Wikipedia talk:Soft deletion

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Deletion
WikiProject icon This page is within the scope of the WikiProject Deletion, a collaborative effort dedicated to improving Wikipedia in toto in the area of deletion. We advocate the responsible use of deletion policy, not the deletion of articles. If you would like to help, consider participating at WikiProject Deletion.

Not related to AFD[edit]

We can have Wikipedia:Transparent deletion without making this an AFD outcome or inventing any new process for it. I agree that deletions should generally be transparent, except in cases of copyvio, libel, etc. Friday (talk) 22:11, 3 May 2007 (UTC)

Also, the software would have to be changed to not treat soft deleted articles like real articles- you want links showing up as though there's no article, and you don't want to hit them with random pages, for example. Many of the concerns are exactly those involved with a pure wiki deletion system. Friday (talk) 22:16, 3 May 2007 (UTC)

Actually, I very deliberately tied this proposed policy to AfD. An earlier proposal I made on the mailing list suggested that such a mechanism could be used outside AfD as well. But I concluded, after the discussion, that such a broad soft deletion mechanism would likely be overused and highly disruptive. Since deletion would become so simple and readily accessible, any user who feels that an existing article is not useful or its subject not notable enough might be inclined to soft-delete it. This could then lead to ugly revert wars, where previously we simply used quality tagging ({{unreferenced}} etc.), proposed deletion, and hard deletion.
I think the current process of forcing people to either identify the problems through tagging, or go through a discussion process wherever there is a a potential for informed dissent, is in many ways more harmonious. The main thing I don't like about it is the unavoidable loss of article histories, and the power asymmetry between admins and regular users. This proposal addresses these two problems.
Even if you feel that soft/transparent deletion could work on a larger scale, this proposal can be seen as a gradual, less disruptive change.
As for the red/blue links issue, I do not see it as a real problem. If the article is deleted, links to it should be removed. One can track those relatively easily by checking "What links here" on the AfD page.--Eloquence* 22:25, 3 May 2007 (UTC)
In many cases, though, the red links of deleted articles are a good thing; many topics ought to have an article, just not that article. Instead of de-linking, perhaps the software should be changed so that a tag or magic word can make a redirect of the soft deletion variety appear as a red link.--ragesoss 23:10, 3 May 2007 (UTC)

Various problems[edit]

  1. It is based on a false assumption. Whilst AfD is not upscaling, there doesn't seem a huge regular backlog in AfD closing. So I'm not sure what this would solve. Solutions in search of problems are always a bad idea.
  2. It is going to be far too easy for a 'keep voter', who loses the AfD, to add in one more source, soft-undelete the article and force another AfD. Since very often that Keep voter will be a 'owner' of the article or someone with a vested interest, that is going to happen regularly. In short, this will put more pressure on AfD, not less.
  3. I'd strongly object to this applying to low notability, badly sourced, bios. Even without being obviously libellous, non-notable bios attract the highest number of justified complaints. And these subjects object to things in the history as well as in the actual article, often with good reason. Having thousands of 'shallow graves' is just more trouble than it is worth. No thanks.
  4. Much easier to have a rule that any deleted content can be userfied on reasonable request.
  5. Bottom line is, we don't have a problem with too little borderline notable material on wikipedia - quite the reverse.

--Docg 22:24, 3 May 2007 (UTC)

1) Beyond quantity, there is always the issue of the quality of the decision-making, the burn-out of the involved admins, and so forth -- difficult to measure. Certainly, all admins would benefit if the workload of the "easy cases" on AfD could be more evenly distributed, which would free more energy to work on complex AfD decisions.
2) If this process results in a gradual improvement of the article, so much the better! The deletion process will organically evolve around such situations if they arise.
3) It may be sensible to exempt BLPs from this process, or to not use it when the subject of the bio favors deletion.
4) I don't see how this is easier or harder, or why there needs to be a "reasonable request".
5) Some content that is worth preserving gets deleted, and some content that is worth deleting gets preserved. The world is not black and white.--Eloquence* 22:34, 3 May 2007 (UTC)

1)I'm not convinced that this would reduce admin workloads at all. AfD closing is a miniscule part of admin workload - and I suspect we'd spend longer sorting out the mess caused by poor closings or the problems caused by pov pushers and owners
2)Doubtless, but what's the pain involved in terms of giving a licence for perpetual warfare. How much time have you spend in deletion areas recently? There's enough tendentious stuff comes to DRV, without this green light. Since you were last heavily active, the number of people and organisations interested in promoting themselves on en.wp has multiplied with our growth and popularity. This is fighting yesterday's war at the expense of today's.
3)It would certainly need to be ALL BLPs and possible more.
4)Well, not an "unreasonable one"
5)Sure. But you move in one direction, you may get more of the other.
6)Didn't we look at pure wiki deletion before? We've been over this ground a year or two ago, and there was nothing like a consensus. With the changes in Wikipedia's difficulties since, I'd say this has zero chance of getting any agreement.

--Docg 22:48, 3 May 2007 (UTC)

Some of the perpetual warfare concerns may be a problem we already have. Editors can already easily blank page content, or place the whole thing with "I am cool!!" People can fight over all sorts of things they shouldn't fight over. We already live with this, and actually it doesn't hurt us much. Friday (talk) 22:58, 3 May 2007 (UTC)
Yes, but when someone creates a crappy article on their band, deletion and salting where necessary tends to end it. This just lets them keep coming. Anyway, it isn't going to happen, we've been here already.--Docg 01:07, 4 May 2007 (UTC)
A PWDS has never been implemented. There have been some experimental hacks, and those generally focused on reforming the deletion process as a whole. This process is much simpler and only adds a new type of outcome to the existing processes. It is therefore designed to be minimally disruptive. Your main objection - BLPs should not be affected by this system - has been addressed. Your other objections are hypothetical (generally based on very pessimistic views of our user community), and the best way to test whether your hypothesis is correct is to put the system into use.--Eloquence* 05:26, 4 May 2007 (UTC)
I see no significant gain from this. And the problems I raise are NOY hypothetical. Check out AfD. Look at DRV. See how many pov-pusher and conflict-of-interest users are fighting like hell to keep their crap in the encyclopedia. These people WILL misuse this system to keep protracted debates going. They WILL appear with year another tendentious source and recreate their article demanding another debate before removal. They currently do it on DRV. This is not 'a pessimistic view of our user community' - it is realising that our user community is already having to deal with too many self-promoters who take advantage of our current systems, and you are simply going to strain the community's patience further. If I am being pessimistic, you are being naive.--Docg 09:43, 4 May 2007 (UTC)

Redirect hinders recreation[edit]

One problem I can see with this is that having the article be a redirect makes it quite difficult for someone not familiar with Wikipedia to work out how to recreate it. I know the template includes instructions, but it's still likely to confuse people. I'm not sure if a redlink or a soft delete is more likely to encourage people to produce a good article. --Tango 22:26, 3 May 2007 (UTC)

That's a feature, not a bug. If someone is too new to be able to work this out, do we trust their judgment on what's appropriate to recreate? I don't. Friday (talk) 22:31, 3 May 2007 (UTC)
Agree with Friday, the barriers in this case are more intellectual than technical. Probably a good thing to require some reading before recreation. :-) --Eloquence* 22:35, 3 May 2007 (UTC)
I have to disagree with the idea that barriers of complexity, non-transparency, and non-newbie-friendliness are a good thing. Most of the supposedly non-notable content that gets deleted without a fight (for example, through Proposed Deletion) simply doesn't have anyone interested in it besides the creator. If another new user comes along and in good faith wonders why Wikipedia doesn't have an article, or how to recreate the deleted article, there is a good chance that the problem was with the deleted article itself rather than the encyclopedicity of the topic. Having new users jump through technical/intellectual hoops does nothing to deter bad editing, only unformatted, newbie-style editing (which we deal with pretty efficiently).--ragesoss 23:27, 3 May 2007 (UTC)
Actually, I've seen quite a few examples of fairly detailed articles about comics, open source programs, websites, and so forth deleted completely on notability grounds (relying only on primary sources). In those cases, an awful lot of pre-existing work gets wiped out by hard deletion, whereas soft deletion would make the history transparent to all and avoid unnecessary duplication of effort. I suppose that in the cases where the article is a stub to begin with, it makes re-creation a little more difficult. If that does become a problem, the policy could be altered to not apply to such pages.--Eloquence* 01:04, 4 May 2007 (UTC)
I agree that keeping the deleted versions available is often a good thing. It's just the idea that a usability barrier is "a feature, not a bug", that I object to.--ragesoss 01:10, 4 May 2007 (UTC)
Please take a look at Template:Soft deleted -- the clearer we can make it the better.--Eloquence* 05:22, 4 May 2007 (UTC)

An example of why this would helpful can be found at User talk:Jacob valliere. Although this editor's lack response to the warnings left on his talk page is a seperate problem; he kept recreating deleted articles because of existing red links. If this proposed solution is not adopted, deleting admins need to be certain to check "What links here" and de-link the pages they delete in these cases.--BirgitteSB 13:37, 4 May 2007 (UTC)

Possible Confusion[edit]

Some people who don't use Wikipedia very often might be kind of confused when they came to a deleted page, if this system were adopted. I think it's an excellent idea, but we should make sure that it's very clear for casual readers. Maybe trying to go to a soft deleted article could bring up a page that explains that the article has been deleted but links to the history and discussion, rather than just redirecting to the deletion discussion. --Ketsuekigata 22:32, 3 May 2007 (UTC)

I think it's mostly a matter of wording Template:Soft deleted as clearly as possible.--Eloquence* 22:35, 3 May 2007 (UTC)
BTW I did this in practice for a while a year or so ago using Template:XD7. Had no real problems. Friday (talk) 22:39, 3 May 2007 (UTC)

This relies on a false assumption[edit]

The proposal assumes that there are cases in which the subject should be included, but the current article should not. But, except for when G11 applies, that is not (or should not be) the case. When cleanup is possible, it should always be preferred over deletion, except in unsourced BLPs. And soft deletion is still deletion. -Amarkov moo! 00:06, 4 May 2007 (UTC)

No - this is for the cases where evidence that the subject should be included has not been established, one way or another, but where common sense dictates that it could be found. Typical examples include web comics, Internet phenomena, online games, etc. It is not intended to modify existing deletion policy only inasmuch as the outcome of some deletions would be a soft, rather than hard, deletion.--Eloquence* 01:01, 4 May 2007 (UTC)
Oh God, please spare us. 01:30, 4 May 2007 (UTC)
I have to agree with Amarkov. Except in G11 cases, an article that can be saved should be saved. When someone votes in AfD to delete an article, that person is saying that the article should not exist in Wikipedia whether it is a good article or a bad article. It is possible that many people vote in AfDs without considering this (me sometimes as well), but in theory, a badly written article that has potential of being a worthy article should not be deleted. Captain panda 03:02, 4 May 2007 (UTC)
It's not so much about those that have the known potential, but those that have an undetermined potential (lack of sources one way or another).--Eloquence* 05:21, 4 May 2007 (UTC)
If nobody can find sources in the five days of AfD, it is a reasonable assumption that no sources exist, so there is little reason to allow it to be recreated easily. And if common sense truly dictates that sources exist, then it shouldn't be deleted at all, just sourced. -Amarkov moo! 14:31, 4 May 2007 (UTC)
I am a firm believer that an article should have the sources added when it is created. If a creator can't even go that far when they bother to create an article, why keep it? In essence, the creator is saying, "Here, I value this subject enough to get other people to write about it and research it, but not enough to do it myself. I can't be bothered to show it is true either." I would love to have this as an option for those articles. Slavlin 18:17, 4 May 2007 (UTC)
I'm still not sure how I feel on that, but I'm not sure why it matters. It might if nobody looked for sources in an AfD, but since people do, it's still a reasonable assumption that no sources exist if none are found. Or am I missing something? -Amarkov moo! 03:11, 5 May 2007 (UTC)
From this official policy, it is NOT the responsibility of the nominator of an AfD to look for sources. Looking at the "in a nutshell" portion, you can see that "The obligation to provide a reliable source lies with the editors wishing to include the material, not with those seeking to remove it." so I think a policy like this would cover those situations where a consensus can't seem to be reached, but still no one bothers to source the article. Slavlin 07:53, 5 May 2007 (UTC)

Redirecting to a project page[edit]

Per Wikipedia:Avoid self-references, I am not sure that having a namespace page redirect to a Wikipedia-space page is a good idea. Does this proposal have any precedent? Cos if I was a first-time anon user, who typed something into the searchbox, I might understand if I go to the page that says "Wikipedia currently does not have an article of that name. If you would like to create one, use the box below". I might not understand if I went to a page that looked different from all the other normal pages I had seen, had lots of abbreviations etc that I did not understand, and seemed to be discussing esoteric matters of "notability". I think this proposal - if I undertand it right - could be enormously confusing for new users. Batmanand | Talk 08:38, 4 May 2007 (UTC)

The AfD discussion should be viewed as an "attachment" to the Template:Soft deleted message, which can get fairly detailed and explain some of the common jargon. IMHO, making our deletion processes more transparent can only be a good thing.--Eloquence*
I don't really have any strong feelings about this proposal, although I don't in general see the need. Regardless of what people think about making our processes more transparent, the silent majority of readers absolutely don't care about our deletion process and would be very confused if they were redirected to an AFD debate after clicking on someone's name in the articlespace. The concept of different namespaces is not something that most readers are even aware of. - cohesion 05:19, 6 May 2007 (UTC)

Been there, done that[edit]

This appears to be simply another attempt to implement Wikipedia:Deletion by Page Blanking (more commonly known by its weasel name), not to mention earlier rejected Wikipedia:Zap. I have not seen a demonstrated practical need for this. >Radiant< 11:01, 4 May 2007 (UTC)

Yeah, I though that, I just couldn't remember where. Has Erik missed these debates? I'm really puzzled as to why this has suddenly been put forward just now.--Docg 11:05, 4 May 2007 (UTC)
No, the system described here is different both in implementation and suggested use. Emphatically, it is not intended to be used outside the AfD process.--Eloquence* 12:13, 4 May 2007 (UTC)
Which is precisely why it doesn't help the main problem, which is the stress on the deletion infrastructure. We've been over this ground so many times.--Docg 12:21, 4 May 2007 (UTC)

We need to separate out the component pieces and consider them individually. Transparent deletion has many advantages and no known disadvantages. This is enough to consider it by itself with no change in process. Friday (talk) 13:57, 4 May 2007 (UTC)

  • I can think of a few disadvantages. It should, for instance, never be used for (1) copyvios, (2) BLP concerns, or (3) vanity, because in all cases it would allow the harmful content to remain in place. Another problem is that it adds a layer of complexity to one of our busiest processes. >Radiant< 14:02, 4 May 2007 (UTC)
Right, nobody I know has ever suggested it should be the only means of deletion. Hard deletion makes sense in those cases. I'm not suggesting making it a "process" thing at all. Just a technical option when deleting. Admins who don't know anything about it would continue to do regular hard deletions, no harm done. Friday (talk) 14:16, 4 May 2007 (UTC)
I've outlined several disadvantages above, which have not been addressed. OTOH, I see no real advantages to this.--Docg 14:18, 4 May 2007 (UTC)
How about we soft delete this page? Hehe, I'm just trolling. 20:09, 4 May 2007 (UTC)
Soft deletion also doesn't make sense when users can link to an old diff to show the page the way they want to. -- nae'blis 17:12, 7 May 2007 (UTC)

Ignoring, for a moment, the why; how?[edit]

Is this a separate AfD !vote (Soft delete) or something that any user can use in lieu of admin deletion? Nifboy 23:10, 4 May 2007 (UTC)

IMHO the latter, under the guidelines established through the policy (when not to soft delete).--Eloquence* 18:20, 5 May 2007 (UTC)
So this would become, in some sense, the "default" outcome of a delete AfD, unless an admin stepped in under policy and did admin deletion? Nifboy 21:33, 5 May 2007 (UTC)

Extension to Proposed Deletion[edit]

I'm going to reword the proposal to include the PROD process as well as AfD; even more than with AfD, a lot of what gets deleted through proposed deletion looks accurate but simply has no sourcing and no one willing to put in the work to bring it up to an acceptable standard.--ragesoss 23:43, 5 May 2007 (UTC)

Now that you've suggested it, is there a reason not to make all PRODs some type of soft deletion? -Amarkov moo! 23:50, 5 May 2007 (UTC)
Yes. Many PRODs are dictionary definitions, personal essays, and other things that we never want an article on (many of might have been speedily deleted if there weren't so many criteria to sort through and if PROD wasn't so easy).--ragesoss 00:10, 6 May 2007 (UTC)
Here is the page I created as a PROD redirect destination: Wikipedia:Soft deletion/Proposed deletion. It could use more refinement--ragesoss 00:12, 6 May 2007 (UTC)
Indeed. If we never want an article on something in the first place, there's no point in "soft deleting". >Radiant< 08:27, 7 May 2007 (UTC)


Redirecting to project pages from article space? Absolutely not – Gurch 17:57, 6 May 2007 (UTC)

It would be useful if we had a magic word the prevented articles from appearing as Random articles. That way instead of redirecting, pages could simply be soft deleted by replacing the page with a template explaining the situation (with a link to the AfD where appropriate). Of course, cross-namespace redirects are hardly unprecedented (I don't know how many exist, but there are at least a handful) and the attempt to get rid of them via policy (WP:CNR) was unsuccessful.--ragesoss 05:59, 7 May 2007 (UTC)
Really bad idea in many many ways. Waste of server space, for one. Plus I agree with Gurch. It would be cross-namespace redirects. A big no no. I just don't think we have enough deleted articles that are savable enough to have this. --WoohookittyWoohoo! 05:20, 8 May 2007 (UTC)
Browse through the PRODs that are 1 day from expiring. Lots of them (half I would estimate) are factual articles that have no obvious problems except lack of sources/assertion of notability, and for many of them, more sources probably exist but no one can be bothered to track them down. As for server space, deleted articles are still saved. Cross-namespace redirects aren't a big no-no, just a medium one. Little details like redirects can be dealt with, either technically or by convention; that shouldn't obscure the relevant policy question of soft deletion. --ragesoss 05:43, 8 May 2007 (UTC)
  • Yeah, I'm with Gurch on this one. In any case, we have enough deletion processes (AfD, PROD, and Speedy - that's three different levels) that adding another is just more instruction creep. If someone wants the old content from a deleted article, they can ask an admin to userfy it or otherwise retrieve the deleted content. I do that from time to time, and I know there are other admins who are not averse to doing so. Redirecting everything marginally important rather than outright deleting it encourages people to create marginal content, with the knowledge that it will still be on Wikipedia even after "deletion". I gotta disagree with this policy. PMC 03:44, 13 May 2007 (UTC)
Server space: this is a false argument, and has been for many years. Deletion does not remove the articles from the server. In almost all cases, every version remains forever. The real argument against this is confusion. Space in human craniums seems to be a bigger problem. DGG 01:19, 14 May 2007 (UTC)
Was that comment directed to me, DGG? Because I didn't mention anything about server space in my argument. PMC 01:28, 14 May 2007 (UTC)


It should be simple enough to retroactively look at what seems to be a key question: How many articles that are deleted are later successfully recreated. I know some are salted but can you demonstrate that this would actually accomplish something downstream? This may be much ado about nothing. JodyB talk 11:38, 8 May 2007 (UTC)

"How many articles that are deleted are later successfully recreated?" And further, of these, how many would benefit from the contents of the delete article?--Docg 11:55, 8 May 2007 (UTC)
And conversely, how many unwanted articles are recreated (and deleted again) using the contents of the earlier deleted article? The list of protected deleted pages is quite lengthy. >Radiant< 11:58, 8 May 2007 (UTC)

Is it possible to get some answers to these questions? At present I am not convinced there is a significant problem warranting a whole new level of bureaucracy. You're suggesting what amounts to a fundamental change in the deletion process. It may be a good idea but I ask that we firm up the numbers and demonstrate that there is a problem to begin with. JodyB talk 13:10, 11 May 2007 (UTC)


Sounds like a good proposal. JediMaster16 20:17, 9 May 2007 (UTC)

Authority / trust[edit]

The key problem I see with this proposal is that the action of closing an XFD debate and redirecting the page can be done by anyone, even anonymous (IP) users. What measures can be put in place to ensure that only trustworthy users close XFD debates under this proposal? I would also propose that the redirect page is protected after a soft delete (perhaps indefinitely, or maybe for 2 weeks or so) to prevent opponents to a successful deletion engaging in a restoration/redelete edit war. Waggers 07:27, 11 May 2007 (UTC)

This problem is something we already have, and I haven't seen that it causes us much trouble. Friday (talk) 14:26, 11 May 2007 (UTC)

Hard deletion offensive to GDFL[edit]

It strikes me that hard deletion is in principle offensive to the spirit of GFDL. The authorship of good-faith contributions should be maintained. Who is to say that material created on that now deleted page is not the source for derivative work elsewhere? I believe that hard deletion should be reserved for removing material that was clearly unwelcome at the time of contribution. Examples include copyright violations, offensive material, and anything that is not a good-faith contribution.

Soft deletion is already in practice, through merge and redirect. For example, Grumpy old man was considered inadequate, was blanked and converted into a redirect to Irritable Male Syndrome. None of the material from the first appears in the second, so it was not really a merge, but a soft delete. This is a good thing, because anyone can retrieve the old material and improve it.

Some kind of explicit soft deletion is needed for pages which, like Grumpy old man, contain potentially good material, but for which a suitable redirect target hasn’t been identified. --SmokeyJoe 14:21, 11 May 2007 (UTC)

  • That means there's no need for this proposal at all, because we don't delete articles after we merge or redirect them. >Radiant< 14:29, 11 May 2007 (UTC)
    • No need for this proposal? Maybe this should be merged to Wikipedia:Deletion policy? Because...? No, the issue is of articles that aren't merged because no suitable target is identified. --SmokeyJoe 15:33, 11 May 2007 (UTC)
      • So you allege that there is a substantial amount of articles that end up deleted because no suitable merge target could be found, is that correct? Evidence please? >Radiant< 15:41, 11 May 2007 (UTC)
        • I think his argument is that material from subsequently deleted articles gets re-used in other articles sometimes when there isn't a valid redirect, so they aren't technically merged and the origin of the content becomes unavailable, contra GFDL. If this does happen and is identified, however, it can be corrected retroactively.--ragesoss 16:00, 11 May 2007 (UTC)
        • I say that unnecessary hard deletion is offensive to GFDL.
        • Example: Paradigm High School was a good faith, sourced contribution. The content could be used to write a better article on the school, or on the town. (as I remember)
        • It's difficult to show useful stuff in the hidden places when it is hidden. --SmokeyJoe 16:05, 11 May 2007 (UTC)
          • I'm sure it was good faith, but it definitely wasn't sourced. It was also lacking in notability and basically consisted of an enumeration of scholarly virtues and dresscode. I disagree that this would be useful to write an actual article, and I disagree that Wikipedia would be improved with an actual article on a non-notable unsourced elementary school. >Radiant< 16:13, 11 May 2007 (UTC)
            • I'm at a disadvantage, not having access to the deleted article. I don't remember the virtues and dresscodes. I remember it being a high school. I did add 1 or 2 secondary sources before it was deleted. There is potential interest in it being one of few new schools in its state. I didn't assert that the school was notable, but the town is, and the building of the new school is an item of interest to the town. You and I might not think that the material was particularly useful, but who are we to decide on what someone else might find useful. Note that I am not arguing that the page should still be present. I am saying that the contents should be available in the history (assuming vanity didn't apply). --SmokeyJoe 16:30, 11 May 2007 (UTC)
              The name "Paradigm High School" strongly suggests that it is a high school. The school's website states it "provides a classical, liberal arts education", and claims to build "a solid foundation in the core fields of language, literature, history, science, mathematics and the fine arts", which sound not quite like an elementary school. You can still find (a version of) the article at, and you may agree that little encyclopedic value was lost by its removal.  --LambiamTalk 00:18, 13 May 2007 (UTC)
              • I never disputed that the article should have been deleted (removed from main space). But I don't see why it needs to be hidden. Even if it is just an example of what not to do next time, it can be useful as a reference. Also, there is no good reason to hide the edit history of the contributors involved. --SmokeyJoe 02:29, 14 May 2007 (UTC)
Well, why couldn't anyone just userfy that particular article during the AfD? If someone thought that the article would be suitable enough for reinclusion later on with sufficient improvments, they would have done so. I personally think that the article failed WP:NOTE, WP:V and WP:NPOV anyway.
I see this policy as a failure, because it simply adds to the deletion process and "curves" around AfD and PROD. If an article is not improved enough with the five days alotted for AfD's and PROD's, and the consensus was to delete (and by this, I mean hard delete), then the admin should delete the article because the community decided to delete, and this should not be held up because one or two users wish to see the history. There is nothing to see because if there was potential in the article and someone saw it, it would have been userfied by then.
If there was to be a compromise to this proposed policy, I would maybe suggest a "Requests for viewing", where a non-admin would request an article to be viewed at the most recent edit before it was deleted, and an admin would copy-paste that version using the restoration function. This would not intervene with the already complex deletion process, and is on a one-to-one basis. Sr13 08:45, 12 May 2007 (UTC)
    • Another reason for this proposal comes from the other side of deletion arguments. There are many articles out there that fail policy (usually WP:V), that according to the rules should be deleted, but an attempt to delete would likely be contentious. And the contention that would result is entirely unproductive. An example is Toko school. It's an interesting article, encylopedic except that it's entirely unverified (I even tried). The external link doesn't support any of the content of the article. Soft deletion of this article would be much more palatable than hard deletion. The existing content could well be improved with the addition of sources. --SmokeyJoe 02:36, 14 May 2007 (UTC)
Why couldn't we just use an unref tag instead? Sr13 02:12, 15 May 2007 (UTC)
Well, we could (obviously). I’m not sure what your point is. Do you believe that the article should stay in main space, but tagged? --SmokeyJoe 02:48, 15 May 2007 (UTC)
Well, tagging seems to be appropriate in the scenario you give. Sr13 05:07, 22 May 2007 (UTC)
Tagging is a temporary measure, and the contributors of the substantial content have not since participated on wikipedia for a long time. Which do you believe: The unref tag can stay forever; or the article should be deleted? If it should be deleted, why would soft deletion not be appropriate? --SmokeyJoe 05:28, 22 May 2007 (UTC)


Some of the problems some people have identified with the current proposal, such as the redlink issue, would be solved if soft deletion of an article John Q. Doe takes the form of moving it (with history and associated talk page) to, say, Wikipedia:Attic/John Q. Doe, and then deleting the resulting redirect page. This is similar to userfication, but then it is more communityfication. No new mechanism or anything needed.  --LambiamTalk 19:21, 12 May 2007 (UTC)

Sounds like a good idea...maybe even better than my suggestion. Sr13 20:30, 12 May 2007 (UTC)
I agree that the idea sounds good. I was thinking Wikipedia:Deleted/John Q. Doe, but I suppose Attic is markedly less dry. The attic article could be linked from the deletion log if cross-space links are too heavily frowned upon. But what is so bad about cross-space links? Mainspace is full of them, in tags. --SmokeyJoe 02:41, 14 May 2007 (UTC)
This pretty much already happens - if an article is deleted but contains useful information, admins can still view it, users can request userfication, and viewing can be requested on DRV. The point remains that the vast majority of deleted articles are not in fact salvageable. >Radiant< 10:54, 14 May 2007 (UTC)
"Soft deletion" as proposed here – meaning that users can inspect softly-deleted articles without admin intervention – should be just one option of keep / merge / atticize / userfy / hard delete. But just as we keep the history as a record for other pages, it should be kept for deleted pages unless there is a specific reason for not keeping it. It has happened to me that an article I had just created was deleted as "recreation of deleted material" – of whose prior existence I was wholly unaware. I could have requested to be allowed to view the earlier deleted article, as well as that created by me (of which I had not kept a copy), so as to verify that they were "substantially identical", or whatever the term is, to a degree justifying the speedy delete, but at the time I was sufficiently demotivated to not go through the effort of requesting anything. If the community consensus is that we should have some form of soft deletion, I think the method I have indicated here is better than the alternatives I've seen.  --LambiamTalk 16:53, 14 May 2007 (UTC)
  • Radiant says "This pretty much already happens." Yes. This proposal represents a small, incremental change.
  • If a deleted article contains useful material, then how does a user know to request userfication?
  • Salvageable material is a small subset of useful material. Material may be useful merely in providing a reference for what is not acceptable. It can also be useful as part of the edit history for the contributor. Vandalism edits are usually kept available in histories. Why should good-faith contributions on deleted pages be hidden? --SmokeyJoe 01:04, 15 May 2007 (UTC)


This might be acceptable for articles that are unsourced but not otherwise objectionable--these are the only cases which really present a real conceptual difficult under the present system: they may be about truly notable subjects, and therefore should stay, but if they are unsourced they cannot prove it, and therefore should go. This is a sort of purgatory, it which they can stay not indefinitely, but for a very long time without being confused with verified articles.

I would oppose it otherwise, for it makes everything harder by needing to do it twice: first to put it in purgatory, and then to dispose of it one way or the other. We have enough problems doing it once. Any other purpose could be served by automatically permitting the userification of anything not libel or copyvio, with a fixed time limit of several months. DGG 01:15, 14 May 2007 (UTC)

I agree with much of this, and with User:Doc glasgow's comments above. AfD doesn't get seriously backlogged; adding another layer of purgatory (as DGG put it) would complicate the process and create another point for backlogs to develop. I think userification on demand solves this problem much more usefully. We can encourage people to check the deletion log prior to creating a new page, to ensure they're not duplicating effort; if a previously deleted article turns up, the AfD can be accessed, and the page itself could be provided (in userspace) by requesting it from an admin, so long as there are no BLP/copyvio issues. Also, we should encourage people to have evidence of notability ready at the time they create the article; far too often an article is created on the premise that sources will be forthcoming, but the sources never appear and the article sits around. MastCell Talk 17:23, 14 May 2007 (UTC)

New Editors Concern[edit]

That whould frighten new editors, who were browsing for a page and then *pop*, a random "This page was deleted" pops up. No, just no. If I whould have seen that in the time I was a New Editor, I whould had left Wikipedia. And what if the subject becomes good enough to make an article with it? This will confuse new editors, and cause them to leave. If someone wants to see why it was deleted, they have to check it manually. WOW! What does it take? 2 Minutes? And then we have more editors who will hopefully contribute more than 2 minutes. I think it is not worth the sacrifice. - Flubeca t 02:39, 14 May 2007 (UTC)

Agreed. We don't want Random article to take you to a deleted page. Actually, I don't like the current Random article taking you to usually a pretty bad bad. Maybe Random article should take you to a rate good (at least) page. Someone somewhere has noted that cleaners aren't bothering with the articles needing clean-up category, but are just using Random article. This says that Random article is serving cleaners, not readers. (sorry to digress) --SmokeyJoe 02:48, 14 May 2007 (UTC)

Soft deletion is appropriate for Prods, by default.[edit]


Prod is a serious problem when it is misused (or abused). The article is gone in a flash if nobody notices. And there is no AfD archive to see why. While creating content (yes, there are editors who actually contribute to the encyclopedia, not just policy and guideline talk pages), I make lots of red wikilinks in the blind. I can't tell you how many times, when following these links, I found an article that I could have used was deleted, sometimes recently, while a search turned up no AfD discussion. I assume these were proded into oblivion. Some may say this is no big deal, just create a new article with references. But when you are creating properly wiki-linked content, you often generate links faster than you can fill them out, and it's nice to have a stub to hang your hat on. And when you do create a stub, some sharp shooter is waiting to "nail it" even before you can catch your breath. And frankly, I am put off from creating a new article when the old one has been deleted--that's just asking for agita I don't need. So this is all very disconcerting to editors who actually contribute material. Frankly, we have too many editors who are self-appointed gate keepers who may in good faith think they are doing good work, but are really hurting the project in the long run. Dhaluza 00:10, 15 May 2007 (UTC)

Soft deletion will enable all users to review these frequent, supposedly non-controversial deletions. This is the wiki way. --SmokeyJoe 04:28, 15 May 2007 (UTC)

Well, how about we move prods to the attic for a given period, say 90 days, after which they are deleted. This allows people to browse the attic, identify articles to work on and improve but also prevents the attic from swelling to an enormous size. I'm not absolutely convinced there is a big problem here but I could live with a limited version of soft delete. However, AfD shouldn't be soft delete candidates. JodyB talk 18:23, 16 May 2007 (UTC)

I'm not sure what this solves[edit]

To me, this seems like a mere technicality; a mechanism so that experienced users have easier access to the history of a deleted article. Everyone else, I think, will stop at the word "deletion", turn around, and walk out. I've seen AfDs ending as merge get treated as a deletion by the relevant community.

We already have a mechanism for viewing deleted histories (besides "ask an admin") in the content review portion of deletion review. Although I wouldn't mind making it more visible (yet another admin backlog: Requests for History), I don't see the value in making all afd-and-prod closers make a judgment about whether the history should be viewable or not. Nifboy 06:54, 15 May 2007 (UTC)

With hard deletes, if you don't know an article was deleted, you are unlikely to find out about its prior existence and use either mechanism to peek back in time. That makes this more than a technicality.  --LambiamTalk 19:25, 16 May 2007 (UTC)

Transparency is good[edit]

Reverted article revisions live on without disaster in public histories, yet deleted articles are hidden with barely a trace. A Jimbo quote from 2003, echos my own sentiments: If a page has to be removed for a legal reason, then it should not be viewable by the general public. ("legal reason" would most often be copyright, but also libel would be possible.) If a page is removed for other reasons, then there is no reason to hide it from users who want to investigate and oversee changes. ( diff ). This process allows deserved public access to removed materials. In this review of an old theme, I support E's focus on avoiding, ...loss of article histories, and the power asymmetry between admins and regular users. (from above). here 09:02, 15 May 2007 (UTC)

Redlinks are good[edit]

I would object to any procedure that does NOT create a redlink in mainspace when an article is deleted.

Redlinks are extremely useful for seeing that "something is wrong" (such as a reference that was to a page that is now gone), or to keep lists of things that are present / not present (such as a lot of the cleanup lists).

That said, I do agree that having noncontroversial-but-unimportant articles be viewable after deletion is a Good Thing. Suggestion for modification of proposal:

  • Move article to Wikipedia:Attic/articlename, with no redirect breadcrumb trail
  • (slightly more programming) Have the "create page" edit box show a prominent link to this page IF it is present. (shouldn't be much more work-at-runtime than the current red/blue links, but requires special coding), thus helping with the "inadverent recreation" problem

My thought. --Alvestrand 04:08, 18 May 2007 (UTC)


  • 1. How and what does this help?
  • 2. Sounds like spilled salt.

I object to the soft deletion process. Cool Bluetalk to me 00:49, 19 May 2007 (UTC)

I agree. There are no benefits coming out of this process. Unless someone can convince me (or Cool Blue) within a week, I will consider it rejected. Sr13 01:49, 19 May 2007 (UTC)
Two major benefits: 1) If an article is deleted because most of its material is not worthy of an article, the remaining portion is lost and unavailable to future authors who want to create a good article. 2) If an article is deleted due to lack of notability, the material will be unavailable if the subject becomes notable. This is a particular problem for "rising stars" such as scientific theories that are not yet generally accepted, small-time politicians before they become well-known, and the like. Wikipedia is not an oracle and these articles SHOULD be deleted, but the information should not be sealed forever. Personally, I think all deleted articles that aren't legally-problematic, privacy-related, abuse-magnets, or vanity-related should be soft-deleted rather than hard-deleted. davidwr 09f9(talk) 21:32, 19 May 2007 (UTC)
To rebut, one could simply ask one of the admins to userfy the page for them. Even then, what are the chances of an article to be created, then deleted, then be created again and be accepted as notable enough? Also, this process is simply a copy of what is already at deletion review. Sr13 04:39, 22 May 2007 (UTC)

Transparent deletion via deletion history[edit]

Today, when you go to a non-existant page, you can view the deletion history:

If a page was recently created here, it may not yet be visible because of a delay in updating the database; wait a few minutes and try the purge function. If a page previously existed at this exact title, check the deletion log and see Why was my page deleted?

Changing this to

If a page was recently created here, it may not yet be visible because of a delay in updating the database; wait a few minutes and try the purge function. If a page previously existed at this exact title, check the deletion log and see Why was my page deleted? or check the previous version history to see previous versions of this article, if there are any.

Previous versions can kept in another namespace, perhaps DeletedArticles:ArticleName/DeletedDate. To prevent vanity, the web server can block "deep linking" into this namespace and search engines can be excluded. Update: I see my proposal is similar to Redlinks are good above. davidwr 09f9(talk) 21:34, 19 May 2007 (UTC)

Vote tally so far - for's lead but no consensus[edit]

I know this isn't a vote, but here's what I have so far:

In favor, for the most part, except for articles that really need deleting like copyright and a few other things: 10: Friday, Eloquence, ragesoss, Amarkov moo!, JediMaster, SmokeyJoe, Lambiam, here, Alvestrand, Davidwr

Against change for the most part: 8: cohesion, >Radiant<, JodyB, Cool Blue, Sr13, nae'blis, Gurch, PMC

Many other commenters had mixed reactions, proposed other alternatives, or I couldn't tell where they stood.

It looks like although the for's slightly outnumber the against's, there is no consensus one way or the other at this time. Furthermore, the clear-cut for's are less than half of the participating editors.

If I goofed and put you in the wrong category, my apologies. Feel free to add, delete, or move your name around and update the total votes. davidwr 09f9(talk) 22:11, 19 May 2007 (UTC)

I wouldn't say I'm against the whole thing, just against a redirect to the discussion page. If it was just a page with the template I'm neutral to slightly positive, at least for a test run. - cohesion 01:25, 20 May 2007 (UTC)
Nor am I absolutely against it. I do oppose it in the present form. JodyB talk 01:59, 20 May 2007 (UTC)
I'd say that it is equally unclear what it is what the people who are counted as being in favour are in favour of. I am in favour of soft delete in the particular form I suggested, and then as one possible option in the range from Keep to Hard delete. I am against the original proposal of replacing deleted articles by redirects to the deletion discussion.  --LambiamTalk 08:41, 20 May 2007 (UTC)
So maybe the question we should be asking is: What is the common denominator here? Where, if anywhere, is there any consensus? We should also appreciate that whatever smidgen of consensus is on this page is not necessarily to be found elsewhere among the general editing population. JodyB talk 12:11, 20 May 2007 (UTC)
My impression is the following.
  • No consensus can be reached to turn deleted articles into redirect pages (redirecting to the page with the AfD debate).
  • There may be a majority in favour of having a lightweight method for getting access to material, including the edit history, that was deleted for relatively innocent reasons (such as lack of notability, WP is NOT a repository of, and insufficient verifiability). "Lightweight" implies that no admin intervention is needed, but the user desirous to view the deleted material should nevertheless do something not particularly difficult but still out of the ordinary to gain access.
  • In fact, it looks like this could be a consensus position with respect to speedies and prodded articles – with exceptions, of course, for things like copyvio, personal attacks, and other egregious violations of propriety.
  • Maybe (or maybe not) agreement can be reached that this should also be an option for the recommendations in a deletion debate regarding the disposition of the material.
I am not purely against the process. In fact, I agreed with Lambiam's suggestion above on his/hers "Attic" suggestion. I want a process in which the purpose, giving users the power to view a delete article and help make it pass notability guidelines, is fulfilled, but not in this way. Sr13 05:16, 22 May 2007 (UTC)

stab at least-common denominator[edit]

I doubt we will ever reach 100% consensus, but I think most people will be OK with the following broad outlines. Implementation details can be worked on later. Please tell me if you have any serious objections to this, and please comment on how to handle WP:SPEEDY deletions and how we should default AfD deletions. Here are the broad outlines I think most of us have agreed on:

  • No redirects. Links to deleted articles get red-lined and in some cases bot-dewikilinked just like they do now.
  • Certain categories like copyright, some BLP, etc. will be hard-deleted. See #Stab at categories to be hard-deleted below.
  • WP:PROD should be soft-deleted unless there's a reason to hard-delete it.
  • I'm unable to determine what we want to do with WP:SPEEDY deletions, most of them are either some form of vandalism, trivially non-notable/near-empty articles, or fall into the one of the auto-hard-delete categories
  • By default, AfD will result in a soft deletion. If it met the hard-deletion criteria it would've already been deleted without an AfD. I'm not sure if we have consensus on this, particularly for cases where "hiding history" is desired by almost all of the participants of AfD. Comment: Personally, I think we should leave Hard Deletion as an option but ONLY if there is a true consensus or near-consensus, not just a majority, on a given AfD.
  • Some automated mechanism to allow users who know what they are doing to access deleted versions without confusing new users. This could be through something akin to userfication or through some alternate "attic" namespace. Bear in mind that there may be more than one soft-deleted version of an article, each version should be accessible on its own or in combination by any editor who wants to see it. Comment: As an example of how to do this: The "article does not exist" page or "deletion history" page could have an unobtrusive link that led to the old articles or that led to instructions to userfy the old articles. Just as a side-note, I'd like to see a "copy this version to your user page" button that would clone a particular soft-deleted version to User:yourname/Articlename/versionnumber where versionnumber is 1 for the oldest version of the article, 2 for the 2nd, etc. or User:yourname/Articlename/deletiondate where deletiondate is the date and time of the deletion. After copy is complete you would have a message on your talk page with a link to the userfied copy. This has the advantage that the master deleted copy is read-only but the user can mark up his version and collaborate with others if he wants to.

davidwr 09f9(talk) 20:12, 20 May 2007 (UTC)

  • Comment AfD's should be hard-deleted in my judgment. I think many people said that above, unless I'm having a senior moment here. JodyB talk 02:10, 21 May 2007 (UTC)
    I'm not sure what you mean. AfD debates should be archived, as they are. Articles debated at AfD debates should only be deleted if the consensus is that they should be deleted. However, there are other options, one important one being "keep". Another one is called "userfication", which is a soft form of deletion. As far as I am aware no-one advocated that this option be abolished. Some people did not see a point in adding another form of soft deletion, while others said that they did not like the specific mechanism proposed here. Since these two positions were not clearly delineated from the start, it is not immediately clear how many people disagree with having any other form of soft deletion. Everyone (I should hope) agrees that if the consensus of an AfD debate is "Hard delete", the article should be hard-deleted. At this moment the recommendation "Delete" is taken to be synonymous with "Hard delete" (although not really in the frequently seen "Merge and delete"); this does not preclude that in the future there might be an additional option "Soft delete".  --LambiamTalk 06:07, 21 May 2007 (UTC)
    • A couple of points here. First, although its been asked at least twice, no one has offered any numbers to show that the current AfD needs changing. How often is an article brought back from deletion, re-worked and then survives a subsequent deletion challenge? I think, although I could be wrong, that we are creating process for the sake of process. I am not arguing to abolish any currently existing option, I just don't see the point in adding soft delete as an option for AfD's. When I said tyhey should be hard deleted I did not mean to imply that we should remove options; just not add new ones for AfD. I should have been clearer, sorry. JodyB talk 11:37, 21 May 2007 (UTC)
      • I don’t see any requirements for changes in AfD for this proposal. “How often is an article brought back…” Well, you shouldn’t expect very often, because deleted articles are hidden from most users. --SmokeyJoe 07:00, 22 May 2007 (UTC)

Responses to davidwr.

  • No redirects. Agreed. Red links are useful. An attic article should never be considered a substitute for a real article.
  • WP:PROD should go to soft deletion. For every reason for hard deletion, there is a reason for it to have been sent for SPEEDY deletion.
  • SPEEDY: Leave it alone.
  • AfD: Leave it alone. If soft deletion works well, then you’ll see “soft delete” appearing in AfDs.
  • Some automated mechanism. No. Just link to the article in the attic from the deletion log. I see no problem with editing in the attic, nor a need for a master copy - the version history will be available. --SmokeyJoe 07:15, 22 May 2007 (UTC)

Stab at categories to be hard-deleted[edit]

I went through WP:SPEEDY and separated categories that I think should be hard-deleted by default and which should be soft-deleted by default. If a speedy-deleted article meets any hard-deleted criteria, it should get hard-deleted. Please comment.


GENERAL: G1: Nonsense; G2: Test pages; G3: Vandalism; G4: Previously deleted, if all edits are exactly the same as a previous version; G5: Banned user; G6: Housekeeping; G9: Office actions; G10: Attack pages/libel; G11:Spam; G12: Copyright infringement;

ARTICLE: A3: No content/contact attempt;

REDIRECT: R1: Points at redlink; R2: To userspace; R3: Unlikely typo;

IMAGE: I1: Redundant, all edits are exact copies of another article's edits; I2: Corrupt image, no uncorrupted version in edit history; I3: Bad license; I4: No License; I5: Copyrighted, unused; 16: Missing fair use rationale; I7: Invalid fair use claim; I8: Identical on Commons, all previous versions are also on commons;

CATEGORY: C1: Empty; C2: Renamed; C3: Templated;

USERSPACE: U2: Nonexistent user; U3: Fair use galleries;


G8: Talk pages without a main page;

A1: No context; A2: Foreign language (transwikied), hard-delete if history also transwikied; A3: No content/contact attempt; A5: Transwikied, hard-delete if history also transwikied; A7: No claim of notability;

I1: Redundant, at least 1 edit is unique; I8: Identical on Commons, a previous version exists that is not on commons;

TEMPLATE: T1: Divisive/offensive template;

PORTAL: P1: Portal Topic Meets CSD for an Article; P2: Underpopulated Portal;

Do not speedy-delete: (requires change to speedy-deletion policy):

I2: Corrupt image, able to revert to uncorrupted version in edit history and "delete" corrupted one;

Case-by-case basis:

G4: Previously deleted, do hard/soft the same as deleted version if there are only minor changes from deleted version. If there are substantial changes then it's a new article; G7: Author's request - soft by default, ask author if he wants hard;

U1: Owner's request - soft by default, ask owner if he wants hard

davidwr 09f9(talk) 20:13, 20 May 2007 (UTC)

Spam should either be soft or case-by-case. It's as subjective as A7 in practice. --badlydrawnjeff talk 19:52, 20 May 2007 (UTC)

  • In the above list, you are apparently not assessing how useful it would be to have the deleted history available to people. In particular, for A1/A3, it would by definition not be (per lack of content). P1/G7 are almost without exception similarly lacking in content. G8/T1 is certainly lacking article content, and T1 has additional problems. I1/I8/A2 all imply we have a better version somewhere else, so the worse version is not useful. A5 already implies that the history is transwikied, that's part of the process. U1 means the users wants the page gone, users can obviously already blank anything in their userspace (this also partially covers G7 again). And G4 follows from the above.
  • In other words the proposal is to add extra process for no tangible benefit. >Radiant< 11:01, 21 May 2007 (UTC)
    • I would not support soft deletion for any CSD case, especially A7. I would estimate that at least 20% of speedy deleted articles are deleted under A7, this would mean keeping all of these around in some form or another in hope that they could be improved. The whole point of speedy deletion is to remove things that absolutely do not belong here. If someone writes an artcle about me, as flattered as I might be, it should still be deleted per A7 as I have not done much to make me notable. Keeping it around won't help anything. If I'm not notable, I'm not notable, unless I do something in the future to change that, in which case an article can be written then. Keeping T1's around is just as bad. If someone creates a userbox saying: "This user provides material support to Al-Qaeda", that should not be soft deleted. Also, for G4, if there are substantial changes, it does not apply anyway. G7/U1 should default to hard. If the user wants it soft deleted, he/she can do it themself, they won't use a template to go through process. Mr.Z-mantalk¢ 06:48, 22 May 2007 (UTC)

An example of why this won't work[edit]

Starslip Crisis was merged and redirected to another article, without protection or deleting the history. Yet, it has been put up on DRV not once, but twice. I see quite literally no benefit to soft deletion when merges get taken to DRV. Nifboy 00:48, 22 May 2007 (UTC)

Your example is a red herring. Benefits:
  • Potentially useful stuff can be reviewed by anyone.
  • Non admins can complete non-controversial deletion processes.
  • Soft deletion of borderline notable articles will likely be less controversial.
  • The conflict between needless deletion of content, discussion and contribution histories and GFDL (authorship information must be preserved) will be solved. --SmokeyJoe 07:31, 22 May 2007 (UTC)
I have seen examples of articles that were keepd more than once. There is no benefit to hard deletion either when articles get kept!  --LambiamTalk 06:06, 23 May 2007 (UTC)

Murky fate[edit]

Many users have varying opinions on the matter. The users for the proposal like the idea, but the proposal is so vague that they all present their own views. The users against the proposal either completely object to the idea or want a different process that will reach the same goal this does. It doesn't seem to be going anywhere at this point. There is also no true consensus on what the precise process is, because the description of the process is so vague. Sr13 05:02, 22 May 2007 (UTC)

Yes, the proposal is still a bit vague. Can someone suggest how exactly to soft delete an article? --SmokeyJoe 07:35, 22 May 2007 (UTC)
I have suggested a specific way of doing it, above at Wikipedia:Attic. I haven't seen objections against it from among those who appear to favour some form of soft deletion.
For a full-fledged proposal, next to the how, we also need to flesh out better when soft deletion would happen.
  • For deletion debates at AfD/CfD/MfD, this seems pretty simple: when the consensus is to soft-delete.
  • For PRODded articles, I'd suggest default userfication if there is a single clearly identifiable main editor, usually the creator, and otherwise atticization (as always, unless there is some common-sense reason to do otherwise). However, this is a minor detail.
  • For speedies, a start has been made above, which would appear to tend towards the conclusion that only a few speedy criteria merit soft deletion, of which A7: lack of notability (by far the largest group) may be the main one. (Again, the default form might be userfication if it is one article with a single main editor.)
 --LambiamTalk 08:23, 22 May 2007 (UTC)
Lambian, I pretty much agree with this. BTW SmokeyJoe, thanks for undoing the "rejected" tag. Radiant was premature. davidwr 09f9(talk) 04:11, 23 May 2007 (UTC)

May I (and others) conclude that it is fine if we (not just me, I should hope!) start modifying the proposal in the direction as indicated above, and try to reach a next more mature and semi-stable state, before we invite more discussion? (I know about WP:BOLD, but this will be quite a departure from the current version, so I'd like to check here first. It may of course still get shot down, but if so, it is better if it is clear that the reason is that people didn't like the idea, instead of perhaps only the mechanism.)  --LambiamTalk 06:02, 23 May 2007 (UTC)

Please, go for it. --SmokeyJoe 06:15, 23 May 2007 (UTC)
Go for it. The proposal in its current form has clearly been rejected, so a new proposal is warranted if the "proposed" tag is not to be replaced with a "rejected" tag. --Alvestrand 07:08, 23 May 2007 (UTC)
The current form is not clearly rejected, but it is quite drafty. What happened to the early proponents? --SmokeyJoe 07:58, 23 May 2007 (UTC)
The current form is clearly rejected, because there is obviously not consensual support for it (and many people don't like it, and the need for it hasn't been demonstrated). Please see WP:POL for details ("Consensus need not be fully opposed; if consensus is neutral on the issue and unlikely to improve, the proposal is likewise rejected.") >Radiant< 08:17, 23 May 2007 (UTC)
"consensus" is likely to improve when the proposal starts to take shape. Therefore, it shouldn't be rejected. You (for example) don't like it, but reasons given don't seem logical. Could you summarise your arguments for opposition? What do mean mean by "need"? --SmokeyJoe 08:30, 23 May 2007 (UTC)
  • I refer to comments earlier on this page by Doc Glasgow, Amarkov, Gurch, Woohookitty, PMC, Nifboy, Alvestrand, Cool blue, Z-man and myself. >Radiant< 08:39, 23 May 2007 (UTC)
    I am not sure what you are trying to say here. Alvestrand is clearly in favour. Most of the other people on your list have voiced objections that do not apply to this modification, such as objections against cross-namespace redirects, or creating an additional layer of process. You may be right that a modified proposal could not garner consensus either, but how can you be so sure? One of your main objections appears to be that the need has not been demonstrated. What form could a demonstration possibly take that would establish such a need to your satisfaction? I can't point at 100 deleted articles with salvageable contents, because I don't even know these articles ever existed. (And "need" is by no means the only argument.)  --LambiamTalk 20:06, 23 May 2007 (UTC)
      • While my original comment was aimed more toward opposing using this for certain speedy deletion cases, I really don't like the idea in general. If a non-admin needs to see the history of a page to see if they can improve it, history only undeletions can be done without a vote at DRV. The only thing I've seen here that I might support is the attic idea, but only with significant restrictions on when to use it. Mr.Z-mantalk¢ 20:16, 23 May 2007 (UTC)
        Could you tell me, if I don't know the page ever existed, so that I don't even know the page title and when it was deleted by whom, what is the proper procedure to ask for DRV?  --LambiamTalk 21:06, 23 May 2007 (UTC)

Anyone who likes transparent deletion can just start using it right now today. Friday (talk) 20:12, 23 May 2007 (UTC)

(a) Which ways of transparent deletion could I use right now? Blanking the page does not quite have the desired effect. (b) How do we get others to use a suitable form of transparent deletion? I'd be happy to give others access to the history of what I delete – in fact, I've never deleted any page and I'm not planning to, but the issue is that I want access to the history of pages others have deleted, if the content is not objectionable, just like I've access now to whole sections others have deleted – actually, even if the content was objectionable.  --LambiamTalk 21:06, 23 May 2007 (UTC)
Use {{deletedpage}} without actually deleting the page first. Not perfect, I admit, but getting any better requires software changes, I think. Friday (talk) 21:14, 23 May 2007 (UTC)
You mean, replacing the content by {{deleted page}}? If someone else did that, I'd consider it a vandal action, just like page blanking. As to software changes being required, I think no such changes are needed for the version I've suggested, and which several people appear to find acceptable: #Wikipedia:Attic. You haven't given a reaction to (b), which is the real issue.  --LambiamTalk 22:26, 23 May 2007 (UTC)
It's vandalism to just skip a step? I don't get it. If it's valid to delete it and replace it with that template, why's it a problem to not delete it first? As for b, I suppose that's the trick. It may well be easier to change the software than to change the habits of editors. Friday (talk) 04:59, 24 May 2007 (UTC)

New, bare-bones proposal: Userfy or attic PRODded articles[edit]

There doesn't seem to be a firm proposal for WP:SPEEDY and WP:AfD actions, but I think we are in agreement on WP:PROD.

Before I propose this formally, I want to see if it needs tweaking:

Proposal: Keep PRODded articles by Userfication or moving to the Attic

  • If a PRODded article has only one major non-anonymous contributor, WP:USERFY it. Otherwise, move it to a new Attic: namespace. In either case, provide a pointer to it in the deletion log.
  • The behavior of the new Attic namespace will need to be fleshed out. For example, will articles be WP:PROTected? Will search engines be kept out? Will Wikipedia's internal search offer the option of searching Attic? These and other issues can be hammered out later.

So, thumbs up? Thumbs down? Discussion? davidwr 09f9(talk) 00:00, 24 May 2007 (UTC)

Proposal, modified: PRODded articles may be kept by moving to the Attic, if there is any possibility of future use of the content, discussion or history, or if there is a possibility of past content now existing in other articles. This proposal does not affect Userfication, which is available on request.

  • Technical question: How can you ensure that links to the article subsequently appear red, and that a link to the Attic article can be found? --SmokeyJoe 01:14, 24 May 2007 (UTC)
I'm ok with the human-judgement on atticization to allow for PRODded articles to be simply deleted. However, I expect pure-deletes to be rare since almost everything that was truly that useless would qualify for SPEEDY. davidwr 09f9(talk) 03:40, 24 May 2007 (UTC)

I concur with the above proposal. I would suggest further that Attic Articles (AA) be unprotected and search engines blocked. Internal searches probably should occur -- I think -- otherwise how would a user know they are there? Perhaps add a banner to the article making clear what happened. We can work those out. And maybe add some of those speedies in too but very selectively. Let's see what others think. JodyB talk 02:07, 24 May 2007 (UTC)

What is the advantage of unprotection? Protection has the advantage that the article is frozen at the time of moving to the attic. Any user is free to copy the article to his own userspace or to make a new, improved article. Also, articles in the attic are unlikely to be watched, which makes them prime targets for vandals or people who might hijack an article to post copyrighted materials. By protecting them, it reduces the load on those of us who watch recent-article changes. I'm leaving speedies out of the discussion for the time being so we can come to a consensus on something. If everyone here agrees, I'm open to adding this to the proposal: "This new POLICY says PROposed Deletions SHOULD be userfied or moved to the attic unless there is a reason not to do so. Editors considering SPEEDY-deletions who plan on moving the article to the attic should consider using PROposed Deletion instead." davidwr 09f9(talk) 03:40, 24 May 2007 (UTC)
I can be happy either way about protection, but I think that the RC load would be manageable. After all the goal is to get the articles back in the mainspace and so we might as well keep an eye on them. I think Mr. Z-Man's comments below about a time limit is good and that keeps the numbers from getting unmanageable. However I'll not be displeased either way on protection. I certainly agree that we can keep speedies out of the process for now. I was just thinking ahead on that last comment. But for now keep it simple and keep it for prods only. JodyB talk 11:29, 24 May 2007 (UTC)

Why should the process used to delete determine whether to retain history? Those seem like separate issues to me. Friday (talk) 04:57, 24 May 2007 (UTC)

(e/c) I would suggest that we not userfy pages. Unless the creator is active, this may result in just throwing pages all over userspace where we'll never be able to keep track of them. I would also suggest a time limit for keeping articles in the attic. If they are not recreated in a better form and there is no activity on the attic-article talk page suggesting that creation is imminent, we hard delete the article after a certain time, 60 days perhaps. This is mainly to keep the size of the attic manageable. Category:Proposed deletion as of 21 May 2007 has 116 articles in it (I assume we won't move userpages to the attic). If that is about average, after 60 days we will have nearly 7,000 pages in the attic. If we have too many, the attic will become so crowded with junk articles that it will be hard to find the few decent ones in it. On a side note, how will we find pages in the attic? Will it use categories, some sort of index system, or just the search function? As for how we will know a page is been in the attic for 60+ days, if this is done as a new namespace, it might be able to be an added feature. Otherwise, it might be possible to have a bot that watches the move log and tags the article (or the talk page if the article is protected) with a dated template. Also, what about the mainspace page? Would it be left as a redirect to the attic or would it be deleted? Mr.Z-mantalk¢ 05:07, 24 May 2007 (UTC)

If a page kept in the Attic is protected and a user makes a copy to work on, then this is effectively a fork: the edit history is lost, potentially leading to a violation of the GFDL. This should be discouraged. The easiest solution is to leave the page unprotected and to ask editors who want to work on improvements to do it there. The state at the time of atticization can be retrieved from the page history; this could be facilitated by a time-stamped (substed) banner at the top of the page.  --LambiamTalk 13:27, 24 May 2007 (UTC)

I see no point in Attic articles being protected. The soft deleted version will exist in the history. Protection encourages forking, which is a bad thing. Protection is an administrative function, and thus complicates the process. I see no problem with articles being edited in the Attic, and I see this as preferable to an article being moved into someones userspace. No one owns things in the Attic, and no one is supposed to own articles. --SmokeyJoe 13:57, 24 May 2007 (UTC)
I can support the present proposal (I tweaked it slightly to make it clear what to delete; this is the only part that needs admin powers, I think.) I would also like the template for "there is no page here" to contain a link to the Attic/X page if it exists - that isn't mentioned now, and may need programming. --Alvestrand 20:40, 24 May 2007 (UTC)
It may be wiser not to propose features in this stage that require programming. There is already a link to the deletion log, which, if you follow it, should alert you to the existence of a deleted page, and moreover give a hint as to the reason for deletion.  --LambiamTalk 22:17, 24 May 2007 (UTC)
  • What problem is this trying to solve? By which I mean an actual problem rather than a hypothetical one. >Radiant< 09:49, 29 May 2007 (UTC)
    • One example. From User:Jimbo’s principle #1: “there must be no hierarchy or structure which gets in the way of this openness to newcomers” Unnecessary hard deletion is not “openness”. Not having access to deleted material is not openness. Having to “ask an admin” for userfication is an unnecessary hierarchy that gets in the way. --SmokeyJoe 08:13, 30 May 2007 (UTC)
      • WP:JIMBOSAID is a very poor argument. Not having access to deleted material is irrelevant to openness because if the material was worth looking at it wouldn't be deleted. I suggest you take some time doing New Page Patrol to get an idea of the amount of nonsense articles we actually get here on a daily basis. >Radiant< 09:15, 30 May 2007 (UTC)
        • The argument in favour is based on openness and transparency, which I think we agree is a good idea, barring numerable exceptions. Your point about the large proportion of nonsense, and other blatantly inappropriate material, is a good one. However, most such material is Speedily deleted, and I argue that Soft Deletion should in no way complicate CSD. I have browsed CSD tagged pages and not found stuff worth worrying about. At AfD, there is material that could be appropriately soft deleted, but I wouldn’t suggest complicating AfD before Soft Deletion is shown to be efficiently workable. ProD tagged articles, however, include a non-negligible proportion of potential developable ideas and reusable material, and I have seen nothing ProDed which I consider offensive, which warrants hard deletion. So, I see no disadvantage in soft deleting ProDed articles, and you and others have not suggested specific disadvantages. --SmokeyJoe 00:33, 31 May 2007 (UTC)


After skimming through 84 KB of discussion, I'm a little confused. I would appreciate it if someone would answer (even if very briefly) to the following two questions:

  1. What is the purpose of this proposed deletion outcome? If it is to give pages a second chance ... how effective is it likely to be if 100-200 or so pages are "soft deleted" every day? Won't the category housing all these pages quickly become bloated, much like Category:All articles lacking sources?
  2. How do we exercise some quality control? Having every page deleted via {{prod}} moved to the new namespace seems counterproductive given that much of what is deleted via {{prod}} is almost unquestionably not worthy of inclusion. Also, how far would we extend the criterion of "could potentially be improved"? Given that we operate in a limited-information environment (i.e., we can never be certain that there are no reliable sources about a subject), how would we avoid the situation where "could potentially be improved" is used to justify the retention of every article on the basis of "reliable sources might exist" or "reliable sources could become available in the future" (i.e., someone might write on the subject)?

If these issues have been covered in the discussions above and I've missed it, please direct me to the appropriate section. Thank you, Black Falcon (Talk) 06:36, 30 May 2007 (UTC)

Purpose: To enable any user to access the history and talk page of a deleted article, if there are no reasons (such as copyright) why that shouldn’t be allowed.

A user might like to do this for a particular article that they searched for, and then found deleted. I don’t see advantage in being able to easily peruse all soft deleted material. I would be happy to be able to find the material on an individual basis; it would be sufficient if the only link to the soft deleted article was to be found in the deletion log.

Quality control. Why do we need quality control? Offensive material shouldn’t be ProDed. Soft deleted articles can just sit there forever, just like the history of every page, worthy or not. I don’t see the Attic as a place for things potentially improvable for the same article. More likely the material can serve as a reference for what not to do again, for use in some other way, or for conversion to a redirect to an article that was not earlier realised to be a suitable target.

A major assumption here is that deleted articles are retained anyway, but are hidden from general users.

What possible problem is there with soft deleting every non-offensive article that has even the most remote possibility of potential improvement with the discovery of new material?

I see soft deletion as a small change with a small benefit. It might be a workable compromise between deletionism and inclusionism. What I can’t understand from its opponents is why they don’t like it. Is it a power thing that admins have over general users? --SmokeyJoe 08:45, 30 May 2007 (UTC)

  • A user might. Et cetera. You're assuming that this will actually be used to some benefit, whereas our experience with backlogs indicates that it almost certainly won't be. Furthermore, the "possible problem" is that this requires more work, more bookkeeping, and more complexity - all for a hypothetical and admittedly small benefit. m:instruction creep. >Radiant< 09:13, 30 May 2007 (UTC)

Yes, I am assuming that there will be some benefit, but Soft Deletion properly implemented would not create, or add to backlogs. Indeed, it should reduce them. If any editor can soft delete a ProDed article, then this reduces routine administrator workload. My technical question, “how can an ordinary user turn links to the soft deleted article red?” remains unanswered.

But I imagine a possible significant benefit: Soft Deletion might be an effective compromise between deletionism and inclusionism. If borderline worthy articles are soft deleted, then they are removed from main space, satisfying deletionists, but content is retrievable, satisfying inclusionists. If the deletionist is right, then the material will sit in the Attic, just as hard deleted material sits in hidden archives. If the inclusionist is right, then the material will be accessed, used or improved. In either case, there is no need for a contentious, principled argument right now, in an AfD debate. I have been studying AfD. It is a dreary place. I disagree with very few of the outcomes, and even then, not particularly strongly. It is a very important place, for it is effectively the enforcement chamber for questions of content in main space. Unfortunately, it is bloated with contentious unproductive debate over unimportant articles. If these relatively unimportant articles could be syphoned off by soft deletion, then this would be a very beneficial thing for AfD. So really, I see a benefit for AfD, if soft deletion can first be shown to be workable.

Soft deletion is a perennial proposal, as you say. The fact that it recurs suggests that there is a longstanding desire for it. Unfortunately, similar past proposals were prematurely abandoned before being thoroughly thrashed out.

As for instruction creep, this is not it. Soft deletion is the winding back of unnecessary hard deletion. If any part of Soft Deletion would create more work, more bookkeeping, or more complexity, then that part need fixing before implementation. --SmokeyJoe 00:36, 31 May 2007 (UTC)

As I understand it, soft deletion can be performed by any user with an account, just like any other move ... is that correct? If it is, then it removes the oversight that currently exists with proposed deletions. Any article that is proposed for deletion is viewed and considered by at least two people: (1) the person tagging it, and (2) the admin deleting it. It may also receive the attention of one or more proposed deletion patrollers. Under the system of "soft deletion", what is to prevent any user with account (which takes seconds to create) from speedily soft deleting an article that does not meet the speedy deletion criteria. Aside from the possibility of misuse or abuse by active editors, there is the potential problem of mass vandalism. It's true that any move can be reverted, but it involves three steps, two of which require admin involvement: (1) deleting the page in the article namespace, (2) moving the soft deleted page back to its proper location, and (3) deleting the cross-namespace redirect that is left over from the move. How, if at all, can this problem be addressed?
Another problem is that the namespace for soft deleted articles may become so bloated as to consitute a parallel article mainspace, at least when it comes to Google results and searching. Although AfD debates can become rather antagonistic when considered in terms of the dichotomy of "keep" or "delete" (although merging is often an option), I think it is an inevitable part of determining content inclusion and exclusion standards for an encyclopedia.
Finally, you note that this proposal should reduce backlogs, but I only think it will consolidate all of the present backlogs (unreferenced articles, articles with subjects of unclear importance, and so on) into one mega-backlog. You note three purposes for speedy deleting articles:
  1. To "serve as a reference for what not to do again" ... this is the purpose of policies and guidelines (such as Wikipedia:What Wikipedia is not and Wikipedia:Notability).
  2. "for use in some other way" ... this is rather vague
  3. "for conversion to a redirect to an article that was not earlier realised to be a suitable target" ... a redirect can be easily created at any time; I do not see how soft deletion would change that.
In short, even putting aside the technical issues, I do not see how soft deletion would be beneficial, except by making AfD less stressful, but in a way that I consider undesirable. -- Black Falcon (Talk) 01:19, 31 May 2007 (UTC)