Wikipedia:Village pump (idea lab)

From Wikipedia, the free encyclopedia
Jump to navigationJump to search
 Policy Technical Proposals Idea lab Miscellaneous 
The idea lab section of the village pump is a place where new ideas or suggestions on general Wikipedia issues can be incubated, for later submission for consensus discussion at Village pump (proposals). Try to be creative and positive when commenting on ideas.
Before creating a new section, please note:

Before commenting, note:

  • This page is not for consensus polling. Stalwart "Oppose" and "Support" comments generally have no place here. Instead, discuss ideas and suggest variations on them.
  • Wondering whether someone already had this idea? Search the archives below, and look through Wikipedia:Perennial proposals.
« Older discussions, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25


Expert review[edit]

Moved from WP:Village pump (miscellaneous): --Pipetricker (talk) 08:06, 1 June 2018 (UTC)

I'd appreciate feedback on an idea. I'm thinking of establishing a free expert review service for Wikipedia featured articles on academic topics - topics well-covered by reliable journals, such as medicine and astronomy. Once an article meets the FA criteria, the world's leading experts on the topic would fact-check it and tell you what they think of its comprehensiveness and neutrality.

I can only offer this service if I'm allowed to put two prominent links at the top of the current article version, one linking the reader to the version that passed review, and the other linking to a simple diff between the current version and the fact-checked version.

The world's topic experts aren't going to review an article if the version they endorse disappears into the article's history in a day or a month. They will if we link to the reviewed version. And the "simple diff" is a service to the reader: it shows them at a glance how the article (and topic) has evolved since the expert-review.

I've thought deeply on this for a long time. I asked BMJ, the publisher of The BMJ, to recruit experts to review Parkinsons disease, and they obliged. One of the reviewers was a main author of the current PD diagnostic criteria and another is the most-published author on the illness. This was a very high quality review. That's the standard of review I intend to maintain.

Do you think rigorous independent expert-review of featured articles is a good thing, and would you support prominently linking to the reviewed version and the diff? --Anthonyhcole (talk · contribs · email) 13:34, 30 May 2018 (UTC)

  • I'd be in favour of this, for what would inevitably be a rather limited number of articles, with some kind of simple control/approval process. In line with WP:MEDRS principles, I think there should a time-limit of up to 5 years set on the links, unless there's some kind of re-review. Johnbod (talk) 13:58, 30 May 2018 (UTC)
  • Not sure who will recall, but there were 2 similar proposals offered back in 2016: User:Atsme/WikiProject_Accuracy, which was presented at meta:Grants:IdeaLab, and a similar project was presented at the same time by another editor: Academic Reviewers. There's also Proposal:Expert review which is along the same lines. Atsme📞📧 14:08, 30 May 2018 (UTC)
How do you feel about those two prominent links at the top of the current version, Atsme? --Anthonyhcole (talk · contribs · email) 16:48, 30 May 2018 (UTC)
I'm fine with it. When I was researching for Project Accuracy, I spoke to quite a few academics (various teaching levels) and explained the significance of the GA & FA symbols on articles. Their responses are what inspired me to design the Seal. I still believe that once an FA goes through the drill of expert/academic review, they should be afforded some protection which makes that "seal" worth something. Atsme📞📧 16:59, 30 May 2018 (UTC)
  • Thank you, Atsme. I'm not sure where I stand on universal automatic protection for articles that have passed expert review. I think I'm against it but need to do more thinking about it. --Anthonyhcole (talk · contribs · email) 17:14, 30 May 2018 (UTC)
  • You're quite welcome, AHC - and if I may briefly explain why I feel some level of protection is needed...once an article has been reviewed to top level, any additions that follow will not have been reviewed; therefore, any newly introduced inaccuracies may be read/cited before the err is caught. The onus will fall on the promoting reviewers (presumably whose links are at the top). At least with some level of protection, it will allow the time needed to review & clear the new material. It is not that we are changing "the encyclopedia anyone can edit", it's simply a brief delay from time of edit to time of publication, but only for those promoted articles. Atsme📞📧 19:45, 30 May 2018 (UTC)
  • I think Dengue fever has been semi-protected since it was published in a peer-reviewed journal in 2014 and that hasn't ruffled any feathers. I see your point about protection. It may further encourage expert collaboration, too. As I say, I'm still making up my mind on this. But it's something for later, anyway. It's by no means a deal-breaker for me. Before I start spending my time and money on this, though, I need to know whether the community will let me do it. --Anthonyhcole (talk · contribs · email) 02:53, 31 May 2018 (UTC)
  • Hmm, the prominent links could be a hatnote, like the one we have linking to introduction articles, e.g like on General relativity. I think that at-least would be accepted by the community, and having a reviewed version does seem good; I'm just thinking - if we're not using the reviewed version as the default, we're sort of un-endorsing it; at the same time the fact that the reviewed versions would get out of date +general principles means we can't keep articles fixed on that. Not precisely related to this, but looking at the Project Accuracy pitch; most readers are not really critically looking at Wikipedia, and thus I don't think having reviewed versions would somehow make Wikipedia more reliable in the eyes of people Galobtter (pingó mió) 17:29, 30 May 2018 (UTC)
Was also here to say this, we already have Wikijournal where people can send FAs to get peer reviewed by professionals. Having one more similar process would drain the reviewer man-power. FunkMonk (talk) 14:16, 31 May 2018 (UTC)
It doesn't have to be separate - it can be coordinated in those topic areas. There's more to WP than just meds and science. Atsme📞📧 14:24, 31 May 2018 (UTC)
I've added more info lower down, but I thought I'd note here that I like the idea of coordinated mechanisms. I think that scholarly journals are an efficient way of incentivising expert engagement (whether WikiJournals or other journals), but I think that multiple mechanisms can work. E.g. an article gets written via WikiEdu, then undergoes GA, then expert review, then journal publication, then FA... etc. NB, There is also a WikiJournal of Humanities in the works. T.Shafee(Evo&Evo)talk 03:20, 15 June 2018 (UTC)

Jens, Mike, Dank and FunkMonk, I've been watching the development of Wikijournal of Medicine since its inception. My model differs in several important ways from Wikijournal of Medicine. The quality of the reviewers I'm offering is the highest possible. That's not the case with Wikijournal of Medicine. I won't be using the same pool of reviewers as Wikijournal of Medicine so I won't be draining that resource. I'm proposing we offer the Wikipedia reader a link to a simple diff showing them clearly the difference between the last reviewed version and the current version; Wikijournal of Medicine doesn't have anything like this in its model. My proposal includes a prominent link to the "reliable" version. The Wikijournal of Medicine model uses a tiny, essentially meaningless little book icon that no readers will understand without clicking and few will click. The names of all my reviewers will be published and prominently displayed on the reliable version; in the WikiJournal model the reviewers may remain anonymous. I'm not proposing to start a new journal to host the reliable version - the reliable version of an article that has passed review simply sits in the history of the article, available to readers who click the prominent link. There are other important differences too but this list should make it plain these are not the same product.

Let me emphasise this important distinction: The traditional academic publishing model relies on the reputation of the publisher, whom the reader trusts to run a high quality review by anonymous peers/experts, and the reputation of the authors, whose names are all disclosed. Both elements - the reputations of the publisher and the authors - are essential to rigorous science publishing. Wikipedia permits authors to remain anonymous and Wikijournal of medicine allows the reviewers to remain anonymous - leaving only the reputation of the publisher as a guarantee of reliability. That's not enough. WikiJournal has no reputation to speak of yet; in my model we use highly esteemed journal editorial boards with an already-established strong reputation for reliability to select only the very best reviewers. But even if WikiJournal were to develop a reputation rivalling Lancet and BMJ the WikiJournal model would still be inadequate. Humans - with careers and personal reputations and egos to protect - need to put their name behind the article. In my model the experts stake their reputations on the reliability of the reviewed article. I can't stress enough how important this particular difference between the two models is. (Although several of the other differences are very significant too, in terms of epistemology.) ---Anthonyhcole (talk · contribs · email) 15:45, 31 May 2018 (UTC)

I'm not saying that WikiJournal of Science meets all our needs and we don't have to consider other journals. I'm saying that there's already precedent for putting a special symbol (the book symbol) at the top of an article to notify readers that there's been some external form of review, and the symbol serves to send them to that paper. - Dank (push to talk) 16:28, 31 May 2018 (UTC)

Would this idea only be for featured articles?Vorbee (talk) 10:15, 2 June 2018 (UTC)

Yep. I only want to submit our very best work to experts (they're busy people) and the FA process is the best system we have for assessing quality. --Anthonyhcole (talk · contribs · email) 11:39, 2 June 2018 (UTC)
  • Support Both the concept of the expert review process as well as the inclusion of a link to the reviewed version. I see a suggestion that it can be done with a hatnote. I agree that it should be something akin to a hatnote but there may be a legitimate argument for making it look a little different than a hatnote as the concepts aren't exactly the same. (Generally, a hatnote is going to direct you to a completely different article, and if I'm looking at an article and it seems to be the right subject, I might not pay attention to a hat note, even though in this case it might be the one I'd prefer to see.)--S Philbrick(Talk) 20:11, 2 June 2018 (UTC)
  • Question – How many articles would this actually affect, in practice? I count 52 FAs in the Health and medicine category, which isn't that many when you think about it. I haven't counted FAs in other "academic" categories like astronomy, but let's be generous and say there are 250 of them. Is it really worth it to create a whole new system for the benefit of 300 articles, many of which figure to be in less need of expert help than B- and C-class articles? It's not like the medical WikiProject is cranking out FAs like crazy; most of the time we don't see any medical articles coming through FAC. Never mind that the vast majority of readers aren't going to bother clicking on a small icon that goes to a potentially years-out-of-date version of an article (they don't do it often for the talk page links to the version that passed a given process), or that experts might propose edits that would damage an article (not knowing our norms for a given topic). Without enough work on relevant articles, I fear that such an idea wouldn't be worth the effort. There's no point in pushing for experts to sign up for a review service if they won't have anything to do. Giants2008 (Talk) 00:49, 4 June 2018 (UTC)
  • Thanks, Giants2008, for your thoughtful response. As Johnbod suggested above, there'll be very few - at least to begin with. There are relatively few medical FAs and very few new ones rolling out.
  • There's no system or infrastructure required to implement this service - it piggy-backs on the peer-review process already in place at all the top journals. It will take a bit of my time to commission each review, and I'll supervise the review to make sure the reviewers aren't proposing off-policy changes (like adding dose information to drug articles). Scan the right hand column of this review, where you'll see this happening. I'm more than willing to put that time in.
  • You mention "... the vast majority of readers aren't going to click on a small icon ...". That's the point of this thread. It won't be a small icon. It'll be a prominent link of some kind. Galobtter, above, suggested a hatnote and Sphilbrck supported a hatnote or something like that. I'm not wedded to any particular format for the links, as long as it's not ugly, fits our style and is obvious to the reader. The hatnote (or whatever it ends up being) will only go up if a version of the article has passed the experts' review. --Anthonyhcole (talk · contribs · email) 10:08, 9 June 2018 (UTC)
  • Mixed Support encouraging such reviews, oppose fossilizing the reviewed version with a diff anywhere, especially in the article space itself. There should be no marker on the article text, and I'm leery of even a talk page notice, which would tend to encourage WP:OWN-type issues and fossilization of articles. I like that experts want to help review our articles, I don't like that someone will use this to prevent future improvement. --Jayron32 01:50, 8 June 2018 (UTC)
  • We're already linking readers to a peer-reviewed version of our articles, where one exists, Jayron23. See Dengue fever. Click the little book icon in the top right hand corner. This proposal is to make that link prominent and obvious when named experts perform the review and the review is managed by an established, highly regarded publisher with a strong reputation for reliability (a publisher who publishes highest quality reliable sources).
  • Experts of the calibre I'm talking about won't be interested in reviewing an article if the reviewed version is going to disappear forever into the article's history the moment another editor saves a revision.
  • Regarding "someone will use this to prevent future improvement", experience does not bear this out. Take Dengue fever, for example. It passed peer-review in 2014. I made this simple diff in 2016. The topic evolved over that time and the article kept up. None of those editors seem to have been remotely concerned about offending the reviewers or messing with a sacred cow. Even if someone does feel that way, Wikipedia's policies and guidelines prevail.
  • As for the reviewed version getting stale: above, Johnbod recommended a time limit on how long we should leave the link up, and I agree with him. --Anthonyhcole (talk · contribs · email) 10:54, 9 June 2018 (UTC)
If they're not interested in reviewing articles that will later get improved, that's what Wikipedia is'. If you want to have some permanent, unchanging encyclopedia written by experts, find somewhere else to do it. --Jayron32 11:55, 10 June 2018 (UTC)
Could you expand on that a little for me please Jayron32? I'm not proposing an unchanging encyclopedia. Editors will still edit the public-facing article. It will evolve just as Dengue fever did after its review. I can see you're strongly opposed to this but can't yet see what your objection is. What's the down side you're seeing that I'm not? --Anthonyhcole (talk · contribs · email) 02:42, 11 June 2018 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────Support: I think that there is room for several models of trying to engage expert review of articles. I agree that attracting high-quality reviewers is essential and collaborating with an established journal such as BMJ is a good way to achieve this. I think a hatnote and category would work well, or possibly a note at the top of the references section (e.g. Rotavirus#References). I agree that just the symbol alone is insufficient (most readers are similarly unaware of the FA star). As well as simple diffs in markup, the visual diffs viewer is pretty good these days or even lust a link to the version after review (same as done with GAs and FAs).

I think that locking or any sort should be handled in the same way as it would for any FA. For example, Circular_permutation_in_proteins has undergone several changes since its publication in PLOS CompBiol. Of course, thew ideal in my point of view would be for BMJ to publish the article if it passes their peer review standards, but that would of course rely on them being happy to publish CC-BY-SA and comfortable with large group authorship attribution, which is uncommon in many journals.[1]

Some possibly relevant links:

I don't think that there will be any great need to lock the pages after expert review (or at least, no more than for FAR). — Preceding unsigned comment added by Evolution and evolvability (talkcontribs) 03:13, 15 June 2018 (UTC)

  • Oppose, with an emphasis on my conflict of interest as creator of Wikiversity:WikiJournal of Medicine, where the WikiJournals can be regarded as competitors to this idea. I do think the WikiJournals serve the main purpose of this proposal already. It is already aiming at having the quality of the reviewers to be "the highest possible". If the problem is that the WikiJournal review symbol is too tiny, I think a better solution would be to make that one more prominent. As for having a latest version and a last reviewed version, I think we already have this mechanism in the form of Wikipedia:Pending changes. And as for a system for reviewers to clearly mark their contributions to articles, we already have Template:External peer review (its usage can be seen on its WhatLinksHere). Regardless, I support having experts review Wikipedia articles, I just don't think we need yet another system for it. Mikael Häggström (talk) 04:45, 16 June 2018 (UTC)
    Mike, I won't support putting a prominent link at the top of a Wikipedia medicine article, linking the reader to a version that has passed review organised by WikiJournal. As discussed recently, here, WikiJournal is not a reliable publisher and its journals are not reliable sources by the standards of WP:MEDRS. --Anthonyhcole (talk · contribs · email) 10:01, 3 July 2018 (UTC)
I accept that, and by the same reason I don't see why this proposal should be allowed to have even more prominent links at the top of mainspace articles. Is there anything making these reviews more reliable than those of WikiJournal? Mikael Häggström (talk) 10:10, 3 July 2018 (UTC)
Here, I see support for prominent links to the reviewed version when the review is conducted under the model described above. If you want to put prominent links at the top of articles, linking to WikiJournal, open a discussion like this one and get input from others. I'll elaborate on my opposition there, if you like, Mikael. --Anthonyhcole (talk · contribs · email) 09:13, 9 July 2018 (UTC)
  • Support: Expert peer review in Wikipedia is currently a marginal practice, and it would be good to test various mechanisms. If there were several mechanisms, they would probably not compete with one another, but rather help one another by making the practice more mainstream. And I do not see what harm would be done by linking to the reviewed version and the diff. This said, I fail to understand precisely what is the aim of the proposal. If it is to improve the quality of science articles in Wikipedia, why start with the ones that a priori need it the least, i.e. the featured articles? (In contrast, the rationale for WikiJournals is straightforward: incite academics to contribute more to Wikipedia, by making Wikipedia-style articles count in publication lists. This is why I am participating in WikiJSci.) Sylvain Ribault (talk) 19:41, 16 June 2018
    A key element of this proposed service is the involvement of the editorial boards of the most prestigious science and medicine journals in reviewer recruitment, and the selection of field leaders and other recognised experts as reviewers.
    They simply won't take on the review of an article that needs a lot of work, any more than they would accept such an article for review and publication in their journals. --Anthonyhcole (talk · contribs · email) 03:36, 18 June 2018 (UTC)
  • Hi Anthony, I have two questions. First, is this proposal for science articles only? Second (the perennial problem), how are you going to persuade the reviewers to do it? SarahSV (talk) 04:03, 18 June 2018 (UTC)
    Personally, Sarah, I'm not taking this any further than medicine. Medicine is a fairly "hard" (as opposed to soft) topic that's underpinned by a great deal of rigorous publishing and robust systems for consensus-building. If it works and is useful in medicine, then it might work and be useful in the hard sciences, too, but I won't be taking it there. I don't know if it will work in softer topics like the social sciences, history and literature, and it won't work for the vast majority of Wikipedia topics that aren't well-supported by academic publishing.
    It's the editor-in-chief of the relevant journal/s who needs persuading. Then their managing editor goes through her Rolodex and offers the gig to the relevant experts.
    The editorial teams at the top relevant journals have the expertise, experience and relationships to do this well.
    IF we begin with and stick with only the most highly-regarded journals, and IF they consistently come through with stellar review panels, I hope experts will soon become proud to be asked to review Wikipedia articles, and it will be something they'll put in their resume. IF that's how it unfolds then, as the reputation of expert-reviewed Wikipedia articles grows, the managing editors may find it easier to recruit the best reviewers. We'll see. It's early days. --Anthonyhcole (talk · contribs · email) 05:08, 18 June 2018 (UTC)
    Anthony, what do you see as the benefit for the editors of the journals? They will have to reward the reviewers in some way. It used to be easier to get academics to do reviews. But the more we grew, and the more money the Foundation became associated with, the less eager reviewers have been to volunteer their time. I can imagine that someone might pay reviewers (e.g. some charitable medical foundation), although there's a risk that the payer would interfere editorially, and we would face the unfairness of reviewers being paid while writers are not. SarahSV (talk) 01:29, 23 June 2018 (UTC)
Sarah, many of the top journals are published by scholarly societies and professional associations that have, as a part of their mission, education and the dissemination of knowledge about their specialism. If we present the reader with a prominent link to the fact-checked version, it'll be fairly easy to convince those journals it is worth the effort to manage a review, I think. The reward for the reviewers is (1) altruism, like you and me, and (2) prestige, per my last paragraph above. This latter isn't an afterthought. It's a key element of the model. I expect a Wikipedia medical article reviewed under this model to be regarded as the most reliable source on the topic, period. And I expect scholars and experts to see an invitation to review as a very visible public acknowledgement of their standing in their field.
As for money:
  • There is a role for a very experienced Wikipedian in each review, liaising between the experts and the writers: (1) ensuring the expert suggestions are compliant with our policies, (2) finding reliable sources that support proposed changes, (3) updating the article in collaboration with other editors in response to the review, (4) re-presenting the updated article to the reviewers for endorsement and (5) formatting the "reliable version" with relevant templates, etc. This is pretty onerous, tedious, exacting work and I can see it becoming a paid role at some point.
  • I would be very disappointed and a bit surprised if it turned out the reviewers needed paying. I think, I'm pretty confident actually, it can be avoided.
  • Any money supporting this effort can't come from the WMF because of perceived (at least) conflict of interest. There are a number of non-profits out there with education in their remit. Anthonyhcole (talk · contribs · email) 09:39, 23 June 2018 (UTC)

I'm thinking about the consequences of Google linking searchers directly to the fact-checked version. That could cost us a lot of editors, like me, who love clicking publish. I'd lose interest in editing if my edits took months or years to appear. A few years ago, Google committed to privileging reliability over popularity in its search results. If this journal-driven expert review service becomes a thing, I'll make sure Google knows the harm direct-linking would do. --Anthonyhcole (talk · contribs · email) 12:49, 25 June 2018 (UTC)

  • Support. Only expert reviewers who spend many hours every week doing research in their field of expertise, who read vast numbers of papers can have a good judgment to weed out subtle problems with review articles written by non-experts. A review article written by scientists is always based on a huge amount of literature research, this is obviously not the way we go about writing articles here. We cruise on autopilot by summarizing the contents of review articles, but this can lead to inaccuracies. The main problem here is caused by our reliance on the top journals like e.g. the BMJ, the Lancet etc., while these journals only publish a small percentage of the research results. A lot more is going on behind the scenes, the vast majority of the relevant research for any particular topic is published in technical journals that we cannot possibly keep track of. While our articles will end up presenting most of the relevant information, things tend to go wrong with giving the right weight to different ideas. Different ideas that are not considered to be equally likely to be true, tend to get presented in a more equal way in the top journals, because one wants to see what the most rigorous evidence tells us and not be biased based on the previously known less rigorous evidence. That's a good way to eventually get to the rigorously verified truth, but it may sometimes mislead lay people about how researchers in the field think about the different ideas. Count Iblis (talk) 15:42, 25 June 2018 (UTC)

References

  1. ^ a b "Wikipedia-integrated publishing: a comparison of successful models". 26 (2). Health inform. doi:10.13140/RG.2.2.27470.77129. 
  2. ^ Mietchen, Daniel; Wodak, Shoshana; Wasik, Szymon; Szostak, Natalia; Dessimoz, Christophe (2018-05-31). "Submit a Topic Page to PLOS Computational Biology and Wikipedia". PLOS Computational Biology. 14 (5): e1006137. doi:10.1371/journal.pcbi.1006137. ISSN 1553-7358. PMID 29851950. 
  3. ^ Maskalyk, James (2014-10-02). "Modern medicine comes online:How putting Wikipedia articles through a medical journal's traditional process can put free, reliable information into as many hands as possible". Open Medicine. 8 (4): e116–e119. ISSN 1911-2092. PMC 4242788Freely accessible. PMID 25426179. 
  4. ^ Shafee, Thomas; Das, Diptanshu; Masukume, Gwinyai; Häggström, Mikael (2017-01-15). "WikiJournal of Medicine, the first Wikipedia-integrated academic journal". WikiJournal of Medicine. 4 (1). doi:10.15347/wjm/2017.001. ISSN 2002-4436. 
  5. ^ Shafee, Thomas (2018-06-01). "The aims and scope of WikiJournal of Science". WikiJournal of Science. 1 (1): 1. doi:10.15347/wjs/2018.001. ISSN 2470-6345. 
  6. ^ Butler, Declan (2008-12-16). "Publish in Wikipedia or perish". Nature. doi:10.1038/news.2008.1312. ISSN 0028-0836. 
  7. ^ Su, Andrew I.; Good, Benjamin M.; van Wijnen, Andre J. (2013). "Gene Wiki Reviews: Marrying crowdsourcing with traditional peer review". Gene. 531 (2): 125. doi:10.1016/j.gene.2013.08.093. ISSN 0378-1119. PMID 24012870. 

Standardizing revert reasons and process[edit]

I'd be interested in people's views on the following idea, to help address a problem I've noticed many encounter

The Problem. Reverting other people's edits is required to deal with vandalism and other issues. But problems arise when this is abused - e.g. when users delete content they disagree with, generating disputes, incivility and edit wars, potentially driving away editors, particularly new ones. Currently there are few guidelines for reverting edits, leaving it mostly to the discretion of editors when to delete other people’s work, contributing to disputes. Editors are merely requested to type their own reason for reverts, some don’t even do that

One Possible Solution. One way to help address this would be to agree upon specific, valid reasons for reverts, and then list these on Wikipedia Diff Pages, for users to click to always indicate their reason for Undo (these are illustrative only):

 Undo(?): Vandalism NPOV Verifiability Copyright Redundant Other 

E.g. clicking NPOV would automatically add “Undid: NPOV” in Edit Summary, to indicate NPOV rule violation was reason for Undo, thus saving typing. Note: Clicking (?) would take user to WP page that explains these agreed-upon revert reasons, including what users should do before, or instead of resorting to undo, where appropriate. For reverts via regular edit, could add radio buttons below Edit Summary box, to indicate delete reason, e.g:

 If you deleted other people’s edits, indicate a reason for delete(?):o Vandalism o NPOV o Reliable Sources o Copyright o Redundant o Other 

Clicking NPOV radio would insert “Deleted: NPOV” in Edit Summary, to indicate NPOV violation as delete reason. There are many possible variations, and slightly different approach could be used for History page Undo and Rollback (latter needs to be single-click)

Benefits. Specifying, reminding and requiring users to give valid revert reasons could reduce abusive reverts, and let people know why their edits have been deleted, thus reducing revert-related disputes, incivility and edit wars. This could also aid in retaining editors, particularly novices, women, etc (Note: above Diff page Undo-reasons would not require any extra clicks, and also save on typing reason. Regular-edit reverts would require 1-extra reason click, but save on typing reason - so almost always would be less effort than currently)--Thhhommmasss (talk) 18:41, 10 June 2018 (UTC)

Wikipedia:BOLD, revert, discuss cycle is a highly influential code of conduct. It states that you can, and should, revert when you don't think that the preceding edit was an improvement. It has worked for us terrifically well. Somewhat paradoxically, it's only by making reverting easy that we can make implementing changes easy.
As for the standard edit summaries, there are tools you can use for this. – Finnusertop (talkcontribs) 06:18, 14 June 2018 (UTC)
That’s fine, but the problem arises when this is abused. See for example the above comment, referring to “instant reverts, threats, insults...”, with the effect that “I have seen more than one editor driven completely off Wikipedia”. Wikipedia, when it works well, as it often does, works because of specific rules (NPOV, Verifiability, etc.) – i.e. people can’t merely exercise their discretion (e.g. “do whatever you think will improve the article” is generally not the Wikipedia way). Yet here, for something that clearly ticks off people – seeing their good-faith edits reverted – a lot of discretion seems to be provided, i.e. revert anything you think does not improve the article, and discuss it out. This can then lead to some of the additional abuse mentioned in the above comment – insults and incivility on talk pages, etc. There are specific, listed criteria for speedy page deletion. This is a similar issue of deleting someone else’s content, thus I think listing comparable reasons for reverts would also be helpful. And I think one can still be bold on reverts, while listing valid reasons, just as listing page delete criteria still allows for speedy page deletion. Also, the above suggestion would in nearly all cases save effort, that is save on typing a revert reason--Thhhommmasss (talk) 19:47, 15 June 2018 (UTC)
"Do whatever you think will improve the article" actually is the Wikipedia way. This principle is enshrined in one of the oldest policies here: Wikipedia:Ignore all rules.
We do have people (new and otherwise) whose idea of improving an article isn't widely shared, with the result that their additions or reversions don't – in the opinion of the rest of us – actually improve the article. But generally, the idea is that you should do your best, and that others should also do their best, and in the end that usually works out. WhatamIdoing (talk) 19:47, 21 June 2018 (UTC)
I still think what makes Wikipedia is its rules, even if occasionally they must be broken. Take away the rules, and you’d have a mess, i.e. people throwing random opinions together. On revert, I’ve seen people come up with their own revert reasons, which I’m sure 99.9% of the people would consider bogus, and then you’re stuck days debating with them over this bogus reason. I think that’s what ticks people off, arbitrariness. I’ve seen this in particular on another language Wikipedia, where people have given up, and the result is every couple of years media articles appear on how biased this language Wikipedia is. Now it’s true that even defined, valid revert criteria, could be misused or misapplied - someone might claim something is a valid delete, e.g. due to lack of Reliable Sources, when it’s not. However Reliable Sources are quite well defined, and it is more difficult to just wing there. In any case, with well-defined, valid revert reasons, the discussion would be much more constrained. The delete reason tag in Edit Summary could link to a description of the delete reason, for everyone to instantly refer to the reason description, use it as an arbiter, instead of having to engage on nonsense, made-up reasons, which take people away from valuable tasks, like improving articles, or as others have noted, leads them to quit--Thhhommmasss (talk) 01:12, 26 June 2018 (UTC)
Editors already widely abuse concepts like vandalism, NPOV, and BLP, as seen in their edit summaries. Formalizing this abuse would accomplish exactly nothing. ―Mandruss  20:23, 21 June 2018 (UTC)
In the limit that could be an argument for no rules, since all rules are likely to be abused. There is a lot of research that people can be "nudged" in the right direction with subtle clues. What no doubt happens in abuse is people get upset by ideas that are contrary to theirs, and instinctively act (i.e. delete first, and rationalize later). Perhaps something that requires them to consider valid delete reasons for just a millisecond before they can delete, can prompt slightly more rational behavior. Ultimately such principles can be empirically tested via A/B testing and other strategies. Often the winning approach can be what we might consider as counter intuitive Thhhommmasss (talk) 21:15, 24 June 2018 (UTC)
The folks willing to be nudged tend to learn from people pointing out to them a particular revert was not proper, or from reading the guidelines, or from seeing what other long-term editors use revert for even without having to think for a millisecond as result of being prompted for a standardized response. The folks who revert things not meant to be reverted while not being open to nudging/correcting are frequently perfectly willing to be deceptive (e.g. claiming they are reverting vandalism—which means they do consider what are valid delete reasons for a millisecond—when what they're reverting is by no means vandalism nor even remotely resembles it).
A similar idea was deployed on the Wikipedia App for general edits, the so-said "canned edit summaries". They are so often used deceptively or dishonestly that I—and probably most vandal-fighters—see use of them by new or IP editors as at best neutral/not effecting the suspiciousness of an edit and at worst an actual indication it might be a good idea to check said edit. If it didn't resolve the issue there, I see little reason to believe it *would* resolve reverting-related issues. AddWittyNameHere (talk) 23:32, 24 June 2018 (UTC)
Nudges work, they work subconsciously, instantaneously, even a small thing like changing button color can impact the likelihood people click it. Here we want to nudge people from automatic, emotional behavior (i.e. delete in anger, rationalize later), to have them first provide a quick, rational delete reason. Hopefully, that can kick them into a more rational mode (e.g. can I defend this delete reason, if not, should I delete?). Studies of negotiation under anger show that for rational reasoning to kick-in, this must be done early in the cycle. Here a delete reason is selected before deleting, which may be more successful than delete first, specify a reason later (which is what canned edit summaries sound like), because once a delete is made, people are more likely to defend it, justified or not. That’s the hypothesis, with some experimental backing (i.e. nudges work, early interventions work), although admittedly those cases are not exactly like this proposal. Thus, I do not know for sure in advance that it will work, just like you can't know for sure that it will not. That is why there are things like A/B testing, so people don't just debate which color they think will encourage most button clicks, etc. Running experiments is also the scientific way – i.e. scientists do not decide merely by debate, or majority vote, which hypothesis is correct--Thhhommmasss (talk) 22:54, 25 June 2018 (UTC)
Yes, running experiments is the scientific way. However, scientists don't generally go from a hypothesis to proposing laws to experiment whether the addition of that law would help solve the problem. Which is the real world equivalent of your proposal.
I'm not saying nudges don't work at all. As you say, there's some research backing that they may work—when someone is acting from automatic, emotional behaviour. However, that is not the premise I am in disagreement with. What I am in disagreement with is the to this experiment fundamental analysis that a large, if not the largest, portion of problematic reverts is the result of people acting from automatic, emotional behaviour to the degree they either lack rational awareness of their actions and/or have said rational awareness overruled by their emotions. (That is not to say that I believe no reverts are done in such circumstances, but rather that from my observation I suspect they account for a far smaller proportion of problematic reverts than you suspect they do)
Rather, I believe that most people who commit problematic reverts are, in no particular order
  1. acting from a misunderstanding of the rules, in which case asking them to provide a rational reason will hardly work, simply because they already are in a rational mindset and genuinely believe they are acting in line with guidelines&policy (good faith problematic reverts);
  2. or acting from a genuine belief that even if this exact situation isn't coded into the rules, it's common sense and obvious that an exception applies/should apply to this particular situation (good faith problematic reverts/misapplication of IAR);
  3. or perfectly well-aware that what they're doing is not in line with the rules, but are either holding disregard for rules in general or are in such disagreement with a specific rule (or rules in a specific context) that they are unwilling to hold themselves to said rule/to rules in said situation (bad faith problematic reverts; this one is partially emotion-based but nonetheless not lacking the rational awareness that is the entire premise of nudges working);
  4. or acting especially because what they're doing is not in line with what should be done. (trolls, vandals, LTAers etc.)
While, as I said above, I don't doubt that some cases of predominantly-emotional problematic reverts occur by people genuinely lacking rational awareness in that moment, even of those it is not a given that they are committed by people who would, if only they were thinking rationally, both realize that what they are doing is not what they should be doing and consider this sufficient reason to thus not do it. Some will be, of course, but I suspect that they make up such a small portion of the problematic reverts that the additional hassle for everyone else using revert in a non-problematic manner is simply not worth it. AddWittyNameHere (talk) 03:39, 26 June 2018 (UTC)
Agree, hypotheses should be questioned, but on the rest I’d disagree. Studies show edit wars are most frequent on subjects that generate strong emotional response – politics, religion, history, gender, race, etc. You probably don’t see heated edit wars on Medford, Oregon or 19th century ballerinas. Psychologists, like Daniel Kahneman, have shown that our much older, automatic, emotional response systems are much more powerful than our recently developed rational systems, and that these emotionally-based systems often lead us astray, with all sorts of biases, etc. Others, like Jonathan Haidt, have shown that automatic, emotional responses are particularly strong on issues like politics and religion, which often go to people’s identity, thus driving bad behavior – incivility, polarization, etc. Cognitive behavior therapies are most successful in dealing with bad behaviors (e.g. anger management) and these seek to put in circuit-breakers, so people aren’t controlled by negative emotions. Of course, that requires individuals taking action themselves, which is unlikely in edit wars, trolling, etc, thus external circuit-breakers are needed. Wikipedia rules (NPOV, Verifiability, etc), are all intended to promote more rational behavior, so people aren’t merely guided by their emotional biases, and these rules have been constantly extended and refined over time --Thhhommmasss (talk) 01:24, 27 June 2018 (UTC)
(edit conflict)Interesting study, though from what little the link you gave shows, it doesn't show that edit wars are most frequent on those topics, but that the highest volume of reverts of any type occurs on those subjects, a fairly significant difference. As these subjects are often also the frequent target of vandalism and test edits, it would be interesting to see what percentage of those reverts is in fact reverting vandalism or otherwise non-problematic reverts. Not every revert is an edit war; not every seeming edit war is a content dispute. (Some of our more prolific LTAers are perfectly happy to keep dumping the same vandalism into articles over and over again until they're blocked. And then again as soon as they can get a new IP)
I'd disagree that one might not see heated edit wars on, say, Medford, Oregon. We've had one at Chicken, Alaska. We've had one over the use of an exclamation mark at Berwick-upon-Tweed(!) Similarly, we've had edit wars on ellipses; on diacritics; on capitalization; on taxonomy; on vernacular names of species; repeatedly on capitalization of vernacular names of species; on whether to call a . a period or a full stop; and on basically every other subject one would think wouldn't be the target of edit wars.
Certainly emotion drives strong impulses and responses, and some subjects are more likely to trigger strong emotions. We disagree on the proportion of problematic reverts that occur from a purely-emotion-driven state, though we both agree it does occur, and probably disproportionally often in case of edit wars as opposed to other types of problematic reverts. As I've already explained in my prior response why I disagree with you on this part, I will not bore both of us by repeating myself. Instead, let's consider those folks that are acting from an emotional state.
As you say, identity drives emotion drives bias. If someone, from a deeply emotional response and lack of recognition of their own bias, is convinced they are dealing with a NPOV-violating edit or article, no nudge is going to make them consider that perhaps the article is NPOV and it's their view that isn't, because if anything, considering such a thing would be a far larger and more direct challenge to their identity than the edit/article involved. They'll remain mostly in their mindset ("wrong, wrong, wrong") and treat the nudge in a similarly-automatic manner as they do the revert itself.
A nudge works on behaviour, by putting people into a rational enough mindset that they see an action they were about to commit is not appropriate for the circumstances without challenging their identity. However, reverts are not by default an inappropriate response to a NPOV-violation. If anything, they fairly frequently are the appropriate response. What is inappropriate is in their identification of NPOV-vio, which a nudge is for the reasons I mentioned above is unlikely to change.AddWittyNameHere (talk) 02:42, 27 June 2018 (UTC)
Psychologists like Jonathan Haidt write that nearly all bad behavior on volatile issues, like politics and religion, is driven by our automatic, impulsive emotional systems, and would likely say that nearly all bad behaviors (e.g. arguing over capitalization), are also driven by our automatic, impulsive systems. Cognitive behavioral therapy is intended to deal with bad impulses, regardless of what sparked them (religion, bad punctuation, etc.), and as mentioned often works by seeking to put in circuit-breakers, so people aren’t driven by negative impulses/emotions. By contrast, an act-first (i.e. revert), and-rationalize-later (put reason in edit summary) approach facilitates impulsive behavior. No process, rule or law will entirely prevent bad behavior, not CBT, not nudges. Yet CBT is effectively applied to deal with highly compulsive behavior – gambling, addiction, violent impulses, etc. Whether above proposed nudge can reduce some bad revert behaviors is unknowable without testing - Google and Facebook constantly do thousands of tests to nudge behavior, precisely because they do not know in advance what will or will not work. Btw, I agree valid NPOV issues should be reverted, but valid, rationally-based reverts would of course still be encouraged, made even a little easier, since part of reason is placed in edit summary via single click, instead of having to type it--Thhhommmasss (talk) 01:31, 29 June 2018 (UTC)
Happy edits are all alike; every problematic edit is problematic in its own way. Having a canned "revert: not compliant with WP:NPOV" edit summary is almost negligibly more useful than no edit summary at all, and it relieves the editor making the revert of the obligation of providing an even remotely situation-specific bit of reasoning. (Hint: virtually every contentious edit – revert or not – is made by an editor who believes they are defending WP:NPOV.) TenOfAllTrades(talk) 01:47, 27 June 2018 (UTC)
As I mentioned to Whatamidoing, I’ve seen editors revert and give ridiculous reasons, and then people have to waste time arguing over ridiculous reasons. If instead people used valid revert reasons - NPOV, Verifiability, Reliable Sources, etc. - these are fairly well-documented, harder to misuse than entirely made-up reasons, and I’m always happy to discuss valid reasons. I’ve also seen where users revert, but not give a WP rule, and then after some back-and-forth, finally give a valid WP Rule (e.g. Verifiability) – again wasted initial effort, compared to always having a revert reason/rule everyone can refer to, without needless back-and-forth. With automated revert-reason tags, users could still be asked to provide further, instance-specific elaboration in Edit Summary--Thhhommmasss (talk) 01:50, 29 June 2018 (UTC)
But providing canned edit summaries wouldn't stop people from making reversions that you disagree with. It would, however, increase the problem of edit summaries that have nothing to do with the actual problem.
I think that you might be interested in https://www.wnycstudios.org/story/trust-engineers/ (partial description) about what happened to Facebook abuse-claims team, on the day after half the world got smartphones for Christmas. They came back to work and discovered huge numbers of photos being reported as hate speech, threats, and nudity. The photos were things like people wearing matching sweaters, or puppies.
It turned out that their friends and family had uploaded a bunch of photos, and people were using the abuse-reporting feature because they were embarrassed by the photos. There wasn't (at that time) another option, so the puppies got tagged as hate speech. The people tagging these weren't deterred by puppies not being hate speech; the system gave them about four options, none of which were "This photo is embarrassing", so they clicked whatever buttons were available, basically at random, and reported the photos that way.
I believe that exactly the same thing would happen with your proposed system: If someone decides to revert an edit, and you force them to choose one of six options to be able to complete the reversion, then they will pick one of the options, even if those options have nothing at all to do with their real reasons.
If you want to deter reversions, you need to be reaching people before they click the undo button, or you need to give them an alternative that is (to their way of thinking) more effective. Your proposal won't do that. WhatamIdoing (talk) 22:07, 29 June 2018 (UTC)
As mentioned in above proposal, users would have to select a valid Undo reason, before undo happens, with the goal of nudging user into a more rational reasoning mode, as opposed to doing impulsive emotional undos, and then typing reasons/rationalizations later (now you can even undo, without typing any undo reason). To your other point, one cold randomly select 100-200 existing undos, where undo reasons have been specified in Edit Summary, to see if these can be categorized into 7 or 8 valid reasons that account for 80-90% of reverts (it may be possible to have more, e.g. 11 canned reasons, but fewer would be better). In any case, it’d be good to make decisions based on such systematic analyses of actual data, instead of what often happens in these discussions – my anecdotes/impressions vs. your anecdotes/impressions. Then for exceptions to canned reasons, one could add an Other reason, with requirement that this be further elaborated in Edit Summary, and that it be used for valid exceptions, not to invent reasons. Further refinements could be done (e.g. later analyze reasons people are putting under Other – if there are repetitive, valid reasons, add these to canned list, but if they’re making up stuff to abuse reverts, give these as examples of what is unacceptable, etc.) Thhhommmasss (talk) 18:40, 4 July 2018 (UTC)
  1. By the time I have asked the software to present me with your list of pre-approved "valid Undo reasons", I have already decided that I want to undo it. As I said, "If you want to deter reversions, you need to be reaching people before they click the undo button." That's "the undo button" that you find in the page history and at the top of the diff, not the "Publish changes" button that you click when you're finished.
  2. What's going to stop the would-be reverter from selecting any of the pre-loaded "valid Undo reasons" at random? Nothing? Then it won't work. Your process is (a) click the Undo button, (b) choose from among a list of pre-approved, guaranteed-to-be-valid reasons, (c) save your change. I'm telling you that nobody who really wants to revert a change is going to be deterred by needing to click on a pre-approved "valid Undo reason". It's barely even going to slow them down. On the other hand, the use of random and irrelevant "valid Undo reasons" is going to be very irritating to people who are trying to review these edits later. No information about the edit is better than wrong information. WhatamIdoing (talk) 02:34, 6 July 2018 (UTC)
  1. As indicated in proposal, on diff page, user would not click Undo, but instead in order to Undo, they would have to click one of given valid reasons listed next Undo, so user would always have to give a reason before undoing. On History page, without getting into design specifics, similar could be done (e.g. as soon as user hovers over Undo, a list of valid Undo reasons pops up, and they must select 1 to do undo), so again would need to select valid Undo Reason before Undo
  2. If user gives random response, then other person can come back with “You’re wrong”. Since people usually do not enjoy hearing they’re wrong, this is disincentive for random responses. Now users can entirely make up revert reasons, and people have to engage with them to discuss these bogus reasons. As to whether a revert is done with no explanation, or with a random canned reason, editors still need to check what was done, and then revert accordingly, so I do not see any difference in effort here. In fact, now when no revert reason is given, I suspect people often engage in additional needless effort to ask why was revert made, or spend time to comment that no revert reason was given, even though it should have been — Preceding unsigned comment added by Thhhommmasss (talkcontribs) 23:25, 8 July 2018 (UTC)
    I assume then - and not entirely facetiously - that one of the pre-approved valid Undo reasons is "You're wrong"? —Cryptic 23:48, 8 July 2018 (UTC)
Also "Test edit".
I just don't see how providing half a dozen Undo buttons (each with its own custom edit summary) is supposed to discourage people from clicking an Undo button. In fact, it would probably result in slightly more reversions, due to misclicks and reminding people of all the many reasons why they could reject someone else's work.
Thhhommmasss, I think this is a bad idea. But in the interests of fairness, if you still think it is a good idea, then your next step is to write this in Javascript and run it as a user script on your own account, to see how it works for you. Alternatively, you could start using WP:Huggle, which already provides this. However, I feel compelled to tell you that previous academic research about Wikipedia indicates that Huggle (and other software like it), which has made it so easy to revert other editors without even needing to hand-write an acceptable reason for doing so, is significantly responsible for the decline in the number of volunteers editing here. WhatamIdoing (talk) 00:45, 9 July 2018 (UTC)
  1. I’m suggesting after user clicks an Undo reason, it puts this reason in Edit Summary, and user can be asked to add more explanation, but at least it fills in part of reason, so there is some saving on typing. User must still click Publish Changes, same as now, for Undo to take effect. So user has chance to look at their Edit Summary, see which canned reason they entered, expand on this, or change canned reason, before they click Publish Changes. So it’s very unlikely they’ll press Publish Changes with wrong canned reason, and should also avoid issues you mention with WP:Huggle
  2. You say Huggle led to increased editor attrition. This goes to core of proposal, confirming that slight changes in UI (i.e. 1-click Revert, vs. having to confirm via at least 1 additional click on Publish Changes), can produce substantial changes in behavior - thus it is not the case that those who behave badly, will behave bad regardless, and UI changes have no impact
  3. On issue of revert reasons perhaps serving as reminder to revert, one simple way to take care of this is to put the Reasons in a drop-down menu, which user sees only when they click Undo, so they are less obvious. Then when they select Reason, it appears both in Undo and in Edit Summary, so there’d be even less possibility for wrong selection, plus this saves space, and is smaller change compared to current UI
  4. Regarding reverting a Test edit, I assume user is reverting own edit. For this there could be a canned Revert reason, e.g. “My”, where they can revert themselves all they want. For Wrong I’d put something like Incorrect (sounds more neutral, means same), and then in documenting valid revert reasons say that for Incorrect they owe more explanation in Edit Summary and/or on Talk page. I’d be glad to debate Incorrect (as opposed to entirely bogus reasons I got on revert, which did not claim my citations were incorrect)
  5. I agree with SMcCandlish, that a good first step would be to document valid revert reasons, similar to how valid page delete reasons were documented, since this is needed regardless of any other possible actions. Is there some way to get 200-300 randomly selected Undos? This would help determine what people give as Undo reasons, to see if these can be grouped into some sensible number of valid reasons. Or have studies been done of Undo reasons? --Thhhommmasss (talk) 02:17, 10 July 2018 (UTC)
This proposal would be very limiting. I would rather see WP:Editing policy be updated with clarity on what are and usually are not valid revert rationales. Many editors (including experienced ones) believe they have a right to revert, no matter what, and to keep doing so (within the limits of WP:EDITWAR), even if they cannot or will not articulate a policy-based reason. They're wrong.  — SMcCandlish ¢ 😼  02:10, 3 July 2018 (UTC)
Agree documenting valid revert reasons, similar to existing effort to document valid page delete reasons, would be good regardless of specific approach. In addition, putting a question mark next to the Undo command that links to these reasons would be subtle, constant reminder of valid reasons. Also “Undid” that now automatically appears in Edit Summary, could link to these same reasons, so everyone, including new users, would have instant access. If revert was done via regular edit, “Reverted” could be automatically placed in Edit Summary with link to same explanatory page (there are many possible tweaks/variations). As to other parts of proposal, I still maintain subtle changes can influence impulsive behavior substantially. That’s why Amazon has 1-click-buying to promote impulsive behavior, since they know just adding one extra click can start people thinking “maybe I should’ve bought the other item”, or “do I really need this?”. While Amazon wants to facilitate impulsive purchases, Wikipedia would benefit by not facilitating impulsive behaviors (i.e. revert-first, rationalize-later), and instead try to nudge people into more rational patterns. Just yesterday i saw someone did a revert on another page from an IP address, giving no reason, because current workflows enable that, and someone else had to go and fix that, spend extra time noting they did not give a reason, and this is no doubt repeated ad infinitum. As to which specific subtle UI nudge is more or less successful, this can only be determined via A/B experiments --Thhhommmasss (talk) 18:40, 4 July 2018 (UTC)

Phasing out human editors in favor of bots[edit]

"If you think I'm going to improve 250 GA articles this evening at breakneck speed, you need to put down the Elon Musk and stop watching Hollywood blockbusters."

Progress in AI technology has led to AI systems that have good language comprehension and skills. Take e.g. a recent IBM test:

https://www.cnet.com/news/an-ibm-computer-debates-humans-and-wins-in-a-new-nuanced-competition/

"We saw computers beat humans at chess in 1997, beat humans at Jeopardy in 2011 and vanquish the world's best human players of the ancient game of Go in 2017. On Monday, a computer edged out a victory over people in a far more nuanced competition: debate."

"To formulate its argument, it had at its disposal a collection of 300 million news articles and scholarly papers, previously indexed for quick search results. But it had to find the information, package it persuasively, listen to its opponents' arguments and formulate a rebuttal."

IBM could probably create a new Wikipedia from scratch that's edited by a similar AI system today. But we would still have an advantage over any such encyclopedia, based on almost 2 decades of editing experience. However, on the long term we'll end up being replaced by autonomous, self-editing encyclopedias unless, of course, we start using AI ourselves. Count Iblis (talk) 16:58, 19 June 2018 (UTC)

  • I, for one, am ready to transcend. --Izno (talk) 17:17, 19 June 2018 (UTC)
Arm yourselves. Tin foil do-rags and SCAR-Ls for everyeditor and their cousin. We will not be oppressed by the lizard men/cabal/Alphabet/Obama/Mnangagwa with their plans to quash the canaille of peasant editors will not be stood for. [FBDB] cinco de L3X1 ◊distænt write◊ 20:54, 19 June 2018 (UTC)
  • I can think of some editors whom I'd like to see phased-out. --Tryptofish (talk) 21:39, 19 June 2018 (UTC)
  • At present, AI systems suck up biased data from the Internet and inherit the bias they find there. White male American chauvinism tends to creep in. And what would an AI make of Elsevier's loony but profitable alternative medicine journals "peer" reviewed by the same community of loony but profiteering academics? General intelligence technology has a high wall to climb before it can relieve me of this editing chore. >sigh< — Cheers, Steelpillow (Talk) 19:02, 20 June 2018 (UTC)
  • Is this suggestion serious? Vorbee (talk) 08:43, 21 June 2018 (UTC)
Is this question serious? EEng 14:31, 21 June 2018 (UTC)
  • Yes - we already have a proposal at Wikipedia: Village pump (perennial proposals) saying that we should have a bot to welcome new users, and a message about why this proposal has been rejected. Vorbee (talk) 15:54, 22 June 2018 (UTC)
  • But I've been activated since ages already! My grammar module seems to be stuck in 'Yoda' mode, though. TP   16:04, 22 June 2018 (UTC)
Surely you mean In Yoda mode stuck my grammar module seems to be? EEng 18:26, 22 June 2018 (UTC)

Mastodon instance?[edit]

Mastodon is a federated social network that emphasizes microblogging. A few large communities & institutions have Mastodon instances of their own, such as MIT, the Chaos Computer Club, and LQDN. I think it would be a good idea to have our own instance, possibly at social.wikimedia.org (or mastodon.wikimedia.org). Thoughts? Enterprisey (talk!) 20:19, 27 June 2018 (UTC)

Could you provide some details regarding what the benefits/impact of doing that might be? --Yair rand (talk) 02:16, 28 June 2018 (UTC)

My first thought was that if one goes to Wikipedia: What Wikipedia is not, one can go to Section 2.5 to see that Wikipedia is not a social networking service. Vorbee (talk) 08:01, 28 June 2018 (UTC)

As the domain name social.wikimedia.org implies, this would not be part of Wikipedia, and instead be another project hosted under wikimedia.org. Enterprisey (talk!) 13:40, 28 June 2018 (UTC)
But why?  — SMcCandlish ¢ 😼  02:15, 3 July 2018 (UTC)
An improved UI and outreach tool for editors and new ones alike to discuss articles would be a worthy endeavor. Talk pages require WikiMarkup knowledge and have very old school forum style. I would be concerned about splitting communication, but the idea of having improved collaboration/communication is a good one. We organize hackathons already to improve content of articles, why not harness that energy online with first class support/search for wiki articles? Something say facebook, twitter, do not sufficiently help with. Shushugah (talk) 12:24, 12 July 2018 (UTC)

New user right request[edit]

I would like Wikipedia have a new user right. I want it called Signature manager, but the user right I want to see on Wikipedia is about the ability to modify other people's signatures. I want to have the modification tool at either Special:SignatureManager or Special:UserSignatureManager and have it logged at either Special:Log/signature or Special:Log/signatureother. 209.52.88.255 (talk) 18:52, 4 July 2018 (UTC)

I'm glad that you saw my response to your message and decided to come here to discuss your thoughts. Why do you believe that this user right is needed? What would having this user right developed and available to be granted to users accomplish? ~Oshwah~(talk) (contribs) 18:54, 4 July 2018 (UTC)
This right might be useful because people can fix typos such as typos in CSS or attributes such as letter case in classes, CSS color, such as ref to red (assuming they're using a QWERTY keyboard), or times when they can't decide between one signature or another, or uses like User:Cyberpower678/SignatureColorKey, where the wrong color is added to the signature in a sign. This user right should also always display a preview of the other's signature first, which wouldn't include the timestamp. 209.52.88.255 (talk) 19:09, 4 July 2018 (UTC)
Hi, first only registered users can have permissions so this would not be something you could use directly 209. Generally "signatures" are just wikitext on pages and are not protected. If someone is using a signature that is causing disruption (such as introducing Special:LintErrors) they can be asked to stop or be blocked under general disruption. Being able to forcibly change someone else's preferences (such as signature) is mostly a non-starter, as it could lead to that person making edits they did not expect to make. — xaosflux Talk 21:09, 8 July 2018 (UTC)
I think phab:T178879 and phab:T140606 are much more likely and reasonable than allowing people to change other people's signatures. --Izno (talk) 21:28, 8 July 2018 (UTC)

Main Page suggestion[edit]

The French Wikipedia's main page currently hides the "Welcome to Wikipedia" banner in the Timeless skin if the screen/window is narrower than 850px, using the nomobile CSS class in Timeless. Would it be a good idea to hide the English Wikipedia main page "Welcome to Wikipedia" banner at small screen widths (for all desktop skins)?

Alternately, only the portal links could be hidden at some screen width, with the "Welcome to Wikipedia, the free…" text becoming aligned to the centre of the banner. (Both options should be possible with TemplateStyles in about a week and a half.) Jc86035 (talk) 11:56, 10 July 2018 (UTC)

When we have template styles, we can fix this properly much easier. I suggest we wait 2 more weeks for that. —TheDJ (talkcontribs) 12:03, 10 July 2018 (UTC)
@TheDJ: Would the .css page be started from scratch entirely, or something like Edokter's main page CSS (from the 2016 redesign) also be used? Jc86035 (talk) 11:27, 12 July 2018 (UTC)
Whatever we want, though we should probably be conservative considering the main page redesign history. Just move part by part into a stylesheet and then slowly rework with the new possibilities we will have. Here is an example of CSS for the main page of Mediawiki.org. —TheDJ (talkcontribs) 11:34, 12 July 2018 (UTC)

WikiProject participant lists[edit]

Rather than have editors self-assign themselves to project participant lists, I think it would be better have an automated tool that listed active editors. Ideally it would be configurable by individual projects (e.g.: the top 50 editors or editors with 10+ edits), but the tool itself would only run monthly, similar to the popular project pages. I have no knowledge of how to make it happen, so I'm here. –Fredddie 01:15, 11 July 2018 (UTC)

@Fredddie: This already exists at e.g. Wikipedia:WikiProject Directory/Description/WikiProject Video games. --Izno (talk) 03:27, 11 July 2018 (UTC)
Wow, I need to get outside of my WikiProject more. Thanks! –Fredddie 05:04, 11 July 2018 (UTC)

Giving schools the ability to access wikipedia with the option of disabling editing[edit]

While talking with a friend who works at a school, he told me how they set up a local wiki (which is view only) since they have students who often vandalise Wikipedia pages otherwise: they are too immature to know better and can't be reasoned with (say below 6-7th grade).

I think it would be great if we, on our part, somehow gave the entire school access to it with an option of disabling editing when needed. Plus point is: If the school feels the need to say, teach them how to edit, they can enable it whenever they're ready. I'd imagine not all schools would do what my friend's school did, they probably block the entire site.

I have not much clue on the technical feasibility of this idea. Ugog Nizdast (talk) 10:44, 12 July 2018 (UTC)

A couple of practical points, how does the systems manager distinguish between a staff account and a pupil account?
When I started WP, anyone could edit and that founding principle established the WP as the worlds greatest source of knowledge. The people who rely on it, and enjoy using it are determined to place restrictions and limitations on who may get involved - it is a case of 'look but don't touch'.
Schools also used to be places where thinking was developed- not where conformity was imposed. Is it correct to teach children that what they see in WP is the only truth, and they haven't the right to be consulted?
If we are talking about selection by maturity, ability, political persuasion I am sure the majority of WP wouldn't trust their own government to run a bath (WP:POV).
Kids experiment, but it isn't too onerous to keep looking at our watch lists. Kids experiment but then start makng responsible edits when they find out it is more fun ClemRutter (talk) 11:42, 12 July 2018 (UTC)
Was just watching this topic, but cannot ignore the silliness there. Any systems manager will be easily able to distinguish between staff and pupil accounts. It has happened in every school I have worked in. (And that's a lot.) And nobody teaches children that what they see in WP is the only truth. HiLo48 (talk) 11:47, 12 July 2018 (UTC)
(edit conflict)Entities (such as schools) that have control of their network can effectively block editing or logging in a number of ways (most of which involve implementing an ssl intercepting internet filter). — xaosflux Talk 11:49, 12 July 2018 (UTC)

Better filtering of "own work" image uploads[edit]

There's lots of images uploaded to en and/or commons which are tagged as "own work", but clearly are not. See this conversation on my talk page for a recent example. In this case, I believe the uploader honestly didn't understand that what they were doing was a copyright violation. In many other cases (paid spammers), they simply don't care. In either case, we need some better filtering.

In most of the cases I see, the uploaded images have no EXIF data. There can be legitimate reasons for this, but in the overwhelming majority of cases, it's a clue that somebody deliberately stripped out the EXIF copyright statement. Virtually every camera in existence puts in at least some basic EXIF boilerplate automatically. I strip out the copyright EXIF data when I upload images to commons, but that's on purpose; I have my camera set up to automatically insert my copyright, but for stuff I'm putting on commons, I want to disown that. I do, however, leave the rest of the EXIF data, including the Author. For example, File:George_Faile_Gravesite.jpg.

So, I'm thinking we should require images which appear to be photographs to have EXIF data. Or, at least, detect that they don't and require the uploader to enter an explanation of why they don't, and flag those for human review as likely copyvios. -- RoySmith (talk) 12:59, 12 July 2018 (UTC)

If it were required spammers would probably just start adding fake EXIF data (or stop stripping it out). It might be better to have the software warn the user, but not tell them why and not prevent them from uploading the image. Jc86035 (talk) 13:20, 12 July 2018 (UTC)

Concept for a new logo for Wikipedia[edit]

Here it is: https://ibb.co/dhQoro Anyway, see you later! Peppa Pig the Second (talk) 14:08, 13 July 2018 (UTC)

Running our own archive service?[edit]

Has any thought been giving to running our own archiving service for references? Right now, some people use archive.org, that doesn't seem very reliable. I can't actually remember the last time I got it to work, and I've given up trying. As reference links go dark, we're losing a major part of the value of the encyclopedia. Even if it was reliable, by having our own service, we would be able to have a cleaner integration with the editing tools. -- RoySmith (talk) 18:49, 13 July 2018 (UTC)

@RoySmith: Does Save Page Now not work reliably? I find it doesn't work on some pages but those usually work with archive.is or https://webrecorder.io, although I am inexplicably only able to access the former through the Tor network (and through Webrecorder). Webrecorder requires an account and content is attached to that account (which has a limited amount of space), but can capture dynamic content like YouTube videos and flash games. Jc86035 (talk) 20:49, 13 July 2018 (UTC)
I honestly don't remember how I used to access the archiver. What is "Save Page Now"? -- RoySmith (talk) 02:25, 14 July 2018 (UTC)
@RoySmith: If you go to https://web.archive.org/save/http://your.url.here, you can save most URLs and most of the embedded content used on those pages. You can also access this through the web.archive.org landing page, by typing in a URL. (Or were you referring to accessing saved URLs? I assumed you were referring to archiving pages.) Jc86035 (talk) 05:48, 14 July 2018 (UTC)
Personally I think that archive.is runs a better service. It saves pages more accurately than archive.org and WebCite and has good uptime. It would be a major undertaking for Wikipedia to offer a similar service, so it looks like we are stuck with third party services for the time being.--♦IanMacM♦ (talk to me) 05:57, 14 July 2018 (UTC)
Managing permissions, rights and licenses for archived material looks like a major headache. (Copyright law varies from place to place.) Wikipedia avoids such issues by having a CC-BY-SA licence for everything. This said, archiving cited material would obviously make Wikipedia much more resilient, so the idea should at least be studied. Aerkem (talk) 11:40, 14 July 2018 (UTC)
If we do not run own service we could improve the page at Help:Archiving a source. It seems like you lot have views on which service are better and they could be included in that article. Emir of Wikipedia (talk) 17:04, 15 July 2018 (UTC) P.S.: Please ping me in your reply. Thank you.
Help:Archiving a source was created a few weeks ago. IMO it's a stub fork of the long-standing WP:Link rot which itself needs an overhaul I started on at Wikipedia:Link_rot/new but haven't had time to finish. -- GreenC 14:37, 16 July 2018 (UTC)

This has been raised a few times, but [at least in those threads I've seen] hasn't led to any sort of formal proposal. It's certainly something that would need to involve WMF tech folks. I'd be surprised if they would be game, since we have such a good relationship with the Internet Archive. I definitely do not think we should be relying on the others, like archive.is. That's not because it isn't a good service, but because it's largely a mystery. AFAIK it's the work of a single person who has expressed that he doesn't know how long it will be sustainable. It has also been blacklisted here in the past (not currently) because of spamming and general sketchiness. Most importantly, we don't know anything about these organizations sufficient to place faith in them to prevent link rot when the Internet Archive, by contrast, has a clear organizational structure, large user base, and is generally well-known as an entity beyond just the service itself. There are a bunch of ways to archive pages and access archived pages, and if those aren't working for you that problem is most likely on your end. It's slow sometimes, for sure, but across all of my computers over the years I don't think I've had any trouble. We also have bots that can do this automatically such that there's limited need for other editors to do so. What's lost in the Internet Archive, however, are those things that prevent archiving via robots.txt. Archive.is ignores robots.txt, and so may make sense to use in those cases when it's the only option. If Internet Archive isn't willing to bypass robots.txt, it's unlikely WMF will be. Scattered thoughts, sorry. — Rhododendrites talk \\ 18:00, 15 July 2018 (UTC)

You bring up a good point about the organizations. I think Perma.cc is probably the one I would put the most faith in due to its backing from universities in multiple continents. {{ping|Chetsford|GreenC|Cyberpower678} You three have discussed on the Perma talkpage so you might want to contribute here. Emir of Wikipedia (talk) 21:13, 15 July 2018 (UTC)
@Chetsford, GreenC, and Cyberpower678: fixing the ping format.
OK, I just tried the suggestion above, with https://web.archive.org/save/http://your.url.here. It did indeed work. I don't remember exactly what problems I had the last time I tried it. Be that as it may, the process is still pretty rough. I remember when I first tried this, I found my way to ((IIRC) Help:Archiving a source. That page doesn't tell you how to do it. It gives you a choice of five ways to do it. So, the newbie user is immediately faced with having to make decisions.
My suggestion about doing our own was not so much about where the data would live, but about controlling the process so we could build simple interfaces. I'm fine with using one of the other services.
But, we should pick just one, commit to it, document how it's used, and build interfaces to it which are dead simple (perhaps, totally automatic, and integrated into the visual editor). -- RoySmith (talk) 00:09, 16 July 2018 (UTC)
Rhodo, IA as of a year or two ago is ignoring robots.txt. --Izno (talk) 00:20, 16 July 2018 (UTC)
Whoa. If that's the case, we should really start to replace archive.is links sitewide. @Beetstra: (someone I remember being part of previous archive.is threads) is this something you're aware of? — Rhododendrites talk \\ 03:04, 16 July 2018 (UTC)
@Rhododendrites: No, I wasn't. Maybe it is then time to replace the links. I am however also very interested in whether we can have our own archiving system (in a way, WikiSource is such a system, for out of copyright stuff that is). --Dirk Beetstra T C 09:59, 16 July 2018 (UTC)
@Rhododendrites: Wayback also ignores robots, except in certain situations so it would be unfair to target one provider or another as Wabyack itself is not consistent ("not consistent" broadly speaking even within its own policies of what to ignore or not). Robots.txt was not designed to determine if a page should be archived or not it was sort of a hack mechanism. Why do we care if a provider ignores robots.txt out job is WP:V and if a provider has the page better for our readers. -- GreenC 16:05, 16 July 2018 (UTC)
  • I just wanted to note that as part of the Knowledge Integrity program at the WMF, we're aiming to work with the Internet Archive to set up a process of near-instant automated archiving of all citation links added to Wikimedia projects this year. You'll be able to follow our progress on this at phab:T199193, though the phab board is still in the process of being fleshed out more fully. Samwalton9 (WMF) (talk) 10:57, 16 July 2018 (UTC)
    • Thanks for that link. I was previously unaware of that effort. -- RoySmith (talk) 11:56, 16 July 2018 (UTC)

Through my bot WP:WAYBACKMEDIC I've worked with dozens of web archives warts and all. There is no single best archive service, they each have pros and cons. All other things equal, we default to archive.org but they don't have everything. The full list of services is WP:WEBARCHIVES, its around 20 or so with more to be added.. we have a rich and diverse group to draw from. I could list many problems with archive.org and I'm concerned about over-reliance of them, but they have the broadest coverage. I would support Wikipedia having its own archiving service, so it can solve the problems other archive services cant/wont solve, so we can tailor the system to the needs and requirements of Wikipedia, to prevent loss of links by providers outside our control, add new features and services. From a technical standpoint the current system of third parties is incomplete, error prone and difficult to manage. The technology in use at webrecorder.io (open source) is probably the future of archiving, there is no way around it with modern websites moving to JS, video, website aging, etc.. each day that goes by without this technology deployed against Wikipedia is a forever loss of cites, it's sad to see no one doing anything about it. User:Jc86035 wrote a good summary of the situation at T199193. Waiting for a third party provider to solve it for us may or may not be the best thing for Wikipedia, another example of why using third party providers is a problem (keeping up to date with technology). -- GreenC 15:55, 16 July 2018 (UTC)


RfC: Can Administrators delete pages under General Sanctions without discussion?[edit]

This RfC is NOT live. Please do not add any !votes. This RfC is being presented for feedback to ensure it is neutrally worded and will produce a usable consensus. The proposed text of the RfC is below, starting at "Background" and ending after "General Discussion".

Background[edit]

Recently, general sanctions were placed on cryptocurrency and blockchain broadly construed. (discussion). An administrator, User:MER-C then deleted Universa Blockchain Protocol, citing only the general sanctions in their rationale. This action proved controversial, and was reviewed at Wikipedia:Deletion review/Log/2018 July 9#Universa Blockchain Protocol. Parallel threads also popped up at the Administrator's Noticeboard here and CSD here. The Deletion Review ended with no consensus, recommending "The community discussion needed to resolve the apparent policy conflict ... needs to happen in a wider venue, such as in a policy RfC". This is that RfC.

Can administrators delete pages under general sanctions without waiting for community discussion?

Please support only the option closest to your views.

General sanctions do not authorize speedy deletion[edit]

IAR notwithstanding, administrators may only delete pages without waiting for discussion when the page meets one or more conditions listed at WP:CSD. General sanctions have no effect on deletions. Deletions in topic areas covered by GS may be appealed as usual.

Support[edit]

General sanctions may be invoked in addition to a valid speedy criterion[edit]

When deleting a page, an administrator may specify that the page is being deleted under general sanctions in addition to a valid speedy deletion criterion. This flags the deletion as an enforcement action. Appeals of these deletions must be appealed as a GS enforcement appeal at the Administrator's Noticeboard

Support[edit]

General sanctions may be invoked instead of a valid speedy criterion[edit]

When deleting a page, an administrator may specify that the page is being deleted under general sanctions instead of a valid speedy deletion criterion. Administrators may, at their discretion, delete pages in topic areas covered by general sanctions, without waiting for discussion and even if no speedy deletion criteria apply. Such deletions are GS enforcement actions, and may be appealed only to the Administrator's Noticeboard

Support[edit]

General discussion[edit]

Discussion on RfC Formulation[edit]

Commentary on how this RfC is presented, especially with respect to its neutrality and the usability of the consensus it will produce, is welcome here. An additional question is the best venue for the RfC, I'm not sure whether WP:VPP or WP:AN is the best location. Pinging some editors who have shown an interest in the discussion: @TonyBallioni:@MER-C:@SmokeyJoe:@Primefac:@Sandstein:@Stifle:@Monty845:. I intend to initiate this RfC in 24-72 hours unless major issues are presented. Tazerdadog (talk) 02:22, 17 July 2018 (UTC)

  • Prefer the RfC to be on its own subpage, for watchlisting, dedicated talk page, and ease of linking. —SmokeyJoe (talk) 02:26, 17 July 2018 (UTC)
  • My preferred position would be: "Unless the community explicitly authorizes special Deletion rules (Speedy or otherwise) as a community imposed Sanction for a given area of conflict, then General Sanctions do not apply to the administrative action of deleting." Disruption effecting deletion discussions would still be subject to sanction, and any deletion appeals could be handled via normal channels, and not be subject to the special appeal rules for Sanctions. Monty845 04:13, 17 July 2018 (UTC)
  • I see no need to have an RfC for this and would oppose every above option as pointless: we already have an appeal procedure that would have answered this if it had been correctly followed, review at AN. Instead we had a DRV with unclear jurisdiction that closed as no consensus. If a deletion claiming DS/GS happens again, then just appeal using those procedures. The community can rule there definitively whether or not it sees it as a reasonable use of sanctions. This was a large part of the reason I opposed the recent DRV: there is a 100% clear appeal provision for GS that could have settled this within a week. Even if one thinks this isn’t a valid use of DS/GS, AN is the place to clarify it via appeal, because that’s where we evaluate claims under general sanctions (and consensus easily could have been no this was not authorized).
    Now we have the option of a complete mess of an RfC (which this will be if it is held) or waiting for a new deletion sanction under this. The easiest way is just to have AN rule on it through an appeal if one ever comes up (and because of the complete mess of a DRV, the current case is too fragmented to be used as a test, IMO.) Wait for a new deletion you disagree with, and appeal it then. If one doesn’t happen, then there’s no need at all for an RfC. An AN appeal will take at most 72 hours compared to over a month of community time being spent on an RfC that could well end in no consensus. TonyBallioni (talk) 04:22, 17 July 2018 (UTC)
    • An RfC is needed. Some, like you, are supporting a deletion power grab by WP:AN, under the guise of Discretionary Sanctions, claiming deriveded authority of ArbCom. Others, like me, do not support a high throughput drama page being able to quickly enact new deletion processed, bypassing Deletion Policy, and in particular WP:CSD. You have recently expressed contempt for WT:CSD, and what you are arguing for is a significant watering down of CSD policy, quoting the long standing opening paragraph as

      "The criteria for speedy deletion (CSD) specify the only cases in which administrators have broad consensus to bypass deletion discussion, at their discretion, and immediately delete Wikipedia pages or media. They cover only the cases specified in the rules here"

      WP:CSD, with its simple and formal objective criteria even for new criteria, and WP:DRV, have long been pillars for calm contemplative discussions for the long term good of the project. WP:AN is in contrast a high speed dramaboard. You seem to prefer bitcoin deletions to be under the control of AN, which might be easier and more effective in the long term, but it is a clear step in the direction of the kangaroo court usurping control of the project. This will be a discussion on long standing principles of deletion policy and review versus expansions of the scope of ArbCom and its delegations, a 1-2 month RfC, on its own page, advertised at WP:CENT is appropriate. In the meantime, WP:CSD#G11 suffices for bitcoin spam pages. --SmokeyJoe (talk) 05:19, 17 July 2018 (UTC)
      • I went ahead and posted the one deletion I made at AN for review. It will both give feedback on that individual action, and hopefully clarify if there is a firm consensus on the issue. If people still think there is a need for an RfC going forward, that is fine, but we do have a fair amount of individual admin actions here, and if we can get a consensus both for them and as a whole, that would also be helpful if an RfC is needed. Also, for the record, I don't have contempt for WT:CSD, I just think it's a horrible place to actually try to make changes to the CSD policy because it tends to lean hyperinclusionist in my view. If one wants to discuss actual changes there, VPP is much better. It is good at getting a read on the most cautious reading of the current CSD policy, which is something I generally advocate for myself in ordinary situations. TonyBallioni (talk) 05:50, 17 July 2018 (UTC)
        • TonyBallioni, I think I see what your saying, but I maybe "hyperconservative" instead of "hyperinclusionist". Hyperconservative at agreeing to any change. --SmokeyJoe (talk) 04:09, 18 July 2018 (UTC)
    • You're begging the question. You've gone ahead and assumed that there is a policy or consensus that deletions under general sanctions are not appealable to DRV, which is wrong. Stifle (talk) 14:27, 17 July 2018 (UTC)
  • Oppose anything that either
    1. Grants authority to create new speedy deletion criterion (in name or in effect) without a consensus formed at a discussion at, or flagged at, WT:CSD; or
    2. Ousts WP:DRV as the venue for appealing improper deletions. Stifle (talk) 14:25, 17 July 2018 (UTC)
  • I don't know where I would come down on the wordings of the three options for RfC, not loving any of them. I do think a CSD criteria around Bitcoin/cryptocurrency is called for as I don't think GS gives that authority. Simplest thing would be to include in A7 but could also support some new formulation (A12). Best, Barkeep49 (talk) 15:32, 17 July 2018 (UTC)
    • It wouldn't work as part of A7, and any new criterion would have to be in the G- series. Fully half of the deleted pages logged at WP:GS/Crypto were outside of the main namespace, with 6 in Draft:, 4 in User:, and 1 in Wikipedia:. —Cryptic 15:50, 17 July 2018 (UTC)
      • That's an excellent point and I've struck that sentence. My larger point that I think CSD is needed but not covered by GS and thus this RfC isn't the right path forward remains. Best, Barkeep49 (talk) 16:10, 17 July 2018 (UTC)
  • Based upon my admittedly cursory review of this proposed RFC, I think it looks fine to me, with the caveat that I support the proposed wording suggested by Monty845.--S Philbrick(Talk) 15:49, 17 July 2018 (UTC)
  • The question "All page deletions may be reviewed at WP:Deletion review", while touched by each of the two questions, should be put as well. Should a DRV nomination should be summarily closed due to an alleged association with a discretionary sanction?
"General sanctions do not authorize speedy deletion" has been answered well enough at Wikipedia_talk:Arbitration_Committee#Can_ArbCom_authorise_arbitrary_deletion_of_articles?. WP:BLPDELETE deriving from WP:BLPBAN has years of precedent.
"General sanctions may be invoked instead of a valid speedy criterion" is tending to moot with early agreement by a small number of people at WP:AN (here and here) that GS associated deletions should mention the GS in addition to the applicable CSD criterion, with the note that In areas where discretionary sanctions, authorized by the arbitration committee or the community exist, administrators have increased leeway in applying the speedy deletion criteria.
Challenges to discretionary sanction go to WP:AN. I think the authors of that expected challenges to focus on behavioural matters? Was the rebuked editor violating a sanction? Was the admin UNINVOLVED. Was the sanction imposed (typically a WP:BLOCK) over-harsh, or to be reconsidered due to wider circumstances or a subsequent apology? Page deletion disputes have always been reviewable at WP:DRV. At DRV, rarely does behavioural issues come into it, except for threshold conditions such as appeals by new WP:DUCK accounts being often summarily closed.
A question that the RfC should consider is whether a DRV nomination should be summarily closed due to an alleged association with a discretional sanction.
A content-based question that might be better reviewed at DRV is whether some valid notable topics are being excluded by misinterpreted and overzealour discretionary sactions.
Admittedly, this is all mostly wonkery. Where a discussion is held should not matter. Any editor may post at DRV and AN. Holding two reviews in parallel would be silly if discussing the same thing. Has there even been a genuine dispute over the deletion per se of page under GS. For the case of Universa Blockchain Protocol, the nomination at Wikipedia:Deletion review/Log/2018 July 9#Universa Blockchain Protocol did not challenge that the page should not have been deleted, but questioned the process.
There could be simple solutions to the silly question. A challenge at forum could be cross-posted to the other. A review at one could be transfered to the other, instead of summarily closed. A discussion page could be transcluded to both places.
My suggestion is that a DRV discussion touching a discretionary sanction issue must be advertised at WP:AN. Probably also at Wikipedia talk:General sanctions.
--SmokeyJoe (talk) 03:22, 19 July 2018 (UTC)

Talk page wikilink[edit]

In the talk page banner the this is not a forum phrase links to what Wikipedia is not. I was wondering would it be a good idea to add another sentence saying for factual questions about the subject see Wikipedia:Reference desk? I put it in here instead of proposals in case someone could improve the idea. Mobile mundo (talk) 19:07, 17 July 2018 (UTC)