Wikipedia talk:Protection policy/Archive 4

This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Since when has indefinite semi-protection been allowed? I thought semi-protection was supposed to be temporary. --ALL IN 02:49, 13 April 2007 (UTC)

Maybe, but unprotected George W. Bush is nothing short of foolish. And Jimbo himself said that indefinite semi-protection could be used on biographies subject to drive-by nonsense, and he's a big promoter of the "anyone can edit" philosophy. You're right about the userpages though, they shouldn't be indefintely protected, but very few people seem to agree. --Deskana (fry that thing!) 03:20, 13 April 2007 (UTC)

I just had added Penis as another example of an indef semi-protected article, but rememberd that it has been briefly unprotected a few times. It seems that articles like that that are blatant vandal targets should be. Why don't we have a list of permanently protected pages, so this can be debated? The way, the truth, and the light 11:52, 26 April 2007 (UTC)

'Indefinitely protected' isn't a formal classification for an article (like say,

Steel

17:14, 26 April 2007 (UTC)

Yes, I was suggesting that maybe it should be. It certainly wouldn't hurt to avoid the disruption cause by unprotecting problem articles, while articles that usually aren't would not have to be protected any longer than they need to be. I'm confused why you would 'do away with' the term, indefinitely protected pages exist whether that's formal policy or not. The way, the truth, and the light 18:26, 26 April 2007 (UTC)

See above. The term implies that 'indefinitely protected' is a class of articles like featured articles or good articles or whatever, when in reality it's simply a protection that no-one feels the need to lift. And this system works well, I don't really see the need for a bureaucratic approvals process to designate articles 'indefinitely protected class'; indeed, this is a textbook example of

Steel

19:16, 26 April 2007 (UTC)

I have a concern that this may contravene the principle of "anyone can edit". I understand the need to reduce vandalism, but the fact is, many of Wikipedia's most popular pages are semi-protected. If the most popular articles can't be edited by anons, we're moving farther and farther away from our values. I feel that we're moving to a Wikipedia where any page that is controversial or popular enough to cause serious dispute is closed to anonomyous editing. A possible solution would be for anonomyous editors to put their edits up for approval, and have the approval threshold be very low, namely "Is this a good-faith edit?" If the approver disagrees with the content of the edit, they can just approve it and then revert it, which is (hopefully) what they would do if a logged in user made the same edit. This seems like the best way to allow all editors to have some influence while preventing vandalism. By the way, I do have an account, but I'm making this comment anonymously to accentuate my point. 76.84.5.138 18:15, 16 August 2007 (UTC)

{{

R!

) 21:04, 16 August 2007 (UTC)

The proposal is at

WP:FLAGGED. I think this should be put in place on a trial basis to assess its utility, but I'm fairly skeptical that it would be any improvement. The two problems are that a lot of vandals either won't know or care that their work is not seen by the public (and it's still seen by the reviewers), and the amount of work involved in keeping on top of everything and avoiding backlogs. Richard001

01:29, 17 August 2007 (UTC)

Talk page protection

Currently

23:14, 16 April 2007 (UTC)

The talk page was protected because people were posting libel on it. I don't think there is an easy solution to this, if a different page is provided for new editors people would just post libel there instead. –

Steel

23:21, 16 April 2007 (UTC)

A solution would be useful, since as it stands now, it appears as though we are actively discouraging participation from people drawn by current events. Zyxwvutsrqp 23:34, 16 April 2007 (UTC)

I'm all ears if you have a solution. –

Steel

23:35, 16 April 2007 (UTC)

How about a policy where the talk page cannot be protected? People come to Wikipedia to view the article, as the article is where the actual information is. Eliminating the possiblity of discourse on the topic by protecting the talk page could have such a chilling effect that the credibility of Wikipedia could be at stake. The talk page is *by definition* not an assertion of truth. Libel, BY DEFINITION, requires that the matter be asserted as truth. No reasonable person would believe that things being discussed in a talk page are assertions of truth. I don't necessarily disagree with protecting the article itself, but the talk page should never be protected. Rooot 20:18, 17 April 2007 (UTC)

This is moot. It's unprotected now anyway. –

Steel

20:21, 17 April 2007 (UTC)

This is not moot because the possibility of this happening again still exists. The issue is greater than one specific incident. Rooot 20:25, 17 April 2007 (UTC)

Talk pages will generally be protected or sprotected in serious situations (like the talk page "rumours" issues). We can't outlaw it because we might need it just in case. -

) 21:53, 18 April 2007 (UTC)

SOMEBODY ERASE THE sack by charles woodson. He never sacked brady during the "tuck game", it was an incomplete pass. Erase the part where it says fellow wolverine charles woodson made the sack.

it is not our duty to make history, but to write history. to write to history and to decide what happened really, historian make researchs and the result must be depended on the real documents not the fake ones secondly citations and documents must not be distorted. as an alternative source of knowledge, wikipedia can share any knowledge that is a fact. however, so called armenian genocide claim has many contradictions. and it has not been proven yet. moreover, the fakeness of the documents that were seen as the proofs of the genocide and the distortion of the documents and the knowledge has been shown on this page. ( http://www.ttk.org.tr/index.php?Page=Sayfa&No=90 ) if you have any difficulties understanding the content of the page, i will be pleased to help. to be fair and to help justice to be served wikipedia must share this information on the web. otherwise, it would be shameful for you. with my best regards

umut azak [email protected] —The preceding

) 17:16, 17 April 2007 (UTC).

It's highly unlikely anyone here can read Turkish, or that they would believe the claims Turks make (who mysteriously are the only ones who deny the genocide). This also has nothing to do with protection policy, which is about protecting pages. Richard001 00:32, 5 August 2007 (UTC)

I know this is strange but why did they change the lock color from gold to silver? Personally, I think gold is more appealing and easier to see. —The preceding

) 21:48, 17 April 2007 (UTC).

true that —Preceding unsigned comment added by 89.148.56.167 (talk) 21:00, 1 September 2007 (UTC)

Special:Protectedpages has been updated, and now include many options for sorting and filtering the list, admins may be able to use this to help resolved improper or forgotten protections now. — xaosflux ^Talk 02:20, 20 April 2007 (UTC)

Finally! A real replacement for

Review Me!

) 02:23, 20 April 2007 (UTC)

I just got 30-40 really old pages unlocked. Also helpful to go and set expirys on pages that are currently set for indef. — xaosflux ^Talk 02:55, 20 April 2007 (UTC)

I have some protection script[1] to help as well.

Voice-of-All

19:18, 21 April 2007 (UTC)

Famous play subject to disfiguring vandalism daily by middle school kids. Please semi-protect.

The little icon on the right corner is too subtle and it is confusing new users or anons. Many anons who don't know why it is protected may ask "Where did that edit button go" since the "edit button" will be replaced by "view source". I think that there needs to be at least some text that at least tells the user that this page is protected and to link to a page that shows the reason why. Maybe a small note beside the icon saying "Page protected see (link to reason) for details" or something. -- _Hdt83 ^Chat 22:56, 20 April 2007 (UTC)

If I recall correctly, the view source page has wording to that effect. –

Steel

17:08, 26 April 2007 (UTC)

Of course, many people won't know what the "view source" button means... Perhaps the words should be changed to "edit this page" but still link to the same page? -- _Hdt83 ^Chat 05:03, 4 May 2007 (UTC)

I agree. I work a lot with computers and after a while I figured out what "View Source" means. However, it would be confusing to many people. I suggest it be replaced with something to the effect of "Page Protected" or "Protected - View Coding" or "AntiVandal Protected". Something that would convey what needs to be understood. Also, "Fully Protected" and "Semi-Protected" and "Temporarily Protected". Maybe "Page Locked" or "Anti-Vandal Lock". There are all kinds of things you could put there. View Source is technical lingo, and should not be used. SadanYagci 18:06, 25 August 2007 (UTC)

As I write, Wikipedia:No personal attacks is locked in a version which includes a contested change. Therefore, the contested version is now policy, in spite of the fact that the change in question has been disputed for over two weeks with no resolution in sight. The disclaimer simply doesn't work in this case.

Policy pages which are locked for edit warring need to be locked as they were before the edit war began. Mangoe 14:50, 27 April 2007 (UTC)

Agreed; unless protected policy pages lose their status while locked, including something disputed in a policy makes little sense. -Amarkov moo! 14:53, 27 April 2007 (UTC)

This may be a bit too idealistic to implement properly. I can see someone reverting then protecting, then another admin reverting the reversion because they see that as indicating a preference. --Deskana (fry that thing!) 15:42, 27 April 2007 (UTC)

No comment on the first half, but the 'locked on the before the edit war' version is suggested here periodically and would be a nightmare to impliment properly, not least because it's not always clear which version is the 'before the edit war' version. A comment I made a few sections up seems relevant here as well: I'm not convinced by the "protect the version prior to the edit war" suggestion. This will encourage people to whinge and wikilawyer over which version is the 'before the edit war' version. A much better use of time and energy would be to discuss the relative merits of the versions and resolve the dispute. (Just quoting myself, not accusing anyone in the WP:NPA war of anything)
Perhaps part of the protection policy should be changed with regards to protected policies, it's not the worst suggestion anyone has ever made, but it'll need to be thought through. –

Steel

15:52, 27 April 2007 (UTC)

That is a new dilemma. I agree with what Steel said though. It's best to discuss the changes on that talk page and get the dispute resolved than try to wikilawyer the

Review Me!

) 02:50, 30 April 2007 (UTC)

I would support that. Actually, I expect it's possible to have the {{

Steel

02:54, 30 April 2007 (UTC)

I assume there's a way, but I have no idea (I'm not the best person when it comes to way complicated template syntax). -

Review Me!

) 03:26, 30 April 2007 (UTC)

I'll see if I can get it working, otherwise we can wheel in a template whiz. The wording still needs to be decided upon, at any rate. –

Steel

15:57, 30 April 2007 (UTC)

• If it turns out that a policy page is protected in a version that obviously misrepresents policy, you should ask a neutral admin to fix that, using {{
  editprotected}}. In the cases where it's not obvious, it's probably not that big a deal to leave it in the "wrong" version for a few days. >Radiant<
  09:49, 1 May 2007 (UTC)

Semiprotection in response to continual low-level vandalism?

When there's an article which has high visibility to people unfamiliar with Wikipedia - like an article on a cartoon fandom, for example - and lots of people are amazed by the ability to edit the article and it ends up with amateurish edits or vandalisms a few times a day, is this a suitable case for indefinite semiprotection? I wouldn't think so, because the policy specifically points out heavy vandalism, and a few vandalisms a day are easy enough to keep up with, though it really is a pain to keep having to edit them out. I'd like to see this situation mentioned in the policy, and for it to clearly state whether or not indefinite semiprotection is an acceptable solution. Right now I'm seeing a large number of articles with indefinite semiprotection in response to really light vandalism. - Brian Kendig 03:13, 4 May 2007 (UTC)

Yes, the concept of indefinite semi-protection is not really defined, and therefore there are no standards for it, other than 'consensus' of admins. I guess you would support my idea that there be an organized list of indefinite-protected articles. The way, the truth, and the light 04:13, 4 May 2007 (UTC)

In cases of light vandalism, just reverting and warning the vandals (read: warning is a very important step) is good enough. If it's the same people then they'll get tired of getting blocked (if they're normal). -

Review Me!

) 04:06, 4 May 2007 (UTC)

That's how you deal with individual vandals. However, many articles are continually vandalized by different people, and that's when the question of semi-protection arises. The way, the truth, and the light 04:13, 4 May 2007 (UTC)

Gasp! Our pages are editable by every random joe that walks by! Every page receives vandalism. Random drive by has no solution, except revert, warn, block. Read

Review Me!

) 04:16, 4 May 2007 (UTC)

I know that. But clearly there's some point at which semi-protection is appropriate; that's already been decided. And a few articles are permanently protected, or nearly so. My suggestion, which the original poster seems to agree with, is that this be formalized. The way, the truth, and the light 04:21, 4 May 2007 (UTC)

• I think the page already says that. "Indefinite semi-protection may be used for ... articles subject to heavy and continued vandalism". >Radiant< 09:08, 4 May 2007 (UTC)

I'm talking about articles that are targets of moderate and continued vandalism, by a wide range of anonymous newbies. The article I'm particularly thinking of as an example is Virtual Magic Kingdom, where different kids keep wandering in and editing it to say "come find me in the game!" or "here's a list of all the events this week!" or deleting sections or slipping in four-letter words or doing any of the numerous things that newbies do that aren't appropriate for Wikipedia articles. If you'll look at the article's history, you'll see that the vast majority of edits are of this type or reverts of them; and since there are many, many more newbies than experienced editors looking at this article, there have been some obscene edits left up there for days, and some vandalism has gone unnoticed for weeks. I would really like to put this under indefinite semiprotection, but the vandalism can't be called "heavy". And if semiprotection isn't appropriate for this article, then it definitely isn't appropriate for articles I see under semiprotection because they've gotten a couple of vandalisms over a couple of days. I just want the policy to be more clear in this area. - Brian Kendig 11:27, 4 May 2007 (UTC)

I'd like to see the group come up with a bit clearer definition - and include it in the policy - of just what "persistent" vandalism consists of. What is persistent and moderate to one person may not be so to another. For instance, 1986 averages 3 to 6 anon IP vandalism acts a day, almost every day. I, personally, would consider this persistent vandalism, and since most of the vandalism is by IPs, sprotection would certainly help in this case. I've gone over the archived discussions here, and cannot find anywhere where we've tried to actually quantify the rather vague standard we have now. I may be all wet, but I think it's time to reduce this policy's vagueness. Therefore, as an attempt to generate discussion on actual thresholds, what would the community think about defining "persistent" vandalism as "an average of three vandalism edits by anon users per day for a period of two weeks or more"? Akradecki 04:20, 5 May 2007 (UTC)

I think that such a low number would result in semiprotection becoming commonplace here on Wikipedia, and I don't think that would be a good thing. I'd suggest a higher bar, something like "ten vandalism edits on most days over a month-long period." - Brian Kendig 16:53, 5 May 2007 (UTC)

It's lucky for everyone then that we have admins who have gone through RFA and are allowed to use their own good judgement to decide a)to protect or not and b)for how long. The fact that it isn't so set in stone is a good thing, because then we can be flexible. As for

Review Me!

) 18:41, 5 May 2007 (UTC)

The problem with giving that article "a couple weeks of sprotection" is that this doesn't address the underlying problem - random newbies vandalizing the article for the fun of it - and that's going to pick right up again when the sprotection goes away. You're correct that the policy is good to be loosely defined, but I believe it's too loosely defined when there are no guidelines for how much vandalism deserves permanent sprotection. - Brian Kendig 20:07, 5 May 2007 (UTC)

What we clearly need here is some kind of

Steel

19:24, 5 May 2007 (UTC)

That's a really good idea, actually - having a procedure in place for deciding what articles deserve more-or-less permanent semiprotction. It would make the decision depend less on the opinion of a single editor. - Brian Kendig 20:07, 5 May 2007 (UTC)

You get a

Steel

23:40, 5 May 2007 (UTC)

I like that that comparison there. But I disagree that sometimes it is based on the opinion of 1 editor (for example, if I come across an article that has massive recent vandalism, I might protect it). But everything is over-turnable too. For example, if someone declines to protect on RPP, then another admin might come along and protect if they think it needs to, and I don't really care if someone overrules me on something like that. -

Review Me!

) 00:32, 6 May 2007 (UTC)

When I said "that's a really good idea," I wasn't referring to that silly bit with all the percentages - I only meant your idea about setting up Wikipedia:Protections for discussion and Wikipedia:Protection review. Right now, if I un-semiprotect a page because I don't believe that the vandalism on it qualifies as "heavy", and somebody else re-semiprotects it because he does feel that the vandalism on it qualifies as 'heavy', it would be nice to have a forum to use to come to a consistent decision, and to see other similar cases all in one place. - Brian Kendig 15:21, 7 May 2007 (UTC)

Each others' talk pages, or if you really can't resolve it, ANI. Really, does this occur so often that we need yet another process for it? –

Steel

15:23, 7 May 2007 (UTC)

I suppose not. All I'm really looking for are some better guidelines for what constitutes the sort of "heavy vandalism" which merits semiprotection. It seems that a lot of articles are semiprotected after a short period of light vandalism, and I'd like to see the guideline be clearer about this. - Brian Kendig 18:02, 7 May 2007 (UTC)

Case in point: Geometry

This conversation is relevant to a discussion at the Math WikiProject right now. User:Arcfrk made a strong argument for protecting Geometry. Of course there are also the foundation principes to consider. I would appreciate hearing others' thoughts on the matter. CMummert · talk 13:30, 11 May 2007 (UTC)

Given that history, I would semiprotect it for a couple weeks. I don't see where it contradicts "current admin practices", because there are no unified admin practices regarding protection (I mean, there's no formula, no requirements than need to be met for each and every case. Each admin may evaluate in a totally different way). The protection log is also clean, so it didn't just come off protection or anything. So I would say s-p is a good band-aid for a little while at least. -

Review Me!

) 20:57, 11 May 2007 (UTC)

There are two ways of approaching vandalism protections. Either we look at the rate of vandalism and ask ourselves whether it's too high to expect volunteer RC patrol and watchlisters to revert it (Eg,

Steel

21:10, 11 May 2007 (UTC)

Thanks for the comments. The only reason I hesitate to protect the article is that the reason to protect it (its high profile name) will never go away, and so the protection would be justified indefinitely. And I'm sure there are a lot of other articles in the same situation; it's hard to protect one and then refuse to protect the others. CMummert · talk 22:56, 11 May 2007 (UTC)

You've got to at least try unprotection for some articles (heck, even Bush gets unprotected a couple times a year). -

Review Me!

) 04:13, 12 May 2007 (UTC)

... or we look at the article itself and ask whether an anon activity in general is likely to improve it or not. However, that's not a criteria for semiprotection - as the guideline says, "Semi-protection should not be used to prohibit anonymous editing in general." - Brian Kendig 04:36, 12 May 2007 (UTC)

I know that; perhaps I wasn't clear. The first approach is current practice and policy, the second isn't, but en.wikipedia needs to at least consider shifting in that direction. –

Steel

16:03, 12 May 2007 (UTC)

Or we should shift in the direction that we use semiprotection to dissuade some people from making repeated vandal edits (the same reason for blocking really). It's kind of like a "smarten up". -

Review Me!

) 19:59, 12 May 2007 (UTC)

That sentence in the policy page ought to be reworded - the entire point of semiprotecting a page is to prevent anonymous editing in general. The alternative is to block IPs one at a time. C Mummert · talk 20:45, 12 May 2007 (UTC)

To temporarily prevent anonymous (and non-autoconfirmed) editing. -

Review Me!

) 21:08, 12 May 2007 (UTC)

The distinction is that you don't want to say, "I don't think anons should be allowed to edit this article." That goes against the spirit of Wikipedia. - Brian Kendig 21:30, 12 May 2007 (UTC)

Although the change to policy because of this talk looks good, I fear that it may lead to more wikilawyering about it (who else but the protecting admin can say what the "sole purpose" was). -

) 02:28, 13 May 2007 (UTC)

I found it very difficult to find a good phrasing for that bullet. If anyone else can think of a better way to express it, please get rid of my phrasing. Anyway, I think admins are less prone to wikilawyering than is generally feared, especially about admin actions.C Mummert · talk 03:18, 13 May 2007 (UTC)

I meant some users may wikilawyer about an admin decision to protect saying that they're intent was to prohibit anon editing or something. -

Review Me!

) 05:30, 13 May 2007 (UTC)

One of the main problems with the protection policy is that some admins focus on the number of vandalism events per unit day. But this is not the critical measure. It should be the fraction of vandalism edits per total number of recent edits. This is because it does not matter how fast the vandalism arrises on a page when you are trying to undo the edits. It just matters how many of them there are mixed in with the productive edits. So I think we could gain a lot of ground by explicitly mentioning in the policy that the rate of vandalism shall be measured per unit edits when we are determining if a page has high low or medium vandalism.

Ttguy 10:25, 16 May 2007 (UTC)

Our current policy says:

Articles subject to heavy and continued vandalism, such as George W. Bush.

I hope we can work out some guidelines for this. I recently indefinitely protected

. The elated response I got from the regular Batman editors makes me think that I was doing the right thing. I've gone over this page looking for some guidelines for making decisions about when to add semi-protection and for how long and found none. I have noticed the following about the three pages I semi-protected, and perhaps this can be the start of some criteria for adding semi-protection:

• The vandalism is often of an offensive nature on high profile pages. This has a detrimental effect on how Wikipedia looks to the public.
• Almost all, if not all of the vandalism is coming from anons. Semi-protecting would effectively solve the vandalism problem.
• Anons are making very few contributions to the article compared to the amount of vandalism coming from anons. Blocking anons would not have a large effect on discouraging their positive contributions.
• There are regularly many new vandals, therefore it would be a huge task to notify and warn all the vandals. This seems to be too large and too dreary a task to expect anyone to undertake.
• The proportion of constructive edits to nonconstructive edits (vandalism and the reverts) by all editors is very low. This means that the regular editors are devoting most of their effort to reverting vandalism.

These seems like a good set of criteria for semi-protection. In addition, I've noticed the following:

• All of the pages have been semi-protected previously. This shows that the problem is on-going, and that temporary semi-protection did not help.
• Vandalism continues very shortly after semi-protection is removed. This shows that the page is a popular target for random vandalism.
• The vandalism is not related to a current event (I'd have to check this for Spider-man). If it were related to a current event the semi-protection could be lifted once the event was out of the public eye, and perhaps it would no longer be needed.

These seem like some good criteria for making the semi-protection indefinite.

What do people think about this? Should these criteria be quantified? I'm new to the semi-protection policy discussion. I'd like to hear from others who have indefinitely added semi-protection, and what criteria they have used to decide. -- Samuel Wantman 01:22, 14 May 2007 (UTC)

I feel that this part of the policy should be as loose as possible to allow for the greatest amount of flexibility by admins. Although several of those criteria that you listed are already de facto when it comes to indefinitely semiprotecting pages. Although I'd like to respond to a couple of them.

The vandalism is often of an offensive nature on high profile pages Not something new or alarming. This happens on all pages really

Almost all, if not all of the vandalism is coming from anons and Anons are making very few contributions to the article compared to the amount of vandalism coming from anons This has already been studied (not vandalism studies, but other ones), and it's already widely known. But it's a Foundation issues (#2)

There are regularly many new vandals, therefore it would be a huge task to notify and warn all the vandals There are many JS's to quickly warn vandals, and there will always be new vandals here. And sometimes I think that people need to be less lazy about it. Sprotection isn't the end all-be all of answers. -

Review Me!

) 01:40, 14 May 2007 (UTC)

Perhaps, some day, after too many pages are semi-protected there will come a cry from the vandal patrollers, "Please unprotect some more pages so we can track down more vandals!" I don't see this as being very likely. The reality is that we have the ability to semi-protect, and admins are using that ability to semi-protect articles. The point of having guidelines is to describe what the current practice is as it evolves. By making it explicit we can discuss the pros and cons of the practice. Are you advocating not having guidelines? If you think there should be guidelines, what do you think they should be? -- Samuel Wantman 01:52, 14 May 2007 (UTC)

Are you advocating not having guidelines? Am I an anarchist now? But really, there's no point in trying to create guidelines for when to permanently protect pages. You'd have to throw in over a hundred "if" situations. Usually permeant sprotection comes about after discussion by admins/editors, or sometimes an admin will just declare when semiprotecting a page "don't unprotect it, it doesn't work", and other admins come around, see the history, and agree. I think we need to trust our admins a little more. We're not perfect, but that's what talk pages are for. -

Review Me!

) 22:28, 14 May 2007 (UTC)

I'm an admin, and I've spelled out the criteria I am using. I want to hear from others about the criteria they use. If there is a consensus about the criteria, we can write a guideline. This doesn't mean that it is rigid, inflexible and incapable of changing. What it would mean is that each admin would not have to reinvent the wheel from scratch. I want to hear from the collective wisdom of the community about this subject. I also don't think it is a good idea for each admin to have radically different criteria for protecting pages. If that were the case, people would shop for a hard-ass protector when they want a page protected and a lenient one when they want it unprotected. A scenario like this sounds like a recipe for frustration. -- Samuel Wantman 23:53, 14 May 2007 (UTC)

Nice to see some discussion of this. Not having guidelines is very frustrating, not only for unsure or inexperienced admins, but especially for editors who want to request protection. They come here, but they get no indication of whether their article is appropriate for protection or not. Additionally, there is inconsistency among administrators - I've had some admins protect some pages with continued but low levels of vandalism, while others have stated they would expect at least 15-20 vandal edits a day before they would consider protection. Besides the number of vandal edits, two other important factors could be considered if proposing guidelines (even just very rough ones) - the percentage of time the article remains vandalized (this is important for reader perception - it isn't quite so bad if the vandalism is cleaned up quickly), and the percentage of vandalism vs. contributive edits (an article that has had little addition for a month is a better candidate than one that is vandalized, but also improved regularly).

I think it would be a good compromise to provide some rough guidelines, pointing out that it isn't simply one factor that is taken into consideration, and making it clear that there is no concrete rule and that administrators will use their own personal judgment as well. Richard001 09:21, 14 May 2007 (UTC)

See related discussion at

Steel

14:02, 14 May 2007 (UTC)

I, for one, like Sam's first attempt at some guidelines. As I've said before, I believe we should have them. Maybe one other thing to consider, at least loosely, is a ratio of constructive vs. vandalism edits from IPs on the subject article. If there's evidence that it's the type of article that IPs regularly edit constructively, then there's a good case for keeping the protections as temporary as possible. On the other hand, if a look back over the previous month or two shows that there's a lot of vandal IP edits and almost no constructive ones, that's a good argument in the other direction. Akradecki 00:10, 15 May 2007 (UTC)

Well then, I must admit that I don't really have a list of criteria that I follow. Every page is pretty unique in the vandalism it receives and the amount. That's why I never evaluate based on the article's topic. I don't know what half the pages I've protected are about. That's because we should be treating every single page the same in terms of vandalism. If it's getting hit just a ton, then it's semi'd. I also don't look at each individual edit to check it for vandalism or good faith. When I look through history, I specifically look for reverts (admin, TW, VOA script, VP, VS, popups, undoes, others). A revert tells me the edit was judged to be bad. If the page has just a ton of edit-revert combinations in the first 50, then it's a excellent candidate for semiprotection (esp if it's over the past couple days). A good candidate for indefinite would have a lot of prior semiprotections in the log, which might have been on and off for a good while (not just a couple months). Other unforeseeable circumstances would have to be taken into effect also. -

) 03:39, 15 May 2007 (UTC)

The process you describe sounds very much like the process I've used. It sounds like you have criteria, you just haven't made them explicit. You are going through a process to convince yourself that semi-protection is needed, and there are no other practical options. Like you, I look at the history and the protection log. I look at the history based on the when the previous protections started and stopped. Comparing the vandalism level before and after gives me a good idea as to whether the protection should be temporary or indefinite. I also check some diffs if it is unclear whether an anon was reverted or not, such as when there are several anon edits followed by a revert. I'm not hearing anyone saying that there should be criteria based on quantities or ratios. I'm fine with leaving that as a judgement call. -- Samuel Wantman 05:07, 15 May 2007 (UTC)

As a non-admin person trying to understand the semi-protection policy I want to say - yes we need a more explicit policy on this. Because it has been my experience that several admins measure the degree of vandalism on a per unit time basis. They will look at the history of a page and say "there is only one vandalism event per day - this is not enough for semi-protection". However this ignores the fact that it does not matter if 10 vandalism edits arrive on a page in one day or over 10 days, if, in both cases, they are mixed in with 2 (say) productive edits. It is just as much work to find the vandalism and undo the edits. So I think the policy should explicity state that the vandalism rate is to be measured as non-productive edits per total number of edits rather than non-productive edits per day.

I have had an admin say that they will not protect a page because the level of vandalism on the page is just the "usual level of vandalism". The problem with this is that the poor admin guy is allways looking at pages that have come up for request for protection and thus she thinks that the degree of vandalism on these nominated for protection pages is the norm. I did a quick little analysis by using the Random page link to see what the rate of vandalism is on a few random pages. You can see the details of the analysis here. But in summary on two random pages we have 100% and 47% of the IP address edits being productive whereas on pages nominated for semi-protection we have

Genetically_modified_food the fraction of productive IP edits is 6%

Geneticaly Modified Organism the fraction of productive IP edits is 16%

1992 the fraction of producive IP edits is 15%

I think it would be a good idea to codify what constitutes "significant" levels of vandalism from IP addresses. And I would suggest that it should be a figure calculated by counting the number of IP edits done in the last 50 edits and calculating a fraction of productive IP edits from this. Then applying an arbitrary number - say 30% - If the % of productive IP edits is 30% or greater then you do not semi-protect the page. If the % of productive IP edits is under 30% you do semi-protect it. We don't need this arbitary % figure. But what we do need is to get away from the vandalism counts PER DAY method. Looking at the ratio of productive to non-productive edits is the way to go and this should be codified into a policy. IMNSHO

Ttguy 14:56, 15 May 2007 (UTC)

Yes, I'm very glad someone brought the subject up. Looking at the

13

01:52, 16 May 2007 (UTC)

Please avoid instruction creep. Need for protection, regardless of the vandalism amount, varies significantly from a page to page, and so is the sensitivity of vandalism and/or disruption. The best thing to do is just trust the judgment of administrators in accepting/declining a nomination. Michaelas¹⁰ 14:08, 16 May 2007 (UTC)

As another non-administrator, I also think some more guidelines should be in place. I had become very frustrated over a period a weeks at the vandalism in an article I was watching. I finally looked up the protection policy and submitted it for semi-protection. I was shocked when protection was declined. What I thought was clear case of heavy vandalism (41 of the last 50 edits being either vandalism or reverts) was viewed differently by someone else. Frankly, I'm still not sure why it was declined. Following this, another editor and administrator got the article protected.

Wikipedia allows articles to be protected. It may as well be used in a consistent way. This would help regular editors as well as administrators avoid frustration and better use their time. TK421 20:09, 16 May 2007 (UTC)

If you can find a diff, that would help us out more. We can't comment on why it was denied or even by who, as there's no log of RFPP, as it would be very, very long. And the best way to understand the policy is to see it in action, not read it. Admins kind of have a de facto standard for protection for the most part (and it's definitely not all about x amount of edits over y time period, but that is a part of it). -

Review Me!

) 22:07, 16 May 2007 (UTC)

(Comment to nobody in particular) I hope everyone realises the impossibility of perfect consistency when we have 1,200+ admins. Concerns about the vagueness of this policy may have some merit, but people should note that the

Steel

22:44, 16 May 2007 (UTC)

At least there are criteria for speedy deletion. We don't have criteria for semi-protection. I'm just proposing we have some common sense guidelines as a starting place. Admins will still have flexibility to interpret them. But at least we will have some common understanding about when and why to use semi-protection. My main impetus for trying to get something more explicit is because there will always be people who are new to the problem of dealing with ongoing vandalism who come to this page looking for advice on what to do. There will also always be admins (like myself) who are just starting to make decisions about whether or not to protect a page. We all need to be brought up to speed to what the normal practice is. Admins have to use their best judgement, but that doesn't mean they will automatically understand all the issues involved and the things they need to consider from the get-go. Editors coming here for help want advice on how to proceed. I'm not advocating creating a specific methodology to apply to all situations. I am advocating creating criteria that illuminate the important considerations that should be taken into account, and by doing so, help explain the rationale behind them.-- SamuelWantman 00:33, 17 May 2007 (UTC)

Well, yes. I wasn't trying to say that bringing people up to speed on current practice is bad, rather that this "I requested protection and it was declined, then next week someone else requested it and had it granted" so-called 'problem' isn't going to be solved by it. –

Steel

23:12, 17 May 2007 (UTC)

Sometimes CSD is a little to restrictive. The one nice part about CSD though is G1. Even though it's a little restrictive sometimes, G1 makes up for a lot of it in it's flexibility. That's kind of the "this page is crap" criteria. -

Review Me!

) 01:57, 17 May 2007 (UTC)

increasingly i find pages that are allegedly "temporarily protected" are simply protected in such a manor to prevent anonymous editing. this is a clear contravention of the use of semi protection policy. this is done to avoid the proper discussion that would be involved in permanently protecting a page.

if an article is indeed going to be temporarily protected. attention must be more prominently brought to this fact. the template should show when the article was protected and should emphasize that it should be unprotected as soon as possible... —The preceding

) 13:44, 14 May 2007 (UTC).

You first paragraph indicates that you don't understand the policy; your second indicates you don't understand how protections and unprotections actually work. At any rate, I encourage you to

Steel

14:02, 14 May 2007 (UTC)

first of all i don't see how your arrogant and dismissive response is likely to help anyone...

please correct me if i am wrong but my understanding of the policy is that it is quite clear that semi protection should not be used "With the sole purpose of prohibiting editing by anonymous users. Protection should be used only to prevent continuing disruption." it is also clear that permanent protection is ONLY to be used for "heavy and continued vandalism." it is clear to me that light or sporadic vandalism is not grounds for permanent protection. the problem i am highlighting however, is that" temporary" protection in far to many cases has become defacto permanent protection.

i think you would do wise to consider

unsigned comment was added by 58.106.71.227 (talk

) 14:19, 14 May 2007 (UTC).

Perhaps you could provide examples. –

Steel

14:22, 14 May 2007 (UTC)

even for an example such as

unsigned comment was added by 58.106.71.227 (talk

) 14:40, 14 May 2007 (UTC).

If an article is protected frequently and generally gets reprotected shortly after an unprotection due to vandalism, people begin to see little value in trying another unprotection and it becomes a permanent protection by de facto. That's the way we do it; we don't have discussions on what articles to designate 'permanently protected'. –

Steel

14:48, 14 May 2007 (UTC)

Ditto to what Steel said. Eventually it becomes pointless to unprotect. If you want to take it up with the foundation, I'm sure they'll side with us on this one (sometimes practicality overrides principle). -

Review Me!

) 22:23, 14 May 2007 (UTC)

I concur as well. Though it may seem pointless, I'd unprotect any page if requested to test whether the vandalism level is low. Better to call it indefinite protection, rather than permanent. You never know, miracles sometimes happen. -- Samuel Wantman 05:13, 15 May 2007 (UTC)

• If you see articles that have been "temporarily protected" for, say, a month or other excessively long time, please visit
  this forum to request unprotection. >Radiant<
  12:10, 15 May 2007 (UTC)

Propose move protect lock and info change

See Template_talk:Pp-move for more details. -- _Hdt83 ^Chat 14:59, 15 May 2007 (UTC)

Could a table explaining the various colors of locks be added to the policy page? Silver, gold, and the greenish gold that is on

don't really emote the particular protection that is in force on a particular page. I'm just thinking something simple like:

Examples

	Gold	Protected from editing until disputes have been resolved Protected from editing to deal with vandalism Protected from editing to prevent [Page name] from using it to make disruptive edits or continuing to abuse the {{unblock}} template Protected from editing It is a high-risk template that has been protected from editing to prevent vandalism.
	Silver	Editing of this page by unregistered or newly registered users is currently disabled to deal with vandalism, It is a high-risk template that has been protected from editing by new or unregistered users to prevent vandalism, A user talk page that has been protected from editing by new or unregistered users to prevent [User name] and other new users from making disruptive edits, Editing of this page by unregistered or newly registered users is currently disabled to deal with vandalism by a spambot, Editing of this page by unregistered or newly registered users is currently disabled.
	Olive	Protected from moves until disputes have been resolved on the discussion page. Protected from moves to prevent page move vandalism.

Dismas|^(talk) 22:24, 16 May 2007 (UTC)

The colours of the padlocks aren't important and we could probably do away with this protection colour coding. The average anon (or

Steel

22:48, 16 May 2007 (UTC)

I know at least

Review Me!

) 23:48, 17 May 2007 (UTC)

• Likewise. >Radiant< 12:47, 23 May 2007 (UTC)

If semi-protection must be used on a user talk page multiple times, yet every time the vandals wait it out and continue vandalizing the page when the semi-protection ends, do you think that indefinite semi-protection might be acceptable? If not, what would be recommended? — Eric Herboso 02:47, 18 May 2007 (UTC)

I would be whole-heartedly against indefinite semiprotection of usertalk pages. If a talk page is getting hit by "new users" who are socks, the best course of action is to file a

Review Me!

) 03:16, 18 May 2007 (UTC)

That would sound reasonable. >Radiant< 12:11, 21 May 2007 (UTC)

Of course we have a definition of heavy vandalism. For "heavy", check your dictionary. For "vandalism",

12:48, 23 May 2007 (UTC)

I hardly think this is useful. Given the above discussions, it's apparent that the problem is that people don't agree as to what consitutes it. The way, the truth, and the light 13:05, 23 May 2007 (UTC)

If as you say that is the problem, then the solution would be to define it, not to point out what we already know. Adding such wording to a policy page just gives scope to wikilawyering about the term. >Radiant< 13:26, 23 May 2007 (UTC)

After reading this and this and this it appears that there is a concensus that

• 1. it might be good to have some guidelines on how semi-protection decisions are made

and

• 2. The guidelines should not be too prescriptive.

Considering this I am proposing that we make the following addition to the Semi-protection section:

To determine if a page is being subjected to a high-rate of vandalism from anonymous IP addresses the administrators shall examine the fraction of productive edits vs non-productive edits made by anonymous users over recent page history and shall include this factor in their basis for their decision.

A page that has not had any anonymous IP vandalsim within the past month shall not be a candidate for semi-protection addtional proposed addition after discussion belowTtguy 14:16, 24 May 2007 (UTC)

The floor is open to discussion. —Preceding unsigned comment added by Ttguy (talk • contribs)

I don't know if there is a consensus. It seems that the admins supporting the status quo are not listening to any criticism of the current state of things.

I personally agree about examining the fraction of vandalism and not just the rate.The way, the truth, and the light 13:09, 23 May 2007 (UTC)

I have no substantial objection to clarifying exactly what admins look at when deciding whether to protect the page other than that it's somewhat trivial and obvious information. "Admins look at the rate of vandalism and whether there are any constructive anon edits being made" - well duh, what else would they possibly be looking for? –

Steel

21:35, 23 May 2007 (UTC)

Steel - If your quote is trying to paraphrase what I am saying in my proposed addition, then you miss the point. I am trying to clarify what is meant by "rate of vandalism". I think it should be measured on a per edit basis not a per unit time basis. When a semi-protection request is denied the admin should be able to defend the decline with a statistic not some vague "not enough recent activity to justify the request". They should be able to say 50% of the anon IP edits over the last 50 edits are not vandalism therefore the page shall not be semi-protected.Ttguy 21:46, 23 May 2007 (UTC)

In practice, both are used. To clarify using your terms: "Admins look at the rate of vandalism" (time basis) "and whether there are any constructive anon edits being made" (edit basis). Using only the per edit basis will result in ridiculous scenarios and people arguing for protection for pages that don't need it. –

Steel

21:53, 23 May 2007 (UTC)

Steel - could you give us some examples where using only the per-edit would result in ridiculous scenarios? Because I believe that pages that have low frequencies of edits (per unit of time) also have low rates of IP vandalism per unit edit. So I would suggest that ignoring the rate per day will not result in ridiculous claims - you are only going to see high rates of IP vandalism per edit on pages that also have high rate of edits (and thus vandalism) on a per unit day. The reason why I beleive this is that I did a quick little analysis by using the Random page link to see what the rate of vandalism is on a few random pages. You can see the details of the analysis here. But in summary on two random pages we have 100% and 47% of the IP address edits being productive (these pages are edited at the rate of 0.5 edits per day ) whereas on pages with high edit rates per day we have

Genetically_modified_food the fraction of productive IP edits is 6% (edited at 3.3 edits per day)

Geneticaly Modified Organism the fraction of productive IP edits is 16% (edited at 3.5 edits per day)

1992 the fraction of producive IP edits is 15% ( edited at 8.2 edits per day)

But if Steel can find a few examples of pages with a high fraction of vandalistic IP edits which are edited in frequently then he will put my case on more shakey ground and we should consider including both criteria in the policy.

Ttguy 09:17, 24 May 2007 (UTC)

I've only scan-read the Kusma discussion but all I really see is another admin disagreeing with you. But anyway, a page whose history is almost entirely vandalism but hasn't been edited for 12 months is obviously not a candidate for semi-protection. –

Steel

12:47, 24 May 2007 (UTC)

No I don't think it would be a candidate for semi-protection. But I chalenge you to find example pages like this. I don't believe you will find pages like this. It is my thesis that pages that are edited at low frequency have a low proportion of vandalsim and my analysis above shows this is the case for 2 randomly choosen pages. I am asking you for counter examples.Ttguy 13:26, 24 May 2007 (UTC)

Unfortunately I don't have counter-examples to hand, but I have seen articles like that in my nine or so months of doing

Steel

13:46, 24 May 2007 (UTC)

You are right. If such a page does come into existance my propose criteria would mean it should be semi-protected. I think it highly unlikely that such a page would get nominated for semi-protection - at least not by a sane person - because they get nominated when some one is sick of reverting vandalism on a page that they watch and so by definition such a page will have been recently edited. However, I guess someone could stumble opon a page that has not been edited for a while and find a large proportion of vandalistic edits in the history and then nominate this page for semi-protection. So ... how about in addition to what I have allready proposed we have:

A page that has not had any anonymous IP vandalsim within the past month shall not be a candidate for semi-protection

Ttguy 14:16, 24 May 2007 (UTC)

So, why exactly do we need that statement? Have there been problems recently with admins protecting pages like that when they shouldn't, such that it needs clarification? –

Steel

16:49, 24 May 2007 (UTC)

(Split for clarity. -

Steel

12:47, 24 May 2007 (UTC)

I have not said you need to calculate percentages. I would imagine a scenario where you just pull up the page history and go down the list counting the number of IP edits that get reverted for vandalism and noting the number of IP edits that stand. Then once you have counted down a few edits on the page you will be able to go "hmm most of these IP edits are good" or you will be able to say "every one of these IP edits is vandalism". Then you make a judgement call.

I guess it is a bit harder than looking at the page history and saying - "no edits today" - therefore I can do a decline with "There is not enough recent activity to justify protection at this time".

But it is not that much harder and it is a lot fairer.

But if the policy in practice is "no edits today" - decline with "There is not enough recent activity to justify protection at this time" then this would be OK - if the criterion was published so that users could look at it and say - "oh that's fair - thats the rule - I understand". But we don't have that clarity either way.

You say "In practice, both are used." Can you write a counter proposal that you believe adequately describes the criteria you use for granting or denying a semi-protection request and maybe we should discuss/use it. There are a few of us that would just like the criteria clarified. They don't have to be changed - just clarified. Ttguy 14:03, 24 May 2007 (UTC)

I will happily respond if you retract the claim that admins don't give due consideration to requests. –

Steel

14:21, 24 May 2007 (UTC)

Not all admins use the same criteria, and some may overlook some. Listing what criteria are used, and preferrably a very rough minimum level (give some examples if you like) must be added to the page in order to help people understand the policy. Richard001 00:19, 24 May 2007 (UTC)

Of course not all admins use the same criteria, this website has 1200+ admins. I hope you're not considering listing every possible condition? Or just the common, obvious ones which are, by definition, common and obvious? –

Steel

12:47, 24 May 2007 (UTC)

Remember, I am only trying to clarify what is meant by "high-rate of vandalism from anonymous IP addresses". I am not trying to define every criteria for semi-protecting a page. So there should not be too many criteria on which to judge a "high-rate of vandalism from anonymous IP addresses". And I think we have hit on the only two possible measures. And all I am saying is that you probably do not need to consider one of them. Do you have a third way to measure "high-rate of vandalism from anonymous IP addresses"? Ttguy 14:03, 24 May 2007 (UTC)

• Perhaps it is a reasonable common-sense definition that doesn't need further clarification. Is there any concrete problem you're trying to solve? >Radiant< 14:17, 24 May 2007 (UTC)

The concrete problem I am trying to solve is the problem that many of encounter when we submit a page that has 80% of the IP edits in the last 50 being IP vandalism and get a "Declined – There is not enough recent activity to justify protection at this time". And that problem is "how can you possibly say not enough vandalsim activity?" when 80% of the activity is vandalsim. If you look through the RPP history you find many disgruntled users because of this. I don't think we would mind so much the decline if we knew how such an assesment was being made. Disgruntled editors is the problem I am trying to solve. This and inconsistent application of semi-protection requests.Ttguy 09:46, 25 May 2007 (UTC)

• I dissent. Any language along the lines of "the admin shall ... " is both overly legalistic, and prescriptive. In this case it really isn't saying much, either. >Radiant< 12:53, 24 May 2007 (UTC)

It is not prescriptive because there is no figure in there saying what is the cut off. It is just saying that the admin will look at this when making a decision. It just seems to me when you get your semi-protection request denied with the stock template answer "Declined – There is not enough recent activity to justify protection at this time" then it is a fair and reasonable question to ask - "what is enough?"

Ttguy 13:18, 24 May 2007 (UTC)

• Just because it's not an exact prescription doesn't mean it's not a prescription. Also, it's not saying much, and it's overly legalistic. >Radiant< 13:54, 24 May 2007 (UTC)

Can you work with it then to make it less legalistic but still give us poor users some idea about what it means when we get our semi-protection request denied with "Declined – There is not enough recent activity to justify protection at this time" . Because at the moment it feels like a the Admin deities decide at random whether or not to grant us our wishes.

I personally would like it to say more and be more prescriptive. But I made it a bit vague because much of the discussion on this issue was about giving admin people discresion. Which I don't have a problem with. But right now the semi-protection for high rates of anonymous IP vandalism seems to be total arbitrary because there are no guidelines. Help us out here with some. Some. Any. Guidelines

Ttguy 14:26, 24 May 2007 (UTC)

• I can understand that, but if we were to add here that e.g. a page is protected if 20% of the edits in the last week are vandalistic, that would (1) not reflect reality; (2) not actually change anything, only giving the appearance of change; and (3) allow for wikilawyering by vandals that pages were protected "out of process". That is why Wikipedia rarely employs such standards. The protection rates aren't realy arbitrary because in general the protecting admin looks over a hundred of such pages a week and protects those with, in his opinion, above-average vandalism. You might try asking individual admins how they deal with it. >Radiant< 14:49, 24 May 2007 (UTC)

• Plus, we need more rules like we need a hole in the head. The whole point is that we expect admins to apply Clue in these calls. Sometimes the best move is to block a vandal, sometimes, to semiprotect, sometimes to protect, sometimes to knock heads together. I for one have no intention of filling in a checklist every time I protect a page for 24 hours. It's bureaucratic, and there's no compelling evidence of a problem needing fixing. Not to mention that the criteria would never get agreed because of some editors' fundamental ideological objections to protecting at all. Guy (Help!) 18:14, 24 May 2007 (UTC)

I guess I'll just reiterate myself (again) and say that I agree with Steel, Radiant!, and JzG. I say this to all users: Stop trying to formalize everything. The definition given above is basically the same as the one we use, but formalizing it is a horrible idea. Non admins will start arguing about what constitutes "above average" and even "rate". I think I'll just say this here now: We are admins. We went through RFA and have the understanding (and of course the rights) to make our own decisions. We take recommendations (

WP:RPP

). But the final decision remains with us (or Jimbo if he ever said so, or the WMF if they ever got involved, which is probably never). If you think an admin is abusing their power, RFC and ArbCom are there to say "yes they are" or "no, there're doing their job".

I also refuse to fill out forms or do anything bureaucratic. People complain about admins not doing anything, then they want to make it harder for us to do anything. I have said it, and I will say it again: Just let us do our jobs. -

Review Me!

) 00:15, 25 May 2007 (UTC)

Yes, we admins are more than capable of ruling over you peons with our superior knowledge and private IRC channels. Policy pages are far too complicated for you to understand anyway, so just trust us when we say that We Know Best and stop questioning our absolute authority. –

Steel

00:40, 25 May 2007 (UTC)

"Trust our judgement"? "We went through RfA and have the understanding"? Or how about this one?

Yes, we admins are more than capable of ruling over you peons with our superior knowledge and private IRC channels. Policy pages are far too complicated for you to understand anyway, so just trust us when we say that We Know Best and stop questioning our absolute authority.

Okay, so that last one might have been

WP:FAITH

issues going on here. TT's opinions are pretty similar to mine, although I might choose a different tack in expressing them. The anti-vandal/protection policy seems so arbitrary that it feels worthless to try and clean up low-edit-traffic pages, unless I put every single page I take vandalism off of on my watchlist.

Most of the dissenting admin opinions seem to be along the lines of "I don't want to do more work". Maybe I'm reading them incorrectly, but if I'm right that's pretty disheartening. edit note: I retract my questioning of good faith after reading Steel's response on TTGuy's talk page. - CredoFromStart ^talk 14:58, 25 May 2007 (UTC)

We are short staffed as it is. I agree to volunteer to help wikipedia by being an admin. The amount of crap we take for being volunteers would have had some people running to the hills by now. The only reason Wikipedia works is because people are willing to give their time to work on articles, administrations stuff, and even the crap. I don't want to fill out forms and ratios ect because its a useless waste of time. I'll protect and unprotect, block and unblock, delete and watch for vandalism. I will do this as long as I can and as long as the community is respectful of it. Do you know how many admins left last year? 20 admins resigned in 2006 alone (that's just personal requests, not involuntary). I know that some of them were basically driven out too. And is there any reason that you can't trust our judgement? (and no, one admins decision's do not reflect on the >1000 other admins) Some people are just so anti-admin. -

Review Me!

) 22:45, 25 May 2007 (UTC)

WP:PEREN

?

Add to

00:42, 25 May 2007 (UTC)

That sounds like a good idea. What wording did you have in mind? –

Steel

00:44, 25 May 2007 (UTC)

Numerical protection conditions

• Proposal: something along the lines of "the ratio of productive to total edits by new and unregistered users", "the number of unproductive edits per day", etc. as a protection condition or factor on protection period.
• Reason for previous rejection: m:instruction creep; situations may vary significantly; admins often use different methods of dealing with disruption per own discretion.
• See also:
  Michaelas¹⁰ 16:59, 25 May 2007 (UTC)

I take it that referencing (rejected)

WP:RFPP, if that is the basis they want considered. The persons requesting protection would than know that they could make their case using that ratio and that admins would consider it as factor for decision No one concerned with upgrading an article can be unaware of the disruption caused by vandalism. The net uproductive Edits (vandalism+reverts)/total edits nicely captures what irks serious editors & slows article improvement, as suggested by the 1st para. of Wikipedia talk:Protection policy#A more explicit semi-protection policy for articles subject to vandalism. --Thomasmeeks

20:33, 26 May 2007 (UTC)

Separate issues: "productive" Edits and "free lance" unprotects

Wording at the top of this section as to rate of "productive" vs. "unproductive" Edits is relevant. How to define "unproductive"? One way is to include not only obvious vandalism but Edits that are shortly reverted. These are arguably equally or more disruptive if one succumbs to the temptation to explain why a bad Edit is bad (which might be a good idea to reduce likelihood of an Edit war but can be very time consuming). My experience is that disproportionate "unproductive" Edits (aside from vandalism) are also from unregistered or new users. As noted above relevant are:

1. the ratio of productive to total Edits by new and unregistered users
2. the rate of unproductive Edits to total Edits by new and unregistered users, say per 100 Edits.

It would help to standardize measures so that Admins would a basis of comparison.

On Steel's nicely ironic point at the 00:40, 25 May 2007 Edit above, it would also be very nice if there was a strong guideline for Admins to inform themselves of Talk page section on vandalism or semiprotection for an article before taking down a semiprotect. It comes across as esp. careless for an Admin to take down a semiprotect without becoming informed as to grounds for putting up the semiprotect (on the Talk page) in the first place. --Thomasmeeks 01:39, 25 May 2007 (UTC)

There are protection summaries, visible in the protection log, that admins are meant to check before unprotecting. I would hate to try to scan a very long talk page for protection reasoning, which won't even be there if the page was protected via a request at

WP:RFPP. —Preceding unsigned comment added by CMummert (talk • contribs

)

This suggests 2 ways to improve semi-protect review quality: (1) leaving a record on the Talk page of the request for semi-protection, so as to build a cumulative history as to the case for not removing semi-protection prematurely & (2) restricting Admins acting as free lancers in removing semi-protection prematurely without going through

WP:RFPP where there is at least the check of having some accessible record that may be referred before removing semi-protect. --Thomasmeeks

11:15, 25 May 2007 (UTC)

I can guarantee you that if you added a few lines to that effect to the policy they will be ignored. There is no requirement for people to discuss things on talk beforehand, and this applies to edits, protections, deletion nominations, etc. Protection summaries are largely sufficient and simply talking to the admin if their action is disputed is much more efficient than obligatory talk page notes every single time someone does anything. –

Steel

11:39, 25 May 2007 (UTC)

On (2) above, I understand you to be saying that there is an enforcibility problem. Given that Admins can remove semi-protect without going through

WP:RFPP who presumably had more accessible info. The whole point of (2) is to encourage transparent due process likely to improve the quality of articles. --Thomasmeeks

12:58, 25 May 2007 (UTC)

You do realise that the vast majority of pages are unprotected by "free lance" admins and the vast majority are uncontroversial? (And the few that are can be easily reversed.) What you propose is

Steel

13:40, 26 May 2007 (UTC)

Thx for your comment. On your first 2 sentences, no, I did not know (but, yes, I was surprised). On your 3rd sent., perhaps you're right. On your last sent., I agree with your second sent. at 11:39, 25 May 2007 and did not mean to suggest otherwise. On your points following that, perhaps you are right, but on one point my experience is different from what your generalization suggests. What concerns me is an unprotect that ignores the record of vandalous or disruptive Edits before the semi-protect or even Talk page sections that document the disruption. --Thomasmeeks 17:55, 26 May 2007 (UTC)

If there's a good reason not to unprotect a page and an admin ignores that, take it up with them on their talk page. There's no lasting harm done if the page is reprotected a few hours later. –

Steel

18:29, 26 May 2007 (UTC)

A middle path?

I believe it is possible to create criteria for semi-protection that are descriptive and not legalistic. There is a need for this. Is the following a possible middle path?

Proposed Semi-protection guidelines

Articles subject to heavy and continued vandalism can be semi-protected. There are no explicit rules that determine the level of vandalism that is necessary to trigger semi-protection. Administrators should use their best judgment to determine if semi-protection is warranted. Here are some guidelines that may be helpful to determine if semi-protection is appropriate:

• The vandalism is often of an offensive nature on high-profile pages. This has a detrimental effect on how Wikipedia looks to the public. A major exception to this is the featured article linked to the main page, which should not be protected, but monitored instead.
• All or almost all of the vandalism is coming from unregistered users.
• Unregistered editors should be making very few contributions to the article compared to the amount of vandalism coming from unregistered editors. The negative effects of semi-protection on discouraging positive contributions should be more of a concern than the positive effect of decreasing vandalism.
• There are regularly many new vandals, therefore it would be a huge unending task to notify and warn all the vandals individually.
• The proportion of constructive edits to nonconstructive edits (vandalism and the reverts) by all editors is very low. This means that the regular editors are devoting most of their effort to reverting vandalism instead of contributing to the articles. If vandalism is being reverted as part of the normal course of writing the article, semi-protection is not needed.

If semi-protection is to be tried, its first application should be for a short duration, a few days or a week. If vandalism continues after the protection expires it can be added for a longer duration. At some point an administrator might determine that the semi-protection should be made indefinite.

• Pages that are indefinitely semi-protected must have been semi-protected previously. This shows that the problem is on-going, and that temporary semi-protection does not have a lasting effect.
• Vandalism that resumes very shortly after semi-protection is removed demonstrates that the page is a popular target for random vandalism. Such pages are likely candidates for indefinite semi-protection.
• If vandalism is related to a current event, the semi-protection should be lifted after the event is out of the public eye. Perhaps the protection will no longer be needed.
• The only way to determine if ongoing semi-protection is still necessary is to remove the protection and see if wide-scale vandalism resumes. For this reason, all pages that are indefinitely semi-protected can have their protection removed from time to time. After removing the protection, the administrator should monitor the page.

Nothing here is quantified, and most is based on common sense. I think this describes the current practice, and having this stated explicitly will probably be helpful to many who come looking for help. --

uelWantman 01:52, 25 May 2007 (UTC)

Why don't we just have this in an essay or something, like

Review Me!

) 22:52, 25 May 2007 (UTC)

You make some very good points. Among them are that:

• unproductive Edits (either vandalism or other Edits that are reverted shortly theereafter) are almost entirely an un- or newly registered user problem (
  Wikipedia:WikiProject Vandalism studies/Study1
  ), so that semiprotect largely eliminates vandalism.
• that the biggest cost of bad Edits may be (proxied by productive edits?) in discouraging productive Edits and thus lowering article quality.

But making explicit the current practice reveals that the current problem, b/c of the the subjective and presumably indefensible variations in practice ("presumably" b/c of dissatisfaction of users whose requests for semiprotect are denied). At least making current practice explicit has the advantage of inviting more explicit criteria for deciding how vandalism is too much or for inviting a change of current practice. As for the term "legalistic" in the current application, if quantitative standard improve the article & stop vandalism, who cares what it's called? --Thomasmeeks 23:50, 25 May 2007 (UTC)

We are not going to explicitely quantify these so-called criteria for protection, not least because it won't work in practice. –

Steel

23:58, 25 May 2007 (UTC)

I concur with Royalguard11, just make it an essay.

WP:PROT describes the exact same criteria in a more general method, which has been proven to be quite effective by now. Why fix if it ain't broken? Michaelas¹⁰

00:31, 26 May 2007 (UTC)

It isn't broken, but it is deficient, and many of the discussions on this page demonstrate that there is room for improvement. I'm fairly new to this semi-protection stuff. When I came to

☑ Sam

uelWantman 09:13, 26 May 2007 (UTC)

Seems like a good argument for

Steel

13:36, 26 May 2007 (UTC)

I have created

☑ Sam

uelWantman 20:40, 26 May 2007 (UTC)

What about article quality? This is a big one for me. It's not so bad if a vandal messes with a disorganized stub, but for our best articles it does a lot more damage to the quality of the encyclopedia. The second important factor here is that the higher quality an article is, the less likely it will be improved by a random edit, and furthermore it is less likely anyone will need to edit it in the first place. For example a featured article may be damaged by a poor quality piece of writing, while a stub may be improved, even if the text may need a little work by other editors. Article quality should be treated as an important consideration when considering whether to apply protection or not. Richard001 11:05, 5 June 2007 (UTC)

I don't support that statement at all. I would be less likely to protect a stub (because it'll probably get improved), but I don't protect longer, better articles because they're less likely to get better. I protect them because of vandalism, linkspam (and occasionally BLP). -

Review Me!

) 22:18, 5 June 2007 (UTC)

I'm not suggesting articles be protected for being high quality, I'm suggesting the requirements for semi-protecting such articles should be set slightly lower. As you said above, you would be "less likely to protect a stub (because it'll probably get improved)", so you are basically in agreement with me here. Richard001 22:28, 5 June 2007 (UTC)

I have introduced a Special cases section detailing the two separately operated policy/guidelines. The

10:47, 26 May 2007 (UTC)

I moved this to a subsection of the full protection part. It seems reasonable to me. CMummert · talk 12:26, 26 May 2007 (UTC)

Steel

13:32, 26 May 2007 (UTC)

After discussions above, I have created

uelWantman 20:42, 26 May 2007 (UTC)

This is good. I think this should be on the protection policy page so that editors know how semi-protection is to be applied Ttguy 13:00, 28 May 2007 (UTC)

I can appreciate from the discussion 2 sections up that there is a desire to keep the protection policy as it is. My goal is to create guidelines that help bring people up to speed on current practice. So I'm O.K. with keeping this on a separate page. If there are no objections, I'd like to add a link to the semi protection section on policy page that says "See also:

☑ Sam

uelWantman 00:58, 29 May 2007 (UTC)

Call for comment on proposed Unprotection paragraph

The following has benefitted from comments of Steel, though he is not responsible for remaining deficiencies. The article section Wikipedia:Protection policy#Unprotection has the following 1st paragraph.

CURRENT version (with sentences numbered for ease of comparison and reference):

(1) With the exception of any pages tagged with the {{

office}} or {{reset

}} templates, any admin may unprotect any page after a reasonable period has lapsed, particularly if there is no discussion on the talk page.

(2) However, unless consensus has been reached, pages should not be unprotected soon after protection without prior consultation with the admin who first protected the page.

(3) This is particularly important in the case of controversial pages, where the conflict may start up again and the protecting admin may be in touch with the disputants.

Provisional PROPOSED alternative:

(1a) Anyone may request unprotection through

office}} or {{reset

}} templates), provided a reasonable period has elapsed.

(2a) For non-

WP:RPP

action, admins should factor in visible Talk-page section discussion as to (un)protection.

(3a) This includes the case of controversial pages where conflict may be renewed with unprotection.

Purpose: to clarify, simplify, and guide admins and non-admins.

Rationales for proposed alternative:

• It is written for non-admins (as well as admins), & not just as an afterthought.
• Non-admins are more alerted to put there (un)protect concerns on the Talk page, so an admin can factor in those concerns.
• Admins are busy people. The end of (3) is too special, roundabout, & uncertain to warrant mention.
• It reflects the current paragraph.

Other advantages of proposed alternative:

There is no added admin burden whatever at

WP:RPP

.

(2a) only states clearly what a prudent admin would be doing anyhow.

Non-admins are more directly informed of what they need to do to make it easier for admins to decide on unprotect.

A 2-day call for comment is requested before submission (although nothing prevents any changes in the current article section before then). One possibility is listing consecutive amendments below (1b, 1c, etc.). --Thomasmeeks 20:31, 27 May 2007 (UTC) (proofread fix) Thomasmeeks 22:38, 27 May 2007 (UTC)

There appears to be a lot of needless rephrasing here without much actual change. In fact, the only change I can see is your (2a). For reasons why this isn't necessary, see

Steel

12:53, 28 May 2007 (UTC)

(1a) addresses concerns of both non-admins and admins; (1) does not. (3a) includes controversial pages. (3) elevates controversial pages for protection. (3) thereby suggests the implausible implication that say high vandalism is a less important reason for protection. There is also the third point of 'rationales' above.

As for (2a), surely it would be prudent for an admin without the benefit of a

WP:RFPP request for unprotect to look at the Talk of the article to see if there is a section on (un)protection, which might have information or discussion that would be helpful in making a final decision. (1) in the current article already refers to the Talk page in passing. (2a) is more explicit. --Thomasmeeks

18:46, 28 May 2007 (UTC)

I think WP:RFPP is a red herring here - there should be no difference in the procedure to be followed whether or not the page is listed at RFPP. Listing at RFPP is just a way to get admin attention or centralize discussion about the unprotection. CMummert · talk 13:30, 28 May 2007 (UTC)

You make an interesting point. But as written the current article para. reads like it is written for non-

WP:RFPP unprotects. What I believe that you are saying is that the paragraph calls admins (whether through WP:RFPP or not) to check the Talk page of the article as a factor in unprotection. Without commenting on the merits of that, surely the current paragraph should be more explicit on the point. --Thomasmeeks

18:46, 28 May 2007 (UTC)

If you don't know about RFPP, then you shouldn't be requesting unprotection anyways. RFPP only exists as a place for people to alert admins to pages needing protection/unprotection. There is actually no policy that states you need to follow that process (if someone requested protection on my talk page for example, I'd have no problem doing it, and it's perfectly allowed). RFPP is not official policy (contrasted to xFD's, prod, csd, and drv which are all part of

Review Me!

) 22:41, 28 May 2007 (UTC)

The subject of this section is about the above article paragraph, which (I believe we'd agree) should be written to inform and guide admins and non-admins.*

WP:RFPP

not being a policy page, but it is referenced on the current article page. I also believe that its use is an efficient, though not unique, vehicle for implementing (un)protection policy.

* I hope that addresses your point. If not, please try again until I get it. --Thomasmeeks 00:38, 29 May 2007 (UTC)

The policy says Any type of protection, as well as unprotection, may be requested at Wikipedia:Requests for page protection, in contrast to

Review Me!

) 03:16, 29 May 2007 (UTC)

On your 2nd to last sentence, I take it you are saying only that an admin could change (un)protect status directly. But if not, please clarify. --Thomasmeeks 11:33, 29 May 2007 (UTC)

• I don't understand what you're trying to say with (2a); it's vaguely worded. "Conflict may be renewed with unprotection" is also vague. You are omitting the point that it may be useful to discuss with the protecting admin, and that protections can time out. Overall this rewording is a net negative. >Radiant< 14:19, 29 May 2007 (UTC)

Thx for comment. Well, there is (1a) as context. Would you have a suggested less-vague alternative ("2b")? --Thomasmeeks 18:03, 29 May 2007 (UTC)

P.S. Of course the relevant comparison is between the current proposed alternatives, e.g. (2a) vs. end of (1) ("particularly . . .") + (2). The latter are, I believe, vaguer still, which does not mean that (2a) can't be improved. Similarly, (3a) vs. (3) . My rationale for omission of the end of (3) is the 4th point of the "rationales" para. Timed expirations are in neither the current nor proposed paragraph. (No reason for not working them in at some point of course.) I'm still hoping for a proposed alternative ("2b"). --Thomasmeeks 18:56, 29 May 2007 (UTC) (typo, line 1)

Everyone, little question: on the tennispage, you can read:
Editing of this article by unregistered or newly registered users is currently disabled until June 8, 2007 (UTC) to deal with vandalism. .

But THIS article tells me that 'newely registrated users' are users with an account less than 4 days old. Still, I cannot edit the tennispage. Anyone knows what I am doing wrong?

15:11, 30 May 2007 (UTC)

It does sound peculiar that you are blocked from editing. You might check that you are logged in. (If you are not logged in, it will say "log in" on the top right hand tab of your Wiki page. My experience is of periodically getting logged out automatically for reasons unknown.) Next you might try Wikipedia:Contact us. This page is mostly used for other things, so don't expect too much here. --Thomasmeeks 16:37, 30 May 2007 (UTC)