Wikinews:Water cooler/proposals - Wikine

Page last updated: Friday 24 at 2115 UTC.

Protecting templates

I am working through the pending changes to templates now. I think it would be wise to protect all templates to 'auto-confirm' accounts as a minimum, not changing any existing protection levels. A lot of the pending changes so far include reverted vandalism (this is good, it is being found and reverted quickly). Many of the vandals are IP/temp accounts, which presumably haven't yet been auto-confirmed.

That could be a lot of tedious work to do up front, but may pay off in the long run.

Any thoughts/objections?Michael.C.Wright (^Talk/_Reviewer) 18:35, 17 February 2026 (UTC)Reply

Do you have a few examples? Gryllida 18:48, 17 February 2026 (UTC)Reply

Feel free to look through the edit history/recent changes. I was typically descriptive using 'vandalism undone.'Michael.C.Wright (^Talk/_Reviewer) 18:49, 17 February 2026 (UTC)Reply

I just hit this one, so here's an example: Template:TScope/ex/doc.

Feel free to take action on it.Michael.C.Wright (^Talk/_Reviewer) 18:51, 17 February 2026 (UTC)Reply

Ah yes. Those were not edit protected, and anyone could edit them. For when nobody looking, that backlog is pretty tame...

I personally would think that edit protect them makes sense only if the proposed change was to go ahead -
but then, I would suggest to make them "reviewer or admin only" as is more similar to like the current setup, where the intent is for the edit to be checked by a privileged user. Relying on recent change patrolling may not work well on a small wiki that is news focused.
In any case of protection I would suggest to work on making it easier to request an edit. Clicking "talk" and leaving a message is pretty hard For a newbie.
Could the software show a banner at top of all edit protected pages with a big fat "request edit" button that takes to new section at talk with relevant "edit protected request" or whatever template already filled in with a space to describe proposed edit and reason. This banner should only show for users who do not have the privilege to edit the page.
If needed i can write a js that protects page, places a banner at page top, and sights the edit.

- Gryllida 19:10, 17 February 2026 (UTC)Reply

> I would suggest to make them "reviewer or admin only" as is more similar to like the current setup, where the intent is for the edit to be checked by a privileged user.

We can't set "reviewer" as a level of protection. It's not available in the protection menu (I'm not sure if we can somehow enable it).

Also, protecting to admin-only would put too much burden on a group that is already underserved and over-taxed. It is further frustrated by the fact that one of our two interface admins (Asked42) is neither a reviewer nor an admin and therefore they would not be able make any changes to templates if we protected them all to reviewer (if we are able) or admin.

A good portion of the changes I sighted were indeed made by Asked42.

I think auto-confirmed as the minimum baseline, with some fully protected (as they already are) is a good compromise/happy medium.Michael.C.Wright (^Talk/_Reviewer) 19:23, 17 February 2026 (UTC)Reply

yes we can add a $wgRestrictionLevels, i guess 'sysop or reviewer or interface admin' could work Gryllida 21:06, 21 February 2026 (UTC)Reply

This would require initiating a separate proposal to add editautoreviewprotected as a protection level and user right (autoreviewed users, reviewers, and administrators). Protection for interface administrators is currently not possible, as regular administrators cannot edit site CSS/JS pages, or CSS and JS pages in any user's userspace (except for their own). Codename Noreste (talk) 21:10, 21 February 2026 (UTC)Reply

Based on this, I think we should protect all templates to a minimum 'auto-confirmed' level, then increase it on an as-needed basis, with the understanding that over-protection impedes regular maintenance.
Protecting all templates to reviewer + admin is in my opinion overkill and prevents our single, active, interface admin from being able to make any changes to any templates.
Very few of our templates are used on published articles, which is arguably our most sensitive namespace as far as vandalism is concerned.
This also requires no changes be made to the overall config.
So to be clear, I would support the following (~~assuming~~ I assume we also do it in the following order):

Protecting all templates to a minimum of auto-confirmed users
Remove/disable flagged revs in the Template namespace

Michael.C.Wright (^Talk/_Reviewer) 23:01, 21 February 2026 (UTC)Reply

that risks all sysops (who may be awol at times) missing an edit an autoconfirmed user made, whereas protecting would prevent this. i think on a small wiki it is better to reduce risk of an edit being made inadvertedly and then being missed. this is not wikipedia where good faith is assumed and infinite resources are available for post-edit page patrol. that would be a low activity backlog anyway seeing that you handled all pending revisions within less than a day - and they were accumulated over course of a few years i believed. Gryllida 00:09, 22 February 2026 (UTC)Reply

Whatever user right, it should be granted based on consensus or a sysop decision, not "this user was on wiki for 3 days therefore not spammer" (whatever that was named), in my view. Gryllida 00:22, 22 February 2026 (UTC)Reply

Oppose applying reviewer and/or admin-only protection to all templates. It will make general maintenance too difficult.
I

Support applying auto-confirm protection to all templates that are not currently protected, or:
I

Support not changing template protection broadly at this time. I don't think this decision should impede the original proposal since it was added later.Michael.C.Wright (^Talk/_Reviewer) 01:11, 22 February 2026 (UTC)Reply

I initiated phab:T418066. Codename Noreste (talk) 20:58, 21 February 2026 (UTC)Reply

Note that we recently had an issue with our Main page not transcluding updates to the lead article templates and the cause is suspected to be the disabling of FlaggedRevs in the template namespace.

See Wikinews:Water cooler/technical#Main page leads not updating properly for details. See https://phabricator.wikimedia.org/T423512 for more details.

If you see similar issues with templates not displaying, please notify everyone at Wikinews:Water cooler/technical#Main page leads not updating properly.

Thanks!Michael.C.Wright (^Talk/_Reviewer) 22:30, 15 April 2026 (UTC)Reply

Options how to protect

[edit]

For templates which are current in control by flagged revs, a reviewer has to sight an edit for it to be approved and become included in other pages. With the flagged revs being disabled for them. There is a need to make a decision what to do to prevent vandalism and other unwanted edits.

Option 1: do nothing; the templates become available for anyone to edit. Another user can undo that. There is a risk that an edit is done wrong; any user can monitor and undo.
Option 2: protect on 'auto-confirmed' level. Accounts with the 'autoconfirmed' privilege can edit. Another 'autoconfirmed' user can undo if needed. There is a risk that an edit is done wrong; an 'autoconfirmed' user or users need to check recent changes and monitor. This inconveniences users who are not 'autoconfirmed' because they havve to open a new section at talk page to request an edit. Work load on 'auto-confirmed' is increased because they have to monitor such requests and recent changes.
Option 3: protect on 'some new privilege name here' level. The privilege is given out to any user who was doing something constructive and is provided by request without needing to do a full vote. A vote may be held to remove it if needed or an admin can remove that privilege if it was misused. Another 'some new privilege name here' user can undo if needed. There is a risk that an edit is done wrong; an 'some new privilege name here' user or users need to check recent changes and monitor. This inconveniences users who are not a'some new privilege name here' because now they have to manually request new edits at the talk page. Work load on 'some new privilege name here' is increased because they have to monitor such requests and recent changes.
Option 4: protect on 'reviewer or sysop or interface admin' level. Accounts with 'reviewer or sysop or interface admin' privilege can edit, and another user with that privilege can undo if needed. There is a risk that an edit is done wrong, however, the pool of users was hand selected and voted on and the risk was reduced. This inconveniences users who are not a 'reviewer or sysop or interface admin' because now they have to manually request new edits at the talk page. Work load on 'reviewer or sysop or interface admin' is increased because they have to monitor such requests and recent changes.
Option 5: protect on 'sysop' level. Accounts with sysop (Administrator) privilege can edit, and another user with that privilege can undo if needed. There is a risk that an edit is done wrong, however, the pool of users was hand selected and voted on and the risk was reduced. This inconveniences users who are not a Administrator because now they have to manually request new edits at the talk page. Work load on 'Administrator' is increased because they have to monitor such requests and recent changes.

Let me know if I missed something. Gryllida 03:01, 22 February 2026 (UTC)Reply

Comments and discussion

I have stronger support for 5 than 4 than 3 than 2. The amount of such changes made in template namespace was tiny and I propose to reduce load on 'contributors, users, reviewers, and interface admins', as they best work on news. Solving vandalism or 'hey lets change this because thats how i did it somewhere else and i thought that is good here' or 'i am confused' issues is a sysop job (ignoring that sysops have huge backlog because I am here and I don't even know that these tasks exist, but still. sysop handle mitigations of abuse and approvals of legit edits, in my view. I do not want to bring up question of sysop backlog size or lack of prior involvement in such issues before as it is not directly relevant to my argument). Gryllida 02:50, 22 February 2026 (UTC)Reply

From Wikinews:Deletion_requests#c-Michael.C.Wright-20260107193100-Israeli_naval_ships_intercept_Gaza-bound_Global_Sumud_Flotilla there is a note from @Ternera "I would even go so far as to propose that we block all AI articles on this wiki.". I agree with this proposal in short term. In long term I think it would be good to allow it after some reliable workflow is developed. BigKrow may be interested in this discussion. Gryllida 21:51, 7 January 2026 (UTC)Reply

Support noting this is for short term only and further effort could be undertaken to see what went wrong and how it could be addressed. Gryllida 21:51, 7 January 2026 (UTC)Reply
I'm unsure about this right now BigKrow (talk) 22:05, 7 January 2026 (UTC)Reply
Oppose greatly, BigKrow (talk) 22:21, 7 January 2026 (UTC)Reply

Hi @BigKrow what would you think of requirement 'ai involved articles are in beta phrase, and are only allowed to be written for events which happened on first day of the month'? or some other random blanket rule that leads to effort managing ai generated content as a beta stage only and not every day. then maybe some understanding could be reached how to use ai more efficiently. please let me know, thanks. Gryllida 13:26, 11 January 2026 (UTC)Reply
im just so used to using AI I personally enjoy it and I think it helps in some ways to news stories. @Gryllida BigKrow (talk) 20:27, 11 January 2026 (UTC)Reply

Yes, short term. When AI gets better or we have more manpower etc., we of course gonna use it. But not for now Sheminghui.WU (talk) 04:45, 9 January 2026 (UTC)Reply
Hi @Sheminghui.WU, what would you think of requirement 'ai involved articles are in beta phrase, and are only allowed to be written for events which happened on first day of the month'? or some other random blanket rule that leads to effort managing ai generated content as a beta stage only and not every day. then maybe some understanding could be reached how to use ai more efficiently. please let me know, thanks. Gryllida 13:26, 11 January 2026 (UTC)Reply
Support a ban on AI generated news pieces. I'm not seeing how it provides value since anyone can ask AI to summarize a news article. Plus, the AI model appears to hallucinate and includes false information or bad sources, so it seems like it creates unnecessary work. Ternera (talk) 21:55, 7 January 2026 (UTC)Reply
I used AI to help with the following published articles, for just a few examples:

I believe they all benefitted from it. I'm not saying they're perfect, but I think they benefitted from AI assistance.

I also use it to check neutrality and balance during reviews and I find it's very helpful in that regard.Michael.C.Wright (^Talk/_Reviewer) 22:59, 7 January 2026 (UTC)Reply

Hi @Ternera what would you think of requirement 'ai involved articles are in beta phrase, and are only allowed to be written for events which happened on first day of the month'? or some other random blanket rule that leads to effort managing ai generated content as a beta stage only and not every day. then maybe some understanding could be reached how to use ai more efficiently. please let me know, thanks. Gryllida 13:27, 11 January 2026 (UTC)Reply

Oppose Each instance of AI hallucinations slipping through was either 1. A violation of WN:AI or 2. Should have been caught by the review process, or both. Maybe if we de-flagged reviewers after three AI slip-ups in published articles, we'd catch more of them sooner.
I think AI should be treated no different than a human editor. That means any hallucinations, which are simply inaccuracies or unsupported statements by another name, are catchable in the review process (but should be caught in the development process). I would imagine if we ban it, we will simply drive its use "underground."
And to be clear; I'm not advocating for AI-generated articles, but AI-assisted articles.
I'd prefer that we enforce the full and complete review process. Pseudo-reviews that skip the verification process predictably lead to more corrections and retractions.Michael.C.Wright (^Talk/_Reviewer) 22:38, 7 January 2026 (UTC)Reply

I think we have yet to update Wikinews:The use of AI in generating content then approve it either as a policy or a guideline. Codename Noreste (talk) 22:41, 7 January 2026 (UTC)Reply

I agree. The proposed guideline variously states that humans are responsible for the state of the article.

"these drafts must always be critically reviewed and fact-checked by human contributors before submission."
"the final decisions on wording and style remain human-led."
"any interpretation or conclusions drawn must be reviewed by human contributors."
"The editorial process, including pre-reviews and final reviews, must always be carried out by human contributors."

Michael.C.Wright (^Talk/_Reviewer) 22:52, 7 January 2026 (UTC)Reply

AI makes stuff up and this is quite dangerous. I asked it about a snow hazard in Switzerland and provided list of 3 sources; it told me the date 13 months off. If AI assistance is to be kept, I would suggest requiring that the author(s) provide a list of prompts to their AI tool on the talk page. This could help to improve quality of reports. Gryllida 02:09, 8 January 2026 (UTC)Reply

> AI makes stuff up and this is quite dangerous.

However, those facts must be present in the listed sources to be included in our article. Bogus facts provided by AI won't be in the listed sources.
A full and complete review will catch that. That's how they were caught in both recent cases where a retraction was proposed. Someone read the article and tried to verify every statement against the sources.Michael.C.Wright (^Talk/_Reviewer) 03:06, 8 January 2026 (UTC)Reply

I was told Grok AI gave information on underage kids pornography so strong delete@Michael.C.Wright BigKrow (talk) 17:55, 8 January 2026 (UTC)Reply

https://www.theguardian.com/technology/2026/jan/05/elon-musk-grok-ai-digitally-undress-images-of-women-children BigKrow (talk) 17:58, 8 January 2026 (UTC)Reply

"Maybe if we de-flagged reviewers after three AI slip-ups in published articles, we'd catch more of them sooner." Should the authors be penalised in any way, if they repeatedly fail to catch AI's hallucinations? I'd like to know your opinion on this, @Michael.C.Wright. Gryllida 13:20, 11 January 2026 (UTC)Reply

Authors are responsible for accuracy. But reviewers bear final accountability for published content because review is a permission granted by the community to publish on its behalf. In these cases, the failure was not the existence of errors, but their approval through incomplete review.Michael.C.Wright (^Talk/_Reviewer) 15:56, 11 January 2026 (UTC)Reply

Hi @Michael.C.Wright

1. I ask to approach both reviewer and author in these cases as they both need to change their approach to ensure issues don't happen again and again. I asked you and you said 'Yeah, except reviewers were elected.' Yeah, sure, but my point still stands. I'm speaking of both.

2. Doesn't answer my question how to handle users who repeatedly insert same kind of issues into articles. That is not even always visible as the review feedback of failed articles gets deleted along with their talk page. I try to at least post a copy of it onto talk page via 'custom message' feature of ezpr2. Make that an official policy, would you? I'm sure you would be interested in this as it would allow to keep track of issues in articles submitted by that user in one spot, and provide a path towards figuring out why they happen and how to solve them, and improve 'article quality'.

3. And the same could go to posting feedback to a reviewer's talk page about post-publish retractions etc. I believe I didn't even get a talk page message about that retraction request; yet I did get an insinuation here. The note that a particular reviewer is being dodgy sits in your mind and doesn't sit on their talk page. I think that's inefficient. The concern about your mentioning of de-flagging lies in unnecessary escalation; the general civil thing to do has been to leave a message at a talk page before taking further action.

Hope it helps. Gryllida 16:14, 11 January 2026 (UTC)Reply

Hi @Michael.C.Wright see w:Hallucination (artificial intelligence) for your reference in case you didn't know, regards Gryllida 13:28, 11 January 2026 (UTC)Reply

Since hallucinations are inaccuracies, they won't be in the linked sources. So a full and proper WN:REV would catch them. Just as they were caught in the previously mentioned articles once a full and proper review was performed.

The reviewer should check that all information in the article is fully sourced,

Seems like we already have the solution; perform full and proper reviews.Michael.C.Wright (^Talk/_Reviewer) 14:21, 11 January 2026 (UTC)Reply

Seems like incomplete solution since users don't fix anything fast. I have seen it tried for 15 years. Gryllida 15:50, 11 January 2026 (UTC)Reply

~~Support~~I don't think we are on the right side of history. But current experience shows that editers and reviewers can not determine if theres any content in an AI generated article. And we "cannot edit after it published", so no remedies can be made. I also believe we should issue an announcement after disabling AI, which would show that we are a responsible media outlet.

Of course, I maintain the stance discussed earlier that, ideally, editors should consciously strive to discern, and reviewers should ensure everything meets the standards, such as accuracy. If the review process can achieve this, it doesn't matter whether the article is written by a human, Shakespeare's monkey, or AI, nor is explicit labeling necessary, because we ultimately get a good article.

I'm not saying people can't achieve this, but it requires too much effort, which is clearly not cost-effective. And if we implement policies similar to this now instead of banning AI, it will lead to situations where submissions written with AI cannot be published due to insufficient and timely review, which is obviously not in our favor.

I apologize for the mistakes in an article I participated in editing. ~ Sheminghui.WU (talk) 04:03, 9 January 2026 (UTC)Reply

> editers and reviewers can not determine if theres any content in an AI generated article.
If that were true, how was it identified for the proposed retraction?Michael.C.Wright (^Talk/_Reviewer) 14:32, 9 January 2026 (UTC)Reply

I'm just coming to this now and have no idea what this question is asking. Is there some background somewhere about a proposed retraction that isn't immediately clear in this thread? What is that phrase referring to? Tduk (talk) 04:13, 10 January 2026 (UTC)Reply

Yes, the proposed deletion linked in the first sentence of this section. Gryllida 13:19, 11 January 2026 (UTC)Reply

Hi @Tduk fyi

for last few months now, partly due to my medical issues and partly due to my preference, i have been publishing articles checking only the basics (5Ws and H) and not checking each sentence thoroughly. i read through sources and ensure that what i have read does not have serious npov issues or contradictions that immediately strike my mind, but don't check fully.
i believe @Michael.C.Wright has made it clear that he is here to uphold the standards to highest thresholds and now this is taking the form of him doing a full review of an article at time of archival. i disagree with this approach as i think it is taking away his valuable time from reviewing stories which are actually fresh. i do not believe anyone ever did a full review of an article at time of archival. this process previously only involved checking for red links, issues anyone previously flagged at a talk page, etc.
this in some cases results in him finding a full bucket of issues that he does not wish to write a 'correction' for because it would require a complete article rewrite. as i am one of the active (and lazy) reviewers here i should normally reply to these requests but i lack the energy to do this as my current state for last few months results in my energy levels being extremely low, at around 40% of what an average person has got. cause is unknown.
yet i keep reviewing articles and keep ignoring these requests to retract articles because it is even harder to reply to those than to one full review in the first place (need to articulate why some of the points being nitpicked were utter bullshit, as i find that i do disagree with some of them).
this leads to me becoming more exhausted and circles back to me not participating properly
there is an ondoing discussion of it in this section (on the very same 'proposals' subpage of water cooler)

hope this helps you to get an idea of the situation at least qualitatively.

regards Gryllida 13:36, 11 January 2026 (UTC)Reply

Hi @Sheminghui.WU, i liked that thought, that catching AI hallucinations is time consuming and may lead to delays in publishing. Gryllida 13:25, 11 January 2026 (UTC)Reply

To make sure I understand the position being advanced here, I see the following claims and implications:

That, for an extended period[2], reviews have been intentionally conducted without full verification, despite policy requiring it.
That multiple proposals have been made to relax or redefine review expectations[3],[4] to align with this reduced, 5W-only level of scrutiny, but without community consensus.
That, following retractions triggered by full verification after publication, the proposed remedy is to restrict or ban AI use rather than to restore compliance with existing review standards.
That AI-assisted content is treated as uniquely suspect, while comparable human errors from incomplete review are framed as understandable due to fatigue.
That thorough verification during review is characterized as counterproductive because it increases post-publication workload, rather than being recognized as work that belongs in the review process.

I disagree with those priorities.

You previously articulated the risk clearly yourself, noting that becoming more liberal in approvals in order to increase output is “disastrous,”[5] and that pursuing readership through speed leads to biased, lower-quality journalism. That analysis remains sound and is manifest here in the number of resultant retractions.

Applying full verification after publication did create additional work, but only because required checks were knowingly deferred during review. The remedy is not to restrict tools, but to align review practice with policy.

The underlying issue is not that policy was followed, but that it was repeatedly and unilaterally dismissed.Michael.C.Wright (^Talk/_Reviewer) 15:42, 11 January 2026 (UTC)Reply

This is too long, but I will answer each point anyway.

That, for an extended period[4], reviews have been intentionally conducted without full verification, despite policy requiring it.
- Yes. I try to do full verification to a degree, but not to each letter. The motivation here is to verify key details and publish while article is still fresh.
That multiple proposals have been made to relax or redefine review expectations[5],[6] to align with this reduced, 5W-only level of scrutiny, but without community consensus.
- That discussion is still open.
That, following retractions triggered by full verification after publication, the proposed remedy is to restrict or ban AI use rather than to restore compliance with existing review standards.
- This is an overloaded sentence for me. For once, it's not only 'review' standards at the plate, it's also the 'write up' standards on the plate.
That AI-assisted content is treated as uniquely suspect, while comparable human errors from incomplete review are framed as understandable due to fatigue.
- I do not know where this came from; please try phrasing it again. I lack resource to attempt to read this more than once.
That thorough verification during review is characterized as counterproductive because it increases post-publication workload, rather than being recognized as work that belongs in the review process.
- I do not know where this came from; please try phrasing it again. I lack resource to attempt to read this more than once.
Pursuing readership through speed leads to biased, lower-quality journalism
- I do not care about readership. I do not know why you think I care of that.
"Applying full verification after publication did create additional work, but only because required checks were knowingly deferred during review. The remedy is not to restrict tools, but to align review practice with policy." is missing a 'In my view' note.
"The underlying issue is not that policy was followed, but that it was repeatedly and unilaterally dismissed."
- I do not agree with this statement. I do not think the current policy requires 100% coverage of fact check.

Hope this helps. Gryllida 15:57, 11 January 2026 (UTC)Reply

> I do not think the current policy requires 100% coverage of fact check.

It is explicit: "The reviewer should check that all information in the article is fully sourced..."

It's been a part of the policy since its inception, eighteen years ago.Michael.C.Wright (^Talk/_Reviewer) 16:04, 11 January 2026 (UTC)Reply

Thanks for that. Posted below. Gryllida 16:35, 11 January 2026 (UTC)Reply

Comment My understanding is that, for a fair amount of time:

There has been a lack of humans, writing articles, editing articles and reviewing them, making it difficult to publish even one article a week;
That for a fair amount of articles "written" with usage of AI, the quality problems were significant - not only in style of writing, but even in "facts" mentioned in the article;
That we currently have a tool to discourage blatant plagiarism from sources - which therefore encourages "lazy" (or tired, or whatever) contributors to use AI as text generator;
AI assisted content is not treated as uniquely suspect. Rather, where text is written by humans, humans are much less likely to generate hallucinations, and much more likely to follow guidelines for writing articles. AI assisted content doesn't know Wikinews guidelines, and even if it were provided with them, it would still be likely to hallucinate;

The proposals in this thread have included:

Temporarily banning AI usage for writing of text in Wikinews articles - which I tend to agree with, though I would prefer a constructive discussion on positive and negative examples of AI generated contributions to Wikinews;
Removing reviewer duties from people who have missed AI hallucinations in articles they have published - which I absolutely disagree with. We need more people in Wikinews, not less.

I would like to add the following proposals:

Possibly, add a tool to detect AI hallucinations - numbers (10 or ten or dozens - format doesn't matter), places, people which are not mentioned in the sources;
Or even better, write a Wikinews-specific (AI-powered? I don't like AI, but if you do, go ahead, contribute to development and testing) tool to generate text based on the provided links to sources;
Remember that having duties of being a reviewer doesn't deprive you of rights of being an editor. If an article needs improvement and editors are not showing up, feel free to improve it yourself - with caveat that you will have to find another reviewer to publish it afterwards.

Reviewers don't have to be perfect. Reviewers need to be there, active and helping. But they don't have to be perfect. They just have to do their best. And discouraging writers and reviewers, penalizing them for their mistakes, can kill the project. This isn't school where kids have to go, whether they want to or not. This is a project based on volunteers. So instead of wasting their time with these heated opinionated threads and further dividing the community, get out there and do what you can. /AI will not write Wikinews for us./ Wikiwide (talk) 07:48, 12 January 2026 (UTC)Reply

> Reviewers don't have to be perfect. Reviewers need to be there, active and helping. But they don't have to be perfect. They just have to do their best.

We hold contributors accountable under our policies and guidelines, and the same applies to reviewers. Publishing inaccurate, biased, or plagiarized content due to incomplete review should not be acceptable. Reviewers volunteer for the role with an understanding of its requirements. Unilaterally choosing to disregard core policy for convenience undermines that responsibility, creates more work for others, and damages the project's reputation.

If reviewers knowingly accept inaccuracies, bias, or plagiarism for reasons of convenience, it raises a fundamental question about the purpose of the reviewer role itself. The review function exists precisely to prevent those failures, not to normalize them.

Michael.C.Wright (^Talk/_Reviewer) 14:49, 12 January 2026 (UTC)Reply

> Removing reviewer duties from people who have missed AI hallucinations
That was not proposed. I forwarded a thought experiment by saying "Maybe [emphasis added] if we de-flagged reviewers after three AI slip-ups in published articles, we'd catch more of them sooner."
There used to be the notion of "easy-come, easy-go" for the reviewer bit. There is a very brief discussion here between Pi zero and Brian McNeil, two contributors who were instrumental in developing the modern review process. That discussion illustrates the level of attention to detail they hoped for from reviewers. Our current policies and guidelines still largely reflect that expectation.
My point is that this is not the way to go about policy change. What has occurred is one reviewer has unilaterally disregarded policy for personal convenience (as stated), and that has resulted in the publication of inaccuracies, bias, and/or plagiarism.
If WN:IAR is used as a justification (and it has not been, so far), even that requires the action to improve the project.
Reviewing as if a policy change has occurred, when it has not, predictably increases post-publication errors and workload. It then becomes circular to cite those downstream burdens as justification for lowering the standard that would have prevented them.Michael.C.Wright (^Talk/_Reviewer) 17:42, 12 January 2026 (UTC)Reply

I don't think there was disregard from one reviewer, that's why I opened RFC below. I believe that doing 100% verification was not common practice. Gryllida 21:04, 12 January 2026 (UTC)Reply

Do we have consensus? @Gryllida, @Michael.C.Wright BigKrow (talk) 21:58, 12 January 2026 (UTC)Reply

Not before a few other contributors weigh in. This may take a few weeks. Gryllida 12:58, 18 January 2026 (UTC)Reply

Support At this time. Let's see what the future holds, though. Maybe let's revisit this next year?--Bddpaux (talk) 16:19, 29 January 2026 (UTC)Reply
Support a ban on unsupervised use of genAI for article creation or expansion, Oppose a blanket ban. Part of why I oppose the "nuclear option" is because it's nearly impossible to definitively know whether a certain sentence was purely AI-generated or merely the signature of a more formal human writing. As long as it's used responsibly, genAI can be a very handy tool for writing breaking news articles, seeking feedback on the quality of an almost complete article, correcting the grammar, translating content, and other purposes. For example, when I wrote "Bangladeshi interim leader Muhammad Yunus returns home after four-day visit to London" and "Australia's truth commission finds genocide against Aboriginal Victorians, suggests practical remedies" (which was deleted due to staleness), I asked ChatGPT very simple prompts such as "Please assess that article" and "Is my article good?", and it gave me some rather useful advice (in retrospect, I admit that I should perhaps have disclosed it on the talk pages). Of course, it's a double-edged sword, and I discarded the recommendations which seemed very nitpicky or would run against P&Gs such as NPOV. I think a responsible (if imperfect) approach would be to give a hard "Not Ready" verdict on articles that look entirely AI-generated from the beginning to the end, grant a bit of leeway for those where only a specific section or paragraph is AI-generated, add one or more sections at Help:AI for guidance on rewriting these sorts of articles, and direct contributors to both that help page and WN:The use of AI in generating content. Redir editor (talk) 01:36, 18 February 2026 (UTC)Reply

Comment In light of a recent suggestion to use AI tools to evaluate or fact-check AI-generated material, and a recent declaration of AI use in generating a published article, it would be helpful to clarify the scope of this proposal.

If the use of AI in some capacity is considered acceptable by the original poster, then the concern being addressed here may be narrower than originally framed. If that is the case, should the proposal be revised or withdrawn so the discussion reflects that distinction more clearly? Michael.C.Wright (^Talk/_Reviewer) 02:36, 10 March 2026 (UTC)Reply

I agree that clarification is needed. AI can be used for many things while writing an article without actually creating the wikinews-authored article text itself. Tduk (talk) 15:28, 4 April 2026 (UTC)Reply

Consensus assessment and possible next steps

[edit]

My read of the consensus to this point is:

Support: 3
Oppose: 2
Support Retracted: 1
No stated position: 3

Support remains conditional and temporary in framing, not a blanket permanent ban.

With only one-vote separation and multiple participants explicitly calling for further discussion, this still does not meet a strong consensus threshold.

The thread shows more agreement on process deficiencies (review standards, workload) than on prohibition itself.

Based on that, and keeping in mind that consensus is not determined by simple vote totals but by working toward a solution that reasonably addresses the main concerns raised on all sides, I suggest the following be considered for further community discussion:

Transition WN:AI from a draft guideline into a proposed policy, so its requirements and enforcement expectations are clearer.
We carefully monitor AI usage to ensure it is 1. transparent, 2. accurate, 3. NPOV. See Category:AI usage declared as just one tool towards this end.
For a limited trial period (for example, three months), apply a standardized disclosure notice on newly published articles that involve AI assistance. (~~someone else~~ Bddpaux proposed this elsewhere[6]). See Template:AI disclosure as a proposed draft.
Reassess outcomes after the trial period to determine whether current practices are improving article quality and review reliability.

Comments welcome on whether this approach reasonably addresses the concerns raised so far.Michael.C.Wright (^Talk/_Reviewer) 15:05, 30 January 2026 (UTC)Reply

Agreed. Works for me. Gryllida (talk) 03:26, 10 March 2026 (UTC)Reply

I have tweaked WN:AI to reflect that it is a proposed policy rather than a proposed guideline.

If we are considering a reader-facing disclaimer placed directly on articles, as I understand this suggestion, I think it should first have clear and explicit consensus.

Please comment below indicating {{support}} or {{oppose}} for using Template:AI disclosure directly on articles for a three-month trial period. You may also provide specific suggestions regarding the reader-facing template. If opposing, please include a brief rationale.Michael.C.Wright (^Talk/_Reviewer) 14:58, 26 March 2026 (UTC)Reply

I did not understand what change is proposed. I am already following that linked guideline and the software is already prompting to declare AI usage at submission. What would "support" change here? Gryllida (talk) 10:07, 27 March 2026 (UTC)Reply

An interesting read at Slashdot, that may be informative to this discussion: Will 'AI-Assisted' Journalists Bring Errors and Retractions?. Original article is here.Michael.C.Wright (^Talk/_Reviewer) 11:00, 6 April 2026 (UTC)Reply

This conversation has been marked for the community's attention. Please remove the {{flag}} when the discussion is complete or no longer important.

The current verifiability policy states "The reviewer should check that all information in the article is fully sourced". In a discussion above, Michael.C.Wright noted that this requires checking each piece in the published article to the letter.

So for example the way I understood it from above discussion, publishing an article with note 'A bus crashed, it carried 45 customers.' with a lot of other newsy information such as 5Ws, H, a driver was drunk, driver survived with one leg injury, etc other details, would require reviewer checking everything; according to this interpretation, just checking everything else while ignoring that piece of detail about '45' (and hence risking it will be published with incorrect number) would be against the current policy. This is purposefully a bit exaggerated example to get my point across. From time to time Michael.C.Wright picked up some more serious issues in my reviews, for example one article I didn't run in copyvio check, and then he went and put a dozen of issues which weren't adequately verified or attributed or some other thing, that I am yet to look at when I am slightly less busy hopefully towards the end of this week, and proposed a retraction. That retraction discussion link is located in first sentence of the above proposal to ban AI usage at Wikinews.

I don't believe this is realistic to expect a reviewer to check 100% of the content in an article unless someone is hired to do such extensive reviews and another user is hired to implement the requested changes quickly (a total of 2 users hired). I believe that requiring full review will delay reviews as they are simply more time consuming for the reviewer; and this will also delay publishing because a long list of issues will be presented to an author, and an author will take longer to address them. I proposed earlier, for example, that only the 5W and H, and plagiarism, should be checked, and potentially NPOV next on the list, but I expect that checking article to the last letter of each sentence is not required. For example being satisfied that around 70% of the article was confirmed in the sources could be sufficient to publish it.

I am seeking comment from each reviewer about this. Please type it into below if you do not mind. Thanks. Gryllida 16:34, 11 January 2026 (UTC)Reply

This section is just for reviewers to say how they are doing reviews or how they think they should be done, whether they should be '100% verification coverage', etc. When posting this I wanted it just to see what each reviewer is doing currently about this. It is not intended for discussion just to see the current status. Then, further down the page, this section is then followed by another section for discussion. Gryllida 16:34, 11 January 2026 (UTC)Reply

Space for @Acagastya comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Bawolff comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Chaetodipus comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Cromium comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Heavy Water comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Gryllida comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @JJLiu112 comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @LivelyRatification comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Lofi Gurl comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Michael.C.Wright comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @RockerballAustralia comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @SVTCobra comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @ShakataGaNai comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Tom Morris comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @Tyrol5 comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply
Space for @William S. Saturn comment (click 'reply' next to my nick and add your answer below) Thanks Gryllida 16:34, 11 January 2026 (UTC)Reply

General discussion. Gryllida 16:34, 11 January 2026 (UTC)Reply

Asked42, not sure you are the best to do this but maybe you know someone who does: need to contact each reviewer off wiki asking to comment here and to advise when available for audio call as i think that is the only way to reconnect reviewers, by having a call at least once each three months. I scream in pain as this software does not have a calendar nor a doodle built into it and I am lost. Gryllida 15:42, 16 January 2026 (UTC)Reply

Noting I tried EmailUser and had no response. Gryllida 15:43, 16 January 2026 (UTC)Reply

General discussion point 2: recent change

[edit]

For a related discussion, see Clarification about post publish and post archive edits

Comment I feel that what Michael.C.Wright is doing, by cross checking the work of other reviewers at time of archival (or insert some other 'eternity later' description here), and by flagging all possible issues that were missed as part of review, undermines two types of relationships: author to reviewer, and reviewer to reviewer. While both of these could be blamed on the reviewer who approved the story that was later scrutinized, I do not find this approach particularly fruitful for the quality of news output nor for the integrity of the project as a whole. I am seeking comment from everyone about this matter. In my opinion lowering the tolerance to omissions, as long as it does not land into plagiarism or substantial bias territory, was previously allowed and the level of cross check in the last year has unnecessarily -- and without a consultation -- increased, putting unnecessary burden on involved authors and other reviewers. I request that this workflow comes back to what it was previously. Thanks for your attention to this sentiment, and I hope to see your comments about it soon. Gryllida 04:00, 9 February 2026 (UTC)Reply
cc @Asked42 @Koavf @Bddpaux, please reply now and tag 3 other users who weren't tagged previously, thanks. Gryllida 04:01, 9 February 2026 (UTC)Reply
There's quite a bit here and I'm not sure which three other users would even see this conversation who weren't already pinged. I can see some value in checking articles at the time of archiving to see if the process was followed optimally and maybe tagging the talk page somehow. This is more-or-less equivalent to how some proper news organizations operate, where they review their own published works and issue retractions, make corrections, etc. (it was only two days ago that I saw my feed for Mother Jones retract a story that they couldn't verify). So I agree in principle that it's worth us having an author-to-reviewer relationship that works at the time of publication and then if someone is motivated, an archiver-to-reviewer relationship. What that looks like may be an open question. —Justin (koavf)❤T☮C☺M☯ 08:59, 9 February 2026 (UTC)Reply
Hey @Koavf My question is whether you have noticed review standards got tougher than usual in the last year? Gryllida 16:28, 9 February 2026 (UTC)Reply
Good question, but that is not something that I have personally experienced. —Justin (koavf)❤T☮C☺M☯ 20:27, 9 February 2026 (UTC)Reply
Thanks. Gryllida 09:54, 10 February 2026 (UTC)Reply

Would you say there has been more discussions about review quality recently, than before? Gryllida 09:55, 10 February 2026 (UTC)Reply

Hello

context

some radio have news highlight every day from 8am to 815am, and
some youtubers have $whatever new upload every week on Tuesday at noon local time. For example.

As a result, Fans sweat sitting on top of their fingers in ancipation of the release on that schedule.

proposal.

Could it perhaps be helpful to identify platform and format and method, to get this done here?
- Weekly, say one entry on blyesky with list of recent news over last week; and one entry via audio podcast i.e. youtube or audible or whatever has audio outputs?

Thanks. Gryllida 11:16, 26 January 2026 (UTC)Reply

pinging @NikosLikomitros @Sheminghui.WU @Sj @Asked42 @Sbb1413 @Michael.C.Wright @Wikiwide @RockerballAustralia @Matthiasb @Vyacheslav84 @Pharos to name a few, as this output could include news from other languages. cc @Revi C., @Base Hoping to hear your input and aim for first date of a first initial post on Feb-01 or thereabouts. Thanks Gryllida 01:45, 27 January 2026 (UTC)Reply

I didn't understand what exactly is being proposed? --Vyacheslav84 (talk) 03:19, 28 January 2026 (UTC)Reply

Comment Not totally following the proposal/question, but I will say this: If an event happens and that event is newsworthy there should be some type of jetsam present that verifies it actually DID happen. Historically, an accredited reporter's notes are valid. e.g. "I watched this program/podcast/live stream on this date at this time and this thing happened then that thing happened' etc. etc. But, again, the EVENT must be newsworthy AND have solid sourcing. Because someone with the YouTube channel CoolDude47 said a thing happened at his neighbor's house, that SOURCE may not be dense enough to be trustworthy. Know what I mean?--Bddpaux (talk) 16:24, 29 January 2026 (UTC)Reply

If I understand Gryllida's proposal correctly, it is that we could (a) post a weekly news roundup to Bluesky or a similar site and/or (b) post a weekly audio news roundup/"best of Wikinews" to YouTube or a similar website. At this stage, though, I don't think we are in a good position to do either of those. I'm concerned that casual readers/listeners might not realize that Wikinews's topic selection is constrained by what our contributors take the time to write as well as what our reviewers take the time to review. We're averaging much less than one new published article per day at this point, resulting in a collection of articles that could be politely described at best as "eclectic". --Metropolitan90 (talk) 06:08, 30 January 2026 (UTC)Reply
Correct; there's always a chance to add one short disclaimer at the beginning of the text or audio piece noting that "<name of the nickname for the Wikinews broadcast> is in beta and audience can add their news by going <here>" where <here> is a user friendly place to start. Gryllida 19:51, 30 January 2026 (UTC)Reply

Hi @Vyacheslav84

I guess you are not familiar with the services similar to the ones i described in 'context' then. here is a wiki page about them: w:PBS News Hour. basically they air themselves every evening at same time.
I am proposing Wikinews produces at least one text output and at least one audio output regularly, for example, weekly, and posts them at the same time of day on a certain day of week consistently.

Hope this helps. Gryllida 19:48, 30 January 2026 (UTC)Reply

I agree, but who will do it? --Vyacheslav84 (talk) 14:43, 16 February 2026 (UTC)Reply

@Vyacheslav84 one person can not do all news media so i suggest to have one person per one or two social media accounts. I created page Wikinews:Social media posters you can add your name there and name of a social media where you are going to post and under what nick. Important not to have 'Wikinews' in the account name there as this is a trademark and requires permissions from WMF. Gryllida 23:49, 16 February 2026 (UTC)Reply

@Gryllida Okay, thanks, I'll take a look! --Vyacheslav84 (talk) 11:16, 22 February 2026 (UTC)Reply

Comment I mean -- it sounds cool, but: getting it done...... Know what I mean?--Bddpaux (talk) 16:13, 4 February 2026 (UTC)Reply

Hi Bddpaux please see Wikinews:Social media posters, if you have a friend with social media addiction, please point them to the page, thanks! Gryllida 00:18, 17 February 2026 (UTC)Reply

Comment It's been 14 days since we had a new article published. If our text and audio outputs were based on our new content over the past week (and they should be if we were to undertake those projects), that would mean we would have had to skip two weeks. With the risk that the Wikimedia Foundation might shut down Wikinews altogether, I think we need to focus on our primary mission of publishing news stories on this website rather than trying to take on projects in new formats. Keep in mind that reviewers/administrators would presumably have to take the time to review the text and audio outputs before they are released, and there aren't enough of them to even have the time to review all the articles that get submitted now. --Metropolitan90 (talk) 07:52, 5 February 2026 (UTC)Reply

14 days??? Why. Gryllida 07:12, 6 February 2026 (UTC)Reply

🫩 Gryllida 07:13, 6 February 2026 (UTC)Reply

Concerns over the inability to regularly put out updates like this are well-founded, but I don't think a monthly update would be too much to ask for, and I believe it would put a sense of relevance back into WN. As said by others, the articles we currently publish are mainly based on contributors' interests and availability, which isn't always the most attractive to a general audience. Even 1-sentence updates or a collection of quotes from media would be better than nothing; there's so much going on in the world right now that there's always something readers urgently need to know, and I see it as a foundational step for building, maintaining and retaining an active readership. I would gladly take up this task, but of course it would be better if reviewer(s) are responsible so that the reviewing process can be skipped/streamlined. Not sure about the audio part, unless software/AI is used to turn a transcript into audio. I personally watch Last Week Tonight with John Oliver and Weekend Update, so I believe in the attractiveness of such updates. HKLionel TALK 12:10, 12 February 2026 (UTC)Reply

Hi @HKLionel

I think submissions sized 100-150 words are the easiest to review, even if they do not possess much coverage value. That's the easiest way to get through the reviewing process. NOT MORE because it is pain to review. And not less because that is too short for pass.
There are three high quality submissions in queue. Each of them would take me a week to 'fully review'. Presumably, if I attempt one of them tonight, that will be an express review...

Regards, - Gryllida 04:59, 17 February 2026 (UTC)Reply

Yes, so I suppose a coordinated effort by reviewers/making it the responsibility of a reviewer would make it more efficient. HKLionel TALK 16:23, 17 February 2026 (UTC)Reply

I just had the thought that we could borrow from Wikipedia's in the news and current events portal. HKLionel TALK 18:12, 17 February 2026 (UTC)Reply

Borrow what? Gryllida 21:35, 22 February 2026 (UTC)Reply

Content for, as I suggested above, a monthly update. HKLionel TALK 01:28, 25 February 2026 (UTC)Reply

I also discovered ja:ウィキニュース:短信 (news bulletins). This is a format I'd like to see emulated. HKLionel TALK 02:39, 25 February 2026 (UTC)Reply

Propose to encourage a user to author and publish, or reviewer to review and publish, at least one story every two weeks. Like a w:streak.

At end of a competition period, suggest for 6 months (12 fortnights), the no. of maximum consecutive fortnights would be computed for each participant. For example, if I started and was major contributor to article on week 1, 3, 7, 9, 11, then my streaks are '1 and 3' (streak of 2), then '7 and 9 and 11' (streak of 3), I would score '3'.

Top 10 participant would get a T-shirt or mug or some other object with a number (i.e. 'This user archived a 6 fortnight streak in English Wikinews in 2026' worded fancily and with fancy picture.) delivered to their desired postal address.

Not sure how to fund this, but I think there may be a few people who want to donate to that purpose. if you wanted to donate please specify payment method. cc Pharos as I think you might be able to get support from NYC for this perhaps.

thanks Gryllida 07:02, 20 February 2026 (UTC)Reply

created page Wikinews:2026 streak competition thanks Gryllida 21:12, 21 February 2026 (UTC)Reply

I think I could attempt this. Cheat Codes 4ever (sandbox | CentralAuth) 20:41, 22 February 2026 (UTC)Reply

Hi @CheatCodes4ever thank you, nice. 🙂 Is 'every 2 weeks' a good time frame? Or better 1 or 4 weeks? Whre do you think it could be funded from? Also, how long to run it for: 12 weeks or 24 weeks? Regards, Gryllida 20:59, 22 February 2026 (UTC)Reply

@CheatCodes4ever please add your name on that section section 'interested users signup' Thanks Gryllida 23:25, 22 February 2026 (UTC)Reply

@Gryllida: Two weeks sounds reasonable to me. Cheat Codes 4ever (sandbox | CentralAuth) 20:36, 23 February 2026 (UTC)Reply

great thanks Gryllida 20:42, 23 February 2026 (UTC)Reply

The prizes would probably need to be worked out later, but this is a great idea and I'm interested. HKLionel TALK 01:47, 25 February 2026 (UTC)Reply

https://www.mediawiki.org/wiki/Extension:InlineComments

Could this be installed here? It would be easuer to use than the current system, where highlight is a pain to delete on visual editor, and comment is hidden on talk tab. Gryllida 10:11, 23 February 2026 (UTC)Reply

I find it unlikely that devs will add it, considering the status of Wikinews. I support you making a ticket at phab: if you feel like it's warranted. —Justin (koavf)❤T☮C☺M☯ 17:23, 23 February 2026 (UTC)Reply

That would require consensus from the community. Codename Noreste (talk) 20:32, 23 February 2026 (UTC)Reply

thatas what i am asking about. would you support installation of such an extension here? Gryllida 20:43, 23 February 2026 (UTC)Reply

Which is why I was trying to provide my vote of approval. I did make that unclear, so I apologize: while I don't know that this will be successfully implemented, I am in favor of you requesting it pending others weighing in and a general consensus that others feel the same way. Again, sorry for suggesting that individuals unilaterally request new features. —Justin (koavf)❤T☮C☺M☯ 21:39, 23 February 2026 (UTC)Reply

yeah i got it Gryllida 22:03, 23 February 2026 (UTC)Reply

lets see if using enwp lever works. Gryllida 20:34, 23 February 2026 (UTC)Reply

They are being quite positive, including PHP maintainership, and suggest development of a gadget as prototype. Regards, -- Gryllida (talk) 02:59, 1 March 2026 (UTC)Reply

Same as Justin, though the WP discussions don't seem promising. HKLionel TALK 01:53, 25 February 2026 (UTC)Reply

Just to mention, it is not exactly like inline comments, but EzPR 2025 will include a feature called "Dynamic Marking" under the Sentence Marking section. When enabled, it will use a global gadget to load and display specific sentences that have been highlighted and marked by the reviewer on the article.

This is not fully implemented yet, the global script integration still needs to be completed. However, the idea is to provide a more dynamic approach to sentence-specific highlights and unsourced sentence marking.The data for each marked sentence, including its associated issue, will be stored on a JSON page. -- Asked42 (talk) 18:21, 28 February 2026 (UTC)Reply

Hi Asked42

I believe the dynamically marked comments, as currently implemented in EZPR2, have two disadvantages compared with the extension. First, they can be difficult to remove when working in the visual editor. Second, the extension provides a more user-friendly interface for replying to comments about a specific phrase in the article without having to switch to the talk page and back.
That said, one possible improvement to the extension would be to post a copy of the inline comment discussion to the article’s talk page when the comment is removed. This could be offered as an optional feature in the extension. At present, the discussion remains only in the article’s edit history, which may not be ideal, as the talk page is generally expected to serve as the central record of discussions about previous revisions of the article.

Regards, -- Gryllida (talk) 22:53, 28 February 2026 (UTC)Reply

@Gryllida: I was not aware that there was interest in developing the extension as a gadget, which is, in my opinion, a much more practical and better approach. When the gadget is ready, I think we could also integrate it with EzPR 2025 to allow marking and commenting on specific text within an article. Therefore, I am dropping the Dynamic Marking approach within EzPR.

Just to clarify regarding your comment: "First, they can be difficult to remove when working in the visual editor." - I believe you may be referring to the current approach of wrapping text with templates. Dynamic Marking was not designed in that way; it was intended to function more like inline comments. Additionally, EN wikinews currently does not have the VisualEditor enabled. So even if that concern were valid, the current approach would not cause issues in this specific context. -- Asked42 (talk) 05:56, 1 March 2026 (UTC)Reply

Actually, in Wikipedia there is a dying backlog queue of drafts for review, to create a new page in Wikipedia, users may often have to wait 2-6 months while page is sitting in Draft: namespace. So, I wanted to suggest the gadget separate, then it can be added to ENWP for draft review. For this space here in Wikinews, having the gadget installed there at ENWP could be a part of leverage to get the extension installed in Wikimedia wikis, including here, eventually. Knowing how ENWP centric the ecosystem is. Gryllida (talk) 07:33, 1 March 2026 (UTC)Reply

Btw, inline comment doesn't show which phrase it was referring to, and neither does an inline citation. Only one cursor position. And inline comments were such a pain to edit in VE... For once they are not even visible until clicked or something. :( Gryllida (talk) 07:34, 1 March 2026 (UTC)Reply

Our Style guide states "Do not over-wikify articles. Link only particularly relevant background material."

I believe the reasoning can be found in this essay, which states "Don't use wikilinks for basic information: a news article should stand on its own, understandable without consulting some other page along the way."

As a community, I believe we have strayed from this practice considerably. A drift towards more wikilinking, in my opinion, is part of our broader drift from being a journalism project first to a wiki project first and a journalism project second.

While over-wikilinking has not been an impediment to publication (that I'm aware of), it is something we should either be actively addressing (reducing) or something we should remove from our policies and guidelines.

Please comment below how you feel about over-wikilinking and if you feel we should keep, remove, or update the recommendation. Justifying your response would be most helpful, but a simple !vote might suffice. Michael.C.Wright (^Talk/_Reviewer) 15:27, 27 February 2026 (UTC)Reply

Some useful polling templates might be:

Keep especially for the lede. A link away from Wikinews is an invitation to stop reading our content and start reading someone else's. I also agree that the quality of our articles should be high enough that readers can fully understand the topic without going elsewhere for additional information.

Michael.C.Wright (^Talk/_Reviewer) 15:27, 27 February 2026 (UTC)Reply

Weak keep I am unsure if I have a passionate opinion here, but I will agree that the wikilinking has gotten a bit out of control. In the very least: most of our wikilinking should simply point to a new place at WN. This isn't Wikipedia. And: I 2000% agree that we've lost our 'scope' of journalism first, wiki second. That is not even mentioning the 3,000 people involved with WMF that can only view us as a wiki ONLY. That has been a never ending war. Even some of us here: Admins etc. etc. -- we (including myself) sometimes get lost in the weeds on what is happening at this place: get the news, write the news, submit the news.--Bddpaux (talk) 16:32, 27 February 2026 (UTC)Reply
Just commenting what I see in this...
- Obvious words like "cat" or "dog" are not supposed to be wikilinked. Not to point of wikilinking every second word in a page.
- More complex names (for me it includes name of any US state, for instance, i never knew where they are and am not going to try to build a US map in my head, and many readers are overseas) that are not understood by foreigner or by someone new to that topic, could and should be wikilinked.
- Adding a brief note to a thing is good. At some point I added a note like ", a state in east coast of United States," for something, for that reason.
- Sport articles, for example, suffer a lot from lack of explanations. :S
- Also this approach gave advantage that items that are wikilinked do, eventually in an utopic world when news writing is an everyday thing, become wikilinks to local categories about that event series or place or topic in news context, with a feed for that. For example there is no category "Dog". Or court. Or something.
- If I removed a wikilink, I should probably feel compelled to either add a brief note or add a local news-feed (category) link.
- A reviewer can do such minor chanhe if needed.
- And I personally do not see any reason to change that policy in its current version.
Gryllida (talk) 05:00, 28 February 2026 (UTC)Reply
Weak delete Admittedly, having too many wikilinks looks bad. But we should not discourage Wikinews writers from including useful links to Wikipedia. Wikipedia is our sister project, not a competitor. --Metropolitan90 (talk) 05:31, 28 February 2026 (UTC)Reply

comment - There are two quotes that were mentioned, the style guide, which says ""Do not over-wikify articles. Link only particularly relevant background material."", and an under construction essay called "Categories", abandoned in 2021, which for reasons I can't quite figure out mentions wikilinks. When being asked "Please comment below how you feel about over-wikilinking and if you feel we should keep, remove, or update the recommendation.", it's not entirely clear what "the recommendation" is. Maybe I seem like I'm being obtuse but I'm having trouble interpreting all the !votes and figuring out where I stand on an issue presented as a simply yes/no question but that lacks the needed clarity. If just a comment is being sought, I can provide that of course. Tduk (talk) 03:47, 10 March 2026 (UTC)Reply

For requests regarding permission, removal of permission, or accreditation, I suggest developing a tool that notifies all eligible users and includes a link where they can provide comments, ask questions, and cast their vote; and with a note how they can opt out of receiving such messages (if desired). I believe similar tools have been used on other wikis, such as Commons or Meta. Since this wiki is relatively small and some recent requests have received very few votes for an extended period, such a tool could help increase participation and prevent discussions from remaining open indefinitely. Gryllida (talk) 02:40, 1 March 2026 (UTC)Reply

At Wikinews:Writing an article#Starting out, it says:

If the story has not been started, you can begin it by entering an appropriate title below and clicking "Create article".

Here are Wikinews article title rules and guidelines.

I suggest that we add something like this to the above: "The article title should identify the news event that the article is about. A person's name alone, or a company name alone, should not be an article title."

I am assuming that some people are indeed creating articles by starting from Wikinews:Writing an article but not reading the linked article title rules and guidelines (this assumption might be incorrect). That seems to be how we get such soon-to-be-deleted drafts as Rumi Von-Clarke, Daud Muhammed, and Zakariah gmb services, all of which were tagged for speedy deletion on their second edits. -- Metropolitan90 (talk) 14:13, 6 March 2026 (UTC)Reply

I support this. That seems like a reasonable assumption. This might pull in a few contributors who get discouraged at first. I don't really follow them; what is the typical path a new editor who creates one of those articles takes? Do they stop editing? Silently? Tduk (talk) 14:39, 6 March 2026 (UTC)Reply

I'm cool with that. I just hope new reporters read it. The event is the thing.--Bddpaux (talk) 14:52, 6 March 2026 (UTC)Reply

Works for me. This could be a draft title, and when the user is ready to submit it for review, they can choose a better, more representative title if it is not already set. -- Asked42 (talk) 14:57, 6 March 2026 (UTC)Reply

I'll try to be quick here: We are a small village. There is a kind of poison that works its way in here at times -- people can (and do) begin treating this place as a general wiki, after they've been here for about 6 months.
I am writing these words as encouragement, not to scold -- please note that. I am going to (for the remainder of 2026) focus very much on Reviewing and encouraging new reporters. This is a news site. We have policies and guidelines. I will try very hard to adhere to those -- and I ask all other people here to focus on that very thing. Good things ahead. Bddpaux (talk) 15:19, 6 March 2026 (UTC)Reply

Hello, a new tool is now available that enables reviewers to attach inline comments to specific phrases in a draft article. This means phrases can be marked by a reviewer to highlight issues without actually editing the article.

Phrases that have been marked will appear highlighted in yellow. Non-reviewers can view them, but they will not be able to edit or remove the attached comments or markings. However, users can mark a specific highlight as "resolved".

This gadget allows users to quickly identify the exact text in an article that needs modification or changes. Reviewers using the EzPR2025 gadget for review will find the "Dynamic Marking" feature integrated into it. However, the tool can also be used independently of EzPR2025.

Full documentation is available at: WN:EzPR2025/Dynamic Marking. -- Asked42 (talk) 08:07, 14 March 2026 (UTC)Reply

there are currently nibe articles in review que going stale. I cannot review any of them because I either wrote them or they're original research. Can someone please get going on a review? This site must publish one article per day to remain a viable destination. Also, I am not going to be able to volunteer any more time writing articles here if everything I wrote this week goes stale for no good reason. Lofi Gurl (talk) 10:41, 20 March 2026 (UTC)Reply

For me, the deeper issue is a difference in what level of quality we are trying to achieve. Our Policies and Guidelines were written by a generation aiming for a much higher standard than what the current generation, broadly speaking, appears willing or able to sustain. The way we are collectively navigating that gap is creating friction (to put it politely) and, in practice, a status quo of low output.

We need to resolve that tension. Either we reform our PaGs to reflect a more achievable standard, or we raise the quality of our work to meet them.

Given the ongoing friction, I find myself looking for other ways to contribute here, if at all. I know at least one other reviewer has expressed a similar view.

I believe this quality–output tension is a central factor behind the SPTF’s recommendation to close Wikinews. I understand the logic behind that recommendation: we have not yet demonstrated that a citizen journalism model can reliably and consistently sustain high output, high quality, and a meaningful wiki process all at the same time.Michael.C.Wright (^Talk/_Reviewer) 16:04, 20 March 2026 (UTC)Reply

If there is a problem regarding the quality of my contributions to this project, then nobody has said anything about that to me whatsoever. I submit my articles in the best shape possible and I still feel like I have to make a scene just to get someone to even look at my submissions. This is starting to look like what I assumed was going to happen if I started reviewing -- me being expected to pick up the entire review workload -- because literally nothing will get published orherwise -- while my own submissions are just forgotten and left to rot in the review que, and hours of my personal time are just wasted. I believe this project has potential but I dont know how to proceed if nothing else is being done to fix the problem that multiple users have been bitching about for months. This project will be shut down if the status quo continues. Lofi Gurl (talk) 23:41, 20 March 2026 (UTC)Reply

There was no specific target in my earlier comment. I was speaking more generally about what feels like a broader tension on the project.

To clarify: it seems we may not all be applying the same review standard in practice. I’m not referring to you or anyone else specifically. There appears to be some preference, broadly speaking, for getting articles published once they are “good enough,” even if not every statement has been explicitly checked against a source or clearly attributed, with the expectation that the wiki process can refine things later.

On the other hand, our existing policies and guidelines lean toward a stricter approach, where each statement is expected to be supported or removed before publication. That gap — between what the standards say and what is realistically being applied — is where I see a lot of the friction coming from.

I may be misreading that, but that’s the issue I was trying to point to.

On the concern about you being expected to carry the review workload, I’m not sure the recent activity reflects that. Publishing has been shared among a small group of reviewers, and you’ve been a significant part of that (thank you!), but not the sole contributor.[7]Michael.C.Wright (^Talk/_Reviewer) 01:12, 21 March 2026 (UTC)Reply

I understand enforcing the rigorous policies for quality control, I just feel there have been too many times where a reviewer has requested revisions after an article I authored has been submitted, the requested changes are promptly made, but the reviewer just goes radio silent immediately afterward and the articles just goes stale waiting for review. It's hard to really feel motivated to make revisions to or gatwick articles I write because they have mostly just been getting left to rot these past few months. I'm not sure the community understands how much human labor power is actually just being wasted due to this failed process. Lofi Gurl (talk) 01:58, 21 March 2026 (UTC)Reply

i published one, others to follow during the week if i manage Gryllida (talk) 13:03, 22 March 2026 (UTC)Reply

it was noted elsewhere that 'a few' reviewers are required for the site to work. like i woukd say around 2-3 is a bare minimum 💀 when only 1 or 2, then, very hard to get things moving. I will also try to write one during the week for you to review! Gryllida (talk) 13:05, 22 March 2026 (UTC)Reply

thanks! Lofi Gurl (talk) 13:11, 22 March 2026 (UTC)Reply

hi @Lofi Gurl what is the issue with reviewing OR? Do you have scoop access? Gryllida (talk) 13:03, 22 March 2026 (UTC)Reply

I dont know how to verify sources with original research. How am I supposed to review an event or interview I did not attend? Lofi Gurl (talk) 13:08, 22 March 2026 (UTC)Reply

you can cross check it against reporters notes. They are shared privately with scoop, an address that accredited reporters and reviewers can read. If you do not have a wn-reporters account, please email me with your desired username, and i will email you the login instructions. Gryllida (talk) 14:32, 22 March 2026 (UTC)Reply

If it helps any, I have two stories currently waiting that are OR. Musique Libre Femmes plays for Women's History Month at Tompkins Square Library in New York City requires scoop access but I tried deliberately to make The Performance Space in New York City holds its annual gala honoring three artists verifiable using only the pictures I provided in the article and a few of the pre-event sources. This is even easier now, as since I wrote the material (immediately after it happened), some genuine sources have come out which also corroborate my OR (which I added).. So that one shouldn't be too difficult. I also made it deliberately short so we could publish it fast... Tduk (talk) 20:35, 22 March 2026 (UTC)Reply

Are we able to publish OR ten days after the focal point event? Lofi Gurl (talk) 20:45, 23 March 2026 (UTC)Reply

Generally, yes.Michael.C.Wright (^Talk/_Reviewer) 01:31, 24 March 2026 (UTC)Reply

How long do we have? Lofi Gurl (talk) 01:40, 24 March 2026 (UTC)Reply

Unfortunately, this is one of the many areas where there is no hard-and-fast rule. WN:Fresh merely states:

Exclusive content has the potential to extend our freshness horizon by days or even weeks, depending on the nature of the original material.

Somewhere there is a relatively recent discussion that involved MeDaWikipedian I believe, in which some consensus formed around eight weeks as a max limit. I'll find it but it might take some time.Michael.C.Wright (^Talk/_Reviewer) 01:50, 24 March 2026 (UTC)Reply

Not much of a consensus for 8 weeks, but it may still be useful to cite as recent precedent when defending publication of material up to that age: Wikinews:Water cooler/assistance/archives/2024/September#Photo exhibition “Ethnic features of the world people” took place in the Crimean capitalMichael.C.Wright (^Talk/_Reviewer) 02:14, 24 March 2026 (UTC)Reply

If you'd like some advice on it, let me know what article you are talking about and what you're thinking as far as why it is still publishable. If you'd rather do it quietly, feel free to use wikimail or email.Michael.C.Wright (^Talk/_Reviewer) 02:17, 24 March 2026 (UTC)Reply

If an article was submitted for review on day N, then on day N+8 could it automatically mark itself as stale? This could help clear out the queue from the old end. I picked number 8 because usually an article does not get published about events over seven days old; I may be deluded there so please adjust the number of days as needed.

Alternatively, the "review" template could change to "abandoned" after 7 days.

I am suggesting to get this decided and implemented some time soon. It would reduce the load on reviewers and sysops, in my opinion. Something similar was done for another template before, but I do not remember which.

-- Gryllida (talk) 08:59, 24 March 2026 (UTC)Reply

I propose increasing visibility and collaboration around drafts and their talk pages so more users can contribute to review and copyediting. That could help authors who aren’t keeping up with reviewer requests and would be more like a scientific journal process where at least two people check a page before publication. Since the software supports pre-reviews, we could encourage everyone to pre-review more broadly.

Proposed actions:

1. Add a friendly site notice inviting users to join a “newsroom” and review at least one article per day. Link to a Newsroom guide explaining categories:

Drafts = need pre-review, expand, submit for review

Queue = possibly ready; check carefully and flag issues

Disputed = issues found; fix and resubmit

That page like Wikinews:Newsroom but a mini-version: with only one paragraph intro linking to key policies; three lists (draft, review, disputed) one paragraph before each; one list (published in last three days) also one para explaining process of improving or cross checking those. That could be more user friendly for newbies than the current "everything and a kitchen sink" current version.

2. On submission, automatically leave the user a message with a request to pre-review at least two pages from the Newsroom, again linking to its over simplified version, perhaps even transcluding it to the user's talk page (via subst, so only then-current version is included in the message) as then they do not need to open the Newsroom link to see the topics being written about in the new drafts.

These steps should speed feedback to authors and surface more issues before publishing, and bring more hands to fix them.

Hope this is helpful. Gryllida (talk) 12:10, 27 March 2026 (UTC)Reply

3. At submission, present user a button "pre-review my own article" which is like pre-review process they do for article they just submitted.

2a. Example: User:Gryllida/MiniNewsroom

--Gryllida (talk) 12:31, 27 March 2026 (UTC)Reply

4. Add button "mark as stale" to the develop template. Gryllida (talk) 12:32, 27 March 2026 (UTC)Reply

Hello everyone, to help increase the number of active reviewers on English Wikinews and support interested contributors in understanding the review process, I would like to propose organizing a structured, documentation-based online workshop. Given that live video sessions may not be feasible for many participants, this workshop would primarily be text-based and hosted on-wiki (with optional use of chat platforms for discussion). However, if any experienced reviewer is interested in contributing via a presentation or live session, that would certainly be welcome.

The workshop would be divided into 4-5 sections, each focusing on a specific aspect of the reviewing process. These may include:

Introduction to the role and responsibilities of a reviewer
Core review criteria (newsworthiness, neutrality, verifiability, freshness, copyvio)
Source verification and common issues in articles
The review workflow (from submission to publication)
Post-publication review and maintenance

Each section would include short explanatory materials (text, slides, or diagrams), followed by practical tasks. These tasks could include answering questions, identifying issues in sample articles, or performing a simulated review.

Participants who complete the workshop should gain a clearer understanding of how reviewing works in practice. Rather than replacing the existing reviewer nomination process, the workshop could serve as a preparation pathway, helping participants build the skills and confidence needed before being nominated. Additionally, participants who demonstrate strong understanding during the workshop could be encouraged or mentored further as potential reviewer candidates.

To keep the workshop effective, participation may be limited to a small group. We would likely need at least 2-3 experienced reviewers to help guide discussions, review responses, and provide feedback. I would be happy to help prepare materials such as slides, flowcharts, and example articles.

I would appreciate the community's thoughts on this idea: Would such a workshop be useful? Would any reviewers be interested in helping organize or mentor participants? Are there improvements or alternative approaches we should consider? Thank you. -- Asked42 (talk) 15:12, 29 March 2026 (UTC)Reply

Support I would've started reviewing much sooner of there was something like this in place. Why there doesn't already exist something similar to this is beyond me.

Lofi Gurl (talk) 16:05, 29 March 2026 (UTC)Reply

Since the site will be frozen after May 4, I would like to suggest that we plan to have the top headline on the last day be something along the lines of "Wikinews closes after 21 years; community moves to Miraheze" or something like that. If the top headline is going to be permanently in place after that, it ought to be something that commemorates the ending of Wikinews. I wouldn't want the headline that the site has to live with permanently to be something like "Small town car accident results in nearly $200 in damage, no injuries".

Of course, this presumes that we can have an article commemorating the site's closure ready and approved on May 3, but it will also require the cooperation of reviewers not to put something else in the lead headline position at the last minute. -- Metropolitan90 (talk) 14:13, 30 March 2026 (UTC)Reply

Support: I think this could be some type of original report. Even if a full article is not possible, I do think we should leave a message on the main page. -- Asked42 (talk) 14:17, 30 March 2026 (UTC)Reply

Support I proposed the same elsewhere (after you did). —Justin (koavf)❤T☮C☺M☯ 00:21, 31 March 2026 (UTC)Reply

Wikinews closes after 21 years. —Justin (koavf)❤T☮C☺M☯ 00:24, 31 March 2026 (UTC)Reply

Support But: it will a long time before I come back here. Not being ugly about it. Maybe Michael can publish it. I'm about to sign out.--Bddpaux (talk) 02:07, 2 April 2026 (UTC)Reply

Neutral Don't care. Lofi Gurl (talk) 02:24, 2 April 2026 (UTC)Reply

Not sure if we 100% are moving to other host, though we should prep the article already, and post by May 3 -- should there be any changes, we do it before the project is RO (read-only). If someone is up for writing it -- I shall be there to review it. PS: please come to IRC, if you all can.
•–• 09:37, 2 April 2026 (UTC)Reply

I think the article would greatly benefit from someone from the previous generation of contributors to write about how the modern review process came about, the various journalism schools/classes we worked with at the time, etc. That time was our peak in regards to readership and activity and is therefore the most consequential for us as a project.

I know @Bddpaux has links to external sites which praised the work of en.WN during those periods. Those could/should be used as sources, quoted, etc.

See: Wikinews closes after 21 yearsMichael.C.Wright (^Talk/_Reviewer) 06:48, 4 April 2026 (UTC)Reply

As per this comment by @Sj, Wikimedia NYC and task force members are discussing the possible stewardship of Wikinews by meta:Wikimedia New York City. Also ping: @Pharos -- Asked42 (talk) 14:56, 1 April 2026 (UTC)Reply

Hello, as resources are required for migration, I am proposing to set a date after which articles will not be published. Maybe around mid-April. Then, a week later, clear the review queue and finish any maintenance tasks and walk away. Then this would leave around two weeks to setup a new site. Otherwise, I am concerned migration will be under-resourced. cc @Lofi Gurl who quite rightfully in my opinion wrote that. Regards, -- Gryllida (talk) 08:06, 5 April 2026 (UTC)Reply

If there is time to get new articles published during the second half of April, I think it should be allowed. The migration may be a more important priority, but not all the admins will necessarily be involved with that nor will it necessarily take up all their time. We are still getting submissions, albeit not many good ones. --Metropolitan90 (talk) 17:23, 7 April 2026 (UTC)Reply
Oppose I agree with Metropolitan90 above. We could continue developing and publishing articles that reflect what we aimed to achieve: accurate, neutral reporting, provided reviewers are willing to perform full and complete reviews. More psuedo-reviews would be counterproductive.

That said, it is already April 13 and there have been no substantive updates on migration, use of the name or logo, or related decisions. If the community waits for those before preparing for read-only, there will likely not be enough time (or community will) to complete the work. For that reason, some of us have already started shutting the lights down. See the following discussions:

Currently I am prioritizing admin/maintenance/shut-down work but am also willing to do reviewer work if no other reviewers are willing. I am also prepared to review and publish the last article.Michael.C.Wright (^Talk/_Reviewer) 19:03, 13 April 2026 (UTC)Reply

Oppose - I'm not sure what this would change in practice anyway. Tduk (talk) 02:50, 14 April 2026 (UTC)Reply

I stopped reviewing anyway, as I cannot allocate resources for the "shut lights down" activities AND for reviewing AND for migration. For any non-reviewing tasks, I am happy to help. Gryllida (talk) 11:15, 15 April 2026 (UTC)Reply

I would like to unblock User talk:Dexbot, to be used by @Ladsgroup to complete our DPL conversion to static lists prior to pending closure of the project.

The bot was blocked almost twelve years ago by Pi for being operated "without authorization in violation of project policy, and operated with an effective bot flag without local control as well as without authorization."[8]

Ladsgroup has used the bot already on other language projects for this exact use; to remove DPL prior to Wikinews being closed. See phab:T421796#11851688.

There are just shy of 3,500 pages remaining with DPL. I believe a vast majority are the Wikinews:Archives.

I propose the following pages and their sub-pages and transcluded templates not be touched by the bot (most are protected any way). These should be manually converted at the last minute (target: April 30) for maximum usage before closure:

I do not believe admin rights will be necessary for the bot (and it doesn't look like they were utilized on other language projects), as the archives appear to be protected only to auto-confirmed, if at all. Flood might be needed, and if a local bureaucrat can't assign that and it's needed, I will request that a Steward assign that flag.

I will make myself available to change protection on a page-by-page basis, as needed to reduce our DPL usage.

This should be a community decision, so all input is welcome.

Also pinging all admin: @Acagastya, Bawolff, Bddpaux, Chaetodipus, Cromium, Gryllida, Koavf, Leaderboard, Leaderbot, LivelyRatification, Michael.C.Wright, Microchip08, RockerballAustralia, SVTCobra, Tom Morris, Tyrol5, William S. Saturn:

Are there any objections to me unblocking this bot and allowing it to perform the action of replacing DPL with static lists? Michael.C.Wright (^Talk/_Reviewer) 16:21, 23 April 2026 (UTC)Reply

@Michael.C.Wright I don't think there's any reason to keep a 12-year-old block, and on that basis alone I've unblocked the bot. A local bureaucrat should be able to assign flood (or bot). Leaderboard (talk) 16:24, 23 April 2026 (UTC)Reply

Great. Thanks. —Justin (koavf)❤T☮C☺M☯ 16:27, 23 April 2026 (UTC)Reply

So we're busy cleaning up Russian Wikinews, and I noticed that our total number of pages is about to drop below what you guys currently have. This was surprising, so I checked and apparently there are 2978k pages, and 2765k of them are user talk pages created by User:Wikinews Welcome. And it keeps creating them. I was shocked to learn that English Wikinews sends a welcome message to every user ever.

Now I'm not part of the local community, but I do think that it would be in the interest of Wikinews as a whole (even in read-only state) to trim the fat as much as possible. Thus my suggestions are 1) that meta:Welcoming policy be adopted and 2) that all pages which don't follow this policy be deleted (that is, if a user didn't create their account here, never made an edit here, and their talk page was never used by anyone else, it's up for deletion). In my estimation, about 100k user talk pages will remain if this is done.

I imagine that the technical staff (like @Ladsgroup) would be interested in having this go through. AtUkr (talk) 18:33, 24 April 2026 (UTC)Reply

Support This is one of the lowest priorities I can imagine, but if someone can do this in an automated fashion, sure. —Justin (koavf)❤T☮C☺M☯ 18:54, 24 April 2026 (UTC)Reply

Wikinews:Water cooler/proposals - Wikinews, the free news source

Protecting templates

Options how to protect

Consensus assessment and possible next steps

General discussion point 2: recent change