Misplaced Pages

Revision as of 17:47, 25 May 2005

Miscellaneous

This policy is ridiculously weighted toward deletion, especially the stricture that undeleted pages need to be listed on Votes for Deletion again. --The Cunctator 15:32, 3 Dec 2003 (UTC)

• Presumably the idea behind putting an undeleted article back on VfD is to give a heads up to people that an article has been undeleted. Is this heads-up necessary? VfDers are likely to be monitoring unVfD too. Also if an article is undeleted does it re-appear on the watchlists of those who were watching it prior to deletion? If the answer to this last question is yes, then I support strongly a change in policy so that it is NOT necessary to relist and undeleted article on VfD. If the answer is no, the I still think a change in policy would be ok, but wonder if we can persuade a developer to introduce the feature (I noticed on User_talk:MyRedDice that User:Tim Starling that the undeletion code is a bit flake-y and needs beefing up anyhow. Pete/Pcb21 (talk) 15:55, 3 Dec 2003 (UTC)
• If you look at the number of votes on each you can see that Votes for undeletion has a fraction of the interest and traffic. I, for one, had no idea that this crap was going to be undeleted or I would have said something to that effect on the page. I suppose I trusted people on Votes for undeletion to be sensible. If they are not going to be then I will have to waste time picking up on people's silly little games. Secretlondon 16:01, Dec 3, 2003 (UTC)

I disagree that the policy is weighted towards deletion. In fact, it is quite the opposite. However, the policy had a tendency to allow for articles to ping pong back and forth, so I added a requirement to the undeletion policy that articles deleted via VfD need a notice on VfD that they are listed on VfU. I think that is a better method than Pcb21's proposal. I would be okay with removing the relisting requirement as well if and only if the notice of a VfU listing were required to stay on VfD for the entire 10 day period. I think it makes sense to have as many people involved as possible to avoid ping ponging of an article's status. Daniel Quinlan 16:26, Dec 3, 2003 (UTC)

If, say, I am interesting in working on the VfD page but not the VfU page, then I don't want the VfD page loaded up with VfU information. Alternatively, and more likely, I am interested in both VfD and VfU then I will watch both pages anyway. In either case repeating the content of VfU on VfD provides with no benefit. Is this a fair analysis? What are the wins from your proposal? Pete/Pcb21 (talk) 18:22, 3 Dec 2003 (UTC)

It's just a reference to the VfU page, like the following:

• undeletion proposal for Santorum on WP:VFU#Santorum, comment there only please

Daniel Quinlan 19:20, Dec 3, 2003 (UTC)

Undeleting selected revisions

Sysops now have the ability to undelete selected revisions of a page, which I have explained (correctly, I hope) on Misplaced Pages:Viewing and restoring deleted pages by sysops. This allows sysops to purge individual revisions from the history without bothering the developers, which can be a good thing (for example if an article started as a copyright violation but was completely rewritten before being deleted) and a bad thing (sysops can now hide stuff they don't like in such a way that nobody but another sysop can see it). I would like to open discussion on:

• Why we have this feature.
• When it should be used.
• What our policy on its use should be.

—Charles P.  13:14, 24 Dec 2004 (UTC)

I think this is highly likely to affect the way we deal with copyright violations, so I have added a note to Misplaced Pages talk:Copyright problems mentioning this (potential) discussion. --rbrwr 13:43, 24 Dec 2004 (UTC)

Could this also be used to remove personal info from things like Talk:Sollog history? Pakaran (ark a pan) 16:00, 25 Dec 2004 (UTC)

Removing personal information and removing copyvios are the only two things I can think of that this should be used for. I don't think it should be used to remove standard vandalism - and it certainly shouldn't be used to remove disputed content (of the sort that is so often wrongly described as "vandalism") -- sannse (talk) 16:03, 25 Dec 2004 (UTC)

Those are the uses I was seeing it as good for. -- Cyrius|✎ 21:51, 25 Dec 2004 (UTC)

I think we really need some policy on this, basically what sannse said. It is now possible for an admin to delete a page and then restore it with one or two revisions missing. It should be made very clear that this is not allowed at all besides for the reasons of copyvios and displaying personal informations (which should only be done if requested IMHO). --Conti|✉ 22:29, Dec 25, 2004 (UTC)

Yes, I agree with Conti that personal information should only be removed on request. A question to consider: is it OK for an admin to act on their own behalf if their own details need removing? Or should they ask an uninvolved admin to do the work? (I'm thinking of the general principle of not acting to block someone you are in dispute with, protect an article you are involved in, etc.) -- sannse (talk) 19:43, 27 Dec 2004 (UTC)

Yes, I think this is a perfectly reasonable restriction well in line with the principle that one isn't supposed to use sysop powers in disputes where one is involved. —Charles P.  22:17, 27 Dec 2004 (UTC)

Some notes on GFDL and this feature, and how this feature should be made use of, in my opinion.

GFDL require proper preservation of revision history of the document. If certain revisions are selectively undeleted, it would look like as if some people never contributed to the article.

The basic principle to follow is "if you keep certain revision undeleted, there should not be any copyrightable contribution by that revision's author."

Some illustrations:

Case 1) Revison 1 (no problem) > Revision 2 (copyvio by User:A) > Revision 3 (copyvio remains, additional contribution made by User:B) > Current version (copyvio removed by User:C).

In this case, in order to remove copyvio from the past revisions, both Revision 2 & 3 have to be removed. But if only Revision 1 and Current version are undeleted, then the contribution made by User:B would look like as if made by User:C. And that is a problem in terms of GFDL. The simple solution in this case, then, is to undelete only the Revision 1, giving up all later Revisions.

Case 2) Revision 1 (no problem) > Revision 2 (copyvio by User:A) > Revision 3 (copyvio removed and additional contribution made by User:B) > Current version (additional contribution made by User:C)

In this case, undeleting Revisions 1, 3 and the Current version is fine.

Case 3) Revision 1 (no problem) > Revision 2 (copyvio and original contribution by User:A) > Revision 3 (copyvio removed by User:B) > Current version (additional contribution by User:C)

In this case, User:A's original contribution is still in the current version, and if you undelete everything but the Revision 2, it would look as if the original contribution made by User:A is made by User:B. Again, this is problematic. The simple solution in this case is to give up everything and Undelete only Revision 1. A better solution, which is not always possible, is to ask User:C to make another edit and remove the original contribution made by User:A, and undelete that latest version along with Revision 1. This way, none of User:A's contribution is reflected in any of the current or past revisions, and therefore User:A does not have to remain in revision history.

Major implication of the relation between this feature and GFDL's history requirement is that "if you find any copyvio, revert to a version before it first happened, do not simply remove infringing part." "Undelete the revisions before copyvio has happened & any versions after the reversion." That way, any contribution made afterwards can safely be undeleted later.

However, there is a tricky exception about this. Think about the following case:

Case 4) Revision 1 (no problem) > Revision 2 (copyvio and original contribution by User:A) > Revision 3 (additional contribution by User:B) > Revision 4 (Reverted to Revision 1 by User:C) > Current version (Non-infringing contributions of User:A and User:B are re-inserted by User:D)

In this case, if Revision 1 (the revision before copyvio has happened) & Current version (a revision after the reversion) are undeleted, it would look as if User:D has made all the contributions that he simply re-inserted. User:A and User:B's names would not be shown in the revision history. Again, this is problematic.

My opinion is mostly formed through discussions at Japanese Misplaced Pages, where I am most active. This feature is developed by a Japanese Wikipedian (Tietew) mainly due to the fact Japanese Misplaced Pages accumulated over 1,000 revisions to be reviewed and selectively deleted/undeleted to remove copyvio and other legally problematic contents. (We asked developers' help to remove specific revisions, but no case was resolved over a year for various reasons, and the issue became a big concern among some Japanese Wikipedians.) Relevant discussions are available at ja:Misplaced Pages‐ノート:著作権/特定版削除向けrevert , ja:Misplaced Pages‐ノート:削除の方針#.E7.89.B9.E5.AE.9A.E3.81.AE.E7.89.88.E3.81.AE.E5.89.8A.E9.99.A4.E3.81.AE.E6.89.8B.E9.A0.86, and ja:Misplaced Pages‐ノート:特定の版の削除.

Tomos 23:07, 27 Dec 2004 (UTC)

That's food for thought. If we want to keep good contributions that are made at the same time as the copyvio or at any time up to the removal of the copyvio, we have to leave the copyvio in the history, despite this new sysop power. Drat.

Anyway, I have a simpler case that I'd like to try this function out on: the copyvio (albeit quite a small one, and marked as "work in progress") was the first version of the article and was properly removed and replaced with a copyvio notice. A second user started a /Temp version, which the anon IP which had done the original copyvio then cut-and-pasted over the copyvio notice, and continued to develop. I would like to merge the histories of the two articles (as per the usual procedure for fixing cut-and-paste moves), and it would make sense to leave the copyvio and the copyvio notice deleted. Would this be controversial? See the histories of Hanna Sheehy-Skeffington and Hanna Sheehy-Skeffington/Temp for the case in question. --rbrwr 17:42, 31 Dec 2004 (UTC)

Copyvio test case

A few days ago I was working on some of the older things on WP:CP, and I came across Alfred Wegener, which was listed due to having copyvio added into the existing article. Using the new powers, I deleted three revisions.

1. the addition of the copyright violation
2. {{copyvio}}
3. revert to last clean revision.

They were all contiguous, no other edits were made between the initial addition and the revert. Any opinions? -- Cyrius|✎ 03:59, 1 Jan 2005 (UTC)

Hmm, maybe admins should be supposed to list such partial deletions somewhere. I know that it's sort of an instruction creep, but on the other hand it's not so easy to follow the actions of multiple deletions and undeletions of a page. I fear that this process could be (ab)used without (almost) anyone being able to reconstruct what exactly happend. --Conti|✉

RFC

Does this page still need to be listed at RFC? Maurreen 06:50, 13 Jan 2005 (UTC)

I think the conversation is dead. The lessons learned shall be integrated into the policy, and it shall be delisted from RFC. So sayeth I, and so shall it be. —Ben Brockert (42) 03:40, Jan 14, 2005 (UTC)

Proposed revision to "Exceptions" clause

Recent experience has illustrated to me that the Exception rule is flawed. In particular, it appears that there is an expectation that the VfU must last five days, even if the deleting sysop states that the deletion was not in accordance with policy. The policy as it currently stands states:

If the page was obviously deleted "out of process" (i.e. not in accordance with current deletion policy), then a sysop may choose to undelete immediately. In such a case, the sysop who deleted the page should be informed of the undeletion and the reason for it. The page then should be listed on Misplaced Pages:Votes for deletion. If you are unclear as to whether it was out of process or not, the normal undeletion procedure should apply.

I would propose the following revision to the policy of exceptions.

If the page was obviously deleted "out of process" (i.e. not in accordance with current deletion policy) and a Vote for Undeletion has not been started, then a sysop may choose to undelete immediately. In such a case, the sysop who deleted the page should be informed of the undeletion and the reason for it. The page then should be listed on Misplaced Pages:Votes for deletion.

If you are unclear as to whether it was out of process or not, the normal undeletion procedure should apply. All articles being voted on at VfU must remain for the whole five days, regardless of the apparent merits of the case.

Support

1. Sjakkalle 09:06, 27 Apr 2005 (UTC)

Object

1. It is generally an informal process. The best method for normal users to let and admin know that there might have been an out-of-process deletion is to list it on VfU. Making it so that listing something on VfU prevents its immediate undeletion is counterproductive. —Ben Brockert (42) 23:30, Apr 26, 2005 (UTC)

It seems that this proposal actually says the opposite: that an article may be immediately undeleted, but the discussion must remain for the full duration.

Comment

Information changed

In my view the undeletion policy should be revised to reflect that a possible reason for undeletion can be that a subject or individual has become more relevant or significant than at the time of his/its deletion. --Daniel C. Boyer 19:11, 27 Apr 2005 (UTC)

That seems more like a reason to create a new article. The old one would be obsolete. —Ben Brockert (42) 21:54, May 13, 2005 (UTC)