talk) 21:48, 29 May 2025 (UTC)[reply]
Is v3-2 less confusing to you, @BigKrow? Gryllida (talk, live chat) 10:18, 30 May 2025 (UTC)[reply]
@Gryllida, looks better thanks. BigKrow (talk) 10:21, 30 May 2025 (UTC)[reply]
Glad to hear it. Let me know if you are planning to use it and encounter issues. Thank you! Gryllida (talk, live chat) 10:22, 30 May 2025 (UTC)[reply]
Will do. BigKrow (talk) 10:24, 30 May 2025 (UTC)[reply]
The idea of having criterion checkboxes that a writer has to tick off is interesting, because perhaps that can be implemented in the mainstream interface and then writers who claim to have fulfilled certain criteria but failed to do so can be penalized after doing so too many times. GreekApple123 (talk) 23:26, 29 May 2025 (UTC)[reply]
I didn't find a convenient way to make tickable boxes on-wiki in a talk page section. Can you help? Gryllida (talk, live chat) 12:48, 31 May 2025 (UTC)[reply]
I just asked here. -- Gryllida (talk, live chat) 12:50, 31 May 2025 (UTC)[reply]