It has always seemed strange to me that we can freely delete articles created by someone banned by the community, but not explicitly for someone banned by the Terms of Use, though G5 could certainly be interpreted in that way. I see two options:

G5. Creations by banned or blocked users

This applies to pages created by banned or blocked users either in violation of their ban or block or by virtue of violation of the Wikimedia Foundation Terms of Use, and that have no substantial edits by others. G5 should not be applied to transcluded templates or to categories that may be useful or suitable for merging. For a banned or blocked user:

• To qualify, an edit or article must have been made while the user was actually banned or blocked. A page created before the ban or block was imposed or after it was lifted will not qualify under this criterion.
• To qualify, the edit must be a violation of the user's specific block or ban. Pages created by a topic-banned user may be deleted if they come under that particular topic, but not if they are legitimately about some other topic.
• {{Db-g5|name of banned user}}, {{Db-banned|name of banned user}}

Gx. Violation of Terms of use

This applies to pages created in violation of the Wikimedia Foundation terms of use, and that have no substantial edits by others.

To qualify, the edit or article must have been made by a user demonstrated or admitted to be in violation of the Terms of Use. Specific categories of violation include:

• Harassing and Abusing Others
• Violating the Privacy of Others
• Engaging in False Statements, Impersonation, or Fraud
• Committing Infringement
• Misusing Our Services for Other Illegal Purposes
• Engaging in Disruptive and Illegal Misuse of Facilities
• Paid contributions without disclosure

The article must have been created by an editor in violation of the terms of use and have no substantive content edits by others. Guy (Help!) 16:59, 23 February 2018 (UTC)[reply]

!votes

• Support as proposer. Guy (Help!) 16:59, 23 February 2018 (UTC)[reply]
• Support long overdue. Legacypac (talk) 17:46, 23 February 2018 (UTC)[reply]
• Oppose for all the reasons from last time. Propose speedy close of this RFC considering the last one ended less than six months ago and this proposal does nothing to address the reasons the last one failed. Regards SoWhy 18:35, 23 February 2018 (UTC)[reply]
• Support extension of G5 to include clear ToU violations by blocked or banned users prior to user's first block. Am in support of retaining the wording "this applies to pages...that have no substantial edits by others". ☆ Bri (talk) 19:17, 23 February 2018 (UTC)[reply]
• Oppose most of these are either covered by existing criteria (G3, G10, G12) or are a bit nebulous for speedy deletion criteria (e.g. point 3 would apply to articles which contain deliberate factual inaccuracies, even if not blatant). The major exception is undisclosed paid contributions, and while I'm sympathetic to a CSD on those grounds if we're going to have that then we should just say it rather than wrapping it up like this. Hut 8.5 19:34, 23 February 2018 (UTC)[reply]

  They are covered if you turn your head sideways and squint. I don't see why violations of the ToU should require creative interpretation of the rules. Guy (Help!) 19:42, 23 February 2018 (UTC)[reply]

  I think most of them are pretty clear:

  • "Harassing and Abusing Others" covers "harassment, threats, stalking" - would be G10 or G3, "spamming" is the definition of G11, "vandalism" is the definition of G3
  • "Violating the Privacy of Others" - most of this is pretty nebulous and covers content forbidden by any applicable laws, which turns us into lawyers if we want to enforce it properly. The one specific case is "Soliciting personally identifiable information from anyone under the age of 18 for an illegal purpose or violating any applicable law regarding the health or well-being of minors" - this would already be nuked from orbit under Wikipedia:Child protection.
  • "Engaging in False Statements, Impersonation, or Fraud" covers "libel or defamation" - pretty much the definition of G10, "posting content that is false or inaccurate" - if blatant that's clearly G3 (hoaxes), if subtle it's something we'd want reviewed through another process, "Attempting to impersonate another user or individual" and "Engaging in fraud" are a bit more nebulous but they don't happen often and I don't think it's much of a stretch to delete either under G3
  • "Committing Infringement" is basically just copyright violations, which come under G12
  • "Misusing Our Services for Other Illegal Purposes" nebulous stuff covering content which violates any applicable laws, except for child porn
  • "Engaging in Disruptive and Illegal Misuse of Facilities" - this is basically various ways of trying to hack the site, which would come under G3.

  So they're all either very nebulous or basically covered under existing criteria except undisclosed paid editing and paedophilia, and I don't think we need a new criterion to get rid of the latter. Hut 8.5 19:59, 23 February 2018 (UTC)[reply]

  I am wondering about "child porn". I've seen deletions on Commons for that reason, sometimes performed by WMF staff and sometimes by volunteer admins (presumably when they make a "better safe than sorry" deletion before calling the Foundation). "Privacy of others" I usually see deleted under the "non-public private information" rationale or some euphemism. Jo-Jo Eumerus (talk, contributions) 20:13, 23 February 2018 (UTC)[reply]

  I am mainly trying to remove the distinction between someone who is not allowed to edit Wikipedia because we say so, in which case their articles get nuked, and someone who is not allowed to edit Wikipedia because the Foundation says so, in which case they don't. That is back to front. As it stands, to delete an article created by an impersonator, we'd first have to decide that the impersonation was harassment. That seems bureaucratic. Or imagine if someone dropped an article on a school shooting survivor who dared to open their mouth. It may not be obviously harassment, it may be a fanboi even, but it could very well still be an obvious infringement of privacy. As I say, the thing that seems inconsistent to me is that if we say you may not edit Wikipedia then G5 applies, but if the Foundation does, we have a potential drama-fest. I do like the idea of making blocks for violation of ToU effectively retroactive, but it may be a bit rouge for some. Guy (Help!) 20:52, 23 February 2018 (UTC)[reply]

  Wouldn't we delete those articles because they are bad articles instead of caring about who created them? —Kusma (t·c) 20:57, 23 February 2018 (UTC)[reply]

• Oppose, "undisclosed paid editing" can't be found out without some digging, and may need to be dealt with through some other process. Everything else is already deletable (G3 and G10 are pretty much a catch all for most abuses and script kiddies). Also, undisclosed paid editing is either bad (and then most of the results is deletable under A7 or G11) or good, in which case other editors might want to adopt the page and do some rewriting. Either existing criteria are enough or speedy deletion isn't a good answer. —Kusma (t·c) 20:56, 23 February 2018 (UTC)[reply]
• Procedural Oppose and also suggest a speedy close per SoWhy. Proposer should consider opening a separate discussion without the straw poll and see what can be crafted, rather than re-arguing the exact same subject from 6 months ago. --Izno (talk) 21:08, 23 February 2018 (UTC)[reply]
• support this is a useful umbrella ToU construction. It would be helpful to avoid having the community waste more time - the lifeblood of this place - dealing with bad faith contributions. The community is moving itself steadily to deal with steady onslaught of promotional articles that flow into WP based on the notion that WP is an essential platform for promotion for companies, authors, actors, celebrities, etc etc. There is no doubt that many people see WP that way. We have been dealing with that, for instance with ACTRIAL, automatically community banning serial socks (which are generally paid editors using throw away accounts), raising NCORP standards and the like. Getting this passed (yes it is somewhat perennial) is another essential tool. What we all want it to spend our time building an encyclopedia and not to waste so much time dealing with industrial waste that has been dumped into our beautiful project. Jytdog (talk) 21:51, 23 February 2018 (UTC)[reply]
• Oppose too much of a wide remit, as an editor blocked for being disruptive may have created some good articles not connected to his block. Would support just adding UPE but these can also be deleted by prod as for example last month I prodded 4 UPE articles and 3 were deleted Atlantic306 (talk) 11:49, 24 February 2018 (UTC)[reply]
• Oppose Basically per my questions below; I understand the theory, but in practice this doesn't make sense to me. Basically, it boils down to two criteria: undisclosed conflict of interest, and illegal activities. The latter does not need a CSD for it to be swiftly dealt with, and the former, as noted, requires significant effort to out, not lending itself to CSD anyway. More to the point, it feels backward. These would apply to pages made illegally/via undisclosed paid editing, but you're talking about the users. It's not like the foundation regularly calls down to enWiki saying "Hey this editor here is trying to do illegal things, y'all should delete their pages." It would be an editor here determining that "this page violates the ToU" and then taking action, but that 1. is already what happens, and 2. does not need a CSD to be effective. On the off-chance the foundation finds something before we do, there's G9. I get the logic, but it doesn't seem to play given the reality on the ground. I suppose I could support adding something (to G3?) like attempts to violate US law but that just seems overly fraught and not particularly helpful. In short, a well-meaning solution to a problem we don't have. One could also make the argument that, once someone violates the ToU, they're "banned" whether they or the community knows it, thus G5 would apply, but that's... weird. ~ Amory (u • t • c) 12:32, 24 February 2018 (UTC)[reply]

Consider undisclosed paid editing, then. It is not permitted per the ToU, but some people will, almost as if to make a point, oppose deletion of "good" (subjectively defined) content because they repudiate the no undisclosed paid editing rule. So actually the current situation causes exactly the issue you identify: it complicates the cleanup of edits that should never have happened in the first place. G5 was written for exactly that. An edit should not have been made, an article was created by a user sh hould never have been editing, so it is nuked. But if the user cannot be tied to some previously blocked spammer, at present, we can do nothing. Of course, quite a few of these "brand new" spammers will be old spammers returning. A motivated spammer will have little difficulty in circumventing CU, even. And under current rules, on;y the WP:OFFICE can speedily nuke articles created in violation of the ToU, unless someone can definitively link the spammer to another spammer already blocked. And even then, if the other spammer was blocked after the first spammer registered, we can only speedy from the first bloc, so earlier spam will remain. Someone can register a dozen accounts and until the first one is blocked, all articles created by all the accounts are not eligible for speedy, even though the abusive behaviour was there from day 1. Guy (Help!) 23:09, 25 February 2018 (UTC)[reply]

I still think you're mixing cases here. If an article is spammy to the point of violating ToU, it's an easy G11. If it's not obvious spam, then not only would these criteria not apply, but we wouldn't even know we had a problem anyway. Like I asked below, it matters whether these are designed for an article or a user violating the ToU. You've said the article, but you're arguing the user. ~ Amory (u • t • c) 01:33, 26 February 2018 (UTC)[reply]

"If an article is spammy to the point of violating ToU, it's an easy G11."??? No, that is not true. It is very easy to make a spammy article G11-proof, ("pages that are exclusively promotional") just add a small amount of properly sourced material that can be re-used when the spam is cut. And violating the ToU, being an Undisclosed Paid Editor, is no impediment to G11-proofing the not "exclusively promotional" spam. The typical ToU violating UPE writes a page using some properly sources facts and pads it with veiled promotion. These pages are not G11 eligible. --SmokeyJoe (talk) 01:55, 26 February 2018 (UTC)[reply]

• Support - it looks like this is just about undisclosed paid editors. UPEs are banned by the TOU from adding anything - they are just not allowed to edit. Not removing their edits once they've been found out just looks like an attempt to nullify the ToU. Why do folks want to do that? Smallbones_(smalltalk) 23:50, 24 February 2018 (UTC)[reply]
• Support. The “Terms of Use” needs teeth. Specifically, “No Undisclosed Paid Editing” needs teeth, and the onus for communication needs to be on the paid editor. I would be happier if this were tied to “promotional content”, specifically for-profit companies, their products, or their CEOs/founders. Leave open undeletion options if the editor(s) subsequently properly disclose, or if an experienced editor in good standing offers to take responsibility for the article and future activities of the paid editor. —SmokeyJoe (talk) 00:49, 25 February 2018 (UTC)[reply]
• Support While many, if not most, ToU violations are addressed by other CSD criteria we should have this which directly addresses breach of ToU. In particular having the ability to reach back to the time of the first breach of ToU and deleteall "fruit of the poisonous tree". Right now bad actors have incentive to violate ToU (particularly UPE) because they still get their articles in. Beyond that we simply need a specific way to enforce the ToU that does not get caught up in all of the 'if, well, AGF, blah, blah' crap that comes up when we try to back-door ToU enforcement with other CSD criteria. Jbh ^Talk 04:05, 25 February 2018 (UTC)[reply]
• Support – in principle – option 2, "Gx. Violation of Terms of use". We have considerable precedent for such speedy deletions, amply sufficient to show that the community does not oppose them in principle. This goes back at least as far as Orangemoody, where a large number of articles were deleted at essentially the same time as the check-user blocks were made (when they were thus not technically eligible under CSD G5). It should be perfectly, dazzlingly obvious to everyone that content created in violation of the terms of use cannot be kept and must be immediately removed, but until we formalise that in local policy the "not supported by policy" argument will continue to be put forward. So it's high time we did this; the precise wording and scope will need to be hammered out. Justlettersandnumbers (talk) 14:06, 25 February 2018 (UTC)[reply]
• Oppose - Criteria should generally focus on page content, not editor behavior, expect when narrowly deemed necessary (e.g. WP:G5 and WP:XCSD); this is much too broad. Furthermore, generally, pages "Harassing and Abusing Others" are eligible for deletion per G3 or G10; pages "Violating the Privacy of Others" are eligible for oversight; in regard to "Engaging in False Statements, Impersonation, or Fraud", "Engaging in False Statements" is too broad and unfortunately subjective; pages "Committing Infringement" are eligible for deletion per G12; in regard to "Misusing Our Services for Other Illegal Purposes" and "Engaging in Disruptive and Illegal Misuse of Facilities", a separate discussion on these two aspects may be due; "Paid contributions without disclosure" was already shot down at Wikipedia talk:Criteria for speedy deletion/Archive 66#New criteria. A new or expanded criterion should not be too broad or overly redundant. — Godsy ^(TALK_CONT) 19:52, 25 February 2018 (UTC)[reply]

Did you notice how the terms of use don't say you can do these things if the content is good? And neither does G5? The whole point of "edits that are not permitted int he first place" is that they are, well, not permitted, regardless. Guy (Help!) 22:47, 25 February 2018 (UTC)[reply]

However, neither the ToU nor the current WP:PAID policy does not contain any language that require deletion of such material. As I said last time, IMHO no change to this policy should be made without first updating WP:PAID as described on that policy's page. Regards SoWhy 13:09, 26 February 2018 (UTC)[reply]

• Oppose This is covering way too much stuff. The G5 proposal really does nothing, because the WMF actually bans people (the WMF itself uses that word) and banned means banned whether WMF or us. The GX mostly duplicates what we already have. The arguable new items (privacy, paid-editing) should have their own stand-alone discussion. Oiyarbepsy (talk) 00:31, 27 February 2018 (UTC)[reply]
• Support, though as UPE goes, we should consider everything a UPE creates to be covered by G5. They were never allowed to be editing in the first place, and were therefore for all intents and purposes defying a ban. Seraphimblade ^{Talk to me} 17:49, 27 February 2018 (UTC)[reply]
• Oppose as redundant and unnecessary. If the content created by a TOU-violating user is bad it is already deletable. If the content isn't bad we shouldn't be deleting it. If it's unclear whether it's good or bad then send it to AfD/MfD. If there is specific content that you think is bad but which isn't deletable currently, then get a consensus about that content - either it will become deletable or you'll find that the consensus is that it isn't bad. Thryduulf (talk) 00:49, 28 February 2018 (UTC)[reply]
• Support, mostly moral support since this is highly unlikely to pass. Of course, deleting anything by UPEs is a necessary part of an incentive system to encourage disclosure. Regretfully, the community has already rejected the idea of an UPE CSD, so not sure if veiling this as something else is helpful to the overall effort. There are many different angles from which this problem can be approached. One is extending the G5 criterion as proposed by Oiyarbepsy below. Another is a TOU Prod that was proposed and gained significant support following the previous RFC regarding a TOU CSD. Rentier (talk) 03:29, 1 March 2018 (UTC)[reply]
• Oppose - this is far too broad, and many of the bullets listed in the proposal are things likely to require a discussion to determine that a ToU violation has occurred, meaning the criterion is neither objective nor uncontestable. I would support a much narrower CSD on deletion of articles created by undisclosed paid editors. Ivanvector (^Talk/_Edits) 14:10, 1 March 2018 (UTC)[reply]
• Oppose Comment per bot request on user page. I strongly oppose conflating editor behaviour with article constriction and maintenance. If an article is poorly written we have means of taking care of that. Some of the language invites harassement, for example, allegations of false statements. Who decides what is false. I see the potential for abuse in attempts to control article space as well as damage to well meaning editors.(Littleolive oil (talk) 16:14, 1 March 2018 (UTC))[reply]
• Oppose the scope of the proposed criterion (Gx) is huge and too large to be effective. In addition, a number of items on the dot point list would likely need discussion to determine whether there actually was a ToU vio. The largest change (as pointed out above) is this proposal is deleting pages created by UPEs which should be proposed separately given that it's a rather significant change hidden in this proposal. For the record, I'm opposed to blanket or speedy deletion of pages created by UPEs. This is for range of reasons, including that whether someone is a UPE or not isn't usually clear so will need more consideration than a speedy criteria, in addition to Thryduulf's point. Callanecc (talk • contribs • logs) 05:07, 5 March 2018 (UTC)[reply]
  • Callanecc, I said "support", but pretty much agree with all of your details. I guess I assumed I was supporting moving forwards for round in which the proposal would be tightened. On "opposed to blanket ...", what would you think of my idea below (23:38, 1 March 2018) of making a repository of discovered UPE product, blanked until the author satisfactorily discloses? --SmokeyJoe (talk) 05:14, 5 March 2018 (UTC)[reply]
    • If there's actually a problem with the content sure, but if there's nothing wrong with what they've written then why would we remove it (it could, I guess, be copied or transcluded into a repository. Deal with the editor's behaviour sure, but why pull the article if there's nothing wrong with it only possibly the person who wrote it. This isn't the same situation as a banned editor who knows that they aren't allowed to edit the project and needs to be shown the door. This is someone who may not know they're doing anything wrong, and if they do would likely be covered by G5 anyway. Before an article is deleted per G5, an admin needs to be sure that the editor was blocked or banned when it was written, it's much more difficult to be sure that someone is an UPE and we shouldn't be deleting/blanking a quality article on the possibility that someone might be violating the ToU when the article is good. Callanecc (talk • contribs • logs) 05:50, 5 March 2018 (UTC)[reply]

RfC discussion

It is weird that an editor who is blocked for some minor infringement will be subject to G5, but an editor who is violating a Wikimedia Foundation mandated policy, for example by impersonating someone, is not. Some of the ToU can lead to speedy deletion:

• Harassing and Abusing Others - G10
• Violating the Privacy of Others - RevDel / oversight, potentially G10
• Engaging in False Statements, Impersonation, or Fraud - potentially G10, may be G3
• Committing Infringement - G12
• Misusing Our Services for Other Illegal Purposes - includes child porn and other issues, not covered by CSD at this time (though undoubtedly likely to be nuked per WP:IAR)
• Engaging in Disruptive and Illegal Misuse of Facilities - e.g. viruses, malware etc, would typically be handled as IAR but not covered by CSD
• Paid contributions without disclosure - not covered by CSD, may qualify as G11 but PR material is often not blatantly promotional.

Violations of the ToU are grounds for indefinite blocking or banning, but we make this worth the gamble because some or all of the articles may "stick", and thus the abuser gets their abusive article, spammer gets paid or whatever. Guy (Help!) 16:59, 23 February 2018 (UTC)[reply]

A few questions to make sure I'm reading this correctly:

1. Would this apply to any page created by such a user, regardless of whether the page itself violates the TOU? The first and last sentences seem to contradict the second sentence on this point.
2. Who would be determining when the editor has violated the ToU?
3. Would this apply only once the user is indefinitely blocked, or before? (relates to the above)

~ Amory (u • t • c) 18:03, 23 February 2018 (UTC)[reply]

I think the page creation would have to be a violation, it's only banned users where we would apply a scorched earth policy. Guy (Help!) 20:56, 23 February 2018 (UTC)[reply]

• Okay, so this is basically a CSD for "paid contributions without disclosure that are not advertising or about non-notable people" (everything else is already deleted quickly and does not require a new CSD). How often do we currently have pages like that in our other deletion processes? —Kusma (t·c) 18:27, 23 February 2018 (UTC)[reply]
  • I've been tagging articles in a current cleanup case as g5/g11, e.g. Matan Gavish. Some interpretations of g5 say that tagging is improper if the creation was prior to the user's first block. I think the gist of this is to get around that technicality and make g5 retroactive if there was clear ToU violation. ☆ Bri (talk) 18:33, 23 February 2018 (UTC)[reply]

Good question. I would say the page would have to be a violation. So: a user that creates a page with an exploit, the page would be deleted and the editor banninated. Guy (Help!) 19:40, 23 February 2018 (UTC)[reply]

• Is it me or didn't we just have this discussion less than six months ago? Regards SoWhy 18:31, 23 February 2018 (UTC)[reply]

Yes, and the tide of crappy PR articles continues to rise. We have had several paid sockfarms uncovered since then. Guy (Help!) 19:40, 23 February 2018 (UTC)[reply]

You might want to present your evidence before starting a new RFC without addressing the reasons the last one failed. Regards SoWhy 20:23, 23 February 2018 (UTC)[reply]

Ah, so you're not familiar with WP:COIN then? I understand now. Guy (Help!) 20:45, 23 February 2018 (UTC)[reply]

Please avoid such comments. People might consider them belittling and that won't help anyone. Instead of just WP:VAGUEWAVEing to another page, show some statistics how there are so many more problematic pages now then there were six months ago that speedy deletion is the only way to handle them.

Also, the last RFC failed for a whole number of reasons and you have not addressed any of them. Before reopening a discussion, it's usually expected that the person re-proposing something that failed previously explains why circumstances have changed in their opinion and the previous reasons to oppose no longer apply, especially if the last proposal was only a few months ago. Regards SoWhy 21:35, 23 February 2018 (UTC)[reply]

Look at COIN. Right now. There are at least five undisclosed paid issues - with sock rings - under discussion there. If you don't frequent COIN you probably won't be familiar with how often this happens, hence the comment. Guy (Help!) 22:56, 25 February 2018 (UTC)[reply]

I did. And I noticed that oftentimes such articles are cleaned up instead of being deleted. Which was one of the reasons brought up as reasons to oppose in the previous discussion five months ago. However, saying that there are now X problematic cases is not the same as demonstrating that those X cases are actually the result of more such problems. And the fact that it takes discussion on how to handle them is basically a reason against any kind of speedy deletion in itself. Regards SoWhy 13:03, 26 February 2018 (UTC)[reply]

• JzG, I've read your proposal as putting forward two distinct options, an extension of G5 or a new Gx criterion. Was that your intention? I ask because others don't seem to have read it in that way, so perhaps it is less clear than it ideally should be. The last discussion on this topic ended as no consensus precisely because of lack of clarity ("arguments in opposition seem to be primarily concerned with the relative vagueness and/or subjectivity"}, and it would be a pity if this one did too. Justlettersandnumbers (talk) 14:56, 25 February 2018 (UTC)[reply]

Yes, two options, presented as A, B or neither. But in the end I think we still have the issue that a small but vocal subset of editors thinks that undisclosed paid editing is fine, even though it is explicitly forbidden. Guy (Help!) 22:51, 25 February 2018 (UTC)[reply]

{{trout}} to JzG for repeatedly lying about his opponent's position. Nobody, and I mean nobody, has said that paid editing is okay. They are arguing that this proposal is poorly thought out, fails to address what was wrong with the last proposal, and that deleting these pages often does more harm than good. Oiyarbepsy (talk) 00:37, 27 February 2018 (UTC)[reply]

@JzG: The problem is not the editors, it's the non-notable/promotional/otherwise undesirable content they create. This content is neither better nor worse than the identical content produced by disclosed paid editors and by unpaid editors. This proposal will therefore not address the actual problem it is trying to solve, which is why I oppose it. You first need to identify the content you want to speedy delete that cannot be speedy deleted already using A7, G11 or other existing criteria. You then need to show that this content can be objectively defined AND that all content that meets this definition should always be speedily deleted. The usual requirements for a new speedy deletion criteria do apply to proposals related to the ToU. This proposal both ignores and fails the objectivity requirement, fails the uncontestable requirement, vaguely handwaves in the direction of the frequency requirement and probably at least partially fails the non-redundant requirement (the lack of objectively defined coverage means it's not possible to be sure about this). Thryduulf (talk) 13:36, 1 March 2018 (UTC)[reply]

The problem is also the editors, though. There are extensive instructions online now about how to write a spam article so it doesn't obviously qualify for A7 or G11, and those same instructions also tell you how to hire someone to create it for you. There are also instructions on how to break the chain so that CheckUser doesn't track you across too many accounts, so in one case four separate C_checked sock rings were uncovered several of which are highly likely to be the same banned user, but not certainly so. I don't care if we allow retrospective G5 for sock farms or if we enforce the ToU, but I thnk we need drama-free way of nuking spam once abuse is uncovered. That's all. Guy (Help!) 18:43, 1 March 2018 (UTC)[reply]

You've declared the problem is the editors, but the actual problem you've described is one of content - i.e. the article is spam regardless of whether it was created by Y (not paid to edit) or Z (paid to edit). Your comments about too many accounts, etc. are things to be considered for the banning policy but are basically irrelevant to content policies like speedy deletion. Thryduulf (talk) 23:18, 1 March 2018 (UTC)[reply]

• Questions.
• 1. The proposal seems to be a bit unclear about what exactly should be deleted. If we focus on the contributor, does it say that from the moment User X violates the Terms of Use, they should be treated like a banned editor, and all of their edits after that point deleted? Or should we delete all pages added by User X since they registered their account? If we focus on the edits, the similarity with G5 disappears, so should we only delete all undisclosed paid edits and keep all unpaid edits?
• 2. What happens if a UPE editor comes clean? Can they just request a WP:REFUND of their contributions if they tag them correctly? Do we want this to be "punishment" or would we prefer to encourage compliance with the ToU? —Kusma (t·c) 14:41, 26 February 2018 (UTC)[reply]

An alternative solution

Instead of crafting this awkward, vague, and problematic new speedy deletion policy, why not simply allow bans to be retroactive. The idea is that the beginning of the ban is the beginning of the problematic behavior, as opposed to when it was actually discovered. It seems that this would address the major problems about the sock farms - with a retroactive ban, you could delete all the contributions back to that retroactive date. Thoughts? Oiyarbepsy (talk) 00:45, 1 March 2018 (UTC)[reply]

Oiyarbepsy This is very true. Making G5 retroactive would be far more effective at combating UPEs than the proposed speedy deletion criterion. I think it would also be far more acceptable to the community. Rentier (talk) 03:36, 1 March 2018 (UTC)[reply]

To be 100% clear, I would absolutely not make the speedy deletion criteria retroactive - only the ban itself. Whether to make it retroactive needs to be decided when the ban is decided, and should only be retroactive with good reason. Oiyarbepsy (talk) 04:13, 1 March 2018 (UTC)[reply]

In that case, what would making a ban retroactive accomplish? Rentier (talk) 04:18, 1 March 2018 (UTC)[reply]

By allowing the deletion of all the pages back to the retroactive date. The idea being that the retroactive G5 deletes only apply if everyone agrees at the time the user is banned. This shouldn't be routine. Oiyarbepsy (talk) 04:28, 1 March 2018 (UTC)[reply]

So are you saying that if user:Example had been contributing since January 2015, and was banned today then the banning discussion could say that G5 applies to all his contributions since say 20 May 2016? If so, I like the idea in abstract, but I think the chance of getting a consensus on a date in individual ban discussions is going to be pretty slim. Thryduulf (talk) 13:16, 1 March 2018 (UTC)[reply]

I think what Oiyarbepsy means is that bans should be made retroactive from the date the reasons for the ban first existed (i. e. ex tunc). Actually, that's probably the only way a speedy deletion of UPE could work objectively (if one agrees that UPE should lead to deletion). The alternative offered is to treat users violating UPE as having been banned from day one because hypothetically, if their UPE had been discovered on day one, they would have been banned immediately. It would not affect other types of bans because the reasons for those bans (like community bans for disruption) only exist from the time the disruption has been determined to be unacceptable. Personally, I believe that content, not contributor should be the deciding factor but iff the community one day decides that UPE creations should be deleted without looking at the content itself, this proposal is the best way to achieve this under the current requirements for speedy criteria. Regards SoWhy 13:30, 1 March 2018 (UTC)[reply]

I'd support this, in fact I think I've proposed it myself in the past. A user violating the ToU ought to be considered banned from their first violating edit, not just from when the community decided to enact the ban. It would simplify our work at SPI in a way I've been looking for: rather than finding a new UPE sockfarm and struggling to determine if they're related to any of the existing sockfarms to determine if they're already blocked or banned and whether or not G5 applies or if we need to AfD all of their contribs or if I can invoke IAR or on and on and on, I can just say this is UPE, nuke their contribs. This needs to be discussed at the banning policy but I would very much support it. Ivanvector (^Talk/_Edits) 14:20, 1 March 2018 (UTC)[reply]

Second that, although I'd like to hear reasons against; it sounds too reasonable. ~ Amory (u • t • c) 16:05, 1 March 2018 (UTC)[reply]

The main reason is probably what I mentioned above, i. e. that there is currently no consensus that content created in violation of WP:PAID should be deleted just because of how it was created, which will inevitably lead to deletion of encyclopedic material created by such editors that otherwise is in line with policy. That was one of the main points brought up last time, i. e. that both WP:PRESERVE and WP:ATD tell us to preserve good content, regardless of how it was created. If the policy is changed, which requires a site-wide RFC imho, we can consider how to implement it. Based on the voices from last time, there seems to be a strong opposition to speedy delete good content this way.

Without abandoning my own position (see above), I think the only objectively fair way to handle such pages - iff their deletion based on creator is agreed upon in the first place - is a sticky PROD like system like WP:BLPPROD that allows any good-faith editor to challenge the proposed deletion based on their belief that the content meets the standards for inclusion. But again, first we need to establish consensus that creator, not content, is a valid reason for deletion. That has yet to happen. Regards SoWhy 16:31, 1 March 2018 (UTC)[reply]

I mean G5 already allows (and deletions occur) of otherwise not policy violating articles because of who the creator is;this would work on a similar principle and thus I don't think would be a great change. Galobtter (pingó mió) 16:35, 1 March 2018 (UTC)[reply]

Similar, but not strictly the same. G5 operates under the principle (or that's at least how I have always understood it) of assuming bad faith. We allow such pages to be deleted because it can safely be assumed that the page creator knowingly violated a ban or block that was put in place to prevent them from creating such pages (i.e. WP:DENY). On the other hand, while UPE is forbidden in the ToU, we can probably safely assume that >99% of all editors have not read the ToU before their first edit (I certainly haven't). So while they act in violation of the ToU, they are most likely unaware of that, i. e. acting in good faith. Thus the difference. Which coincides with the fact that we currently have four warning levels ({{uw-paid1}}, {{uw-paid2}} etc.) that should be applied before an editor can be blocked for UPE. Hence such a change to G5 would make this criterion out of sync with how WP:PAID is applied in the rest of the project, which explains why I advocate a site-wide discussion of the underlying question before we consider implementations. Regards SoWhy 17:02, 1 March 2018 (UTC)[reply]

Not assuming bad faith nor denying recognition, but banned means banned; i.e. if we ban a user but allow them to contribute with a different account, in any way, then bans are meaningless. As for PAID violations being innocent I very strongly disagree: while there are no doubt some users who edit afoul of the policy inadvertently or innocently, and then self-correct when they're advised, they're an indescribably minuscule speck in the vast galaxy of deliberately malicious users who know exactly what they're doing: throwing a continuous torrent of uselessly promotional content at Wikipedia and getting paid when it sticks. The easier we can make it to deal with that problem, the harder we make it for spam to stick, the better for Wikipedia. It's incredibly unlikely that any useful content would be inadvertently removed if we did this, and for that there's WP:REFUND. Ivanvector (^Talk/_Edits) 17:24, 1 March 2018 (UTC)[reply]

If we need to protect the good faith UPEs who don't read the ToU and only start to cooperate after the fourth level warning, the retroactive G5 will be still very effective against UPEs if it is limited to users blocked or banned for sockpuppetry. It should be possible to come up with objective criteria that cover the UPE-sockfarms but exclude users like SwisterTwister whose articles we obviously don't want summarily deleted. Rentier (talk) 17:46, 1 March 2018 (UTC)[reply]

• I like this idea. When a user is banned, we should say when the ban starts (which may be now or in the past). In practice, this means we are giving responsibility for G5 deletion (or not) of UPE to the ban discussion, which seems to be a good place for it. —Kusma (t·c) 19:47, 1 March 2018 (UTC)[reply]
  • This must apply to both blocks (in particular indefinite CU blocks) and bans. It's of no use otherwise. Rentier (talk) 20:23, 1 March 2018 (UTC)[reply]
  • Doing this would require an RfC to change the banning policy to allow bans to be retroactive. Thryduulf (talk) 23:20, 1 March 2018 (UTC)[reply]

    Why? There's nothing in the banning policy that forbids this. Hawkeye7 (discuss) 23:37, 1 March 2018 (UTC)[reply]

    • • @Hawkeye7: The banning policy doesn't mention retroactive bans at all, and the clear implication is that they are not. This is backed up by the wording of WP:CSD#G5 "A page created before the ban or block was imposed ... will not qualify under this criterion.". This wording will require tweaking for clarity if bans become retroactive, but that is likely to be uncontroversial if the change to the banning policy has a clear consensus (it might be worth mentioning in the RFC intro). Thryduulf (talk) 23:49, 1 March 2018 (UTC)[reply]
        • I don't see how such a conclusion can be drawn from the banning policy when it doesn't mention it at all. And G5 could also be read to allow it. Hawkeye7 (discuss) 23:56, 1 March 2018 (UTC)[reply]
          • Whereas I don't see how the banning policy can be said to allow retroactive bans when they are not mentioned at all. G5 is very clear that it doesn not apply to page created before a ban or block was imposed - there is no way the current wording of that would support deleting any page created before, at the earliest, the discussion authorising the ban (for a community ban)/the arbcom case (for an arbcom ban) was closed (and I'd be sympathetic towards pages created between that point and the later of the user being notified and the block being placed). Thryduulf (talk) 02:51, 2 March 2018 (UTC)[reply]

• I agree with Thryduulf; G5 is normally taken to mean that creations by a user prior to their ban are disqualified from the criterion. While there's nothing in the policy currently forbidding a retroactive ban, there is also nothing permitting it, and no precedent really either way. The question needs to be asked, we can't just make up interpretations of policy. Ivanvector (^Talk/_Edits) 16:52, 2 March 2018 (UTC)[reply]

• • • An RfC to establish that a Undisclosed Paid Editor (UPE) ban is retroactive is a good idea. I think it is obvious that it must be, but the community needs to be brought along with the decision making process. "Product of an Undisclosed Paid Editor" carries little weight at AfD, I think because the wider community is not up to speed with what a great problem it is.

Someone asked: What if the UPE later discloses? In that case, REFUND is a possibility.

However, is there an actual need to delete? How about moving all UPE product into a repository and blanking it. Leave it available for Wikipedians to review, but unpublished as far as the UPE sponsor is concerned.

Has anyone else begun to suspect that most of the SPA probably UPEs seen submitting drafts are sockpuppets of a much small pool of actual people? I think keeping a repository of discovered UPE product will help shine a light on the networks. --SmokeyJoe (talk) —Preceding undated comment added 23:38, 1 March 2018 (UTC)[reply]

If G6 were being proposed as a new criteria today it would be shot down for failing both the objective and uncontestable requirements as is so broad. Per the section above and several previous discussions it is by far the most misused of the current deletion criteria, being used by some administrators as a way to delete anything that doesn't fit another criteria, regardless of whether it is uncontroversial maintenance or not.

To rectify this, I propose that we repeal the whole criterion and replace it with a set of objectively defined critera that cover what G6 was intended for but not what it wasn't. I believe the following set covers all the frequent legitimate uses of G6:

G14 Temporary deletions

For example to merge page histories when fixing cut-and-paste moves.

G15 Pages unambiguously created in error or in the incorrect namespace

• This includes redirects created when moving pages to the correct title or namespace only when the old title/location is obviously implausible - if there is any doubt nominate the redirect at WP:RFD.

A12 Disambiguation pages

• which disambiguate one extant Wikipedia page AND whose title ends in "(disambiguation)"; or
• which disambiguate zero extant Wikipedia pages, regardless of their title.

C3C4 Empty dated maintenance categories

This does not apply where the date refers to the current period of time or to the near future.

R4 Redirects blocking page moves

• Administrators must be aware of the proper procedures where a redirect holding up a page move has a non-trivial page history.
• An administrator who deletes a page that is blocking a move must ensure that the move is completed after deleting it.

R5 Redirects in the "File" namespace with the same name as a file or redirect at Commons.

• This does not apply if the redirect has any file links (unless the links are obviously intended for the file or redirect at Commons).

T4T5 Templates orphaned as the result of a consensus at WP:TfD

The deletion summary should link to the TfD discussion.

G16T6 Articles or Templates being replaced by drafts or rewrites

This applies only to pages in the article or template namespaces where the draft or rewrite is unquesitonaly better than the exiting page or there is explicit consensus for the replacement.

U6 userspace drafts containing only the default Article Wizard text

This applies only if the user who created the page has been inactive for at least one year.

Almost all the language used here is taken directly from the existing examples, sometimes tweaked for tense or different phrasing (e.g. R5 has been inverted). The exceptions are the bullets for G15, which are a codification of existing practice, G16 which is brand new (based on a couple of recent uses of G6), and the exception to C3 which is hopefully common sense. If there are other proper uses of G6 not covered here then they can be added to the list before we enact it, or they can be proposed as new criteria if and when it becomes clear they're needed. Thryduulf (talk) 00:23, 7 March 2018 (UTC)[reply]

• There've been previous, now-repealed criteria numbered C3 and T4.
  I don't think we need G14, and have never thought we needed TFD deletions here. WP:CSD is not WP:DP; if you're deleting something and immediately restoring it, it's no more a deletion than reverting yourself counts toward WP:3RR, and TFD is the opposite of speedy. U6 is really a limited expansion of G2 into userspace, and I'm undecided whether it would be better as a separate entity or not. And your G16 should be separated out into another proposal, especially for articles. —Cryptic 00:52, 7 March 2018 (UTC)[reply]
• (edit conflict) @Cryptic:
  • C3 renamed C5, T4 renmaed T5.
  • G16 limtied to templates and moved to T6 - if it isn't an uncontroversial replacement for something covered by G6 currently it doesn't belong in this proposal. I'll leave a separate proposal for someone else.
  • re U6 vs G2 - the choice here is a longer list of simple criteria vs a shorter list of more complicated ones, and when it comes to CSD I'm very much in favour of the former.
  • re G14 - I think it's much better to be explicit that these are allowed, and it will allow a simple entry for the deletion summary.
  • T5 (was T4) I'm ambivilent about, but it is currently an explicitly listed criterion and it is certainly uncontroversial. It's a bit different to other deletions following discussions as the orphaning process may take significant time. I'll leave it in here for the time being. Thryduulf (talk) 01:29, 7 March 2018 (UTC)[reply]
• I think the general concept is a good one but it could use some tweaks. My thoughts:
  • G16 (replace with drafts/rewrites) - Absolutely not. Deletion is not required to do this. A copy/paste will do the trick, and is better as it leaves the page history available to everyone. If people are deleting pages to do this, it must stop now.
  • C3 (empty maintenance categories) - Needs mention that some categories will be empty by design and should not be deleted if empty. Strike that, misread it the first time
  • T4 (orphaned categories after TfD) - No speedy deletion required, there was a deletion discussion, just delete. Or, create a Gxx that says any page with a deletion discussion saying to delete where the page wasn't actually deleted.

Oiyarbepsy (talk) 01:23, 7 March 2018 (UTC)[reply]

• • • Re G16 (now T6), hmm, I'll think more on that you make a good point. Re T4 - see my comments to Cryptic. I decided against a Gxx as everything else should just link to the deletion discussion. Thryduulf (talk) 01:31, 7 March 2018 (UTC)[reply]
      • Your current T5 would need to be under a G criterion to catch all the uses for {{db-xfd}}. -- Tavix ^(talk) 01:39, 7 March 2018 (UTC)[reply]
        • It would, but is there actually a need for a db-xfd criteria? I don't see that there is - the deletion is done according to the consensus of the discussion - all that is needed is something to flag there is consensus to delete something but it hasn't yet been deleted, which is not a speedy deletion issue. Thryduulf (talk) 01:50, 7 March 2018 (UTC)[reply]

• Well yes, there has to be a criterion that matches the db template, especially when thinking about what is currently being used by Twinkle. For example, I can remember several times I've used G6 to tag a page where an admin has closed a discussion as delete and forgot to delete it. Other uses I've seen for it include WP:CFD, where Marcocapelle has been a de facto admin there for a while, and uses the template to tag for deletion categories that have been emptied and ready for deletion by the CFD process. Mirroring the language from db-xfd and G6, it could be:
  • G16: Deletion discussions
    • Deleting pages where consensus has been reached at a deletion discussion, but it has not yet been deleted, including templates orphaned as the result of a consensus at WP:TfD. -- Tavix ^(talk) 01:59, 7 March 2018 (UTC)[reply]
      • But that's not a speedy deletion - it's just a deletion, for which there is explicit consensus at an XfD, that has not yet been performed for some reason. All we need is a template that says "There is a consenus to delete this page at [link to XfD]." and I see no reason that {{db-xfd}} could not be used to do exactly that. Thryduulf (talk) 02:32, 7 March 2018 (UTC)[reply]
        • If that's the case, you shouldn't have a T criteria for exactly that either. Take your pick. -- Tavix ^(talk) 02:38, 7 March 2018 (UTC)[reply]
          • I've explained above (in reply to Cryptic) why I see orphaned templates as different to other cases but to repeat and expand on that: The decision to "orphan then delete" is different to "delete", the orphaning may take some considerable time and this would be an explicit signal that it has been completed and the second part of the consensus can now be implemented. Thryduulf (talk) 02:58, 7 March 2018 (UTC)[reply]
            • But that's not a speedy deletion - it's just a deletion, for which there is explicit consensus at an XfD, that has not yet been performed for some reason. -- Tavix ^(talk) 03:03, 7 March 2018 (UTC)[reply]
              • I think we'll have to agree to disagree about that - you see them as identical, I see them as qualitatively different. I'd prefer T5 to having neither, but doing away with both is preferable to your G16. Thryduulf (talk) 03:40, 7 March 2018 (UTC)[reply]
• Prefer to label G14 as G6a, for example. It is squarely in the family of G6 deletions. --SmokeyJoe (talk) 03:02, 7 March 2018 (UTC)[reply]
  • That's actually not a bad idea to label each of them as G6A, G6B, etc. (like the C2 criteria). That way we aren't creating several "new" criteria, but re-purposing a familiar criterion to more finite "sub-criteria". -- Tavix ^(talk) 03:06, 7 March 2018 (UTC)[reply]
    • I thought about that, but these really are not a "family" of criteria, they are a hotchpotch of completely different ones. Part of the purpose of retiring the G6 label is to reinforce that it is not a dumping ground for things that do not fit elsewhere. Thryduulf (talk) 03:40, 7 March 2018 (UTC)[reply]
      • The family I think should be taken as the family of speediable pages that have trivial page histories. --SmokeyJoe (talk) 04:05, 7 March 2018 (UTC)[reply]
• Prefer G15 to not be explicitly listed, but to be a speedy WP:Move follow by WP:CSD#R2 (or "redirect suppressed" if the editor has that permission). --SmokeyJoe (talk) 03:02, 7 March 2018 (UTC)[reply]
• Oppose A12 dot point 1. One target DAB pages should be redirected per policy at WP:ATD. --SmokeyJoe (talk) 03:04, 7 March 2018 (UTC)[reply]
• Oppose A12 dot point 2. Zero target DAB pages should be taken to WP:AfD for discussion to educate the creator, and including options to WP:TROUT them for incompetence or WP:BLOCK them for disruption. --SmokeyJoe (talk) 03:06, 7 March 2018 (UTC)[reply]

Well, that seems awfully bureaucratic and WP:POINTY. If you need to educate the creator of a certain dab page, drop them a note at their talk page. -- Tavix ^(talk) 03:11, 7 March 2018 (UTC)[reply]

If it happens once, it is not frequent enough for CSD. AfD is for once-off mistakes. Sure, explain stuff to the user on their talk page, ask them to G7 it. However, Titling / DAB pages is an area fraught with strong POVs tipping into active disruption, and I read this CSD as providing a stick to be used in DAB page policy battles. These battles, whether rare or frequent, should go to AfD. --SmokeyJoe (talk) 03:31, 7 March 2018 (UTC)[reply]

@SmokeyJoe: one-target DAB pages that end with "(disambiguation)" are not suitable for redirecting to pages that are not disambiguation pages, and no other otne-target DAB pages are within the scope of either the present G6 or this proposed criterion (the language is essentially identical) for preceisely that reson. I agree with Tavix regarding zero-target DAB pages - that would be disruping Wikipedia to make a point, it's not even close to the purpose of AfD. Thryduulf (talk) 03:40, 7 March 2018 (UTC)[reply]

I'm not convinced that one target DAB pages can't redirect to the unambiguous target, but if people think this is a bad idea, then treat them as zero target DAB pages.

Taking bad DAB pages to AfD is disruptive to make a point, as an argument for a new CSD criterion? Um, I think not. Oppose new CSD for bad DAB pages. The onus is on the proponent to establish that it meets the four new criterion criteria. --SmokeyJoe (talk) 04:03, 7 March 2018 (UTC)[reply]

• Oppose. I wanted to give this a chance, but it's obvious from the above that it quickly becomes too complicated. What this is basically doing is removing the This is for uncontroversial maintenance, including: part from G6 (with my emphasis), and replacing each bullet point with new criteria. The bullet points currently listed are consensual examples of uncontroversial maintenance, but they are not the only forms of uncontroversial maintenance. Trying to develop new criteria and figuring out which forms of uncontroversial maintenance should be made into new criteria is over-complicating the issue for little gain. While it's true that G6 has been misused and that should be reigned in a bit, this is not the answer. If a specific "thing" is being misapplied, such as "pagemove clean-up", more guidance on that specific issue can be hashed-out. But that doesn't mean we need to blow the whole thing up and start from scratch with at least nine(!) new criteria. -- Tavix ^(talk) 04:02, 7 March 2018 (UTC)[reply]
  • Nine new criteria that are objective is much better than one old criteria that gets abused. Thryduulf (talk) 04:21, 7 March 2018 (UTC)[reply]
    • Please provide specific examples of abuse of this criteria, and we can go from there to find ways to limit it. The solution is not to remove the criteria all together, that is too limiting. As a parallel, editors often claim A7 is abused, specifically the phrase "credible claim of significance". Instead of removing the criterion, an essay was written to explain what is and isn't a credible claim of significance. Perhaps all that is needed is a page called WP:Uncontroversial maintenance, where we can flesh out what is and isn't uncontroversial maintenance. This would be a lot less blunt than removing what I feel is a necessary criterion and then massively ballooning the number of criteria we already have. -- Tavix ^(talk) 04:33, 7 March 2018 (UTC)[reply]
    • I fear that you may be retrospectively authorise past bad speedies, such as speedying disagreeable DAB pages. Have people been speedying disagreeable DAB pages under G6? Are we all aware of different interpretations on TWODABs, and whether a broad concept article on a tingle topic should be called a disambiguation page? --SmokeyJoe (talk) 04:36, 7 March 2018 (UTC)[reply]
      • G6 currently contains the following language: Deleting a disambiguation page which either: disambiguates only one extant Wikipedia page and whose title ends in "(disambiguation)"; or disambiguates zero extant Wikipedia pages, regardless of its title. WP:TWODABS is a specific case of a disambiguation page that disambiguates two extant Wikipedia pages, so WP:TWODABS are never speediable under G6. -- Tavix ^(talk) 04:39, 7 March 2018 (UTC)[reply]
        • Indeed, the disambguation pages that can be speedied under G6 and the disambiguation pages that could be speedied under this proposal are identical. Any disambiguation page that has more than 1 target is never speediable under G6 and would not be speediable under this proposal. Thryduulf (talk) 11:30, 7 March 2018 (UTC)[reply]
          • Yeah sorry my mistake. —SmokeyJoe (talk) 11:42, 7 March 2018 (UTC)[reply]
• I propose we add the nominator's CSDs and eliminate G6 as a CSD but expand the PROD system, made for uncontroversial deletions, into where G6 would have applied. Jjjjjjdddddd (talk) 10:41, 7 March 2018 (UTC)[reply]
• Oppose This is all just useless bureaucracy and solution to non-existent problem. The "uncontroversial maintenance" in G6 really means any and all procedures that lead to negligible information loss, typically by being applied to pages with trivial histories, and that are sufficiently similar to listed cases of uncontroversial maintenance. Some examples: over year old userspace drafts containing the default Article Wizard text corrupted by few random characters or containing a broken redirect, less than year old userspace drafts containing only the default Article Wizard text created by indef blocked users who have zero good edits, any other suspicious userspace artifacts created by indef blocked users who have zero good edits, pages that simulate mediawiki interface elements for malicious or unknown purposes, *.js or other pages that try to exploit mediawiki technical limitations, pages designed to trick users to compromise their account security somehow, duplicates or near duplicates of existing pages with weird unicode characters or HTML entities substituted in page name or content, and so on. jni ^(delete)_{...just not interested} 21:49, 7 March 2018 (UTC)[reply]
  • Most of those shouldn't be deleted using G6 anyay - the userspace drafts that are not identical to the default text have no consensus for speedy deletion currently and so should not be being deleted at all. Pages created by indeffed blocked users are G5 if created after the block and not speedy deletable currently if created afterwards unless they are G2, G3 or similar (G6 is not similar). Anything created for malicious purposes is G3 vandalism or G10 attack page depending on how it is mallicious, likewise those that try to exploit mediawiki limitations or try to trick users are G3. Duplicates of existing pages are G2 and/or A10. Your comment is exactly why we need to get rid of G6 as it's being used incorrectly. Thryduulf (talk) 22:47, 7 March 2018 (UTC)[reply]
    • All examples in my comment are speedily deletable by common sense and existing practice, if not G6 or some other criteria. With duplicates I was mostly thinking duplicates in namespaces where G2 or A10 does not apply. Often a nomination to MfD (outcome of which is easy to predict) is needed, but not always. Capricious namespace restrictions in other rules is one reason why G6 is occasionally needed. Your proposed "unambiguously created in error" could be used for userspace dupes, I guess, especially when extending the interpretation of "unambiguously" as G6 is stretched today. Suppose indef'ed vandal account has created a short page in their userspace with some garbled sentences in some foreign language google translate does not quite grok but there is at least one swear word in it. To what extend do admins need to analyze random textdumps left to random locations by obvious non-contributors to determine if they are G10 attack pages? In practice these kind of artifacts get speedily deleted based on behavioral evidence and reasonable man standard ("no reasonable man would extend effort to decipher it further"), even if it cannot be really proved they are G3 or G10. jni ^(delete)_{...just not interested} 19:26, 8 March 2018 (UTC)[reply]
• Oppose the whole purpose of G6 being broad is that it allows for technical deletions that no one would ever oppose and that people would do anyway by finding a way to fit them into the other categories. I don't think I have ever seen a contested G6 on another admin's talk page or my own. It is the embodiment of IAR for housekeeping when it comes to deletions, and gives it a policy basis in the deletion policy. Examples that wouldn't be included here that we routinely do: G6ing NOTNOW RfAs if someone asks. G6ing RfAs where the candidate never consented to the subpage being created. Vindictive SPIs. Basically all of the already existing redirect criteria that no one ever remembers or gets right when trying to select them from the drop down menu or in Twinkle. Joke AfDs that no one finds funny enough to comment on when it is April Fools Day.
  I think I have a pretty strong grasp of the deletion policy as is, and I can't remember most of the non-G and non-A criteria, and I'm willing to bet that most admins can't either, but they can easily spot a G6 case and delete it, and when they do, it's likely one of the already named criteria in another area. If anything, I'd support getting rid of all of the R and T (and possibly C) criteria and wrapping them into G6 rather than expanding them.
  Also, to the above point in the other thread where people are complaining about this not being the intended purpose of G6 even though it is the commonly accepted practice: that's 100% fine. We don't change accepted practice to match policy. We change policy to match accepted practice. If there is something in the current criteria that prevents it's use as a catchall (I don't see it, but if there is), that should be removed, not the other way around. TonyBallioni (talk) 23:51, 7 March 2018 (UTC)[reply]
  • Also, an example from my recent deletions of G6 that isn't covered by any other criteria listed above or existing: User:Doveanupam. Someone got it in their head that the user was a sockmaster without an SPI, started going around changing CU confirmed tags, and eventually created a userpage for the master with a sockmaster tag. I had to block and bother a CU to talk the user down. This is an obvious deletion (blanking would be pointless, and it shows up in the history, which isn't ideal). It isn't G3 as it isn't vandalism, and I certainly wouldn't call it G10. G6 fits the bill nicely here. TonyBallioni (talk) 00:20, 8 March 2018 (UTC)[reply]
    • That's a great example! I recall at some point we had hundreds of redirects with letter "p" added at the end of title by a software bug in the bot (or whatever entity) that had created them. G6 to rescue and they were cleaned away! jni ^(delete)_{...just not interested} 19:29, 8 March 2018 (UTC)[reply]
• Oppose, needless bureaucracy. I trust admins to use it properly for 'uncontroversial' technical deletions at admin discretion. If an admin isn't using it uncontroversially, then that is a separate issue, but adding a dozen more criteria isn't a solution. G6 is the net that catches everything uncontroversial that falls through the cracks of the other criteria and is, as TonyBallioni put it, "the embodiment of IAR for housekeeping". — Insertcleverphrasehere ^{(or here)} 00:44, 8 March 2018 (UTC)[reply]
• Oppose - the housekeeping needs can change, and policy needs to reflect this. For example, when C6 was established, there was no way to move categories; we would copy them and delete the originals. Now that we can, I have been fixing these copy&paste moves - and there is always a revision at the source which needs to be deleted (or the move needs to be done without leaving a redirect, which may only be done if the redirect would be speedy deletable). None of these explicit criteria cover this, although the task is clearly uncontroversial maintenance. עוד מישהו Od Mishehu 14:31, 8 March 2018 (UTC)[reply]
• Oppose, per TonyBallioni and others above. I'm not seeing a real problem that all this fine tuning needs to fix. If there are situations where an admin is incorrectly applying G6 in a material way, then raise the concern with the specific admin.Mojo Hand (talk) 17:08, 8 March 2018 (UTC)[reply]
• Oppose G6 is a useful tool for situations where deleting a page would be utterly uncontroversial but where it doesn't fall under any explicitly enumerated situation. This proposal would get rid of that. In practice I suspect people would continue to delete pages like that under IAR or some misapplied criterion instead of waiting ages for PROD or AfD, which is bad because policy is meant to reflect practice. Hut 8.5 18:30, 8 March 2018 (UTC)[reply]

• Comment I've used G6 as a low profile way to delete pages containing personal details of a minor. We don't want a specific criteria for these cases. I do support splitting out some of the most common G6 issues as a way of encouraging more editors to tag these common problems. Most editors are not CSD experts so making CSD as simple as possible is a good thing. Legacypac (talk) 20:02, 8 March 2018 (UTC)[reply]
  • The low-profile way to delete minors' personal details is to mail oversight. This is usually faster, and there's sites that preferentially mirror enwiki pages that are tagged for speedy deletion. —Cryptic 20:13, 8 March 2018 (UTC)[reply]
• Support in spirit; try a footnote. I agree that G6 is the most easily and most often abused criterion (and yes, the singular is criterion; criteria is plural). However, it's not rampantly abused, and forking this into a bunch of new numbered rules doesn't seem to be getting any support. It's an overly complicated approach. It would probably be more practical to include these as a list of examples of what G6 encompasses, in a footnote.  — SMcCandlish ☏ ¢ 😼  12:20, 10 March 2018 (UTC)[reply]
• Oppose. We already have enough criteria and Wikipedia is not supposed to be a memory gymnastics exercise for admins. G6 is fairly rare and I cannot recall having ever seen one contested, but there is always WP:DELREV if someone were to disagree with one - after discussion with the admin of course. Let's guard against making bureaucracy for the sake of bureaucracy until at least G6 gets abused to such a vast extent that something would really need to be done about it. And I can't see that happening any time soon. Kudpung กุดผึ้ง (talk) 01:00, 11 March 2018 (UTC)[reply]
• Oppose A well-intentioned proposal in search of a problem. When G6 is abused, it's a problem with the admin abusing it and not the criterion itself. Cryptic for example has mentioned a number of examples in the previous section and it's clear that the same admin performed five of those seven deletions he mentioned as problematic, so there seems to be a problem that can and should be addressed instead. Regards SoWhy 13:50, 12 March 2018 (UTC)[reply]
• Comment - Seems pretty snowy here: time to close? Narutolovehinata5 ^tccsdnew 14:41, 12 March 2018 (UTC)[reply]