Wikipedia talk:Pure wiki deletion system/Archive 1

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 1

Copied from meta

Please can we keep all discussions on this page, and reserve the "proposal" page for a detailed specification of the proposal. Hopefully the proposal will grow and evolve to take account of any ideas and criticisms that emerge on this talk page. It's clear that there are several users who oppose the whole idea in principle, and we should accord their views due respect. However, the purpose of this page is to work out the details of this reform, not to discuss whether we need it at all. A more appropriate place for discussions of those wider questions is Talk:Strengths and weaknesses of the current deletion system GrahamN 14:43, 31 Oct 2003 (UTC)

"Hard delete"/"Soft delete"

  • 'Soft delete': Blanking of page, links to page appears as a non-existent page link
  • 'Hard delete': Deletion as it is carried out now, only sysops can delete/undelete
Is this a meaningful distinction? As I understand it, the contents and history of even "hard deleted" pages remain in perpetuity somewhere in the database, so it's just a question of who is permitted to access them: standard users, administrators, or developers. GrahamN 14:43, 31 Oct 2003 (UTC)
I think this is a common misunderstanding. The contents and history of deleted pages do remain in the database temporarily, and this is why a deleted page can be viewed and undeleted by administrators. But database management can and does remove these deleted pages. After this, the page and its history should still be recoverable from the backup taken before the database reorg, but that's a developer's job. Andrewa 00:01, 2 Sep 2004 (UTC)
So where's the misunderstanding? That's exactly the point I was making. In the present system nothing is deleted irretrievably. Some information can be retrieved by anybody, some only by administrators, and some only by developers. The idea of the reform is to make all the information available to everybody. How could that be a bad thing? GrahamN
Well, the point probably is that database management should be able to permanently remove pages that have been "really" blanked - that have been gone for a long time. Otherwise the page "Sirmob is the coolest man alive" that I created (not really) would never be able to be cleaned up, and it and thousands of others would crowd the database for all time. I've added a point to this effect in "Possible Changes." Sirmob 01:45, 4 August 2005 (UTC)
Basically, soft delete leaves the history of every edit ever made in easy reach of anyone who wants to read or restore them. Hard delete moves them someplace where they're hard to access. Hard deletion has to be left available in some form to save room on the places that are easy to access, and we can't allow regular users to read, restore, or otherwise play around with these 'hard deleted' pages, since it takes many more resources (in extreme cases, a developer manually accessing the backup) to dig them up. Thus, the creation of an intermediate 'soft deleted' state--in this case, blanking--which pages can be moved into and out of as easily as a normal edit. Aquillion 23:12, 26 August 2005 (UTC)

Pros and Cons (please add/edit this)


  1. Because soft deletion can be trivially reverted, it is less of a big deal. All users could do it, dramatically increasing openness and democratic nature of WP.
  2. Decisions to soft delete could be made the same way as all other wikipedia decisions... unilaterally if the change is 'obvious' or through discussion on the talk page if more tricky.
  3. The whole community can come together to deal with pages created by vandals, test pages, and other detritus.
  4. The workload of administrators will be dramatically reduced, allowing them to get on with the real work of writing brilliant encyclopaedia articles.
  5. Less strain on sysops to be in the "front line" against vandals.
  6. The current need for sysops to excercise their best judgement in judging consensus is divisive and creates arguments, suspicion and resentment, placing additional strain on sysops
  7. The central VfD page gets lots of edit conflicts, discouraging discussion and discouraging listing of pages that should be deleted.
  8. The lack of space on VfD hinders proper debate and discussion of options available, leading to stereotyped keep/delete arguments and no real attempt at consensus.
  9. Deletion discussion could take place on a page's talk page.
  10. Someone who finds the information useful (either for another Wikimedia project or for a non-Wikimedia project) can still use it. In addition to the direct benefits to Wikimedia and society, this will have the indirect benefit of lessening the animosity between many of the inclusionists and deletionists.


  1. More potential for damage? Vandals who decide to blank pages wreak slightly more havoc as they cause red links to appear on other pages.
    • Vandals already blank pages. The high number of red links would cause a more rapid response, thus reducing the potential for damage, not increasing it.
    • A Special page listing of recently blanked pages could be kept.
  2. More strain on server, with more pages for it remember/linkify
  3. The disappearance of VfD would amount to the loss of a real community page where lots of users interact, and get to hear about recent developments. A lot of discussion on VfD gives people a good idea of what should or should not appear in Wikipedia.
    1. The best thing for an open content encyclopedia is to encourage as many people as possible to contribute their knowledge. While communities are nice for the people that live in them, by definition they are exclusive. So the more "real community pages" there are, the less effective Wikipedia will be.
  4. May add extra strain to another oft-edited page, the Village Pump, as people rush there to say... blanking of X is being considered... please come and support/deny it at the talk page... maybe we'd end up with a pages where blanking is being considered page which could (given how built-in the VfD way is) become a de-facto VfD
    1. Mitigating factor: The number of controvertially deleted pages would presumably be significantly less than the total number of deleted pages, so this discussion page would be more wieldy.
    2. Restoring a page you believe should not have been deleted would be as trivial as reverting an edit you believe shouldn't have been made. You don't get people rushing to the Village Pump to say "the word English has been used on such-and-such a page instead of British. Please come and help me change it." There's no need. You just do it yourself.
    3. Yes, but as with any significant change, there might need to be consensus reached in order to decide whether the page should live as a blank page or as a page with content - and deletion isn't like English/British, it's a big thing. But that said, yes, another page might become a de-facto VfD. But that's displaying more trust in the wiki system than de-jure VfD.
  5. Some developer work required.
  6. Pages begun as patent nonsense, test pages, and vandalism would never be permanently deleted, even though they will often never yeild a real article.
    1. Deleted pages would not be visible to visitors, so this wouldn't matter.
    2. Blatant cases could still be "hard deleted". Besides, harddrive space is cheap (says the guy who doesn't have to pay the bills).
    3. Note that the possible change where long-blanked pages are hard-deleted automatically would address this concern (although it would also limit some of the advantages.)
  7. Opens door to severely disruptive varient of the edit war: the delete/restore war.
    The 3RR etc would still apply. We have edit wars over sentences. At least with a whole page, the issue is much clearer.
  8. Articles that one or two people feel very strongly should exist and that most editors consider useless and deletable but not a great evil (i.e., most vanity pages) would be harder to remove. If one nut keeps on restoring a blanked article, concerned editors with other things to do may give up.
    If someone loves the article more than you hate it...
    Anyone repeatedly reverting blanks of their vanity page is likely to see that page proposed for hard deletion on grounds of edit-war resolution long before their vanity page becomes permanent through their own persistance exceeding that of the editors who are deleting them. —Preceding unsigned comment added by (talk) 21:43, 2 May 2009 (UTC)
  9. Content in blanked pages is not accessible to users, so consensus will still not be possible in disputed cases
    The content is in the article history and can easily be retrieved.
  10. Previous content in blanked pages is inconvenient for editors to access, making the task of keeping an eye on deletions more arduous.
    It is surely trivial for there to be a log or for page blanking to be flagged in RC? No one needs to "keep an eye on deletions", do they?

Do a Test Case

This proposal, to me, seems elegant, and quite possibly the "last, best hope for peace". The main objection to its implementation is that if it blows up, it'll blow up big. So why not do a test case?

I mod for a MediaWiki-based wiki. We are beginning to need a policy for deleting borderline case pages. I think we'll keep hard deleting obvious cases, but I'd prefer to skip a VfD system for the borderline ones, given the heartache it has caused here. We're a special topic wiki, which means we have quite a few pages that are marginally offtopic.

We have not yet come to a decision regarding policy on this issue, but I believe I could convince them of this proposal. But the technical implemntation may be more than our coding staff would wish to bite off.

Assuming that I could get approval, would somebody be willing to implement this proposal at our wiki? Crazyeddie 18:57, 1 Dec 2004 (UTC)

Critique by Cyan

I believe the main advantage of the deletion process as it exists is that people can't get into deletion wars. In my view, edit wars represent a failure of consensus: disputants become emotional and cease communicating in a productive manner. The protection system was designed to force people to take their argument to the Talk page, so that an agreement can be worked out. The present deletion policy effectively makes all deletions protected automatically, so deletion wars can (in theory) never occur.

To move to the proposed wiki-based deletion process would have an effect that I believe the author of this proposal has not considered: protection by sysops would be able to prevent deletion/resurrection of articles (see footnote). And since the protection system does not operate on consensus, but rather on each sysop's own best judgment, this proposal actually increases the power of sysops to affect the content of Wikipedia!

While some may feel that such deletion wars are simply consensus in the making, I personally feel that edit wars actually block the development of Wikipedia. For me, the proposed deletion system doesn't make sense: sysops are already responsible for cooling off edit wars, and the issue of which version of the article is protected can be contentious; are sysops now expected to make interim decisions about whether an article can exist or not?

On a different topic, I feel that the present system is more transparent than the proposed one, for the simple reason that substantive deletions can be seen and discussed by all users in one place. Bureaucratic? Yes. Complicated? Yes. But you can't beat it for transparency.

Footnote: actually, sysops already have the power to use protection to prevent article resurrection. But it is rarely needed under the present system

-- Cyan

Admins can already delete pages at a whim. They would just loose their privileges for it (I'd hope). It would be similar with blanking pages and locking the page: it would be an unacceptable, disciplinable action.
Edit wars and aggressive use of VfD are two sides of the same coin. There's nothing to recommend VfD in that respect. In fact it's worse, as you can enforce your decision by contacting a few people on IM if the objector can only work on the article occasionally, and hasn't time to persuade people to save their article or make a case within the time limit. 12:49, 19 August 2005 (UTC)

Reply by Crazyeddie

As many others have pointed out, the number of controversial deletions is far less than the number of total deletions. The effect of the current VfD systems is to make every deletion a controversial one. Under the proposed system, the few controversial "deletions" could be settled under the systems already in place for dealing with edit wars.

Further, the purpose of the sysop's power to protect a page is to buy time, to allow cooler heads to prevail and a true consensus to develop. Any sysop who used this power inappropriately would highly visible, and would probably be removed. By contrast, the VfD system operates by majority vote (granted, usually a 2/3s majority), not by consensus. Furthermore, the VfD system is said to be controlled by a "shadow tyranny" of regulars, who might not truly represent the will of the Wiki.

Under the proposed system, "deletions" would be visible to the people who actually use the page, and could be "undeleted" by any dissenter. Under the current system, "wrongful" deletions can not be so easily undone.

I would say that the burden of proof lays on the side that wants to delete a page. So in case of a delete war, the sysop should protect the undeleted version, until a consensus can be reached.

Crazyeddie 18:46, 1 Dec 2004 (UTC)

General discussion

I think it's worth a try. We already have edit wars, and we live with them because we can't have wiki without them. This proposal makes things more wiki and less cumbersome, but potentially more contentious. But most deleted pages wouldn't end up in a deletion war. The ones that did would then become protected and go through the VfD process. So, we'd be no worse off than today.

-- Axlrosen 17:51, 14 Oct 2003 (UTC)

So VfD wouldn't disappear, but would be reserved for problem cases. Hmm... I could deal with that, if others strongly support the notion. Transparency (at least my version of it) would still be a problem, though. GrahamN seems to feel that the present VfD system removes power from John Q. User, but I don't see it that way. It's centralized in the sense that everybody who wants to comment can do it at one place, but that's not the same as a centralized power structure. -- Cyan 21:34, 14 Oct 2003 (UTC)

The VfD system as it stands is working, in my opinion. When this whole discussion started on another page here at meta, it seemed as though the objection to VfD was that we shouldn't be deleting things. But the proposed new system, in my opinion, will lead to much higher amounts of "soft deletion". I've seen many pages that I think are bunk and should be blanked: many have been submitted to VfD only for me to find that others like them and would prefer that they be kept. I could easily have blanked these pages under the proposed system, forcing someone else (note--not en:User:Someone else) to come along later and undo my well-intentioned misdeed. Under the current system, I post a little note at a central location which hundreds of people check each week--at any point during the week, a couple of them can object, and then there is no problem. For some reason, GrahamN thinks VfD's system is clunky. I think it's actually a nice model, and far less clunky than the inevitable soft deletion wars (and if we think deletion wars will be rare, I think we're kidding ourselves). Until we see VfD being horribly abused, with a cabal of sysops wreaking havoc (something I've not seen), I don't see any reason to change the system. That's my two cents. -- Jwrosenzweig 22:23, 14 Oct 2003 (UTC)

If a user tried to blank a page, they could be presented with a warning first, i.e. are you really sure you want to do this? Have you read the relevant policy pages? Did you provide a good explanation in the summary and/or talk page? etc. The idea would be to give the user pause, so that they think, well maybe this will be controversial so I should list it on VfD instead.
I guess that's a new proposal for this system. How about this: if something violates an explicit Wikipedia policy (e.g. anything on What Wikipedia is Not), then you can blank it. Otherwise (e.g. if you think it's "unencyclopedic"), then you list it on VfD. I kind of like that idea. Most (though not all) of the controversies on VfD are about whether something is not important enough (or otherwise unencyclopedic) to be in WP. So we'd keep the discussion for controversial topics but punt it for uncontroversial ones, thus streamlining the process.
Axlrosen 14:54, 15 Oct 2003 (UTC)

I strongly support the notion that we should devolve power on deletion and undeletion to everyone, rather than restricting it to sysops. There are lots of deletions that are wholly unremarkable, and Wikipedia would be more efficient (and sysops less burdened) if everyone could do them:

  • deleting a page in order to perform a name change.
  • deleting nasty pages created by vandals
  • deleting test pages created by newbies
  • etc

Currently only sysops can do this, and that gives us sysop burn-out: I'm thinking of en:user:Zoe here, and all the people who spend so much time fixing Michael's created pages, and so forth. If we enabled everyone to help with this kind of problem, then rather than having a few sysops "on the front line", it could be a problem that the whole community can barn raise and solve together. --mrd

I had the occasion to arrive at a plan very similar to this one myself. I am in support of creating a soft-deletion option parallel with the hard-deletion option. -Smack 04:26, 8 Dec 2003 (UTC)

  • I like the idea of the soft-deletion system. It would be sad to see the votes for deletion page go. But would it have to?
  • As someone previously stated (i think), why not just use the current vfd to discuss articles that should be blanked? It's good to have a central place (rather than doing it on individual talk pages).
  • Hopefully the blanked page will have a quick link to see the previous contents (i think it does), which will make it easy for users to see what's been deleted.
  • I don't see a reason for hard deletion to coexist though. A deletion's a deletion.
  1. Hard deletion would still be needed for two main reasons. First, it would provide a way to resolve delete wars. It is also important to distinguish between things we delete because we don't want them here and they don't add productively to the wiki, vs things we must delete because its illegal to keep them (copyright issues and libel, for example). The former would be well served by a simple soft delete, except in cases of controversial deletions, which would still go through VfD. Legal issues, however, can remain a problem even in a user-accessible history page, even if not stated on the front page of an article; for those cases, the difference between a soft delete (history available to all) and a hard delete (history available only to sysops) is legally very significant. (talk) 22:05, 2 May 2009 (UTC) 09:38, 15 Dec 2003 (UTC)

In recent changes, we'd need to be able to display only edits containing DELETED or REVIVED or either, to keep better track. siroχo

Wikipedia is a Wiki. The idea of a Wiki is that anybody can change things, or change them back. Why do so many people who evidently don't believe in that principle spend so much of their time here? GrahamN 01:58, 30 Apr 2005 (UTC)

In response to Cyan's critique - the fact of the matter is that right now, the English Wikipedia system appears to be tilting towards a Request for Deletion system that gives all admins the power to delete pages. So yes, the Pure Wiki Deletion System might be giving admins a little more power, or asking them to use it, by more often having to protect deletions/undeletions of articles - but it's not that much more power, and the possibility that admins might be given much, much more power due to the issues with Votes for Deletion makes it worth it in my mind. Sirmob 02:01, 4 August 2005 (UTC)

Central pillar unnecessary

I'd say the so called "central pillar" of this proposal is actually rather inconsequential. The vast majority of pages which are listed on VFD and deleted either don't contain any links to it or have only links to it which are removed after deletion.

The real central pillar is simply a policy change. We no longer allow admins to delete pages, and we allow and in fact encourage page blanking in certain situations.

The technical details are what makes this paletable: blanked pages are kept out of the search engines, out of search, and out of random page.

Tuning up?

I like the proposal. I still think a few small things should be tuned up:

  • After some period of inactivity (say 1 month) soft deleted pages should become hard deleted - save space
  • Severe copyright violations, etc. should be hard deleted through the existent process
  • Should be a policy that in the case of deleting wars the article should go through the existent process (say after three deletion and reverting by at least two different people on each side).
  • Some articles (e.g. feature articles) should be delete protected. 03:59, 2 August 2005 (UTC) (w:User:Alex Bakharev)

Space is not an issue. Dan100 12:33, 8 August 2005 (UTC)

It's not about space being an issue. It's about information that is dead should become "really" dead because it makes it more tempting to do a vandalize-undelete if the old information is there. Let the dead pages die (those poor soft deleted pages in their persistant vegitative state...) Sirmob 04:37, 27 August 2005 (UTC)

For the third concern: Articles with that kind of dispute would effectively go through the wiki's existing dispute-resolution process, just like any other revert war. I think that's fai; nobody has yet given me an argument as to why disputes over the deletion of an entire article should be treated any differently than disputes over, say, the deletion of a section in that article.

And for the last concern: There's no need to give feature articles and the like special protection from "deletion-vandalism". If someone tries to delete them without consensus, it will be treated just like any other vandalism on a featured article. Aquillion 23:20, 27 August 2005 (UTC)

Central pillar is good

"Links to blank articles will appear the same as links to non-existent articles." Makes sense.

  • {db|reason} tag should still be used - but instead of an alert to an admin, it should work like any other dispute - policy should be take it to the talk page.
  • If the criteria for CSD are met, the policy should allow the ordinary user to blank at once.
  • If CSD criteria are not met, user should be required to wait a period of time (1 week?) after tagging it, and then may blank if there are no objections.
  • In either case, if there is objection to the blank, keep (or restore) the page, and try to reach consensus.
  • Final step in dispute resolution, keep a VFD process, but the vote takes place on the talk page. A VFD page similar to the RFC page can allow pending deletes to be listed with a brief description and link to the talk page.

This should take some of the burden off of admins. But they will still need to be involved in some cases where there are disputes, and may at times need to lock pages, as in other dispute situations. Acerimusdux 04:35, 2 August 2005 (UTC)

Brilliant idea

Why would anyone who thinks they're working in a wiki oppose it? Yes, you might have deletion wars -- especially with pages created by anons -- but it would allow unloved pages to slip away. -- Grace Note 4:45 AEST August 2, 2005

I like this, although I would still keep hard delete. It would be used for existing speedy categories on WP:EN, for example. —Theo (Talk) 17:49, 2 August 2005 (UTC)

This proposal will make it harder to find vandalism

A lot of page blanking vandalism occurs, which can be found fairly easily with en:Special:Shortpages. How would that vandalism be found if this system were implemented? It would make en:Special:Shortpages useless. Angela 18:59, 2 August 2005 (UTC)

Shortpages is a db query on article size less than x. It's trivial to change that to a query on article size greater than 0 and less than x. --Martin
I think we would also need some kind of "Special: Recently blanked pages" or something like that. In fact, that really, really needs to be part of the proposal... Without it, there will not only be no easy way to track page-blanking vandalism, but no easy way to track page-deletion vandalism. This will address the above concern, since people could track the recently-blanked pages list to hunt both kinds of vandals. Aquillion 23:35, 26 August 2005 (UTC)

Policy implications

It makes no sense to implement this before any policies are proposed. If someone keeps blanking/deleting a page, when is that regarded vandalism and when is it regarded as part of the new deletion process? When can someone revert the blanking? Would they have to get consensus before doing so, or after? The suggestion at en:Wikipedia:Requests for deletion that this could be implemented within a week is a very bad one if it implies such a thing would be implemented with no policies in place, and no agreement that blank pages should appear this way. Angela 19:04, 2 August 2005 (UTC)

It would be considered vandalism if there was no edit summary explaining the action and/or if they page was "obviously" not suitable for deletion. Anyone can revert the blanking if she thinks it was inappropriate - just like anyone can revert any change in Wikipedia. - Haukurth (en:wikipedia) 17:08, 5 August 2005 (UTC)

I think that a blanking/deletion is a revert, so 3RR applies. Do it often enough and someone will arbcom you for it. It should be reversible straight away, just like any edit. "Consensus" should be found in exactly the same way it is for edits. Disputes that last longer than X (where X is a number of days to be determined) should be referred to VfD for a binding decision. I think a blanking of useful content without an explanation is vandalism. The explanation needs to be consonant with the policy on deletion. "Consonant" means that the explanation must follow the policy, not that it must necessarily be correct. A blanking log for pages that are not speedy deletion candidates would probably be a good idea but I'm not sure it should actually be compulsory. I think you'd have a lot less fighting than you think. Grace Note 05:24, 10 October 2005 (UTC)

No minimum edit summary

One change is necessary, I think:

"A page cannot be blanked unless the user enters an edit summary of at least twenty characters."

I appreciate the thought behind that rule, but it will only lead to blanked pages with an edit summary of "xxxxxxxxxxxxxxxxxxxxxxx" or "blankkkkkkkkkkkkkkkkkkkk". --DavidConrad 09:32, 3 August 2005 (UTC)

Another Special

We would need something like Special:RecentlyResurrectedPages (shortcut S:RRP) as well. -- ~~~~


I really like this idea, especially considering that so many of the other proposals are so convoluted. Controversial deletions should work just like other disputes, and perhaps VfD can have a new life as a more specific version of RfC.

And maybe this will even get people to pay attention to other kinds of disputes. VfD makes it easy to get the community involved when someone adds an unencyclopedic article, but it's hard to get the community involved when someone adds content that is unencyclopedic to an existing article. By making the processes more similar, we can remove this strange dichotomy. w:User:Rspeer 23:36, 15 August 2005 (UTC)

why does the community need to get involved when someone adds content that is unencyclopedic to an existing article?--Alhutch 07:31, 30 October 2005 (UTC)


I removed this para: Deleted pages may be moved just like live pages, except that there will be no re-direct created. By this means, a live page can be moved on top of a deleted page without losing the deleted page's history.

It seems to require a lot of complicated coding, and is not particuarly necessary; IMO, moving a page on top of a PWD(Pure Wiki Deleted) page should be the same as it is now to move on top of a page with a history; you need an admin to merge the histories. We should address this in the Rebuttals section, but not in the introduction.

and this one: The system will be de-centralised and self-regulating. There will be no "votes for deletion" procedure, just as there is currently no "votes for re-naming" procedure, no "votes for blanking" procedure and no "votes for this particular edit that I want to make here" procedure. There will be no deletion log, just as there is no "re-naming log", no "blanking log", etc.

AFAIK, many of the current supporters of PWD don't want to remove Full Deletion(i.e. hiding the revisions from the general public, conventional deletion), so this paragraph is just wrong. We could have an argument about this on this talk page, if people want to.

and the whole Possible changes section - we should not have "possible" changes listed on the main proposal page. They should be discussed here, and only added if and when they are agreed to. Here's the excised section: Possible changes

  • Administrators' powers to prevent deletion/resurrection of articles will be rescinded.
    • Unfortunately, administrators need the ability to prevent the resurrection of articles, as the WP article on the "popularity of Adolf Hitler" - "a great man, admired by many Jews" - shows. It was created, and created, and created, and finally blocked from being created.
  • Pages that had been blank for a long time (a month? a year?) should be able to be "cleaned up" by the database, if necessary for the good of the database.

and the bits about the edit summary: *A page cannot be blanked unless the user enters an edit summary of at least twenty characters.

  • When a page is blanked, the word "DELETED" is automatically added as a prefix to the edit summary.
  • When a page is revived, the word "REVIVED" is automatically added as a prefix to the edit summary.
IMO, this is redudnant with the log of blanking(and unblanking); we don't use technical measures to require a reason for anything else, and as said above, they wouldn't work even here. The auto edit summary bits are redundant with the log. 07:40, 7 October 2005 (UTC) (actually en:User:JesseW) (Yes, that was me, in case anyone was worried.) (Now signed in at meta) JesseW 07:55, 7 October 2005 (UTC)

After copy

Demotion, not deletion

I think this proposal is passing up the opportunity for a much more useful reform which could give everyone more-or-less want they want, which is a system of demotion, not deletion. "In-between" pages which inclusionists feel belong in a comprehensive encyclopedia, but which deletionists feel don't belong in a general encyclopedia, could be "demoted" with consensus and appear as red (maybe some other color) links by default. They would still be easily accessible to anyone who needed them. Pure wiki deletion wouldn't prevent fighting over this kind of in-between page, because a page which is blanked is inacessible and useless. Actual junk which no-one feels should be kept could then be disposed of easily, with consensus, by AFD or any other method. Kappa 00:58, 10 October 2005 (UTC)

The beauty of the PWDS is that it is so simple, does almost nothing that can not already be done, and requires very minimal changes to the source code. Your idea lacks those features, so it seems to me to be just another deletion reform proposal like the dozens that are already proposed that won't get much support. RSpeer 02:28, 10 October 2005 (UTC)
  • Technically the only difference is using some kind of tag to indicated a "deleted" page, instead of it being blanked. I think the software can check for a tag almost as easily as it can check for an empty page. Kappa 02:59, 10 October 2005 (UTC)
So long as you can do a Special:Blank pages, it's no problem to find out what's been blanked. Deletionists will only war over schools until they're sick of being reverted. At first, they'll delight in expunging articles they don't like but they'll grow tired of having to defend each deletion ex post facto. Grace Note 05:26, 10 October 2005 (UTC)
  • Empirically people never get tired of trying to get schools deleted for being "non-notable", there would just be continuous revert wars instead of continuous useless Afds. Kappa 12:55, 10 October 2005 (UTC)
How does this work if an editor decides to do a merge and then blank the page? The information is now in another article, so if someone decides to undelete we have the information in two places. Of course another editor could revert both changes. I think if someone took the time to do the merge, reverting the deleted article should be a little harder to do. Vegaswikian 07:47, 10 October 2005 (UTC)
Merges already operate on a pure wiki system, so I don't think it would make any difference. Kappa 12:55, 10 October 2005 (UTC)
Clarification - because the usual thing to do after a merge is to leave a redirect (NOT a blank), correct? Sirmob 05:47, 17 October 2005 (UTC)
Yes, I don't think it's an issue. Merges can be undone anyway as things stand, but they rarely are. What Vegaswikian doesn't understand is that editors such as Kappa and I don't vote to include schools just to piss him off but because we think they should be part of the encyclopaedia. We're not waging a war to have lots of crap school articles. If you merged the small articles quietly and without fuss, I don't imagine you'd meet any opposition. It's the offensive and arrogant repeat listing of them on AfD that causes the problems. Grace Note 01:20, 18 October 2005 (UTC)
Oh god. AfD can keep the school debate for all I care. (And if AfD ends up being used only for controversial deletions, it will, because every school deletion is controversial to a ridiculous extreme.) Just don't bring it here. RSpeer 05:26, 18 October 2005 (UTC)
It's something that's a concern. You don't own this page and you can't direct what people may or may not be concerned about! Grace Note 07:14, 18 October 2005 (UTC)
Calm down. It's off-topic, that's all. RSpeer 01:05, 19 October 2005 (UTC)
I'm perfectly calm. So far as I can see, the topic is whatever we're talking about, which was deletion. -- Grace Note.


Anyone interested in trying to get more people to support (or at least comment on) this idea? What would be good ways of doing so? To me, the argument for pure wiki deletion is obvious: anyone can edit Wikipedia, and removal of undesirable content is a type of edit, so anyone should be able to do it. Making deletion/undeletion a special case instead of just another edit is a clear case of instruction creep. The "burden of proof" should be on those who want to do it any other way.

Has it occurred to anyone else that we could (sort of) start using pure wiki deletion without waiting for software changes? The software doesn't support it directly, but we could attempt to approximate this kind of system. Friday (talk) 09:04, 30 October 2005 (UTC)

WP:XD#XD4 is a workaround that is very similar to the PWDS - but it works only for logged-in users. -- grm_wnr Esc 21:09, 30 October 2005 (UTC)

I think that a good way to promote PWDS for now is to experimental delete anything you would once have AfD'd, and add something like "See Wikipedia:Pure wiki deletion system for how this process can become smoother." to the edit comment. What I'd like to see is an XD template (XD6? XDPW?) that actually includes a link to the PWDS. RSpeer 02:19, 31 October 2005 (UTC)

I've added such a comment to the the {{XD4}} template. —Jwanders 20:38, 4 November 2005 (UTC)


To get the process started, I filed a MediaZilla request for enhancement at MediaZilla:3843. I haven't done this before, so I hope I did it right. RSpeer 02:39, 31 October 2005 (UTC)

Looks good to me. I went ahead and voted on it, which I assume makes a difference. Those interested should visit and vote on the bug as well to increase it's priority. here 05:27, 31 October 2005 (UTC)
Good idea. I went and voted on it too. Other people who want to see the PWDS implemented should do the same. RSpeer 15:26, 31 October 2005 (UTC)</nowiki>

Moving Forward

It seems we have very strong core support for this project but are lacking many proposals of what to do next. I figure we should consider:

  • Making the XD4 blanking system above an official project strategy (i.e. moving it to the project page).
  • Adding something to on deletion reform to indicate the level of support, with the view of moving this towards official Wikipedia policy.

As I'm not sure how either of these would be received by the rest of the deletion reform community and am loathe to start yet another rambling discussion, I figured I'd ask for other suggestions and find a consensus here before proceeding. Thoughts? —Jwanders 07:58, 5 November 2005 (UTC)

I went ahead and consolidated the XD options into one to be refined. Hopefully well received, and a step forward to using this more widely. See WP:XD. here 18:43, 5 November 2005 (UTC)
Sounds good to me. Friday (talk) 18:45, 5 November 2005 (UTC)

Simple vandalism versus content dispute

There doesn't seem to be a good answer to this in the proposal itself, so maybe someone can clear this up.

Under the present system, blanking a page is almost always vandalism, with everything that entails—we can use rollback on it, block for it, etc. This is true regardless of what the edit summary may contain.

Under this new system, blanking a page is equivalent to an AFD nomination. It may be removed if added in obvious bad faith, but is not, generally speaking, blockable. How do we distinguish between a good-faith attempt at deleting an article by blanking it, and simple vandalism that also entails blanking it.

In other words: if a user were to blank a hundred articles under the current system, they would be blocked in short order; under PWDS, what recourse, beyond the disruption clause, is there for dealing with them? Kirill Lokshin 23:53, 13 November 2005 (UTC)

Well, the edits would be easily reverted, so undoing it is no more difficult than dealing with any present forms of vandalism. But, as far as blocking a problem editor, a string of bad-faith deletions would be pretty obviously a type of vandalism. The only difference I can think of is that page-blanking would not be automatically considered vandalism as it is now. Yes, making the distinction between good-faith blanking for deletion and bad-faith blanking for vandalism would require human judgement. But, pretty much anything we do here requires human judgement, so I don't see that this is a problem. Friday (talk) 02:33, 14 November 2005 (UTC)
How would they be easy to revert, though? Granted, someone blanking United States is probably a vandal—but what about someone blanking a (potentially) deleteable article? If someone blanks "Exampleville High School" or "Random Pokemon #28781", is the deletion automatically considered good-faith or not?
This problem seems to be aggravated by the idea below—that un-blanking an article would cause it to become a "contested deletion" and would require the un-blanker to complete an AFD nomination for it. This would make reverting blanking vandalism much more time-consuming than causing it. Kirill Lokshin 14:24, 14 November 2005 (UTC)
Well.. As for good/bad faith, I wouldn't think in those terms. You don't HAVE to speculate about the motive of the editor. You can simply decide whether you agree with the edit or not. If someone blanks a deletable article and you agree, you do nothing, except possibly note your support for the deletion on the talk page. If you disagree, you would start a discussion on the talk page about why you think it shouldn't be deleted. And, of course, you could revert the deletion, at your discretion. I do not agree at all with the idea that an article should be sent to Afd the minute there's disagreement, because, as you point out, this would be completely silly in many cases.
Ideally, people would do what they should have been doing all along: use the talk page to resolve disagreements. If you see a page you think shouldn't exist, but you think there might be disagreement, explain on the talk page what you think should be done with the article. This is how people (ideally) decide on things like merges/redirects and it works there. Why should deletion be different? Friday (talk) 14:58, 14 November 2005 (UTC)
My apologies, I wasn't being quite clear. When talking about good/bad faith, I was referring not to the blanker's motivation for wanting the article deleted, but to whether the blanking was an attempt at deletion in the first place. In other words, there should be a distinction between "I am blanking this article, and by doing so I am invoking the PWDS deletion procedure on it" and "I am blanking this article because I'm bored", and it should not be made according to how inclusionist/deletionist the person examining the edit is.
I would be a lot more comfortable with the proposal if language to the effect of "blankings by anons are not considered acceptable" and "blankings without proper edit summaries are not considered acceptable" were added; this would, at least, limit the use of PWDS as an excuse for otherwise unrelated blankings. Kirill Lokshin 16:29, 14 November 2005 (UTC)
I agree that a proper edit summary is highly important. Maybe something could go in there about "don't expect people to take your deletion seriously if there's not a reasonable edit summary." Friday (talk) 16:42, 14 November 2005 (UTC)

Clarify place of PWDS

Most of our recent opposition seems to have a different understanding of where PWDS fits into the overall deletion system. As I understood it, we weren't suggesting that either AfD or Speedy delete be abolished, but that PWDS be added as a third option. Then patent non-sense and copyvios go to speedy delete, other deletes use PWDS unless they're contested, in which case the page is unblanked and sent through the current AfD process. This saves the overhead of AFD for articles which would be annanomously deleted while keeping the centralised discussion process for borderline cases.

If no one objects, I'll clarify the proposal page to reflect this understanding. —JwandersTalk 06:52, 14 November 2005 (UTC)

So how much "contest" does it take to go to AFD and place the PWDS out of line? And, given that links to the blanked article will simply "go red", what exactly will you encounter when you follow the red link? Will it be clear that there previously was an article here? Otherwise I can easily see a scenario where someone deletes a valid but obscure article, and someone quite innocently starts a stub from scratch, unaware that there was quite recently an article on the topic. -- Jmabel | Talk 06:59, 14 November 2005 (UTC)
Second question first: Newcomers to a "blanked" article would see exactly what they do now: the info-box regarding new article creation with "View X deleted edits?" above it. Do you think that would be enough to prevent most people from unknowingly recreating an article?
Regarding the amount of "contest": Any user would be able to revert the article to pre-blanking, add the AfD template and list it under AfD. Once this has been done, deletion by blanking would no longer be acceptable. This errs on the side of sending articles to AfD too often rather than not often enough, which I believe is where we want to be.—JwandersTalk 07:24, 14 November 2005 (UTC)
Jwanders, yes, please do clarify the proposal along those lines. I had thought it was mentioned, but it does seem clearly insufficient, so adding more to make it clearer is important. Re: Jmabel's comments: Jwanders answered this pretty well, but I have no problem with making the link larger if needed. The "contested" level absolutely should be as Jwanders said - PWDS should only be for things that no-one but the proposer cares enough to notice. OTOH, I'm leaning towards the view that things are working OK as is, and the PWDS might not be all that needed... JesseW, the juggling janitor 09:59, 14 November 2005 (UTC)

Maybe it's just me, but in the case of a contested deletion, I'd rather see people discuss it on the talk page than on Afd. Maybe a category could be made for ongoing deletion discussions, to help those people who like to find them. Also, I'd hate for one revert of a deletion to make the article go into a special state where you're not supposed to pure delete it again. A blatant vanity article could be unblanked by its newbie creator who's unfamiliar with how things work here, yet it's obviously unneccessary to send such a thing thru Afd or other long discussion. Friday (talk) 14:40, 14 November 2005 (UTC)

After any revert, an attempt at discussion should be made. That does not necessarily mean AfD, but certainly a stab at the talk page and at least 24 hours of wait time. Obvious cases, likely determined by current speedy standards, might be again simply deleted with an additional note on the talk page. here 05:11, 15 November 2005 (UTC)

Thoughts on benefits of PWDS

Benefits of PWDS: For deletionists:

  • More stuff would be deleted
  • Because:
    • More people able to do it (rather than just nominate it to be done)
    • Easier (one step vs three steps)
    • Less work overall (one step vs 10 or so steps for AfD)
  • AfD would be smaller, and therefore easier to view

For inclusionists:

  • Much less stuff would be hidden from public view
  • Because
    • PWDS deleted stuff would still be in the history
  • AfD would be smaller, and therefore easier to view
  • A large amount of deletion could be canceled easily - just undo the blanking, and leave it alone - or list it on AfD to prevent it from going through PWDS again.

(most of these may also be regarded as disadvantages ;-) ) Comments, arguments, spelling corrections? (sorry, it's late at night) JesseW, the juggling janitor 10:22, 14 November 2005 (UTC)

On edit wars

The main complaint, and particularly the main one that makes sense, is that this will increase the number of edit wars. Possibly dozens per day! But here are two responses:

  • Dozens of edit wars per day would be a drop in the bucket for Wikipedia. AfD is not the entirety of Wikipedia discussion.
  • I've experimental deleted numerous articles over the last few weeks. Only one was ever reverted. I sent it to AfD where it was dealt with without conflict. I think that it removes conflict to not draw undue attention to every stupid page.

rspeer 14:17, 14 November 2005 (UTC)

Response to Greg

Responses to Greg's opposition:

  • A perfectly easy way (a Special: page) to see all recent deletions and recreations is a part of the proposal. That's how you would know.
    Right, but I mean on the non technical side. Right now you can be almost sure that an article being blanked is vandalism, but with this, it takes alot more time. Yes, I know, we shouldn't all be computers, etc. But it puts alot more strain on RC Patrol.
    You can be sure that an article being blanked without a good edit comment is vandalism, though. rspeer 21:33, 15 November 2005 (UTC)
  • I have no idea how a decentralized process like PWDS could be a "cabal", so what do you mean by that word?
    Maybe that wasn't the best choice of words, what I mean is the proccess for deletion is less transparent, right now you can check AfD and KNOW what articles are up for deletion, with PWDS there would be no way to know. It isn't that big of a negative, but I like the way I can see what is up for deletion right now.
    To be blunt about it, I'm sure you do, but I don't like the way you can see what is up for deletion and all the discussion. This is what gives AfD regulars their disproportionate influence. I realize this comment won't help me sway you or anything.
  • If an article that is "totally worthy for inclusion in Wikipedia" is deleted, and nobody cares enough to bring it back, is it such a loss? How does it affect you if an article you never paid any attention to gets deleted by people who edit in that topic area? Why do you know better than them? I believe that people on talk pages make thousands of article content decisions that you disagree with every day.
    I conceed this, you make a good point.
  • What kind of "serious problems with unilateral deletion" could arise? If it can be reverted, it's not any more serious than any other disputed edit on Wikipedia.
    What if someone doesn't notice it? Like I brought up below, it just creates more dramatic edit/deletion wars.
    The article could have just as easily have been vandalized and stay that way, if there isn't anyone watching it to notice. At least with deletion you'd get an entry in the Deletion Log for people to notice. Not that I think that part of the proposal is strictly necessary - after all, we get by just fine without a Paragraph Deletion Log now - but it seems necessary to reassure people.
  • PWDS gives users no additional rights, so I find a comparison to giving everyone developer rights to be completely absurd.
    It gives them the right to delete pages
    Okay, but not in the sense of "hard deletion" we have now, because you're not removing information from the database. It's much more similar to the existing ability of a user to blank or WP:XD a page. rspeer 06:25, 15 November 2005 (UTC)
  • The reasonable objection, I'd say, would be the creation of some more edit wars; as I've said above, though, this would be a drop in the bucket. rspeer 14:33, 14 November 2005 (UTC)
    I don't think it would be a drop in the bucket though, we currently have edit wars going to ArbCom, deletion wars would be a problem. I think the current AfD system has some minor flaws, but I disagree with this change. -Greg Asche (talk) 22:00, 14 November 2005 (UTC)
    Does the revision to the proposal (send all PWDS conflicts to AfD) address this concern? rspeer 21:33, 15 November 2005 (UTC)


First off, I'm not sure who thought of the name "pure wiki" but it is obviously a peacock term because it isn't necessarily more "pure" or more "wiki" than anything else we use or have proposed. I'd suggest renaming it to what it actually does, e.g. "deletion by page blanking".

Second. The PWDS cannot deal very well with 1) copyvios, 2) linkspam, 3) vanity, 4) attack pages and 5) anything supported by socks, and it has been established several times in the past that allowing anybody to see a deleted page goes against the very point of deletion. Acrimony of AFD notwithstanding, deletion is sometimes necessary, and it is with good reason that it cannot be reverted by just anyone.

And third, PWDS would make it easier to delete pages on relatively obscure subjects, since unless anybody is watchlisting them, the deletion will simply go unnoticed. This is a bad thing. Presently, such either show up on AFD, or on the Logs.

With speedy delete it just disappears if the deleting party doesn't put a note on the page.--Gbleem 03:43, 10 December 2005 (UTC)

Radiant_>|< 17:35, 14 November 2005 (UTC)

Regarding copyvios, linkspam, attack pages, and perhaps even vanity: Your point is that those need to be permanently deleted and hidden from general view in the history, right? But anyone, today, can replace the content of any article with copyvios, linkspam or attack pages. When the article is reverted, the linkspam or attack pages are still in the history. What's the difference?
Similarly, for the relatively obscure pages that nobody is watchlisting: right now, today, anyone can replace those pages with a 5000-word essay on gorilla genitalia, and unless anybody is watchlisting them, such changes will simply go unnoticed. —Bunchofgrapes (talk) 17:46, 14 November 2005 (UTC)
  • Yes, they can, and that's bad, and PWDS will encourage this behavior by basically telling people it's okay to blank pages. Radiant_>|< 17:51, 14 November 2005 (UTC)
No more so than WP:BOLD encourages vandalism. Also note that the proposal includes a special page showing all recently-blanked pages. —Bunchofgrapes (talk) 17:55, 14 November 2005 (UTC)
(via edit conflict) I don't think anyone is suggesting that we wouldn't also have a "hard delete" option as we do now, even if most normal cases would be handled by PWDS. My opinion is that the new problems created by PWDS are not in fact new- they're problems we already have, and already deal with. I don't see how PWDS would encourage people to put junk into articles. The people putting blatant junk into articles would not likely be aware of PWDS, just as they're probably not aware of our current deletion methods right now. Friday (talk) 17:57, 14 November 2005 (UTC)
  • Noone will be telling anyone it's ok to blank any page you like.
  • Obscure pages are already vulnerable to vandalism. This places them no more at risk.
  • PWDS does not encourage bad behavior in any way. Blanking is already available to idiots who feel like being destructive.
  • In response to inevitable, this will clutter the rc feed: No, it won't. Proper deletions using PWDS should include references to PWDS in their edit summaries. n00b vandals won't know to reference this, and experienced editors will learn or face the wrath of the community if they cuase trouble blanking.
here 05:23, 15 November 2005 (UTC)

a few refining points

The new system looks great, but a few modifications should be made in order for it not to run astray. Those I see are:

  • The talk page should definitely be available, not only the history, because the talk page is likely to contain the debate and vote that led to the deletion, if any.
  • Only logged-in users should be allowed to blank pages, because this would make impeding the vandalism much easier. And people who haven't been long enough on wikipedia to bother logging in have no business deleting pages anyway.
    • This seems to complicate the proposal somewhat. What should happen when an anon blanks the page? Does the software not let them save it? What if they type a single character and then save it? If this really does accomplish something, why don't we prevent anons from blanking now? I think that anons should simply be held to the standards of edit comments that everyone else is. rspeer 06:29, 15 November 2005 (UTC)
      • well, I think at least it would prevent vandals from making red links appear all over the place, which they can't do for the moment. We don't prevent it for the moment because it doesn't have such dramatic effects. I can imagine someone going through all the main page links to blank the pages (I'm sure the bot to do that is quite easily made and incredibly fast), making the main page all red... We reserve moving pages to logged-in users, so why not deleting? It's not like we'd be removing a privilege to anyone, we'd just not extend them to everyone (from admins-only for the moment) Jules LT 17:10, 15 November 2005 (UTC)
        • Why is red links bad? at least you can see there is a problem and fix it, rather than leave it for someone trying to use the encyclopedia to get annoyed because a link just gives him a blank page. Sandpiper 23:48, 16 November 2005 (UTC)
          • Maybe they should be a different shade of red (to distinguish them from pages that simply never existed), perhaps a sort of reddish-purple. I have to agree with Jules LT, though. If a vandal starts deleting pages, it is easier to talk to/block a user than a dynamic IP address / someone using proxies. Armedblowfish 17:35, 18 April 2006 (UTC)
  • We should remember to ask for blanked pages not to be listed anymore on Special:Short pages, because otherwise it's going to be flooded with them. Jules LT 21:28, 14 November 2005 (UTC)

Start using it!

Lets start using it regularly and see what happens. Category:XD is growing slowly.

newpages at least 20,000 old (roughly a week at the moment) is a nice spot I feel comforatable finding pages needing initial filtering. Wikipedia:WikiProject_Stub_sorting/Stub_types is useful to help sort pages not warrenting deletion. here 09:43, 15 November 2005 (UTC)

  • No. Until there is consensus for this method (and there presently isn't, per significant objections), blanking pages constitutes vandalism. Radiant_>|< 10:33, 15 November 2005 (UTC)
  • People are using WP:XD right now. That's a poor man's approximation of PWDS. It's not vandalism, it uses a template explaining what's going on. Outright blanking would in fact be a very bad idea until the software changes are in place to make it function as described. Friday (talk) 14:29, 15 November 2005 (UTC)

For the record the support/oppose count is 23/9, with the first 3 in opposition more of the opinion that no reform is necessary, not outright opposition to the proposal. Feel free to stop by Category:XD to see what has been done. As usual, much more stub sorting than deletion, and all controversial cases are sent to AfD. I have had not a single revert or comment on the deletions I've made using this process.

I realize that I'm biased in favor, but I do not see any standing objections to this proposal. Perhaps you could pull your specific remaining objections from Wikipedia:Pure_wiki_deletion_system#Common_objections_and_responses so that we can try to address them? In using it on actual pages, I have definitely found a few procedural improvements to the process which may even effect code changes and will help refine the process. Thanks for the thoughts. here 17:48, 15 November 2005 (UTC)

  • The support/oppose count is not particularly valid. The 'oppose' section was not present for the first couple of weeks, this proposal was not widely advertised, and I see there are almost no admins (or indeed, users I've heard of) among support. The last may be subjective but I don't think this has had input from most of the community. Radiant_>|< 18:07, 15 November 2005 (UTC)
    • Radiant, your contributions seem to be primarily to administrative things and deletion pages. Where would you expect to have heard of users who aren't admins and don't participate in the current deletion system? I can't help feeling like you're dismissing the validity of opinions held by users outside of your circle. rspeer 19:43, 15 November 2005 (UTC)
      • I'm sorry if it sounded that way. My point was, rather, that since I know a large group of editors, if I find no editors that I know on a certain page, then I doubt that said page has had much publicity. I'm not saying that the proposal is bad, I'm saying that the claim that it has consensus is premature. Radiant_>|< 23:44, 15 November 2005 (UTC)

Perhaps an external request for comment is again warrented, where would be most appropriate? Otherwise, can you help us focus on remaining issues in the #Common_objections_and_responses section of the project page? here 18:19, 15 November 2005 (UTC)

  • Yes, let me think on it some more. I'll be back tomorrow. Radiant_>|< 23:44, 15 November 2005 (UTC)

Change to the proposal

Because it seems to be favored by the proposal's promoters, and it addresses a large concern of the opposition, I've edited the proposal to more explicitly say that PWDS disputes should be sent to AfD. This acts as a "circuit breaker" for PWDS edit wars. I hope this addresses a lot of people's concerns. rspeer 16:40, 15 November 2005 (UTC)

Don't get me wrong, I'd love to see this go into effect, but I am just convinced that, without propper safeguards, this has great potential to be abused. I hadn't noticed your reply to my comment on the MediaZilla feature request page, but I feel that this has the potential to be unnoticed. For articles that might not be watched, but have been vouched for their notability should be disallowed from deletion. For less notable articles, or articles that aren't needed as much, I think it would be okay to have them be deleted. -Mysekurity 21:17, 15 November 2005 (UTC)
If we have a good article that isn't watched by anyone, it's not going to stay a good article for long due to vandalism. I don't think there are very many of those in reality. But in case there are, there's the deletion log. Preventing blanking in certain cases seems to me like an unnecessary complication on the proposal that will almost never come into effect. rspeer 21:30, 15 November 2005 (UTC)
What I still don't understand from the critics is how implementing this proposal adds anything to the vandals toolbox? Blanking is already easy to do anonymously! Furthermore, even if a PWDS deleted page is out of line, eventually someone who cares will notice and simply revert the page, no real harm done. (See Request for Rollback for related discussion on how and why every user should have rollback ability.) here 22:48, 15 November 2005 (UTC)
I think the difference being pointed out is this: blanking currently only affects a single page. Blanking under PWDS would affect every page with a link to the blanked page, because those links would turn red. Big deal, not a big deal, I'm not sure— but it is a difference. —Bunchofgrapes (talk) 23:21, 15 November 2005 (UTC)
A similar, existing thing has proven quite useful in fighting vandalism. If a link on the main page (like from the featured article) suddenly turns maroon, it's probably been vandalized; a featured article probably wouldn't link to a stub in its first paragraph. So I see deleted links turning red as a good thing: more information. rspeer
That's not on by default, right? So your typical reader (including every non-signed-in individual) doesn't see the maroon links at all (though a lot of editors do, and it's a useful tool for them). One could argue it's a good thing the typical reader doesn't see the bad links; they would see the red links created by blankings under PWDS, though. I do not think this is a big deal, but I want to make sure we think about it with precision. —Bunchofgrapes (talk) 23:51, 15 November 2005 (UTC)
If you are thinking in terms of an encyclopedia, then a red link is a flaw. Why have a link to a page with no content? Arguably, an apparently valid link is even more of a flaw, since if someone uses it to look something up, they will waste their time changing pages, just to see nothing. So although a red link highlights a fault in the encyclopedia, at least this invites someone to do something about it. Either add content, or remove the link. I don't see any way that advertising the fact that a page is blank is a bad thing. The problem I see is the reverse case, a vandal persistently recreating a worthless page. Now, the question would be how serious is this? Sandpiper 23:37, 16 November 2005 (UTC)

Posted on Village Pump

I think we're ready to publicize this proposal more, so I put it on the Village Pump. rspeer 01:52, 16 November 2005 (UTC)

I like, but

I blank a page and think it's gone (say I'm not an admin or it's not SDable). Then it's remade. The article won't appear on newpages (since it's not new) and on RC it will only appear as an edit and I didn't put it on my watchlist. How do I know that it was remade? When someone creates an article on a blank page it should appear on newpages (except in the case of senseless blanking where it is reverted). Both cases are bad. Either things aren't caught in new pages or new pages is flooded with simple vandalism reversion. Broken S 03:00, 16 November 2005 (UTC)

Just to correct a misunderstanding - the remade page would appear on Newpages(quoting from the proposal: "unblanking of a page would be logged"); the deletion log is already "flooded with simple vandalism"; this proposal would not change that at all, either way. JesseW, the juggling janitor 22:58, 16 November 2005 (UTC)
Having remade pages appear on Newpages seems like the right choice to me. It's consistent with the idea that blanked pages act deleted. Newpages isn't going to be "flooded" any more than it is now - we get 3,000 or so new pages every day, and not nearly as many instances of blanking vandalism. rspeer 23:32, 16 November 2005 (UTC)
yes, but the amount of blanking vandalism is prone to change if the policy is adopted. Imagine a willy like vandal blanking tons of pages. Broken S 04:07, 21 November 2005 (UTC)
Willy was an issue because it was harder to revert than it was to do. This is not the case with this proposal. Other than the lack of reversion problems, it was just the same as any other vandal bot(or spam bot) - block, then revert the damage. Why would this be any different? JesseW, the juggling janitor 22:52, 21 November 2005 (UTC)
Yes, but a mass blanking vandal would flood the new articles page beacue all the reversions would be counted as the creation of a new article. This would/could make new page patrol more difficult. This is more of an issue than say flooding recent changes or the move log because there are far fewer pages made each day than edits. Perhaps rollbacks should be excluded from the page. Broken S 03:16, 22 November 2005 (UTC)
To me, that sounds like a solution in search of a problem. I really don't think this will be such a frequent occurrence that it requires another special case in the MediaWiki code. rspeer 06:21, 22 November 2005 (UTC)
I agree that Special:Newpages does not currently contain the reversions that Special:Log/moves and Special:Recentchanges contain, and that with this proposal, it would. However, it would be a trivial job to write a user script to hide rollbacks from the newpage log, if and when the problem came up. It certainly shouldn't be a requirement of the proposal that this be done. JesseW, the juggling janitor 07:38, 22 November 2005 (UTC)

XD4 simulates Pure Wiki Deletion

Note that experimental deletion method XD4 simulates PWD using a clever trick in your mediawiki user settings. This allows you to try it out right now!

If folks have other ideas or improvements, please mess around with XD4 and modify it as nescesary, or make a new experimental deletion method at Wikipedia:Experimental Deletion

I noticed Here had done some consolidation, but his changes sort of missed the entire 12 byte stub limit trick entirely. (That and the idea is to host other experiments too, PWD is not the only way to do deletion eh? ;-) )

I'm back from wikibreak, so probably I should do more maintenance on XD :-)

Kim Bruning 08:09, 16 November 2005 (UTC)

Any solution needs to include average users. Average users do not edit their red link threshold. Glad to see you back ;) here 08:21, 16 November 2005 (UTC)
DUH! This is a simulation that can be tried out by advanced users, as all XD experiments are. It is experimental. It is NOT the actual PWD, which will require additional PHP code to be implemented properly (whereby scalability is important). Kim Bruning 08:27, 16 November 2005 (UTC)

Uncontested deletions vs. Pure wiki deletions

see discussion Wikipedia_talk:Deletion_reform/Proposals/Uncontested_deletions#Uncontested_deletions_vs._Pure_wiki_deletions.

there... here 08:19, 16 November 2005 (UTC)

Let's see... these are some problems with Deletion By Page Blanking (DBPB):

  1. Do not assume that the devs will implement something. In fact, since they're generally busy with performance and protection, assume they will not. A proposal that requires code changes to work properly is therefore not a very good idea.
  2. DBPB feels like sweeping dust under the carpet, rather than throwing it out. Okay, that may not qualify as a good reason, but I'm sure some people feel that way.
  3. Some people want a centralized list of deletions so they can look at them and detect inappropriate ones - a list like AFD has been called for at UD. Otherwise, anything that isn't heavily watchlisted can be deleted without anybody noticing.
  4. DBPB makes is more difficult to see what a deletion actually was about. For AFD/CSD tagging, you can go to the page and see its text (until it's actually deleted, of course). For DBPB, you'll have to go into the history, which rules out casual editors (you'd be surprised as to how many editors don't know how to use the history)
  5. Anything supported by sockpuppets cannot be deleted by DBPB.


This rename is a bad thing, IMO. I think it should go back the way it was. Friday 17:22, 16 November 2005 (UTC)

I agree. The bugzilla entry and the meta page meta:Pure wiki deletion system (proposal) both use the Pure wiki deletion name. Setup a redirect if you like. here 19:15, 16 November 2005 (UTC)
Done.. though I do like the other name. If you care enough, you could propose a name change on meta? here 22:21, 16 November 2005 (UTC)
  • I don't care overly much about it, but the PWDS name is obviously a peacock term (as it is on Meta). Policy proposals should describe what the intended policy does, not refer to it in POV terms. There is nothing about this proposal that is more "pure" or more "wiki" than most other deletion systems. I suppose people could drag WP:RM into it if they want. Radiant_>|< 23:54, 16 November 2005 (UTC)
    • I don't think there's any rule against POV in proposals; in fact, POV seems to be an inherent part of making a proposal. If you didn't want to convince people it was better than the other options, you wouldn't propose it. For examples of POV in names of project pages, note that there has been a deletion-reform proposal (now defunct) called "the butt-simple option" which wasn't even particularly simple, and there's been a WikiProject called "Wikipedians for Encyclopedic Merit". I don't think those things should be renamed; I think it's far more important to have a consistent label for the proposal than to avoid POV in the name. rspeer 00:18, 17 November 2005 (UTC)
      • Are you then familiar with the large amount of controversy associated with the WPFEM project, which was in part due to its euphemistical name? And that WPFEM was already renamed? Using a misleading name will not cause people to take your proposal more seriously. And I'm afraid you misunderstand what POV means - a proposal should be written objectively so that people can evaluate for themselves if it's an improvement. Radiant_>|< 00:27, 17 November 2005 (UTC)
        • I have no problems with renaming the proposal, but I think consistency is imporant. A large number of people are familiar with the specifics behind this name. Your change was to Deletion by Page Blanking, any other name change ideas? What do the call it on nl:, where I hear they actually use a system like this. If we can agree on a name to change to, I'd support changing all of the pages (meta, bugzilla, here, others?). here 00:45, 17 November 2005 (UTC)
          • BTW, they don't seem to use it on nl:, at least I found a page that suggested they use a similar system to the current one here... JesseW, the juggling janitor 01:42, 17 November 2005 (UTC)
    • If "Pure Wiki" is pro-this-system POV, "Deletion by Page Blanking" is anti-this-system POV. Of course, that's easy for me to say since I like both the system and the "pure wiki" name for it, and the "pureness" and "wikiness" of it are just a matter of opinion.  :-) To me, this system simpler and wiki-er than the bizarre rituals we call AFD/DRV (or whatever they're called this week), but YMMV. However, before changing this away from the name it's been known by, I'd like to hear suggestions and get consensus first, as suggested above. Friday (talk) 01:24, 17 November 2005 (UTC)
  • I have no particular objection to renaming, but I thought I ought to at least mention why the proposal can be considered "pure wiki". Wiki is generally understood to mean "a group of Web pages that ... permits others (often completely unrestricted) to edit the content."(quoting from wiki); the gist of this proposal is that reverting new pages is not different than any other edit - i.e. that deletion be done in the same way as other edits, i.e. in the way of a Wiki. So, wiki deletion system. Pure(which is something of a peacock word), is just to distinguish "a system for deleting pages which treats deletion as just another edit" from "any system used by a wiki to delete pages". That's a possible reason why the term is accurate, however, except for the consistency issue, I have no objection to renaming the proposal. In any case, we can now refer to it as "WP:DPB" without trouble, as the redirects have been created. JesseW, the juggling janitor 01:36, 17 November 2005 (UTC)
    • Right. It would be a problem if the name were misleading as Radiant suggests it is, but I just don't see it. The connotation of the name is simply "we think our deletion system would be a good thing," which is obvious, and "because it's more Wiki than what we have," which is a defining feature of the motivation for the proposal. And sorry for picking a horrid example (WPFEM). rspeer 02:01, 17 November 2005 (UTC)
  • For the record, the Dutch wiki uses a system not unlike Countdown Deletion. Let's just rename it to Deletion System Without An Interesting Name. Radiant_>|< 12:30, 17 November 2005 (UTC)

A response to Drangonflights objection?

I'd like to open a discussion among proponents of PWDS regarding Dragonflights objection, i.e. as I understand it, that the "real problem" of the current system is the polorizing effect, and that PWDS would just make this worse. I'd like to get a nice rebuttal (assuming there is one) for this, but I'm not sure what to say. Thoughts, suggestions, anyone? JesseW, the juggling janitor 06:38, 24 November 2005 (UTC)

That's easy. The reason why the current system is so polarizing is because it feels final; it's very hard to reverse. The deleted page is made inaccessible to normal users, and everything contributed to it is gone. Look at the comments some of the most ardent inclusionists make; what burns them up most about deletion is the way they feel it practically incinerates an article, making it inaccessable and virtually closing off all future discussion once the voting period has ended. (Even reviewing the deleted material later requires administrator intervention, something that many of them have been focusing on recently.) AfD forces debate to occur very quickly, in one brief vote; it strongly discourages discussion, and encourages one-line comments with no further elaboration or consideration. Worst of all, AfD rewards factionalism; since people can easily drop in and make a one-line vote, and since the votes made during a brief AfD listing are generally the sole deciding factor in an article's future, people who want to influence deletions are encouraged to form organizations and hit lists to accomplish their goals. Indeed, most of the nastiest inclusionist/exclusionist divides have their roots in such vote-stacking organizations.
By comparison, vote stacking would be ineffectual and silly-looking under the Pure Wiki Deletion System, having little practical effect, while forming factions generally be far less effective. Fortunately, since deletion wouldn't be settled in forced, ugly five day rushes of madcap voting, those factions would be unnecessary anyway. I think it is clear, especially from that last point, that the Pure Wiki Deletion System would greatly reduce the polarization and acrimony of deletion debates. --Aquillion 08:20, 1 March 2006 (UTC)


I realize people are concernned about developer time, but I'd like to suggest a nontrivial code change, which would probably resolve most other objections to PWDS:

  • Compute the average edit hamming distance & standard deviation, and automaticallly tag anything above or below one half standard deviation with a green - for minor or red + for major. All blanking could be marked with a red 0, irrespective of hamming distance.
  • You could change the "This is a minor edit" check box to "Significant edit", thus allowing people to declare their small edits to not be minor, i.e. no greeen -.

Sure, these are code changes which require a number of tweaks in several places, although none are too serious really. But, they would provide significantly more information to those who look at edit histeries. JeffBurdges 10:32, 24 November 2005 (UTC)

P.S. PWDS is probably less work on the serve eventually, as pages which have been blanked for a long time can be pushed out of the picture automatically, if the server does that sort of thing to deleted pages now.

Oh well

It looks like en.wikipedia isn't ready for this proposal. Any support we can give it will be drowned out by the knee-jerk reactions ("but but but VANDALS would take time out from changing numbers and facts to be wrong, and start blanking more pages than they do now! And that would be REALLY BAD!")

I'd love to see it implemented on lower-traffic places like Meta, which has enough cleanup problems without every deletion being a multi-step process.

Maybe for now we should get behind Uncontested deletions, which has some of the same benefits, but also has enough ceremony that it doesn't threaten the AfDers' outlook on life. rspeer 06:36, 28 November 2005 (UTC)

Well put. It amazes me the number of opposition comments which seem to be based on a gut feeling of aversion (or fear?) rather than any concrete criticism. I think uncontested is a nice middle ground. In the meantime, lets try and clean up the outdated and messy Wikipedia:Deletion reform pages and try out variants at WP:XD. While not proposing Experimental deletion as a strict alternative to WP:AfD, I would love to see more active participants in developing and using the experiments. XD5 and XD6, among others, reasonably approximate the two most active proposals ( Uncontested (delayed) and Pure wiki (blanking) ) here 07:20, 28 November 2005 (UTC)


I've written up a rationale for an expansion of the power to revert which I think people involved in this page might find interesting. Just comments for now. At User:JesseW/First version revert proposal. Thanks! JesseW, the juggling janitor 10:02, 28 November 2005 (UTC)


Radiant already mentioned that copyvios were a problem for this proposal. I'd like to amplify that with a legal consideration: as it stands, all previous versions of pages are freely served up by wikimedia when given URLs of the form, and hence pure wiki deletions mean that we would still be hosting copyvios. Only if old versions are blocked to most users can we claim to be respecting copyright, and that means this proposal must exist alongside some proper AfD system.

These URLs need not be a problem for linkspam, since the server can index these pages noindex,nofollow. --- Charles Stewart 23:03, 7 December 2005 (UTC)

isn't it policy to revert copyvio additions to articles (not delete the reversions fromt he history)? So it seems to me people don't have problems with copyvios in the history (I think, though, that we do delete copyvio edits if requested by the copyright owner). Broken S 23:05, 7 December 2005 (UTC)
This is so. WP:DRV, though, expressly treats copyvios differently for the purpose of content review (ie. you can ask to look at prior versions of most deleted articles, but usually not copyvios). I'd put this down as a potential legal vulnerability in the way we handle copyvios. --- Charles Stewart 23:18, 7 December 2005 (UTC)
Just for clarity, I thought I'd state the current procedure on these things (I didn't write it or anything, but it seems reasonable to me):
  1. If all the revisions in an article are copyvios it is deleted.
  2. If all the revisions from the start to some point in it are copyvio the history is truncated.
  3. If it is possible to revert to a copyvio free version, we do so, and the copyright owner must contact the Wikimedia Foundation to have those revisions removed. (Though it is obviously possible for an admin to remove them anyway.)
  4. We do not restore copyvios on WP:DRV as has been observed, since we have no right to do display the material and nor do we restore them to histories under rewrites, mainly by commonsense convention.
Thus, we do retain copyvio revisions in histories sometimes. I presume that reverting away a persistent copyvio exempts one from the 3RR at present and would do so under PWDS. If necessary, the article could be protected on the copyvio tag, as currently happens from time to time.
I suspect, although I do not know, that the non-removal of individual copyvio revisions from histories is a matter of ease-of-use, since it could be extremely problematic on an article like, e.g., GWB if a copyvio chunk of text were pasted in for a single revision. I suppose this is why we say they must actually make legal noises in such cases and so we do seem to implicitly acknowledge the incorrectness of retaining these revisions. Whether retaining substantially many more of them under PWDS would make the lawyers nervous or not, I cannot guess — IANAL. -Splashtalk 23:42, 7 December 2005 (UTC)

Interesting proposal

Excuse me if I'm wrong, but I don't think this proposal would cause more vandalism, because, with the right tools in place, we could easily see vandalized pages. However, I'm not entirely sure how blanking a page is very different than labeling a page "nn-bio" or "db|blah" except that with those, a second party (the administrator) can come in and verify the speedy and simply delete the page. If I'm thinking straight, under the current system, the vandal has to rewrite the entire page which might deter him from vandalizing in the future, but under the PWDS, a vandal might be more inclined to simply revert the page back to his edit (which would still be in the page history. I don't really see a big problem with the current system of administrator power in the speedy delete and AfD process, especially since most articles that don't achieve consensus on AfD can simply be relisted to achieve it. 95% of the time, articles either achieve a solid yay-or-nay (or some variation thereof). This proposal seems like a whole lot of work to maybe improve on a system that I don't think is really broken. JHMM13 (T | C) 02:15, 30 December 2005 (UTC)

What does this replace?

Maybe I missed it, but I didn't see how this proposal would affect speedy deletion. Would speedy deletion be eliminated? The proposal did say that AFD would still be used. I don't like that. If most of the time there is a delete/undelete/AFD sequence, then it doesn't really solve the problem. It could even create more work.

If this is an add-on that doesn't replace any existing procedures, then it seems like just a complication with little benefit. Mirror Vax 05:22, 9 January 2006 (UTC)

Pages could still be hard deleted, so there'd be no affect on speedying. The reason, I think, that AfD would be kept would be to sort out disputes. I agree that it's probably not a good idea to keep it as anything more than an RfC type list, because discussion could of course occur on the articles' talkpages, which would still exist.Grace Note 06:22, 9 January 2006 (UTC)

Yep. I, for one, would heavily encourage the use of talk pages (and probably a category to help people find deletion disputes), rather than something resembling current Afd. Pure wiki can be used with or without our current version of Afd. Friday (talk) 06:26, 9 January 2006 (UTC)