Wikipedia talk:Criteria for speedy deletion/Archive 45

• Less of the patronising tone please. I realise that I am suggesting a departure from what is "traditional" (BTW I dont know of any traditions in real life that needs to be kept). Whether it is radical is a matter of opinion. I am tiring of the beating my head against a cyber brick wall. I seems impossible to change the status quo in order to improve WP. -- Alan Liefting (talk - contribs) 23:43, 18 February 2012 (UTC)

This isn't a case of institutional inertia blocking a new idea, it is a case of a proposal so grossly out of line with what CSD is intended for that literally nobody has as yet shown the slightest agreement with either the proposal itself or the stated reasoning behind it. Things can and do change here, but if you want to make a change as fundamental as this one you need a very compelling, quantifiable reason to do so. Not seeing that here, and apparently neither is anyone else.

talk

) 23:54, 18 February 2012 (UTC)

Re:Your cmt right above Beeblebrox's outdent: Sorry, but I simply don't agree with you (and given that not even BLP's were allowed to be speedied, I don't think many others do either) so I'll close with this: "bad editors" are not the ones creating one line stubs on topics that cannot be deleted via another CSD category. I agree with you about the...err..."trash" at NPP, but your proposal would simply remove many of the few decent ones submitted there.

^{Talk·Contribs}

00:00, 19 February 2012 (UTC)

(In response to Alan's response to me above) Yes, consensus can change but (and that's the important part) you need to make a convincing argument why it should. And that would have to be one that would not only convince everyone who previously opposed speedy-deleting BLPs without references but also those who supported it limited to BLPs only. So far I (and apparently almost everyone else here) does not think your argument is solid enough. I understand that you believe it to be convincing but since you don't seem to have convinced anyone with it, you might want to reconsider whether it's really as convincing as you think it is. I'm happy to change my mind when someone shows me that my ideas are stupid (and I assume most people here think the same) but so far I have to agree that your proposal will be more harmful than good to the project. Regards SoWhy 09:41, 19 February 2012 (UTC)

• Oppose the problem in the Seigenthaler incident was that someone introduced false information into the pedia. Some people have jumped from that to the assumption that reducing the amount of unreferenced information will protect us from having false information in the pedia. There is an alternative theory that the more bitey we make our processes the more retaliation we can expect. My preference is that we keep our systems as non-bitey as possible, and that we focus on problematic areas such as contentious BLP statements, death anomalies and information sourced from partisan sources. If we really want to reduce levels of vandalism in the pedia all we need do is introduce flagged revisions and at least make sure that every edit by a newbie or IP is looked at at least once..... ϢereSpielChequers 17:48, 19 February 2012 (UTC)
• Another question. If the intent is force editors to cite sources from the start, are there possibly other ways of doing that short of threatening speedy deletion? ~ J. Johnson (JJ) (talk) 21:34, 20 February 2012 (UTC)

{{

talk

) 23:34, 20 February 2012 (UTC)

• Oppose It is a bad idea is as well stated by Hut and Rossami. It is not a good idea to chastise new editors by immediately deleting their work. Far better is to advise them on the importance of verification and to allow the Afd process to work, if appropriate. I have been looking at the list of unreferenced articles from October 2006. Of the first six that I looked at, five had good, easily found, references. (I admit that I avoided looking at the song titles and the Lusophony Games.) One article was more difficult, and I had to add a brief "History" section to get a reference that would apply well. But, if we have outstanding unreferenced articles from October 2006 for which it is easy to find appropriate citations, then finding those citations must not be a priority for Wikipedia editors. If any of you believe that this is a problem, I suggest that you fix a couple of those unreferenced articles. Many hands make light work. --Bejnar (talk) 08:36, 21 February 2012 (UTC)

G7 currently reads:

G7. Author requests deletion. If requested in good faith and provided that the only substantial content to the page and to the associated talk page was added by its author. (For redirects created as a result of a pagemove, the mover must also have been the only substantive contributor to the pages prior to the move.) Note that this does not apply to user talk pages, which are not deleted except under very exceptional circumstances: see

WP:DELTALK

. If the sole author blanks a page other than a userspace page or category page, this can be taken as a deletion request.

However, as a former

Well, I've certainly marked articles as G7 in the past where it was clear that the reason for the blanking was that the creator wanted the article deleted. Indeed, it isn't a rare occurrence, and many new users seem to think that by blanking the article they are actually deleting it. It's not difficult to find cases where blanking is done to try and get the article deleted, after a few minutes searching I found Freedom, Inc from a couple of days ago. - Kingpin¹³ (talk) 08:56, 19 February 2012 (UTC)

But usually when this happens, it was already tagged under a different CSD tag, right?

csdnew

09:19, 19 February 2012 (UTC)

Not always but often, yes. But in most of those cases, it's usually correct to assume that they wanted to delete the article because they felt intimidated by the deletion tag or because they were informed that their page is not within our guidelines and they wanted to do the right thing. For me it's a part of AGF to assume the latter if I have no reason to believe otherwise. Regards SoWhy 09:30, 19 February 2012 (UTC)

I suspect that many editors understand that there is a difference between blanking a tag and blanking the whole page. Blanking the very work you are trying to save seems like a very odd error to me. I've deleted lots of G7s and I've never had an editor come to me and say they just blanked it so they could start again and they didn't meant to get it deleted. I have declined a G7 where the author blanked, someone tagged it G7, the author then replaced that with a revised version of the article and the tagger restored the G7 tag. But thankfully such incidents seem rare and are best dealt with by declining the speedy. ϢereSpielChequers 19:22, 19 February 2012 (UTC)

When a blanked page is tagged G7, the reinstatement of the article in a "revised" version is a strong assertion against the G7 speedy, though other criteria might still apply. However, we need to wait for the article creator to reinstate the article after the G7 tagging to see that indeed G7 does apply. If no action follows the G7 tagging, then it's a clear-cut G7. Otherwise, nothing prevents the article creator from resurrecting the article, and since no talk page message is issued for a G7 speedy, the creator might not even know that someone manually deleted his article. -- Blanchardb -^{Me•MyEars•MyMouth}- timed 15:00, 21 February 2012 (UTC)

Non-free images or media that have been identified as being replaceable by a free image and tagged with {{di-replaceable fair use}} may be deleted after two days, if no justification is given for the claim of irreplaceability. If the replaceability is disputed, the nominator should not be the one deleting the image.

What number of days is an appropriate delay if a justification is given for the claim of irreplaceability and the replaceability is disputed? If the answer is still two days, I suggest writing this more clearly as:

Non-free images or media that have been identified as being replaceable by a free image and tagged with {{di-replaceable fair use}} may be deleted after two days. If the replaceability is disputed, or if justification is given for the claim of irreplaceability, the nominator should not be the one deleting the image.

If the answer is seven days, I suggest writing this more clearly as:

Non-free images or media that have been identified as being replaceable by a free image and tagged with {{di-replaceable fair use}} may be deleted after two days, if no justification is given for the claim of irreplaceability. Otherwise they may be deleted after seven days, and the nominator should not be the one deleting the image.

Regards, Martin (talk) 20:20, 22 February 2012 (UTC)

Earlier today I discovered a redirect page whose target had been deleted (Southern Border (disambiguation)). "Surely," I thought, "there must be a speedy deletion criterian for this." So I came and looked. I found the section on criteria for Redirect pages. I noticed that there was no "R1" and wondered what it might have been. I was puzzled to find that nothing seemed relevant to the "deleted target" scenario. I checked the "invalid criteria" section to see if maybe R1 had previously been relevant, but had been stricken from the list of acceptable criteria. And then I had used up all the time I had to wonder about it, and I went ahead and put the page on RfD using Twinkle. Only after an administrator closed it speedily and cited G8 did I figure out what had happened. I see a lot of past discussion about this, so I can only assume that the current state of affairs reflects consensus. However, I will just say it would be a lot easier for someone in my position if the Redirects section could include some sort of explicit statement like "R1 depricated; see G8". At least then I'd have known where to go looking. Just one guy's two-cents' worth. --Rnickel (talk) 00:30, 23 February 2012 (UTC)

We already have that information; it's in the "Deprecated criteria" section. I've gone through the existing criteria sections and added a little thing for each repealed criterion: "[criterion name]. Repealed; see the deprecated criteria section below." Nyttend (talk) 17:27, 23 February 2012 (UTC)

Outstanding! That would have helped me a lot to find what I was looking for. Thank you very much. :-) --Rnickel (talk) 17:36, 23 February 2012 (UTC)

Watchers here may be interested in Wikipedia:Village_pump_(idea_lab)#Allow_any_logged_in_editor_to_delete_their_own_U1_and_G7_cases it is similar to a proposed change from some time back that I think got consensus but would have required a bot. ϢereSpielChequers 21:29, 23 February 2012 (UTC)

I am over 30k edits, and still I have to search at least two pages to get a speedy template right. Why cannot I find the procedure & templates in a single page? WP, stop telling me I am stupid. -DePiep (talk) 22:17, 25 February 2012 (UTC)

Have you seen

WP:FIELD? →Στc

. 22:24, 25 February 2012 (UTC)

You mean: search three pages to get a speedy? No I haven't. -DePiep (talk) 22:25, 25 February 2012 (UTC)

May I suggest Twinkle? It gives you an easy dialogue box on the page from which you can select the correct template, which is automatically added to the page. That's assuming of course that you have a basic grasp of the different criteria and can remember which one goes where with a one line description...very useful for me.

^{Talk·Contribs}

22:28, 25 February 2012 (UTC)

Of course you may suggest Twinkle. And I decline the suggestion right away. I know you are in (A)GF, but this is not what I asked for.

WP:SPEEDY should do the job, in single page. Is not TW an outside thing? Like AWB, HOTCAT, and every other app while I ask for a basic thing: how to do a speedy? The SPEEDY page cannot explain how to do a speedy? -DePiep (talk

) 22:52, 25 February 2012 (UTC)

Oh in that case I completely agree with you. The only policy-based page that is concise and informative enough for me is

^{Talk·Contribs}

02:12, 26 February 2012 (UTC)

Never saw VRS, but usualy I look up

WP:42. Or WP:UAUQLUE, as I remember it by easy mnemonics. Still, after some deep thinking, there is nothing speedy in there. -DePiep (talk

) 02:45, 26 February 2012 (UTC)

What is wrong with the instructions at

Wikipedia:CSD#Criteria? Assuming you know how to place a template on the page they should be clear enough. Yoenit (talk

) 15:04, 26 February 2012 (UTC)

I researched the history of G7 as part of deciding how to vote in a recent DRV. Because some might be inclined to do the same in the future, I'm posting this on this page to save those users the trouble. Of course, there are probably some regular readers of this page who will enjoy reading this anyway. A Stop at Willoughby (talk) 03:33, 6 March 2012 (UTC)

Ratification and early development of G7 (2005-2006)

G7 was not one of the original criteria for speedy deletion. It originated as one of many potential criteria proposed for addition in January 2005. Of those, only three – the items which would later become G7, A3, and A4 – were ratified by the community. G7 was ratified by a vote of 156-21. The newly added criterion provided for the deletion of:

Any article which is requested for deletion by the original author, provided the author reasonably explains that it was created by mistake, and the article was edited only by its author.

On January 16, 2005, SimonP did the honors of adding the three new criteria, including the earliest version of G7. Our understanding of G7 can be informed by some of the instructive comments which were made around the time of the vote.

• MarkSweep wrote, “If you see a valid article listed for speedy deletion, you can try to prevent its deletion by editing it (it may still get deleted for unrelated reasons). Also the original author nominating it is supposed to 'reasonably [explain] that it was created by mistake'. Ideally, trying to explain that a perfectly valid article should be deleted would not come across as a reasonable request, so it cannot be deleted under the proposed policy item.”
• Dori wrote, “This is a matter of courtesy. Even if the article is perfectly good, if no one else has edited it, and the author regrets his decision to put it here, we should honor it. It's just a matter of being a good host of information. People should want to contribute here, it should not be trap of any sort. Once someone else has contributed to it, even if to correct a single spelling, then we can no longer delete it as a matter of courtesy. It doesn't have anything to do with 'owning' it. The author has the copyrights. Due to the GFDL we are legally OK with keeping it. I argue that it would not be morally OK.”
• Thryduulf noted, “If someone else was going to add something to it, they can recreate it themselves.”
• Skysmith explained, “If the writers realize their mistakes and want them corrected, all the better.”
• Isomorphic noted, “Of course, we are not obligated to delete an article just because the author requests it.”

As these comments demonstrate, the crux of G7 when it was ratified by the community was the “mistake clause,” which read, “provided the author reasonably explains that it was created by mistake.”

Even in 2005, there were already complaints about the CSD being too byzantine. On November 24,

On January 9, 2006, R3m0t added a blanking provision to G7 for the first time, after this talk page discussion. This marked the end of the early development of G7; it would be static, with both the mistake clause and the blanking provision intact, for over a year.

Major changes without discussion (2007)

On February 11, 2007, Steel significantly changed G7 by removing the mistake clause. Steel did not explain or discuss this edit anywhere on-wiki, nor was the edit prompted by any on-wiki policy discussion that I can find.

I looked at Steel's deletion log to find an impetus for his edit. On February 11, 2007, he deleted the following pages under G7:

• 19:17, February 11, 2007 Steel (talk | contribs) deleted "Eric Grete" ‎ (CSD G7)
• 19:19, February 11, 2007 Steel (talk | contribs) deleted "Greenfield school" ‎ (CSD G7)
• 20:22, February 11, 2007 Steel (talk | contribs) deleted "File:Antigay.png" ‎ (CSD G7)

That last one preceded his edit to WP:CSD by only 13 minutes, making it the likeliest candidate. As best as I can tell without the ability to view deleted edits, the image was associated with a userbox which Steel was also deleting. At 19:49, he deleted Template:User homophobia under T1, “divisive and inflammatory.” He informed the userbox’s creator, User:PatPeter, of the deletion, prompting a hostile discussion. It appears that PatPeter recreated the userbox in his userspace (at User:PatPeter\User homophobia); Steel deleted it under T1 at 20:08. It appears that PatPeter then recreated the userbox a second time in the same location. What happened next I cannot say without access to deleted edits; however, at 20:22, Steel deleted the userbox under U1 (user request) and the associated image under G7. It’s possible that PatPeter tagged the pages for deletion; it’s possible that he blanked one or both; and it’s also possible that Steel interpreted this comment as a request for deletion. Regardless, after these deletions, Steel made no further edits and took no further admin actions before making his edit to WP:CSD. So why did he make the policy edit? My educated guess is that the file he had deleted did not strictly meet G7 because it was not “mistakenly created,” so he sought to change the policy so it would be less restrictive. His edit was not reverted, and thus it became ingrained in the policy.

On the next day, February 12, 2007, Ais523 added to G7 a requirement that deletion must be requested “in good faith.” According to his edit summary, he did this “to address the reason why the bit just removed from G7 was there in the first place using a different method.” It does not appear that this was discussed, but it was not reverted and thus became ingrained in the policy. It’s safe to assume that “in good faith” was meant in the standard Wikipedia sense of the phrase, i.e. without malice towards the project.

In a brief April 2007 discussion, an editor expressed concern about the removal of the mistake clause. Two administrators responded, indicating that they had no problems with the “in good faith” wording.

Exceptions to G7 are carved out (2008-2010)

Without the mistake clause, G7 was at least technically far more expansive than it had been when the community ratified it. In the period 2008-2010, five exceptions were carved out of G7.

• On July 24, 2008, Rossami added an exception for redirects created after page-moves.
• On August 9, 2008, Ned Scott added an exception to the blanking provision for pages in userspace, in the wake of these two discussions.
• On November 20, 2009, Dank added an exception to blanking provision for categories.
• On December 12, 2009, Davidwr added an exception to the entire criterion for articles with other substantial contributors to the associated talk page.
• On May 27, 2010,
  JamesBWatson added
  an exception to the entire criterion for user talk pages.

• November 2008: A group of administrators concludes that admins can decline to delete under G7 articles which were not contributed to Wikipedia by mistake. They extend this principle to allow the undeletion of articles which were previously deleted under G7, upon request of a potential contributor to the page.
• February 2009: There is a general sentiment that G7 should not be used in instances where deletion is contested. However, the examples used in the discussion are bad-faith requests, so the problem of contested good-faith requests is not really examined here.
• March 2009: Again, there is some sentiment that G7 deletions are a “courtesy” that does not have to be extended to contributors. However, the ethical problems with “forcibly tying [a contributor] to a biography unwanted by the subject” are raised, and some editors acknowledge the appropriateness of deletion in such cases.

Since 2010, G7 has been more or less static.

TL;DR

The original intent of G7 was to allow contributors to request the deletion of pages they created but now regard as mistakes, and even though that clause was removed from the criterion in 2007 (by a single administrator, without discussion or explanation), it’s still an important undercurrent running through our understanding of G7. If an editor adds an article to Wikipedia and comes to regard that as a mistake, administrators are encouraged – but not obligated – to kindly extend them the courtesy of deleting it. A Stop at Willoughby (talk) 03:33, 6 March 2012 (UTC)

Since office actions are the basis of the G9 speedy deletion criterion, why do we list them in the intro as a method of deletion besides CSD?

Good question. It didn't even make sense since the section is headlined as a list of non-speedt forms of deletion. I have boldly removed, added in sticky prods in its place since that appears to belong on that list, and took the opportunity to list all the deletion discussion individual pages in short form next to the entry for deletion discussions.--Fuhghettaboutit (talk) 06:06, 7 March 2012 (UTC)

Personally I think we ought to remove G9 instead. Office actions are performed under Foundation-backed

WP:OFFICE or state that the deletion is being performed under a DMCA request). Hut 8.5

10:38, 7 March 2012 (UTC)

Probably G9 should be changed to "recreation of content previously deleted via office actions". This concept is similar to G4, though office actions (as I get it) are not the results of prior publicly accessible discussions, so adding them to G4 would dissolve the original idea. — Dmitrij D. Czarkoff (talk) 10:36, 12 March 2012 (UTC)

CAT:CSD has links to the following categories:

• Category:Rescaled fairuse files more than 7 days old
• Category:Non-free files with orphaned versions more than 7 days old

Why do we have both? Seems that the first could easily be merged into the second, but I don't want to cause confusion with a CFD if there's an obvious reason that I'm overlooking. Nyttend (talk) 13:04, 13 March 2012 (UTC)

They differ in criteria: the first is G7, the second – G5. — Dmitrij D. Czarkoff (talk) 13:33, 13 March 2012 (UTC)

Huh? Neither one has anything to do with uploader-requested deletion or deletion because of creation by blocked/banned users. If you mean F7 or F5, the two categories have equal effects: either way, the images have been tagged because they have old versions that haven't been used for at least a week. Nyttend (talk) 13:42, 13 March 2012 (UTC)

Yes, it was a typo. Anyway, it may make sense to keep 1:1 category:criteria ratio. — Dmitrij D. Czarkoff (talk) 14:44, 13 March 2012 (UTC)

Do we normally delete the subpages of an editor that has been indefinitely blocked? See User:TBrandley/Disney Channel (Kazakhstan)‎. GB fan 03:30, 14 March 2012 (UTC)

Drafts from sockpuppets can be deleted under G5, but besides that I don't think so. After all, indefinite =/= infinite. Have you asked Fastily about this deletion? Yoenit (talk) 08:49, 14 March 2012 (UTC)

No I hadn't asked, because when I asked it hadn't been deleted yet. I had declined it once before basically saying the same thing as indefinite =/= infinite. The nominator then relisted it the same way so I asked here if that was normal protocol because I had never seen it before. Fastily must have deleted it between me asking and you answering. GB fan 14:29, 14 March 2012 (UTC)

Just so others know what I am talking about, it was nominated

G6 with a rationale parameter of "Userspace of a blocked user". GB fan

14:33, 14 March 2012 (UTC)

My guess would be that, in this case, it was deleted as a violation of TBrandley's unblocking conditions. He had agreed not to create any more articles or redirects until April 4th, but went on to create several in the last few days. It probably should have just been tagged G5, but that works all right. The Blade of the Northern Lights (話して下さい) 17:13, 14 March 2012 (UTC)

He agreed to stop creating articles and redirects yes, but I can't find anything forbidding him from creating subpages, so I don't see how G5 is applicable either. I have little doubt the page was useless, but in that case it should be deleted under

wp:IAR rather than a random criteria. Yoenit (talk

) 17:29, 14 March 2012 (UTC)

Thats not the case. Schools are exempt from speedy deletion criteria A7, if they meet other criteria they can be deleted under that, especially if its a copyvio--Jac16888 ^Talk 12:08, 15 March 2012 (UTC)

Almost all my speedy deletion requests regarding schools are turned down, even whit other criteria as reason. Effectively, they are protected for speedy deletion.

But it still leaves the question, why are they exempt from A7?

_talk

12:19, 15 March 2012 (UTC)

Well then the editors who declined those speedy deletions were incorrect, if there are any copyvios that weren't deleted I suggest you retag them rightaway. As for A7, have a look throught the archives of this page, [5]--Jac16888 ^Talk 12:29, 15 March 2012 (UTC)

A quick look echoes the opinion of Wikiproject Schools, who desperately try to protect school articles from deletion, no matter how bad and unsourced the articles are. Still, it makes no sense to exempt school articles from A7. A statement as "The idea is that simply being a school is good enough to no longer make A7" sends me the shivers over my back.

_talk

12:36, 15 March 2012 (UTC)

I suspect this has been discussed many times before, but you're free to bring it up at

WP:VPR to get it changed--Jac16888 ^Talk

12:44, 15 March 2012 (UTC)

@Night of the Big Wind. You may be approaching this from the wrong angle. Remember deletion is not for articles that can be improved by normal editing. If your concern is with articles being kept "no matter how bad and unsourced the articles are", you need to try sourcing them. Those that you can source will no longer be unsourced, those which you can't find sourcing for can be uncontentiously deleted by prods. ϢereSpielChequers 00:00, 18 March 2012 (UTC)

The reason schools are exempted from A7 (though not the others) is simply that the community remains conflicted on whether schools are inherently notable. If you look at the guidelines at the top of this page, you'll see the admonition that a "it must be the case that almost all articles that could be deleted using the rule, should be deleted, according to consensus." When school articles are proposed to AfD on the grounds of notability, they are sometimes deleted and sometimes not depending on circumstances. That diversity of result says that we can not rely upon the rule for a speedy-deletion criterion. Personally, I agree with you that schools should be held to the same standards as any other organization. The community has not yet reached consensus on that point, though. Rossami (talk) 13:24, 15 March 2012 (UTC)

There is another reason.

Usual practice in the case of non-notable schools is to merge and redirect to the article on the school district or the place where the school is located rather than just deleting the article. The speedy deletion process can't do this. Hut 8.5

13:27, 15 March 2012 (UTC)

The points of Common Outcomes and "inherently notable" are under fire by now. I am just trying to close an RfC on

_talk

13:56, 15 March 2012 (UTC)

CSD should not be used for anything that has even a slight chance at being notable, schools or otherwise (barring, of course, that other CSD criteria don't apply).

The logic, however, is that if there can be a change to affect the notability of schools elsewhere (basically removing the OUTCOMES bit), then I would suspect that this must necessitate the change here at A7, since the exemption right now points back to OUTCOMES. --MASEM (t) 14:02, 15 March 2012 (UTC)

NOBW does have a point. As the rules are currently written I could build a shack in my brother's back yard, call it "The Ritzman School of Hard Knocks", write an article about it and it wouldn't be subject to A7. --Ron Ritzman (talk) 19:12, 17 March 2012 (UTC)

Situations as blatant as that one are the reason why we have IAR. (In fact if you were to write an article about this shack which did not claim it was an organisation it wouldn't be subject to A7 anyway.) Hut 8.5 23:11, 17 March 2012 (UTC)

I would think that the above example is G3-able. →Στc. 00:29, 18 March 2012 (UTC)

Even without IAR, in the most extreme cases, the article can and sometimes has been deleted for no encyclopedic information, or lack of context if we cannot tell where the school even is. And many of the weakest articles are good G12 copyvios. I am perhaps the most inclusive admin on schools, but I delete in those sort of situations. DGG ( talk ) 07:56, 19 March 2012 (UTC)

Yes, I've read the archives, and I understand the arguments for keeping books out of A7 and/or A9 (basically, it's harder to tell from a book stub whether or not notability is met, so we give them the benefit of the doubt - which is sensible). However, I propose that we consider creating a new criterion, or adding to an existing one, to allow the deletion of self-published books with no claim to notability. With the ready availability of ebook publishing through such services as CreateSpace, literally anyone can publish a book and make it available on Amazon or elsewhere. The growing popularity of these and similar services for print (such as Lulu and XLibris) mean that more and more fledgeling authors are looking to promote their work online; Wikipedia seems to be a popular place to do this. A few self-published books are notable and get articles; however, the vast majority get taken to AfD and deleted under

Having a speedy deletion criterion for self-published titles with no claim of notability would reduce the workload at AfD whilst avoiding books that have been published by actual publishing houses (which, even if sources aren't readily apparent, are at least likely to have been given marketing support and had review copies sent out, meaning sources may exist). Checking whether a book is self-published or not is the work of moments via Googlebooks or Amazon. I would suggest exceptions for titles which have a claim to notability in their text (similar to A7); books self-published by notable authors, or that claim to have generated controversy or other newsworthy attention would not be covered.

If there is support for this, I'd be happy to draft text for the criterion. Thoughts? Yunshui 雲‍水 15:09, 15 March 2012 (UTC)

Most of time it is not obvious whether a book is self-published or not, so we would need to maintain a list of self-publishing publishers as well, which seems a bad idea for a speedy deletion criteria. Secondly, I am not convinced we really have that many articles on selfpublished E-books. Could you link to some recent AFD's for such books? Yoenit (talk) 16:02, 15 March 2012 (UTC)

Here are a few I found through searching:

• Wikipedia:Articles for deletion/Red Snow (Novel)
• Wikipedia:Articles for deletion/Natural Execution
• Wikipedia:Articles for deletion/Beyond Cops and Robbers
• Wikipedia:Articles for deletion/Exercising the Penis
• Wikipedia:Articles for deletion/Elephant Parade
• Wikipedia:Articles for deletion/NOVA: The Seven
• Wikipedia:Articles for deletion/Mary King (novel)
• Wikipedia:Articles for deletion/The Magic of Rhomm
• Wikipedia:Articles for deletion/Stick Man
• Wikipedia:Articles for deletion/Aamoo the Aam
• Wikipedia:Articles for deletion/100 New York Mysteries
• Wikipedia:Articles for deletion/Divinities (book)
• Wikipedia:Articles for deletion/Abattoir Blossoms : Dark Poems and Fiendish Delights
• Wikipedia:Articles for deletion/Your Limit is home
• Wikipedia:Articles for deletion/The Violin Diary
• Wikipedia:Articles for deletion/20 Years After The Zombie Apocalypse
• Wikipedia:Articles for deletion/My Dear Sweet Heart
• Wikipedia:Articles for deletion/Killing of the night (Book)
• Wikipedia:Articles for deletion/Heart Chakra: A story of one man's persuit of happiness.
• Wikipedia:Articles for deletion/Children of Rhatlan
• Wikipedia:Articles for deletion/Liber HVHI
• Wikipedia:Articles for deletion/Christmas Down Under
• Wikipedia:Articles for deletion/Speak of the Devil (book)
• Wikipedia:Articles for deletion/Beyond the Mist (2nd nomination)
• Wikipedia:Articles for deletion/Waking Rose
• Wikipedia:Articles for deletion/Joseph: A Victorian Fairy Tale
• Wikipedia:Articles for deletion/Promise me not
• Wikipedia:Articles for deletion/The Children of The Law of One and the Lost Teachings of Atlantis
• Wikipedia:Articles for deletion/Ugly Water
• Wikipedia:Articles for deletion/Zeegpaw
• Wikipedia:Articles for deletion/Is This My Ride? Seeking Divine Guidance in Our Lives.
• Wikipedia:Articles for deletion/Hail Satan (book)
• Wikipedia:Articles for deletion/Hard Work: Success Made Easy (second nomination)
• Wikipedia:Articles for deletion/Velvet Elvis: Repainting the Christian Faith (2nd nomination)
• Wikipedia:Articles for deletion/His Story: The True Biography of Anton Szandor LaVey
• Wikipedia:Articles for deletion/A Barbie's Life For Me
• Wikipedia:Articles for deletion/The Girl Who Believed in Fairies
• Wikipedia:Articles for deletion/Zanimana: The Blue-eyed Cat
• Wikipedia:Articles for deletion/Tomorrow Took Too Long
• Wikipedia:Articles for deletion/Benny Buggles
• Wikipedia:Articles for deletion/Journal of Thelemic Studies
• Wikipedia:Articles for deletion/Savage Earth

I'll grant that it isn't always easy to establish whether a book is self-published or not (being in the book trade does give me an advantage in terms of personal knowledge), but aside from setting up a personal publsihing company, there are really only a few major places to go: Createspace, Lulu, Authorhouse and XLibris are far and away the most common, so the proposed guideline could be made to cover just these four. Yunshui 雲‍水 09:45, 16 March 2012 (UTC)

If this goes through, add Gyan to that list; we run into that a lot in Indian articles, especially on castes. Besides being self-published, they tend to be copyvios of other self-published works as well. The Blade of the Northern Lights (話して下さい) 23:09, 16 March 2012 (UTC)

That's quite a list. But the first is from 2009 and the second from 2008. If you need to go that far back to compile such a list then I'm not convinced this is a common enough occurrence for such a complex addition to speedy deletion. A once a month scenario where we get a clear decision to delete, belongs at AFD not CSD, AFD can easily handle a deletion a month where the article is on a self published book. Its not as if these are going to be contentious AFDs, or that anyone will be paying spammers to market such books. Also notability is an AFD matter, not something that should be decided at speedy deletion. ϢereSpielChequers 22:26, 18 March 2012 (UTC)

Sorry about that. Rather than go back through AfDs day-by-day and entry-by-entry (which was the only other way I could think of to compile such a list of specific AfDs) I just ran searches on "~Lulu" and "~CreateSpace" in Wikiproject space. Foolishly, I made the assumption that results would appear in date order, so didn't check the dates; evidently they do not. Yunshui 雲‍水 09:25, 19 March 2012 (UTC)

No problem. If these were easy to identify and there'd been this sort of number in the last two or three months (or to make things easier half a dozen in the last fortnight) then I'd be more willing to be convinced that this is frequent enough to be worthwhile. But speedy deletion is already very complex, and for it to be worth making it more complex we need a small rule change that prevents a lot of clearcut AFDs. ϢereSpielChequers 11:29, 19 March 2012 (UTC)

• I have always opposed adding an equivalent of A9 for books, but i think self published books might be an exception. I think the number of these submitted here is very likely to increase, and is already enough that a criterion would be useful. I would make the criterion Self published or by a vanity publisher, by an author who does not have a Wikipedia article, none of whose books have ever been published by a conventional publisher, & where there are no references to reliable sources for independent reviews. The only case I can imagine of such an article being kept is where there are in fact reviews, but the author of the article is not aware of them. This is often the case for naïve articles on childrens' books and the like, written by children who do not know to look for reviews, but will not be the case for a vanity publication. In this connection, when using A7 on an author, the assertion that someone has published a book is a reason for possible importance that defeats speedy, except where the book is self published. When the only book or books are self published I always use A7, & I cannot remember any such deletion being successfully challenged. As the book is likely to be even less notable than the author, I think this criterion is a safe one. DGG ( talk ) 07:53, 19 March 2012 (UTC)

I wouldlike to see __NOINDEX__ added to all of the speedy templates. Google, and maybe other search engines as well, are quick off the mark at finding a new WP article. I have seen them turn up in google search results within minutes. Adding __NOINDEX__ to speedy templates means that visitors who come via search engine links will not see half-baked articles and often outright rubbish. Article marked with a speedy deletion tag can sometimes hang around long enough to be chanced upon. Once the speedy tag is removed the article is then visible to search engines. Sounds like a nice idea to me... -- Alan Liefting (talk - contribs) 00:20, 17 March 2012 (UTC)

Sounds like a decent idea. (Though I'm open to other perspectives on this.) I think Template:Under construction might be another one for this. - jc37 01:04, 17 March 2012 (UTC)

I proposed this some time back and it was explained to me, here, that the NOINDEX function is disabled in the mainspace so we can add it all we want but it won't function. It's possible the technical barrier has been removed, but I thought I'd provide the benefit of past discussion.--Fuhghettaboutit (talk) 04:58, 17 March 2012 (UTC)

Looks like that barrier is removed. See

Wikipedia:NOINDEX#Individual_pages_2. -- Alan Liefting (talk - contribs

) 08:11, 17 March 2012 (UTC)

This has already been done, on {{db-meta}}. →Στc. 08:15, 17 March 2012 (UTC)

Great, so why aren't articles with the speedy templates put in Category:Noindexed pages? Is the noindex done but not the categorising? Which is fine by me. As long as the articles don't get picked up by the search engines. -- Alan Liefting (talk - contribs) 08:36, 17 March 2012 (UTC)

Noindex doesn't work in mainspace, as there are many ways malicious users could use it to screw around. There may be a way the devs can get it to only work in certain circumstances, but it doesn't now. The Blade of the Northern Lights (話して下さい) 15:58, 17 March 2012 (UTC)

I'd be uncomfortable with this, not least because if an article gets tagged G10 and blanked I would prefer that mirrors picked up the tagged version rather than kept the attack. My preference would be that all unpatrolled pages are noIndexed, yes that would require a change in policy to allow NoIndex in mainspace, but I think that would be a good change, and as long as we kept it to unpatrolled=noIndex then it would be hard to game. ϢereSpielChequers 18:51, 17 March 2012 (UTC)

I like the idea of unpatrolled=noIndex. What happens after the 30 day limit on the patrolling? Does the unpatrolled status drop off (and therefore NoIndex)? -- Alan Liefting (talk - contribs) 01:14, 18 March 2012 (UTC)

We've been discussing it in the context of the new tool being developed for newpage patrol. The idea is that either the 30 day limit would go or it would become 60 days. If the latter then logically they should cease to be noIndex at 60 days, we may not have spelled that one out, but if we haven't done that we'd have some mainspace articles that are noindex without anyone having a way to spot or change that. Related to the NoIndex until patrolled idea is that you would no longer patrol articles when you tagged them for deletion, tagged for deletion would become a third colour along with yellow and grey. So the attack pages and so forth wouldn't be picked up by search engines and mirrors during their time at cat:Speedy ϢereSpielChequers 12:03, 18 March 2012 (UTC)

Given the fact that the vast majority of valid CSDs are deleted within minutes, and we have no information or guarantees on how quickly a search engine may notice the removal of a no-index tag (If search engines take their time try try to reindex a page which have been explicitly marked to be not indexed, all you need to do is to CSD tag a popular page to remove it from the world), I don't see that this proposal is a net positive. henrik•talk 18:59, 17 March 2012 (UTC)

Sigh... More complications because WP is too easy to edit for the vandals, and the bad editors, and the good faith bad editors, and the mischievous students... -- Alan Liefting (talk - contribs) 01:14, 18 March 2012 (UTC)

The timing is pure serendipity I think but check it out: Wikipedia:Requests for comment/NOINDEX.--Fuhghettaboutit (talk) 04:02, 21 March 2012 (UTC)

Criterion G12 says to delete:

"Text pages that contain copyrighted material with no credible assertion of public domain, fair use, or a free license, where there is no non-infringing content on the page worth saving."

An article I created, X Lossless Decoder, which was a lightly edited version of this GFDL'd article, was speedily deleted (I didn't know GFDL content wasn't allowed, and still don't quite understand why it isn't, but that's a separate issue). So, unless GFDL is not considered a "free license", either the wording here needs to be changed or that wasn't a case for speedy deletion. Which? False vacuum (talk) 00:19, 20 March 2012 (UTC)

The wording definitely needs changing. I'd agree that GFDL is a free license but we can't use material that is only licensed under it, or indeed several other free licenses. I'll go and boldy change it as it is clearly currently wrong. Dpmuk (talk) 02:08, 20 March 2012 (UTC)

Well, that's an answer. And a very efficient edit, too. False vacuum (talk) 04:42, 20 March 2012 (UTC)

For future reference, Dpmuk added a link to

WP:Compatible license, which redirects to WP:FAQ/Copyright#Can I add something to Wikipedia that I got from somewhere else?. Flatscan (talk

) 05:12, 21 March 2012 (UTC)

Please leave in the note that redirects are not deletions. This point is often confused. Jehochman ^Talk 21:26, 18 March 2012 (UTC)

I agree. It's not overkill to clarify something that editors are "getting wrong" on a regular basis. Jclemens (talk) 22:29, 18 March 2012 (UTC)

[G4] also excludes content which was redirected or merged after a deletion discussion, undeleted via deletion review, or.... I thought that if the article was redirected, then it is no longer "sufficiently identical" to the revision that was taken to AfD, and thus not able to be deleted as G4. →Στc. 23:14, 18 March 2012 (UTC)

This edit seems to stem from WP:Articles for deletion/The Shrike (2nd nomination). The previous AfD WP:Articles for deletion/The Shrike was closed as redirect without deletion. Flatscan (talk) 05:09, 19 March 2012 (UTC)

That is not necessarily the case: delete and redirect is both a deletion and a redirect. I know that there's disagreement over whether deletion should be used before redirecting, but that is a separate issue. I think the addition is redundant to the base criterion, "deleted per a deletion discussion." Flatscan (talk) 05:09, 19 March 2012 (UTC)

Deletion before redirection as an AfD outcome serves one of two purposes:

1) If the content is unacceptable and shouldn't remain accessible to non-admins (BLP, copyvio, promotion, attack), then deletion is protective.

2) Otherwise, it's simply a way of preventing non-admin editors access to not-currently-meeting-standards content. If the issue is people keep undoing the redirection, that's what (semi-) protection of the redirect page is for.

Needless to say, I think option 2 is far more common than 1, and redireciton without deletion is more appropriate in those cases. When I saw old, previously deleted yet inoffensive redirects like that, I used to go through and restore the history under the redirect, just in case someone feels like sourcing and expanding the content. Haven't done it in a while, but would be happy to do so on request... Jclemens (talk) 07:05, 19 March 2012 (UTC)

We discussed a piece of this issue at User talk:Flatscan/RfC draft: Restoring to merge after deletion at AfD. If you would like to deprecate delete and redirect except in those limited cases and permit unilateral restoration under redirects, I can dust off the draft for consideration at WT:Articles for deletion. Both deletions and restorations under redirects are being done, as of a few months ago. Flatscan (talk) 05:11, 21 March 2012 (UTC)

How often is G4 incorrectly invoked for an AfD closed as redirect or merge? I agree that users conflating discrete outcomes is disappointingly widespread, but I don't see it as specific to G4. I suggest publicizing WP:Guide to deletion#Recommendations and outcomes. Flatscan (talk) 05:11, 21 March 2012 (UTC)

Seeing no responses, I reverted to the stable version. I am open to adding something like "merge, or redirect without deletion" to the existing footnote. Flatscan (talk) 05:03, 23 March 2012 (UTC)

Is there any reason why F9 exists as separate criteria and why it couldn't be unified with G12, and than F9 depracated? Armbrust, B.Ed. ^{Let's talk}_{about my edits?} 15:10, 20 March 2012 (UTC)

uh... good question. They have a unified user warning ({{

Nothanks-sd}}) and the textual differences are minor, so I don't see why not. The only problem is that the wall of text below G12 on this page will become even bigger. Yoenit (talk

) 15:27, 20 March 2012 (UTC)

It would seem that one of the original reasons for G12 was actually images (see

CAT:CVSD. Dpmuk (talk

) 18:24, 20 March 2012 (UTC)

Image deletion & article deletions are different. and usually worked on by different admins. In particular Article G12s require a different set of judgments about whether there is non i fringing content of an opportunity for stubbification, none of which is present with images DGG ( talk ) 00:17, 25 March 2012 (UTC)

G12 and F9 already use the same speedy deletion category (Category:Candidates for speedy deletion as copyright violations), so I don't see how a merge would affect that in any way. Yoenit (talk) 13:24, 25 March 2012 (UTC)

One of the primary reasons for the difference is the complexities of fair use rules which exist for files but not for non-file namespaces. We can host exclusively unfree content on a file page; this isn't the case for any other namespace (textual fair use is never its own page). We also have an entirely different licensing system for the two; text added to Wikipedia is automatically assumed to be GFDL or CC-BY-SA-3.0; files uploaded must be explicitly given a license and source - there are a number of different processes for handling incorrect licensing issues (in total, I count 10 different processes at

WP:FFD). That said, I entirely support deprecating F9 and merging it with G12, so long as we definitely keep F9's caveats intact. Magog the Ogre (talk

) 03:47, 26 March 2012 (UTC)

The reason that it applies to recently created pages that were moved is because a startling number of articles are created at absolutely hideous titles. Two such examples in my move log are

WP:DAFT). Even beyond those, there are plenty of instances where articles are created with quotes in the title, with some sort of weird typo (I remember one which had w0hen instead of when), with foreign language and Latin characters put together, or missing a space between a letter and parenthesis; those are routinely deleted per R3, and a quick glance at my move log from when I was on the non-admin side of NPP will give you plenty of examples. The Blade of the Northern Lights (話して下さい

) 21:54, 12 March 2012 (UTC)

(edit conflict) Unfortunately, I have just as many examples of bad R3 nominations (many of which are not being properly vetted by the deleting admins). For example, the current wording is incorrectly baiting editors to tag and delete a redirect under R3 because the redirect was "recently created" by the pagemove process despite the content having existed at the old title for several years. And while I agree with the implausibility of most of your examples, we have debated and kept several meeting your descriptions if they were not harmful - typos that appear implausible to you (and sometimes to me) turn out to be entirely plausible to users of alternate keyboards, for example.
If the R3 criterion had been proposed today, I seriously doubt that it would be approved. The term "implausible" is too open to interpretation and is not leading to nominations in which "most reasonable people should be able to agree whether an article meets the criterion." Rossami (talk) 22:23, 12 March 2012 (UTC)

I agree that R3 is frequently misapplied. Maybe we can make it clearer. Would it help to expand the criteria with something like "does not apply to redirects created when established articles are moved"? - Eureka Lott 22:44, 12 March 2012 (UTC)

Don't think so. If the title is really implausible, it can be nominated for

WP:RFD then picking a clear wording and watching its usage. — Dmitrij D. Czarkoff (talk

) 08:54, 13 March 2012 (UTC)

As one of the normal RM closers, I will say that I don't always remove unlikely redirect pages. There is enough work fixing templates, and not free files and other cleanup. Also the inbound links can take a while to clear so at the time of close it is difficult to determine if the redirect should remain. So a review after allowing a day or more for cleanup would seem reasonable. But a default keep of all of these does not seem reasonable. Vegaswikian (talk) 22:21, 12 March 2012 (UTC)

I see three legitimate uses for R3 after page moves:

1. If the page was moved to the wrong title accidently (for example, Wikipedia:Social Media and Television).
2. Viable articles recently created under extremely bad titles (examples given above)
3. Meaningless filenames (for example, File:H65446wecdp.jpg).

Examples of R3 being misapplied are easy to find, see for example the logs of

The problem is that the criterion is being misapplied wildly, and even some of the examples you draw above may be worth keeping under certain circumstances (off-site linking, establish use patterns, etc.). Effectively there is no proper way to decide on whether such redirect should be deleted or kept without having the clean at least month-long stats. Given the ease of misapplication of criterion, I still think that forbidding the speedy deletion of the redirects after move is the best solution. — Dmitrij D. Czarkoff (talk) 10:24, 13 March 2012 (UTC)

• redirects are cheap and it is very difficult to identify redirects which no-one is going to find useful. Obviously redirects to pages deleted on notability grounds can go, and redirects where the creator made a typo and is asking for it to be cleaned up. But many redirects are deleted quite unnecessarily. There is also the problem that many redirects result from people correcting newbie errors, and if such redirects go then the newbies my think their creation has been deleted rather than merely renamed. ϢereSpielChequers 14:36, 13 March 2012 (UTC)

Given the number of times that CSD#R3 is properly applied compared to the number of times that it is mis-applied on any given day, I would rather see the criterion deprecated altogether. The increased load to RfD would be manageable. Failing that, I think we need much clearer wording such as

R3. Implausible typos
Recently created redirects from implausible typos or misnomers. Implausible redirects are ones that no reasonable, good-faith editor would knowingly create or support. Redirects from

US President

are implausible. Redirects from common misspellings or misnomers, some redirects in other languages, redirects ending with "(disambiguation)" that point to a disambiguation page or redirects which bring a pagetitle into compliance with the Manual of Style, on the other hand, are are not implausible. This criterion may also apply to redirects created as a result of a pagemove but only if the moved page was itself recently created at the implausible title. "Recently created" is measured from the creation of the title and is generally understood to be the last few hours or days at most. Pages with history (articles or stubs that have been converted into redirects) are ineligible for this criterion.

It's a lot longer than I'd like it to be and deprecation would be easier but I think that wording would clear up the worst of the confusion. Rossami (talk) 02:21, 14 March 2012 (UTC)

I wouldn't oppose such explanation. — Dmitrij D. Czarkoff (talk) 06:39, 14 March 2012 (UTC)

Are there any objections to the proposed rewrite? Rossami (talk) 21:53, 19 March 2012 (UTC)

I don't think the "Redirects from apple to orange or from profanity to US President are implausible" sentence is needed. I.E. it doesn't clear anything up (and indeed those examples could be considered vandalism). Thus, I suggest dropping it.--ThaddeusB (talk) 23:22, 19 March 2012 (UTC)

As I said above, I think that R3 should be changed, but your proposed wording probably is too inflexible. An example:

Roberto Hernández (baseball, born 1980). The editor making the move at first accidentally moved the article to Roberto Hernández (baseball born 1983), but rectified his mistake within a minute. I deleted the misleading redirect under R3, and it existed for less than half an hour. We need to retain some flexibility for administrators. - Eureka Lott

23:24, 19 March 2012 (UTC)

I do not think that the proposed wording gets in the way of situations like

US President but it is the self-evidently different content of apple to orange. I say self-evidently because his correct birthyear was shown in the article. Had the birthyear in the moved title been correct, it would have had to be a different person.
But maybe I'm missing something. What interpretation do you see that would get in the way of deleting that particular example? Rossami (talk)

04:18, 20 March 2012 (UTC)

Your proposed change says that it "may also apply to redirects created as a result of a pagemove but only if the moved page was itself recently created at the implausible title." Even though it only existed at the deleted title for one minute, the article was created in 2005. Under a strict interpretation of your wording, the page would be ineligible for speedy deletion. - Eureka Lott 13:54, 20 March 2012 (UTC)

How about "but only if the moved page was itself recently created at or moved to the implausible title". Yoenit (talk) 15:10, 20 March 2012 (UTC)

I see the problem now. Yoenit's fix seems to resolve it. Thanks. Rossami (talk) 15:25, 20 March 2012 (UTC)

Any additional thoughts? Are we ready to make the change yet? Rossami (talk) 13:48, 2 April 2012 (UTC)

A3 includes, "Similarly, this criterion does not cover a page having only an infobox, unless its contents also meet the above criteria." I don't understand: if the page has only an infobox, how could it meet other criteria? And is that supposed to mean other A3 criteria or other non-A3 criteria? Please clarify. —

I think the point of this sentence is to say that an article which only consists of an infobox is not eligible for A3 just because it only consists of an infobox. People do sometimes nominate articles consisting of just an infobox for deletion under A3, even when the infobox contains meaningful substantitive content, because they think that the lack of non-infobox content makes the article eligible. Hut 8.5 11:00, 31 March 2012 (UTC)

This edit a few months ago sneaked through a significant change to the policy: instead of saying that a page which survived the most recent deletion discussion cannot be speedily deleted except for newly discovered copyright violations, it now says that such pages should not generally be deleted. The edit summary for the change says it was made as a result of this RfC, but this particular aspect of the change was not part of the proposed wording in that RfC (or even mentioned there) and it doesn't seem to have been discussed on this page. Speedy deletions are supposed to be uncontroversial, if a page has survived the most recent previous deletion discussion then its deletion is at least somewhat controversial and individual admins should not be allowed to overrule the community. I propose we change the lead back by removing the word "generally". (There is some related discussion here).) Hut 8.5 10:08, 2 April 2012 (UTC)

• I agree completely - having 'generally' there just invites abuse.--Michig (talk) 10:20, 2 April 2012 (UTC)
• Support: I see neither sense in this change nor consensus behind it. I wanted to start a discussion on actually setting a timeout that would discard the previous deletion discussions. — Dmitrij D. Czarkoff (talk) 10:22, 2 April 2012 (UTC)
• Support revert of change. I see no consensus for the addition of the word "generally", and we get enough abuse of CSD already without inviting more. -- Boing! said Zebedee (talk) 11:04, 2 April 2012 (UTC)
• Support – Because there are few checks and balances on speedy deletions, CSD is one of the few areas of policy where the community has continuously supported tight, restrictive wording; throwing the word "generally" in there is not in keeping with that. Paul Erik ^{(talk)(contribs)} 11:17, 2 April 2012 (UTC)
• Support removal of generally, and that the additional language added directly to G4 be changed to a footnote, as I originally suggested at the RFC, which appears to have been what was agreed to and expected by others.--Fuhghettaboutit (talk) 11:59, 2 April 2012 (UTC)
• Support Per above. There was no consensus for this particular change and if it were discussed, I'd strongly oppose that wording for the abuse it invites. If a page that previously survived XFD should have been deleted, that's what DRV is for. If consensus has changed in the mean time, then it should be rediscussed. There is no possibly scenario I can think of in which a previously kept page now qualifies for speedy deletion (with the exception of G12 of course). As soon as the community decided to keep something, it's no longer in the hands of any admin to decide otherwise. Regards SoWhy 12:01, 2 April 2012 (UTC)
• Support per above. Rossami (talk) 13:27, 2 April 2012 (UTC)
• Support I have seen articles be speedied despite having previously survived an AFD, but I can't remember seeing one that was correctly speedied despite surviving an AFD. ϢereSpielChequers 13:38, 2 April 2012 (UTC)
• Support per above. BOZ (talk) 14:53, 2 April 2012 (UTC)

 Done. — Dmitrij D. Czarkoff (talk) 16:12, 2 April 2012 (UTC)

I think we need to clarify

F2. Corrupt or empty image.

Files that are corrupt, empty, or that contain superfluous and blatant non-metadata information.[8] This also includes image description pages for Commons images.

• If it is a description page for a Commons file that consists entirely of categories and/or information not relevant to any other project (like {{
  FeaturedPicture
  }}), it should not be deleted.

I don't much care about whether or not categories is in the final version; I just would like some clarification. Magog the Ogre (talk) 03:38, 26 March 2012 (UTC)

If this conversation sits here much longer, I'll take the silence as lack of disagreement and I'll just change it. Magog the Ogre (talk) 21:56, 3 April 2012 (UTC)

Hi, I just proposed a change to A3 to clarify it shouldn't be used when there is content in history. This was reverted with a note that all speedy deletion criteria are invalid if somewhere in the articles history there is a good version. This seems sensible to me, yet currently this isn't specified on the project page, and criteria G1, G10, G12 and A10 name them explicitly. Would it be a good idea to include a blanket statement in General and Article, or on the project page to indicate it doesn't apply and remove them from the individual criteria? Martijn Hoekstra (talk) 15:13, 2 April 2012 (UTC)

Well,

WP:CSD#Procedure for administrators says that admins should "[...] check the page history to assess whether it would instead be possible to revert and salvage a previous version [...]". I reverted your addition because I don't think there is reason to add something that is already a general principle to every criterion individually. I think the current inclusion in the criteria you mention is due to historical reasons only. I would support including a clarification in the introduction and removing the superfluous mention in G1, G10, G12 and A10 though. Something like "Users should not tag pages for speedy deletion if there is a version of the page that can be reverted to that does not meet any criterion for speedy deletion; likewise, admins should not delete pages where reverting to a prior state is an option." should do the trick. Regards SoWhy

16:52, 2 April 2012 (UTC)

Huh... I wonder when that changed. The page used to be explicit that both the current content and every preceding version had to qualify for speedy-deletion before CSD was allowed. (That's not to say that every version had to qualify under the same criterion - a vandalism page overwritten with A7 content is still speedy-able.) We should make that explicit again here. Rossami (talk) 17:21, 2 April 2012 (UTC)

Something like this?

I would prefer this version. — Dmitrij D. Czarkoff (talk) 20:58, 2 April 2012 (UTC)

BTW, may be

I disagree. If there is no harm in keeping the revisions that meet speedy deletion (!= the same as all newer revisions), just reverting to the old version is usually enough. In fact, if the revisions in questions ought to be deleted, they have to qualify for

WP:REVDEL imho. Admins are not expected to selectively delete revisions that meet any CSD using REVDEL, so they shouldn't do it just because it was tagged. As for the wording, I don't like your wording nor Martijn's. Imho, it should mention that the revisions are not eligible for any criterion (your wording sounds as if the same criterion has to be met) and that people should revert to the old version instead. Regards SoWhy

21:18, 2 April 2012 (UTC)

I decided just to kick it off with an edit, edit away if you don't like something. Martijn Hoekstra (talk) 21:20, 2 April 2012 (UTC)

Personally, I prefer the wording from 2006 or even better from 2004 which read
Always check the

restored, such that deletion is not necessary.
Those versions were more positive in tone - "here's something you must do" rather than "thou shalt not". I like the 2004 wording better because it put the burden on everyone - adminstrators to check history before deleting but also regular editors to check history before tagging. Rossami (talk)

22:15, 2 April 2012 (UTC)

I don't mind the current revision, but like to note I'm firmly in the 'whatever, as long as it is clear that pages with valid history are ineligible' camp. If you feel another wording would be better, I believe {{sofixit}} would firmly apply here. Martijn Hoekstra (talk) 13:24, 3 April 2012 (UTC)

Suggest that G4 (Recreation of a page that was deleted per a deletion discussion) be extended to incorporate articles that were deleted under G7 (Author requests deletion). Alternatively, G7 cannot be user on 2 substantially similar articles (but is that really checkable?). In the case of Robert Young (Pro Cyclist & Triathlete) we are currently in the 4th incarnation of an article on a sportsman who fails any notability standard, but for whom the various guises of the creator responds to the inevitable AfD listing by blanking the article, then re-creating it with a slightly different disambiguator a few days later. Thus AfD, and the threat of speedy deletion of a recreation under CSD G4, is evaded. This loophole should be closed. Kevin McE (talk) 09:14, 3 April 2012 (UTC)

Simply let the AfD play out—don't accede to a G7 request either explicit or by blanking, which is what is happening right now at the current AfD. The situation is uncommon and for that reason I don't see any need to consider changing any criterion to address it.--Fuhghettaboutit (talk) 09:27, 3 April 2012 (UTC)

I agree: although I don't see it anywhere in

afd1}} template says "the article must not be blanked... until the discussion is closed", and that is good enough justification to revert any blanking and decline any G7 while the AfD runs. There might be situations where it would be sensible to accept a G7 and close the AfD on an IAR basis, but there is no need to do that if it seems to be being used to evade scrutiny. JohnCD (talk

) 09:44, 3 April 2012 (UTC)

Indeed. We can even let the AfD run its course if the article is G7'ed halfway through the deletion process too; I don't know of any rule that disallows that, and if there were, we can safely

ignore it, because it would be a bad idea to follow it. Martijn Hoekstra (talk

) 10:03, 3 April 2012 (UTC)

Agree. A call for G7 (or U1 at MfD) should be taken as a highly significant !vote, and not be allowed to sidestep completion of a community judgement. --SmokeyJoe (talk) 10:44, 3 April 2012 (UTC)

• Just let the AfD run its course; the advantage of using an AfD in a case like this, is that subsequent re-creations can be treated by G4. The G7 suggestion should just be treated as a !vote for delete. DGG ( talk ) 04:43, 5 April 2012 (UTC)
• I was thinking about this thread when I closed this AFD. I think what I will do in cases like this is reopen and relist the original AFD and let it run its course. --Ron Ritzman (talk) 14:00, 7 April 2012 (UTC)

• I think that in that case, you should have deleted without reference to G7. --SmokeyJoe (talk) 00:29, 10 April 2012 (UTC)

• You're probably right and I might have if the 2 delete !votes weren't "per noms" but this way if someone writes a new article that is "referenced" and "accurate" (the nominator's concerns) it won't be subject to G4. In any case I'm watching the red link and if it's recreated in the same state that it was nominated in, I'm going to break out the NUCLEAR TROUT. --Ron Ritzman (talk) 01:59, 10 April 2012 (UTC)

A9 seems oddly specific, and seems like it was just created to combat people using Wikipedia to promote their garage band. Shouldn't this rule apply to any work of art? --WikiDonn (talk) 17:17, 8 April 2012 (UTC)

CSD criteria are deliberately specific, and shouldn't be construed broader than the letter. Is there an actual need to have a possibility to speedily delete other works of art? Martijn Hoekstra (talk) 17:21, 8 April 2012 (UTC)

I think you hit the nail on the head. Many new editors write about their favorite band (or their own band), and then they start writing articles about the band's records. — Malik Shabazz ^Talk/_Stalk 17:23, 8 April 2012 (UTC)

If I recall correctly A9 was created with the thought that if a band is non-notable, it is unlikely that its musical recordings will be notable. →Στc. 20:02, 8 April 2012 (UTC)

Yes, that's correct. Which is also the reason why it does not cover any other works of art. Regards SoWhy 20:05, 8 April 2012 (UTC)

One of the problems that we are building up over time is an accumulation of IP pages with warnings. This doesn't matter for userpages because the user page should relate to the same person, just maybe a decade older. But IPs get reassigned, and a page full of warnings from 2003 isn't likely to be relevant to an editor on that IP in 2023. I had thought that we were deleting some of the old and out of date IP talkpages, but it doesn't seem to be under any of the codes. Is this an acceptable use of G6? I was thinking maybe we should courtesy blank some of them, but presumably that would result in a "you have messages" bar going to the current user of that IP. ϢereSpielChequers 16:53, 7 April 2012 (UTC)

If I'm not mistaken, minor edits from accounts with the bot flag don't throw up the orange bar. Couldn't that be a way to blank some of these? — Preceding unsigned comment added by Ron Ritzman (talk • contribs) 18:13, 7 April 2012‎

See also: SharedIPArchiveBot. →Στc. 07:42, 9 April 2012 (UTC)

I don't see a problem with G6 deletion of stale dynamic IP Talk pages - we don't need them for anything, and all they do is confuse newcomers reusing the same IPs. -- Boing! said Zebedee (talk) 10:52, 9 April 2012 (UTC)

This request is coming over and over (please see the archives). There is a bot going around to archive them (SharedIPArchiveBot). There is absolute no reason to delete these pages, deletion does hide in some cases important tracks in vandal fighting, or important discussions which nonetheless were blanked. Deletion of talkpages is certainly not non-controversial - admins have been blocked for this. --Dirk Beetstra ^{T C} 11:04, 9 April 2012 (UTC)

OK, if they get archived, that's cool - but I do see a very big difference in importance between registered user ans static IP talk pages, and dynamic ones (though I guess the latter can be useful for investigating disruption from IP ranges). -- Boing! said Zebedee (talk) 11:22, 9 April 2012 (UTC)

Things that I encounter include that I find an IP spamming a certain link. Sometimes, another IP has already been warned repeatedly that their link-additions (of exactly the same link) are considered spam, and have been told what would be the following steps. Since it is very likely, that both IPs are related to the same 'organisation', and often are the same physical person (in the

best case, they both work for the same company, and the manager who suggested the first employee to spam the link did not tell that to the manager of the second employee - so maybe the second employee was not aware of the earlier spamming), I dare with the new IP to start immediately with stronger warnings (even if they are two independent employees of the same company, so not meatpuppets in a strict sense of the word, it is still company responsibility that they do not spam). If I find several {{uw-spam4im}} in series of some time ago, I will now issue a {{uw-spam4im

}} immediately (or even blacklist) - if I can't find those I would start from the beginning. Also, that several such warnings to different editors have been issued at some time is used as evidence that blacklisting was the only option. When such talkpages were deleted, one loses the easy access to such tracks, and might start over (and the link may escape a well due blacklisting ..).

Although I would not object to deletion of talkpages of editors who were only warned for petty vandalism, there the problem is that quite often people revert spam and warn with the vandalism templates ({{

uw-vand1

}}, e.g.) - still those editors should see that as a warning that the edits they perform are not wanted, and a follow-up vandalfighter may use that as a reason to start higher.

Then you have the problems with editors removing warnings, but those are still found in the old revids of a page. If I feed you a list of 20 spammers, and you have 20 redlinks, you'd take a long time to find whether these spammers were warned before - with blanked or archived pages that goes already much faster (and one does not need an admin bit for that action, do note also that many of the cross-wiki active vandal fighters are not admin here).

All in all, some of those IP talkpages with warnings may be of interest, and it would take quite some work to see which ones. Quite some are related to petty vandalism and those are generally useless in the future (but as you said, IP-range vandalism might those also interesting tracks). Others are generally quite useless edits to talkpages and those could also be deleted. But it takes a lot of common sense to see what can be deleted and what not. One admin recently did a short run of deleting 20-25 pages, of which I immediately undeleted one since it did relate to a quite recent spam case, and 10-11 after a short discussion since I felt that they were not non-controversial deletions. I would therefore not delete lightly delete user talkpages under CSD criteria .. --Dirk Beetstra ^{T C} 12:04, 9 April 2012 (UTC)

But how many years would you go back and still find old IP messages useful? Three? Four? I'm certainly not interested in getting rid of such pages if they may be useful, But surely there is a number of years after which we can agree that such things are stale and best archived or deleted. If we have a Bot doing this uncontentiously for shared IPs after a matter of weeks, then surely after some period of years we can do the same for IPs that are not tagged as "shared". ϢereSpielChequers 12:14, 9 April 2012 (UTC)

Yep, I can see there are good reasons for not deleting, and I agree that bot-archiving is the way to go. But I also like the idea of archiving very old IPs not marked as shared too - the vast majority of even "static" IPs do eventually get re-allocated, even when one doesn't (eg a company address), there's a good chance it's a different person using it now. -- Boing! said Zebedee (talk) 12:20, 9 April 2012 (UTC)

WereSpielChequers, I have seen links being blacklisted after massive sockpuppetry by users, which were over a year later de-blacklisted (by an unknowing admin who needed one link on the site), resulting in a restart of sock-spamming only a couple of weeks later. Spammers make money with their links - that is a massive incentive to do it. There is no real gain in deleting these pages, it does not free serverspace, it might take away some 'you have new messages banners' (if that IP was indeed never used in those years to browse Wikipedia). And that while there may be some cases where there is interest in that old data. --Dirk Beetstra ^{T C} 12:19, 9 April 2012 (UTC)

Hi Beetstra, I'm not trying to hide messages that are recent enough to be relevant. The gain is that the current user of the IP probably has nothing to do with warnings from many years ago. So getting rid of such old warnings may make the site more attractive for potential IP editors. I'd be happy to exempt IP pages with spam warnings or put the test at a long enough time period that spammers would have lost patience - Would you be concerned if this was only being done to IP pages that had not been edited in two years? Or if we had some formula such as "no talkpage activity for a year and no spam warnings for more than four years? ϢereSpielChequers 14:04, 9 April 2012 (UTC)

No, you'd have to exempt spammers, not talkpages that have a spam-warning on it (see above). And if the messages are archived, then they are not bothering a new user. However, even if they are three years old, those messages may be evidence for e.g. blacklisting, or may be a reason for blacklisting. And note, spam is an example, there can be other reasons as well. Believe me, even deleting talkpages older than 5 years will result in cases which should not have been deleted because there is information there that is needed. And on top of that, it does not gain a lot, especially not when the page is already archived and replaced with a suitable, general template. --Dirk Beetstra ^{T C} 14:08, 9 April 2012 (UTC)

Oh, and spammers don't loose patience ... they earn money with it. --Dirk Beetstra ^{T C} 14:09, 9 April 2012 (UTC)

OK so can I take it that you'd be OK with extending the archiving from shared IPs to all IPs, providing we don't actually delete warnings? ϢereSpielChequers 14:38, 9 April 2012 (UTC)

GOOD GOD DO NOT.

Do you all have any concept of the size of the English Wikipedia's database at this point? Over 64GB of uncompressed wikitext! That archive bot just creates an additional page for every single IP talk page. You're quite literally suggesting making the problem twice as large as an effective solution. The answer is to delete the pages. End of story.

Dirk has been going on about the spam problem for years, but it's a red herring in the context of this discussion. Over ninety-nine percent of these IP talk pages have nothing to do with spam (or any pattern of behavior) and can safely be deleted. I know this from sampling. Dirk knows this from sampling. It's completely outrageous that a vandal edit five years ago still warrants a separate user talk page that will be kept around indefinitely. It's patently absurd to suggest that it should now have two pages devoted to it. --MZMcBride (talk) 14:49, 9 April 2012 (UTC)

MZMcBride, yes, but get to a good and strict set of rules. I've seen an admin doing 25 deletions a couple of weeks ago - one was an active ongoing case of spam. 10 of them contained discussions where the editor was in an active discussion. I doubt if you get to 99%. You have to get the rules for deletion to a strict set, and I doubt that you can do that with a speedy. I am fine with the deletion if there is a strict subset, but look further than only spam pages. And I tell you, deletion of certain of these tracks does hamper our efforts. --Dirk Beetstra ^{T C} 05:38, 10 April 2012 (UTC)

I encourage everyone to look through Wikipedia:Database reports/Old IP talk pages. Sample the collection and see what percent of these pages need to be kept. Does someone adding this vandalism really need to have an IP talk page forever? This is insane, and of course the database report only shows the tip of the iceberg. We're talking about hundreds of thousands of pages like this. --MZMcBride (talk) 14:56, 9 April 2012 (UTC)

If that was a userpage for a registered account then I'd not lose sleep about it being there indefinitely. But IPs shift so the someone who has that warning now is unlikely to be the same someone, and yes I agree that such talkpages should go, and whilst we need safeguards to keep track of spammers do we really need three year old talkpages of warnings for juvenile vandalism? ϢereSpielChequers 15:43, 9 April 2012 (UTC)

I don't have a strong opinion on the deletion of IP talk pages that contain nothing but very old warnings, except for:

1. Their speedy deletion should be authorised from WP:CSD (not WP:UP, or anywhere else);
2. There should be a new criteria, not G6, because G6 is already very broad. If G6 were to cover old talk pages, I expect some will believe that G6 can cover nearly everything.
3. "Old" would need a precise definition. I would be confortable with "one year of no edits from the IP since the warning". (If the IP is editing, replace the old warnings with a welcome).
4. There seems to be a case for the deletions/blankings to depend on the warnings. High level spam warnings seem to have more historical interest. On this point, should the highest level of spam warning call for inclusion of the name of the current owner of the IP? I also guess that the historical activity of the IP is an important factor. If the IP inserted a hundred spam links, I imagine that records would be useful. If the IP made a few isolated vandalism edits, long term records seem of less value. If the IP appears to be a static IP associated with a company, I'd want to keep the records. If the IP belonged to a library or school, years old warnings will be of little continuing value.

As per Dirk, I am not convinced of the need. Are there so many that there is an actual, described problem? Why is bot-blanking not sufficient?

I really do not see the point of archiving (copying than blanking). I thought a starting point to the logic here was that the warnings are not associated with the current owners of the IP. If so, what is the purpose of the archive in addition to the talk page history? --SmokeyJoe (talk) 22:11, 9 April 2012 (UTC)

I completely agree. I think we need to get back to the reason we are here which is to build an encyclopedia. Wikipedia is not an experiment in permanently archiving IP talk page warnings. To this end, I have personally blanked tens of thousands of IP talk page warnings, or at least replaced them with the "older warnings" template. However, to the extent that I have seen many thousands of pages consisting of nothing but one or two warnings from 2005 or before, for pages that have never been blocked or identified as shared IP addresses. I think the most sensible practice is to delete such pages outright, and there should be no procedural impediment to doing this. bd2412 T 22:47, 9 April 2012 (UTC)

I agree. Archiving to a subpage makes no sense here (and actually makes the problem worse, as discussed above). Blanking the page is a step in the right direction, but what I think people fail to realize is that keeping these pages around does have an actual cost. If this were a few hundred pages, it would be negligible. But we're talking about hundreds of thousands of such pages. The people working on database dumps most certainly notice these pages. It's simply good housekeeping to delete these pages after a specified amount of time, with no recent edits from the IP, etc. --MZMcBride (talk) 22:59, 9 April 2012 (UTC)

Hundreds of thousands of such pages impacting people working on database dumps sounds a fair reason to delete instead of blank. I have started a draft criterion page here, & see below, for others to edit directly: --SmokeyJoe (talk) 00:25, 10 April 2012 (UTC)

Draft Old IP talk page criteria

I have some qualms about the proposed criteria. Many of the oldest IP talk pages pre-existed bots and warning templates, and consist (or consisted before blanking) of nothing more than an editor literally saying, "Hey, you, stop vandalising". Even human to human contact, even a short conversation, is not worth preserving if it is several years old and represents the only activity relating to that IP address, for precisely the reason that IP addresses are dynamic, and can not be presumed after several years of complete dormancy to have any connection to the person whose activity prompted the original message. It would probably be useful to get a list of IP talk pages indicating the date of the last editing of that page, last activity by that IP, and last warning or block (if any). If we can find common patterns, we can craft a policy specifically addressing what those patterns show. Cheers! bd2412 T 03:01, 10 April 2012 (UTC)

I wouldn't consider "Hey, you, stop vandalising" on its own to be human-to-human contact worth preserving, but getting into this sort of detail will cause the criterion to fail new CSD criterion #1 "Objective" (top of this page). Possibly an important factor is whether the IP ever signed their talk page. I would recommend beginning cautiously. Start with the deletion of IP talks pages on a tight criteria, and then see how many are left. --SmokeyJoe (talk) 03:10, 10 April 2012 (UTC)

I agree that it makes sense to begin with tight criteria and knock out what fits there (which will be, obviously, the least worthy of preservation); I think we still need data to generate metrics, and that the measuring stick should be the length of inactivity on the page, and of the IP. bd2412 T 03:34, 10 April 2012 (UTC)

If the IP ever made constructive edits that stuck, and the IP contributed to its own talk page, then I don't think the talk page should be deleted, ever. --SmokeyJoe (talk) 03:59, 10 April 2012 (UTC)

Yea, I don't follow what the problem is, here (Sorry, I probably missed this somewhere up above). What's wrong with simply blanking the page?
— V = IR (Talk • Contribs) 05:24, 10 April 2012 (UTC)

A single piece of vandalism five years results in a talk page that exists forever. These pages don't need to sit around indefinitely. They serve no purpose and can create problems when there's such a large volume of them. --MZMcBride (talk) 06:03, 10 April 2012 (UTC)

MZMcBride - I do not have a problem with simple vandalism. The problem that I have with it is that this is very, very difficult, if not impossible, to write down as a speedy-criterion. I know that the larger volume of these pages are either 'hi, I exist' edits to talkpages, or single on-off warnings for poop-vandalism. Those can be deleted. I only object to the deletion of those talkpages which are related to any form of systematic vandalism (and spam is the easiest example of it). That makes these deletions controversial. Although that the error rate on blind deletion of every talkpage of any IP that has not been used for longer than 1 year is relatively small (I guess in the order of 1-2%) .. it is that 1-2% that makes me uncomfortable to have it as a speedy. And do note, I know that ~75% of the talkpages of IPs with a spam warning are actually more of the type 'no, your YouTube link is not necessary here' .. and even those could go. I do see that these pages can go, but there is a small, though significant and necessary, percentage of pages which should not go. --Dirk Beetstra ^{T C} 06:10, 10 April 2012 (UTC)

It seems we overlapped in posting replies. :-) I just posted below. I think you may be right about articulating this as part of the speedy deletion criteria. --MZMcBride (talk) 06:12, 10 April 2012 (UTC)

No, I still disagree. The point is not who is now on the page, the point is the editor that was using the account. And the point is not that the editor has a spam3 or spam4 on their page. 10 spammers with a spam1 on their page are also a reason to blacklist, and 10 spammers with a vand1 on their talkpage are also a reason to blacklist. Also, an IP shifting copyvio vandal is also a reason to have higher sanctions (which need to be visible for transparency), same goes for a NPOV-vandal. 12 months is not enough, spammers are sometimes active for years, spammers come back soon after years long blacklistings are undone. These tracks are necessary for a plethora of reasons. You are also ignoring all users who have a block-log, which are also cases which may return and which are getting hidden by deletion. Deletion is harming more than that it would help. If people are loading the database, they can ignore talkspace, or can run a deletion script on their resulting database, but I, still, strongly object to deleting any talkpages with warnings where the warnings are for spam, npov, copyvio, coi or any other form of non-petty-vandalism-vandalism, whether after 5 years, even after 15 years. --Dirk Beetstra ^{T C} 05:33, 10 April 2012 (UTC)

I believe this almost amounts to creating shrines for vandals. In a lot of cases, you can see every contribution made by the IP address and evaluate based on this. If the edits are clearly simple vandalism, does the talk page need to exist forever? --MZMcBride (talk) 06:06, 10 April 2012 (UTC)

No. --Dirk Beetstra ^{T C} 06:10, 10 April 2012 (UTC)

• Blank as absolutely necessary, I suppose, but talk pages shouldn't be speedied. - jc37 05:34, 10 April 2012 (UTC)

  Why's that? --MZMcBride (talk) 06:06, 10 April 2012 (UTC)

  I think deletion of a talk page should always require discussion, unless privacy/blp issues necessitate it.

  I understand you're trying for

  WP:DENY, but I've always been of mixed minds on that. We should have a way to retain the history for future reference, without setting up "shrines", as you call them. And so to me, deletion would seem to be counter-intuitive. And at the very least should need a discussion first. - jc37

  06:18, 10 April 2012 (UTC)

I would be very, very comfortable with a replacement of the contents with a kind of welcome template, telling in a small text at the bottom that there is data in the history of the talkpage for those who are interested in the data. I even would not care if that is done after 30 days of inactivity from the IP / 30 days after the last edit by the IPto the talkpage of the IP. But I have no faith in having speedy criteria for the deletion of these pages. --Dirk Beetstra ^{T C} 05:41, 10 April 2012 (UTC)

I'd be happy with that. My preference would be after a somewhat longer period, I think my home IP is often stable for a year at a time, whilst when I use mobile broadband I suspect it is very unstable. So blanking a stale IP talkpage and replacing it with a welcome after a couple of years inactivity would make sense to me. ϢereSpielChequers 11:07, 10 April 2012 (UTC)

I have been collapsing old warnings (especially for shared IP's from educational institutions and very dynamic IP's (starhub)) for a while now when I happened across them (mostly through AIV). Shouldn't something like that be sufficient? How about a bot that specificaly archives warnings templates once they are older than x (say, a year, or some other metric, I could live with leaving at most 10 warnings, and archiving everything over a year old), and creates a link to the 'warning archive'? Is there any reason for deletion then? MZMcBride, you say these can create problems. What kind of problems? Martijn Hoekstra (talk) 11:47, 10 April 2012 (UTC)

I should start out by noting that I'm rather deletionist. ;-)

When I said "problems" above, there are a few different things that I was referring to.

First, just in terms of sheer size, these old IP talk pages make up a significant percent of the pages on the English Wikipedia. Using the most recent stats and a back-of-the-envelope guess, I'd say we're talking about somewhere between 500,000 and 1,000,000 pages. As someone who's recently been investigating ways to search through all current wikitext in a fast manner, this quantity of pages has actual impact on what's possible and at what speed.

Second, these pages are largely unpatrolled and come with the same kinds of problems that unpatrolled pages have elsewhere. They can be filled with test edits, vandalism, attack pages, copyright violations, and the like. Most are automated or semi-automated warnings, but having gone through a good number of these, some are not.

Third, while this isn't necessarily a "problem," I'd say that deleting these pages after a certain amount of time is just good housekeeping. Any database accumulates over time—that's the point, after all. But it's perfectly healthy to perform maintenance and cleanup on the database routinely. While MediaWiki doesn't completely delete the data (it's moved to the archive table and a logging row is created), it will remove these pages from subsequent database dumps, which will in turn remove them from mirrors and other third-party uses. Even internally, when there's less data to output, things go faster. That is, with enough maintenance and deletion, the time it takes to create database dumps could markedly decrease. This is the same reason that database maintenance is done elsewhere with other types of databases. Cleanup improves performance.

I'm in complete agreement that these pages should be kept if there's any question about their future value. While I personally don't find value in them, I recognize and appreciate that Dirk and others doing anti-spam work and other good work do rely on them. However, both in my opinion and what seems to be the general consensus, a lot of these pages are simply not valuable. It's not worth blanking them and having millions of blank pages lying around. They're almost all the result of scripted warnings from tools such as

Hmm, maybe you're right regarding a speedy deletion criterion. This may be something best left to admin discretion. I don't think trying to prescribe a set of strict criteria will ever work. The draft criteria above are already very complicated. It's ultimately about using good judgment and evaluating the contributions of the IP. Perhaps a guideline could be written instead? --MZMcBride (talk) 06:11, 10 April 2012 (UTC)

Sounds like a way forward. --Dirk Beetstra ^{T C} 06:13, 10 April 2012 (UTC)

I rewrote Wikipedia:Old IP talk pages just now. It doesn't incorporate all of your points (or mine), but it's a start. I'll take another look at it tomorrow when I'm not so tired. --MZMcBride (talk) 06:34, 10 April 2012 (UTC)

Hello. I've made some edits to the 'essay'. I hope they're agreeable, since I can't say I've read everything above. A couple of things I'll mention. I have quite an interest in old pages like these, both in terms of good housekeeping as MZMcBride puts it above, but also tracking long term patterns of various sorts over multiple ranges. I would generally agree with Dirk, whatever he's saying now. For me however the ones you can't delete are IPs explaining why they made an edit, on their own talk page. Some of them are priceless. I've seen a few get deleted in the past, probably with a bot and a sql query. Another thing I'll mention is that sockpuppetry was previously explicitly an issue; this time it's hardly mentioned. Half of those sockpuppetry user pages should go, imo, but that's another issue. Oh and these criteria look very similar to what they would be for indef'd vandal accounts. Just saying. -- zzuuzz ^(talk) 07:06, 16 April 2012 (UTC)

an idea

I dunno if this would fly, but... why do we have IP talk pages anyway? Yes we're the encyclopedia that anyone can edit. But one can edit without needing a talk page.

IPs choosing to not register really probably don't need a personal talkpage.

As an option, when asking for the removal of IP talk pages, we could ask the devs to open access to Project:Unblock requests as a special page for anyone who is blocked. (Though that isn't entirely necessary as we have a specifically purposed mail list.)

It's not as if they have a watchlist or any of the other features that come with an account. And collaboration can happen on one of the myriad other talkpages, whether in article space or wikiprojects, etc. - jc37 06:31, 10 April 2012 (UTC)

I think the biggest reason for having IP talk pages is so that IPs can have the "new messages" bar when there's a message delivered to them. This is particularly important in the case of behavior warnings. --MZMcBride (talk) 06:35, 10 April 2012 (UTC)

(ec)Hmm, no, some of us do choose to edit as an IP, and not register. I know some who are very active editors, and we do need to communicate with them too (and we need records for that). Moreover, where would you leave a warning if such an editor needs a warning? --Dirk Beetstra ^{T C} 06:36, 10 April 2012 (UTC)

What might be worth pondering, is that an IP-editor should not be allowed to edit until they have read the message on their talkpage (or at least, loaded their talkpage, as we don't know whether they actually read what is added).. --Dirk Beetstra ^{T C} 06:38, 10 April 2012 (UTC)

Stale warnings may not be relevant to the current editor on that IP. For example I've now stopped logging in from insecure IPs. I recently spent a quiet hour in an airport terminal fixing typos and in similar situations in the future I think I'll stick to IP editing. If someone needs to send me a message whilst I'm editing that way then fair enough and they'd need an IP talkpage to do so, but I wouldn't need to see old IP messages. ϢereSpielChequers 11:21, 10 April 2012 (UTC)

I was considering the issue of warnings. and I'll just say: I'm not sure. Though I suppose we could do it in some way similar to the "you have new messages" banner (having it link to the warner's talk page, and maybe a button confirming the warning has been read). And the warning would be logged like the block log (an IP warning log). Simple, standardised, and saves on kb : )

And I disagree that IPs "need" a personal talk page for collaboration, any more than they "need" a watchlist. - jc37 06:46, 10 April 2012 (UTC)

• Shared IPs such as schools definitely need a talkpage, if only as a place to notch up the warnings and past blockmessages. But there have been longterm valuable members of the community editing from stable IPs and using the talkpage, and of course there have been instances where IPs have been incorrectly warned and need a place to respond. ϢereSpielChequers 11:21, 10 April 2012 (UTC)

• All of this seems a bit OT here, but what the hell, right? The way that I've always thought about things, IP talk pages really belong to the "community" anyway, so I have no qualms about imposing any sort of standards on them, in terms of content or deletion or anything else (I'd put known static IP addresses in that category as well, by the way... although we could certainly be more lenient with them, I think that they still belong to the project rather than the unregistered user. Those who have a problem with that can simply register a user name). I can at least understand the perspective that brought all of this up in the first place now (the whole DENY aspect is a perfectly reasonable concern), although it's nice to see this moving away from being a speedy deletion criteria. I don't particularly care if "old IP talk pages" are deleted or not honestly, but I didn't understand the desire to do so at all, prior to asking. I still think that blanking is perfectly sufficient (I think that adding some sort of standardized IP talk page template, similar to what is done with known school IP pages, is something to seriously consider), but if you guys want to delete them... I'm certainly not going to stand in your way. It's your time to use as you like, after all. *shrug*
  — V = IR (Talk • Contribs) 20:17, 10 April 2012 (UTC)