Wikipedia talk:Semi-protection policy/Archive 7

Archive 1

←

Archive 5

Archive 6

Archive 7

Archive 8

New users

Splash, why do you want to tell people how long they have to wait before they can edit? The legit users will find out soon enough by simply continuing to try; the non-legit ones (and I suspect most of those we help with this edit will be non-legit) will know to prepare themselves with sleepers accounts four days in advance. Also, why do you want to add that sprotection should be lifted within hours? It almost never is, nor is there any reason it should be. SlimVirgin ^(talk) 09:15, 7 July 2006 (UTC)

Why should we not tell them? This is an open editing project, and smoke and mirrors are not helpful.

I didn't add it; it was the result of the discussion above. It is the case that protection is usually removed within a few days, however. Why did you remove that? -Splash - tk 19:25, 10 July 2006 (UTC)

Sprotection should be lifted when the protecting admin thinks appropriate. Sometimes that's a few hours, sometimes days, sometimes longer. SlimVirgin ^(talk) 20:58, 10 July 2006 (UTC)

Trial run of permanent semi-protection of Wikipedia:Blocking policy

Would permanent semi-protection be appropriate for Wikipedia:Blocking policy? It is a very important official policy page and I feel that vandalism on it is more harmful than it would be on an ordinary page. It is not vandalized constantly, but it is vandalized regularly, about once or twice a day.

I'm not basing my whole argument for semi-protection upon similar vandalism, as it is much rarer than regular vandalism, but I reverted this edit the other day. What is disturbing about it is that it is an apparent attempt to change policy without anyone noticing rather just deleting the article or adding "you suck". Someone reading the page who is unfamiliar with Wikipedia policy might not realize that the article has been vandalized and think that it is okay to evade a block that you think is unjust or that anyone with a high edit count can become an admin (see the links on this page to many unsuccessful RfAs of editors with high edit counts). If they don't come back to reread it and they are not informed otherwise, they may retain those beliefs indefinitely. Also, if someone had made a regular edit or two soon after, without noticing the vandalism, it might have gone unnoticed for days. As it was, it was about a half hour before I found it. Semi-protection would stop the everyday vandalism and the sneakier, potentially more harmful vandalism like this edit.

I'm strongly opposed to unnecessary protection, especially permanent protection. However, semi-protection only blocks anonymous users and very new registered users, most of whom are unfamiliar with Wikipedia policies and ill equipped to make substantive edits to the policy anyway. Also, they would still be able to suggest changes on the talk page or sign up and wait a few days so that they could edit the page themselves.

Semi-protection would stop virtually all vandalism of the page, and similar pages, because the vast majority of vandalistic edits on it are made by anonymous users and the rest are almost all new users. There are some reverts of edits made by established users, but those are just disagreements rather than vandalism. Semi-protection of an official policy page is less harmful than the protection of an article, like Jew and George W. Bush, which are given as examples of continuously semi-protected articles.

We might need to modify Wikipedia:Semi-protection policy first, or at least eventually, but I suggest we just try semi-protecting this one page for a couple of weeks, if there is a consensus to do so, and see how it goes. If semi-protection is implemented permanently, I think there should either be no notice on the page itself (with an explanation given when you click to edit the page) or a discrete notice, like a padlock image that sends you to the semi-protection policy page when you click on it. If the trial run is successful and the semi-protection usage is expanded, I would suggest not even using it on all guideline and official policy pages, only those that are regularly vandalized and/or those that can cause serious problems when they are vandalized. I don't think most of the guideline pages would qualify for the second one, but a few might. A few Wikipedia namespace pages that are not policies or guidelines might benefit from permanent semi-protection, too. Still, I think that it would be limited to a dozen or two pages. -- Kjkolb 08:55, 15 July 2006 (UTC)

Extend policy

I know that Jimbo Wales' views are final here, and I know that lots of other admins and editors are very attached to building a "free encyclopedia that anyone can edit" but I can't see the advantage of restricting semi-protection to the limited cases that it is.

I have just asked for List of terrorist organisations to be semi-protected. I have contributed quite a bit to this article and have it on my watch list. However, most of the changes that are done are vandalism by unregistered users. But apparently there's not enough vandalism to merit semi-protection!! I was advised to add it to my watchlist and just revert vandalism when it occurs. Why should I bother - it's not my encyclopedia and the powers that be don't seem to care!

Perhaps people don't fully appreciate the impact that vandalism has to wikipedia:

- it is demoralising to editors - particularly new editors - to have their contributions damaged

- it makes wikipedia look foolish and amateurish have readers come across vandalism all the time

- it diverts resources away from building a good encyclopedia towards chasing after vandals.

- many new editors don't know how to revert (I only found out after several hundred edits)

What is the disadvantage with semi-protection? I dont understand the hesitation at all. AndrewRT - Talk 16:14, 12 August 2006 (UTC)

As one who was in your situation in the past, and had a similar opinion at the time, here are some reasons I can think of today, with more experience, why having a more liberal semi-protect won't work:

1. Anyone who is hell-bent on editing to the article will just set up a dummy one-time account, let it stew or use it randomly a few days (so it won't be brand new) and easily get around s-protect.
2. WP's openness is a natural magnet to both good and bad people - that's life. There is no magic solution - the good have to outlast the bad, just like in real life - that's why you need to monitor your watchlist and revert vandalism like most of us here
3. If there is a real peak of vandalism, virtually any admin will gladly help - if one refuses, re-think your reasons and his/her refusal, and if you are still convinced it's needed, try another

Crum375 17:32, 12 August 2006 (UTC)

Thanks for your thoughts. In essence you seem to be argueing there's no point in trying because it wont work. This certainly wouldn't be the case with my article. That was a case of "drive-by" vandalism - lots of different people coming along at different times and thinking lets do something funny today - not determined users trying to push a POV. Semi-protection would have been a very effective solution to this problem.

In "real life" the bad guys lose out because there is an effective deterrent against people doing bad things. If protection was more widely used, vandals blocked a bit more readily, and maybe reported to their employers/colleges or even prosecuted on occasions, I'm sure you would find vandalism becoming the exception rather than the rule as at present. AndrewRT - Talk 23:51, 12 August 2006 (UTC)

You may be missing something. The 'bad guys' don't have to decide to pick on your article only. They can decide to (and often do) pick on WP as a whole, attacking random articles. So they could still set up as many handy dummy accounts as they wish, let them sit a few days to satisfy any 'newness' criterion, and then attack your article, at random. So again, s-protect is not the magic cure you are yearning for - it is a temporary crutch for some specific situations, to be carefully used along with other tools such as IP blocking or article protection. The bottom line remains the same - you need to watch your article, if it is very contentious you may need to resort to s-protect to at least reduce the pressure. Remember that all admins are here to help you, plus you can go to the page protect request board. Don't take no for an answer if you feel the amount of vandalism from multiple sources is excessive (single source vandals should be blocked by IP). Crum375 01:05, 13 August 2006 (UTC)

What Crum375 says is not borne out by the facts (at least not of the articles I've been analysing):

1. Anyone who is hell-bent on editing to the article will just set up a dummy one-time account, let it stew or use it randomly a few days (so it won't be brand new) and easily get around s-protect.

Perhaps - but during the several weeks that both Computer and Automobile were semi protected, the amount of vandalism on both articles dropped from several per day to absolutely ZERO. There certainly are determined vandals - but they are a teeny-tiny minority.

2. WP's openness is a natural magnet to both good and bad people - that's life. There is no magic solution - the good have to outlast the bad, just like in real life - that's why you need to monitor your watchlist and revert vandalism like most of us here

The good have to outlast the bad - but having the patience to open an account so you can edit semi-protected pages is something where the good will also outlast the bad. What vandalism does is to drastically reduce the amount of time serious editors can spend on writing an encyclopedia - and that's never a good thing. It also degrades the quality of the encyclopedia more than anything else. When you open a page to sit with your child and read about Automobiles together - you aren't generally pleased to discover the page is full of obscenities. It would be interesting to know how many readers (who are actually more important than contributors) are put off by vandalism.

3. If there is a real peak of vandalism, virtually any admin will gladly help...

The problem isn't peaks of vandalism - it's continual, ongoing half a dozen random attacks per day vandalism. It doesn't "peak" and then go away - it just goes on and on forever at about the same level.

The PRACTICAL truth is that semi protection works like a dream. It shuts off almost all vandalism and deters almost nobody who really wants to contribute. The two 'almosts' in that sentence in no way detract from it's validity.

SteveBaker 05:49, 4 October 2006 (UTC)

Main Page Linked Articles

I think the policy could do with a bit of clarification here. Forgive me if it has come up already. But at the moment this page gives no advice specifically on pages linked to the main page. Given that the protection policy does do this - stating unequivocaly that main page linked articles should not be protcted - I think there is a bit of confusion over the issue. Taking Steve Irwin as an example, this was protected/unprotected thirty times on September 4th. There was no wheel warring - though maybe if different admins had been invovled there was certainly potential for it - just a general confusion/fudge over policy and precedent here.

I would propose that this page states that 'With the exception of the day's featured article, articles linked to from the main page can be semi protected. However, this must not be done as a pre-emptive action and semi-protection must not be maintained on such articles for more than x hours.'

Thoughts? --Robdurbar 07:02, 5 September 2006 (UTC)

No. Certainly not in the sweeping sense you just added it to the project page at the very least. Articles linked from the Main Page should not be protected except in dire circumstance. -Splash - tk 20:32, 9 September 2006 (UTC)

OK, well its a shame that I had to try and add this to the article before anyone decided to respond. I accept the idea of protection of linked articles, but note that many articles linked to from the main page do get semi-protected at the moment. I'd also reject that 'as a temporary measure and last resort' is paticularly sweeping.

I can see both the sides of the argument for and against semi-protection of linked articles and, though I'd probably weigh down on the side of 'semi-protect for a short time where necessary' view, I'd happily go along with any consensus.

However, my experiences suggest that the current situation - where linked articles are mentioned on the protection policy but not the semi-protection policy - simply causes confusion. Some people appear to be labouring under the impression that semi-protection of any linked article is also prohibted under Wikipedia policy; at the moment it is not. Others follow this as a personal policy - fair enough, but this could be frustrating for other users who are not admins, see nothing in the policy to prevent semi-protection, and yet do not see it happening. Many others (myself included) appear to be in a state of confusion and this leads to situations where pages such as Steve Irwin are protected/unprotected thirty times in a day - a waste of admins' time, and it looks messy.

Therefore, what is needed is for this page to say something on main page linked articles. Now, what it says - as far as I'm concerned, anyway - is of less importance, but at the moment it creates confusion. This is why I created a section on the talk - I wanted to know what people wanted! Do people want it so that policy prevents main-page-linked semi-protection? Do they want it so that it is a 'dire circumstances' solution? Do they want it so that it is introduced as with other articles but maintained for a restricted length of time, or a restricted proportion of an article's time on the main page? Are there any other solutions out there to the problem that people have thought of? Robdurbar 21:19, 9 September 2006 (UTC)

Maybe this is the case. I've always mentally operated on the principles laid out in User:Raul654/protection and the fact that the policy here makes clear that semi-protection is a last resort. On Main Paged articles, we should be prepared, IMO, to go much more than just the proverbial "extra mile" and where we'd normally lock an article down if it were somewhere else, we should be striving to keep the first article(s) a user sees good to the claim "...which anyone can edit". It is disingenuous of us not to. The trouble of course is that the IRC channels make the Main Paged articles look positively terrifying in vandalism terms, when in fact the good edits are usually also rising in proportion to the edit rate (various statistics have shown this). Given that, and given the fact that these are show-cases of one form or another, we should be reluctant to semi-protect these articles.

The reason I reverted your particular change was that it simply said they "may be protected" when really the case should be that they "can be protected in emergencies/dire circumstances/things are totally out of control/under deliberate targetted attack" or some similar phrasing that puts sprotection beyond the reach of over-enthusiastic vandal fighters. (Who have the best intentions but who only see the bad edits, and rarely the good and have no counterweight to the hit-by-hit scrolling of the bots' lists.)

I don't think that a specific "x hours" thing is likely to fly because we'd be unlikely to find agreement on x. If some approximate codification of what amounts to current practise — which does indisputably agree to keep Main Paged articles unprotected as a working assumption — can be found, then perhaps it can be added. It's hard to see how to avoid creep with that, though, so no promises in advance... -Splash - tk 01:38, 10 September 2006 (UTC) (PS. I saw your message, forgot about it, carried on, and then got reminded by your project page edit.)

How to block vandals without accounts?

The article says: in the case of a few static IP vandals hitting a page (blocking the vandals is a much better option than semi-protection)... but there is no indication how this is done. Is it possible to block all users 58.160.185.* ? There has been persistent and sneaky vandalism to the Alexander Downer page, from users 58.160.185.97, 58.160.185.80 and 58.160.185.32 (obviously all the same person). Rocksong 00:10, 22 September 2006 (UTC)

Hmm. He's struck again from 203.5.217.3 . At a different time of day to all his other edits. I'm guessing that's his workplace, since WHOIS suggests Adelaide where it's 9.45am. Rocksong 00:17, 22 September 2006 (UTC)

Never mind. I got an answer at Wikipedia:Help desk. Rocksong 05:17, 22 September 2006 (UTC)

Yep. Range blocks. But honestly, they are only useful for very short times. Any longer than say 30-45 minutes and collateral damage is caused by them. --Woohookitty^(meow) 09:16, 22 September 2006 (UTC)

Semi as a defense vs linkspam?

If you look at history of Kraków (edit | talk | history | protect | delete | links | watch | logs | views) you will notice that it is hit with linkspam on the almost daily basis from various IPs. I looked through last 50 edits. Out of 24 anon edits - 23 were linkspam. Out of non IPs, 11 were lablelled as linkspam removal. That's means 2/3 of the edits in edit history are linkspam and it reverts! This leads me to a conclusion that we should consider pernamently semi-protecting this article; and further, that we may want to semi-protect other articles targeted by linkspam (which I'd assume would include cities and other places that are touristical attractions, if they are also target by linkspam vandals in such a manner). Comments?-- _{Piotr Konieczny aka Prokonsul Piotrus | talk}  23:58, 24 September 2006 (UTC)

How many thousand articles are you proposing a blanket semi-protection for? -Splash - tk 22:48, 3 October 2006 (UTC)

As many as needed.-- _{Piotr Konieczny aka Prokonsul Piotrus | talk}  19:29, 5 October 2006 (UTC)

What is the point of short term semi-protection for long term vandalism?

I can understand the concept of applying semi-protection for a limited time in the event of very specific vandalism, edit wars, etc. But if an article is vandalised a dozen times a day (as for example Computer and to a lesser degree Automobile are) - then the present policy seems to be to grant semi-protection for a while and then to withdraw it some weeks later (as specifically happened with both of those articles). Why is this a good idea? If vandalism is continual and ongoing from a wide variety of addresses then either semi-protection should be a good thing and therefore permenantly applied - or a bad thing and thus never applied. There is simply no logic to applying a short-term fix for a chronic problem.

Having recently experienced the benefits of semi-protection - (zero vandalism in either article) - only to have it yanked away again (half a dozen vandals in a day) - I would very much like for protection to be permenant.

I know the arguments against this - vandalism is typically corrected very quickly - and sometimes valuable contributions come from anonymous users and sometimes editing anonymously is important. Well, I think that the need to edit articles about (say) Terrorism in America might require some anonymity - but we're talking about automobiles and computers - hardly things where anonymous posting is essential for the continuing freedom of the author! So I'm going to dismiss the "NEED" argument and test the other two reasons. I decided to test those two hypotheses by looking at the actual edit histories (it's hard work - and perhaps a little subjective - but the numbers are sufficiently compelling that a wide margin of experimental error doesn't change the conclusions).

I found that of the 500 edits over the three months prior to semi-protection, Automobile was edited about 200 times by anonymous users and 300 by named users. Of those 200 anonymous edits, all but EIGHT were vandalisms of one kind or another and of the 300 named user edits, (inevitably) nearly 200 were reversions of anonymous vandalism. So we have a 200-edit effort in anti-vandalism paying for the rights of anonymous users to make just eight valid edits (against a background of 100 or so 'real' edits by named users). What of those eight? Well, six were VERY minor - to the point where, frankly I wouldn't have bothered. Only two added any significant material at all...and both of those changes were eventually overtaken by other edits and no longer contribute to the article as it is today. In all likelyhood, some of those anonymous edits were probably from named users who had simply switched computers and forgotten to log in or something - but that's just speculation.

So - if we had had semi-protection over those 500 edits, what would we have lost? NOTHING WHATEVER. What would we have gained? That's harder to quantify - but 200 wasted reversions represent a significant amount of editor's time that could presumably have been spent improving this or some other article. Furthermore, what about "vandalism is swiftly corrected" - well, that's something of a myth too. I didn't check over the entire three months or so of those 500 edits - but over a week or so, the article was screwed up maybe 35% of the time...and in almost every case the vandalism was some sort of obscene text. So - in addition to the wasted efforts of all of those editors, we can imagine that close to 35% of our readership who wanted to read about automobiles came away with an eyeful of crap. That might be acceptable for articles about very obscure subjects - but for Automobile and Computer, I think not.

I think there is a more subtle balance to be struck here. The number of vandalisms needs to be weighed against the number of regular editors who are reverting those vandalisms. In some of these unglamorous but 'headline' articles like Automobile, we have a small number of active editors - but a huge burden of vandalism. There are other articles which have a lot of vandalism (perhaps a lot more than Automobile - but if they have a large community of people watching them, then correction will none the less be swift.

IMHO, the case for long-term semi-protection of continually and frequently vandalised pages is extremely compelling (unless the pages are of a nature where anonymity of legitimate contributors might be required for compelling reasons of possible retribution or persecution of named contributors).

But if the decision is that semi-protection cannot be applied as a permenant solution to a permenant problem - then we should stop short-term semi-protecting for the purposes of long-term vandalism because short-term protection doesn't come close to fixing the long-term problem.

SteveBaker 05:27, 4 October 2006 (UTC)

Without wanting to comment too directly on the arguments of this piece, I think its worth remembering here that 'anonymous' editors who show their IP address are in fact far less anonymous than us registered users who work under a pseudonym and need not give any information about ourselves. --Robdurbar 08:31, 4 October 2006 (UTC)

We must continue to remain committed to openness for all, even for those who don't chose to log in or register. You cite the specific examples of automobile and computer above, but didn't acknowledge the many intangibles that semiprotection would have caused. For example, many of our users and editors were first attracted here because of the openness and the lack of a need to register at first. I, myself, first came here after editing an article, and then only after that became a registered user. I'm sure many of the other administrators and Wikipedians here came through the same process. By permanently semi-protecting these articles, we not only lose the valuable contributions of some non-logged in users (and even though an edit may not stand today, it has built the foundation upon the article that is here today), but lose many potential contributors. In addition, we must also recognize that vandalism will be inevitable - at what price are we to limit progress in an attempt to stem vandalism? Finally, also consider the argument that we could theoretically fully protect the vast majority of articles, most of which are rarely edited, to stop vandalism. Of course it would be effective in stopping vandalism - the articles can no longer be edited except by admins! - but at what price? At ending the open model that served us so long and so well. Permanent semiprotect would upset that balance of openness and editing that we have come to establish. Thanks! Flcelloguy (A note?) 13:53, 4 October 2006 (UTC)

You are arguing in absolutes - I am arguing from practicality and the balance of benefits. If EVERY article - or even 10% of articles were semi-protected, I agree that we'd be losing something important. If 0.1% of the most vandalised articles are semi-protected, I think we lose a negligable amount - but win enormously. The pragmatic balance is between the loss of a tiny number of edits to a tiny number of articles versus the saving of immense amounts of frustration on behalf of a large fraction of our existing editors. If you deal only in absolutes then the front page should be unprotected.

Certainly full protection is theoretically needed in order to avoid vandalism completely - but 99% or more of the vandalism goes away with mere semi-protection. If you deal with absolutes - your argument is convincing - but if you are pragmatic about things, then IMHO, it's entirely wrong. I also started with an anonymous edit (Red Squirrel if memory serves) - but those 'leaf' articles are not at issue here - they are almost never vandalised. It's the 'trunk' articles - with wide scope and obvious names like Computer and Automobile - that need protection. SteveBaker 19:54, 4 October 2006 (UTC)

Agree with Steve; see also my post above.-- _{Piotr Konieczny aka Prokonsul Piotrus | talk}  19:31, 5 October 2006 (UTC)

Addition

"This reasoning does not apply to pages that aren't articles, e.g. templates or pages in the Wikipedia namespace. Indeed, it may be useful to permanently (semi-)protect pages that would require familiarity with Wikipedia to make true improvements to, e.g. high-risk templates." It appears that people have been semi'ing HRTs recently so I figured it wouldn't hurt adding that to this policy. It seems quite reasonable to me that the maxim "everyone can edit the encyclopedia" does not necessarily apply to complex or frequently-used templates that require a bit of expertise to edit. >Radiant< 08:10, 16 October 2006 (UTC)

High risk templates are normally inexistent. There are high use/high visibility templates. For the high use, I would say over 10'000 transclusions is definitely a case for full protection (but this is not the sole criterion). An unprotected 10K template is a high risk. As such, high risk templates are now inexistent on this wiki. Semi-protection for <10K templates is a reasonable measure, as long as it is not advertised (please don't list them, please don't tag them). Mainting a page/category/tag-template that lists semi-portected templates is a high risk. Otherwise, I don't care what Splash reverts or doesn't revert on this page :-). Just use common sense and, yes, take a dip from IAR. --Ligulem 08:40, 16 October 2006 (UTC)

• Of course, the people who have been semi'ing HRTs (I'm not involved) have also been adding a "semi-protected" template to them. So assumedly they can all be found from there. Do you think this is risky? >Radiant< 08:51, 16 October 2006 (UTC)

• As I wrote above: Yes, this is risky, please don't tag them (that's why I've put template:Sprotected_template on tfd). --Ligulem 09:01, 16 October 2006 (UTC)

But your addition went considerably beyond just templates: it said that it does not apply to pages that are not articles. That covers some 12 or 13 (?) namespaces and is sweeping in its assumption. Freely-editable applies to all pages without exception, save for those that are mission critical (the Main Page, copyright pages etc). It applies also to templates. That a sort of quivering-in-fear-of-an-edit approach means we'd prefer to see no improvement to certain parts of particular pages does not result in an extension to everything that is not an article. I am also a little disappointed at the advocation of smoke and mirrors in what is supposed to be an open project. Those who are the target of the smoke and mirrors will quickly see through them; those that cannot see through them are likely to be those with benign intent who are thus unable to contribute where they otherwise might. -Splash - tk 17:09, 16 October 2006 (UTC)

• I see that my wording was overly broad. However, nobody is quivering in fear of edits. Rather, it is acceptable (within strict boundaries) to place limits on what certain parties can edit, in instances where the edits of said party have been shown to do (far) more harm than good. It is for this reason that anons cannot create pages, that new users cannot move pages, and that IP ranges can be blocked. It is quite reasonable to disallow (through semi-protection) new users to edit pages that require familiarity with the wiki to edit properly, in particular oft-used templates. >Radiant< 13:36, 17 October 2006 (UTC)

Protecting user talk pages

I have removed this provision, as it is absolutely an unacceptable provision. No user who is dealing at all with vandals, new, or inexperienced editors should make it impossible for those users to contact them with questions or for an explanation. Phil Sandifer 18:59, 16 October 2006 (UTC)

Your version would also make protection of user pages which the users remove warnings unprotectable unless you block th users...You may want to modify the blanking of warnings template which warns of protection of the user page. See Template:Wr, and what about users who are not admins that are being targeted by vandals, your version would prohibit this protection as well... You therefore may want to modify your blanket change about no protections. --PinchasC | £€åV€ m€ å m€§§åg€ 23:42, 16 October 2006 (UTC)

"Should not be used to protect user talk pages that are the subject of vandalism" work for you? Phil Sandifer 14:28, 17 October 2006 (UTC)

Editorialising policy

Things like [1], [2] are highly editorial changes, to the extent that the edit summaries are baldly misleading. SlimVirgin, what was the reason behind your rewrite? -Splash - tk 19:29, 23 October 2006 (UTC)

It wasn't a rewrite, but a copy edit. Some of the writing in the version you restored was poor and unnecessary, or entirely subjective. What is "serious" vandalism from multiple IPs or accounts, as opposed to vandalism from multiple IPs or accounts? Saying admins have the ability to "temporarily" sprotect, when what you mean is they have the ability to sprotect. Saying it is a temporary measure," whereas in fact it is "usually" a temporary measure: the continuously protected pages are testimony to that.

It seems to me that it's you who wants to editorialize and prescribe how sprotection ought to be used, rather than simply describing how it is in fact used. But regardless, the page should be written tightly if possible. SlimVirgin ^(talk) 07:50, 24 October 2006 (UTC)

It wasn't a copyedit, it was a rewrite. It might (now) describe how you use it, but that's fairly idiosyncratic. Most admins don't protect articles permanently, and most consider the tag more or less par for the course. Articles from the main page are not protected just because they get attacked; only when the attack is extraordinary. The reason the stuff about durations got added in (I forget by who; it's probably further up here somewhere) was because last time you took it out, the page left no guidance of any kind about what non-admins might expect/request nor to new admins about what to apply. To answer your edit summary, I already had put a note on the talk page, so I'm not sure why you asked me to do so. -Splash - tk 13:09, 24 October 2006 (UTC)

It's not a rewrite at all (no substance is changed), but the writing needed tightening and so I did it. You were complaining recently on AN about people not using the tag. The fact is that the tag is ugly and it's not good to have it on pages protected for longer periods, as was discussed on the mailing list; there was consensus to change it to an image of a small lock or similar, but a couple of editors opposed it and the others couldn't be bothered arguing, so the ugly templates remains and is therefore not popular. I don't use sprotection at all idiosyncratically; could you say why you think I do? Please look at WP:PP and you'll see how it's being used. This policy should describe that. SlimVirgin ^(talk) 13:17, 24 October 2006 (UTC)

The hoo-haa over what the template should or should not contain bemused me not inconsiderably; I'm surprised you're still on that. But I quite often see people go and add it to sprotected articles that didn't get tagged because, well, I'd be guessing, but I suppose they feel it appropriate. Mailing list consensus, shmailing list shconshenshush. It most certainly is a substantive change to say that anything on the main page can just be locked down, to say that if more than one static IP is hitting a page, it should now be protected rather than them blocked, to go from not pre-emptive to pre-emptive when you feel like it (someone else took that out before, too), and to go from heavy vandalism to merely vandalism.

Idiosyncratic is protecting pages that get literally a handful of bad edits, and leaving them protected for ever. -Splash - tk 13:57, 24 October 2006 (UTC)

I reversed those things I identified above but left what you term your tighter phrasing in place. It's not ok to widen protection policy one edit at a time in the name of improving the grammar. I don't see WP:PP describing the way you wrote it; and even if it is, that would be descriptivism gone wrong. A policy shouldn't be amended to describe practises that are wrong just because someone/people happen to misunderstand sometimes! -Splash - tk 14:21, 24 October 2006 (UTC)

I'm fine with your changes, except the talk page thing. User talk pages are sprotected if they're being vandalized or people are being harassed. SlimVirgin ^(talk) 15:47, 24 October 2006 (UTC)

I took it from the section above this. Phil Sandifer has a good point. Admins who are regularly blocking new users/unregistered users should not have their talk pages protected. -Splash - tk 15:59, 24 October 2006 (UTC)

Splash, what does this sentence mean exactly? "Those of users engaging with new and/or unregistered editors should be protected sparingly to permit communication." If someone is being harassed, are you saying that, before we help them by sprotecting their talk page, we first of all have to check (a) whether the victim is an admin, and then also (b) look through his block log to see whether he's blocked a lot of unregistered editors and new accounts? And if he has, he has to tough it out? SlimVirgin ^(talk) 16:05, 24 October 2006 (UTC)

Along those lines, yep. If you're interacting with the users who cannot talk to you, particularly those who you are preventing from editing, you need to be available to talk to them. This is why people insist on email enabling during RfA and the like. Vandalism isn't the same as harrassment though; I guess your talk page is permanently protected because of harrassment rather than because a bunch of silly newbs were playing with it. -Splash - tk 21:12, 24 October 2006 (UTC)

People can always e-mail admins if their talk page is sprotected. I don't think female admins should have to put up with being called c***ts and being told what sexual practises they're going to be subjected to, just because they blocked someone's sockpuppet. That's the kind of situation, whether you call it vandalism or harassment, where talk page sprotection is important. SlimVirgin ^(talk) 21:20, 24 October 2006 (UTC)

That would be harrassment, not simple vandalism, and would not have arisen as a result of interactions with new or unregistered users. -Splash - tk 21:27, 24 October 2006 (UTC)

Well, of course it might. A lot of it comes from vandals who don't like seeing their nonsense rolled back. SlimVirgin ^(talk) 23:34, 28 October 2006 (UTC)

Handling apparent abuse of this feature?

I'd like to see at least a quick note in the article on "what to do in cases of apparent abuse of semi-protection." Thanks. -- Writtenonsand 11:24, 24 October 2006 (UTC)

Post about it on WP:AN/I.Voice-of-All 00:59, 25 October 2006 (UTC)

Proposed amendment to this policy

I submit that this policy should be amended, to improve the integrity of official policy pages by permanently semi-protecting them, thereby preventing much vandalism and many good-faith but ill-advised edits against consensus by new and unregistered users. Please see Wikipedia:Permanent semi-protection of official policy pages. John254 03:25, 25 October 2006 (UTC)

Consensus having been reached on Wikipedia talk:Permanent semi-protection of official policy pages to enact this amendment, I added it to this policy on October 28, 2006. [3] John254 02:52, 28 October 2006 (UTC)

I have restored this amendment. As JYolkowski explained on Wikipedia_talk:Permanent_semi-protection_of_official_policy_pages: "It is absolutely not a requirement that changes to policy pages be discussed beforehand [at all]" [4] Consequently, purely procedural objections to the manner in which Wikipedia:Permanent semi-protection of official policy pages was enacted are unpersuasive, especially since Wikipedia is not a bureaucracy. Only one editor objected to the policy proposal, and all objections were rebutted. Furthermore, five established editors (including myself) supported it. Additionally, the arguments in favor of semi-protecting official policy pages were far stronger than the arguments against it. I don't see why this is a "drastic new policy proposal" [5] -- it merely creates a small exception to the general theme of this policy for official policy pages, which new and unregistered users are highly unlikely to edit in beneficial manner. Finally, the decision of two administrators to act on this policy amendment (see [6] and [7]) provides further evidence that there really is consensus for this amendment. John254 05:58, 28 October 2006 (UTC)

There is patently not consensus on that page (I objected, for example), and there hasn't been anywhere near enough time or discussion to form one. JYolkowski // talk 13:50, 28 October 2006 (UTC)

One solitary substantive objection to a policy proposal does not constitute a lack of consensus where there is otherwise substantial support for a proposal. The use of administrative rollback in content disputes is not an appropriate use of administrative privileges: per Wikipedia:Administrators,

Administrators have a faster, automated reversion tool to help them revert vandalism by anonymous editors. When looking at a user's contributions, a link that looks like: [rollback] – appears next to edits that are at the top of the edit history. Clicking on the link reverts to the last edit not authored by that user, with an edit summary of (Reverted edits by X to last version by Y) and marks it as a minor change. One-click rollback is only intended for vandalism, spam, etc.; if reverting over disputed content, it should be done manually with an appropriate edit summary.

JYolkowski's claim that determining consensus requires lots of time and discussion contradicts his claim on Wikipedia_talk:Permanent_semi-protection_of_official_policy_pages that "It is absolutely not a requirement that changes to policy pages be discussed beforehand [at all]" [8] I additionally note that JYolkowski has yet to respond to the rebuttals to the arguments he presented against this amendment on Wikipedia talk:Permanent semi-protection of official policy pages, even after removing it from the policy using administrative rollback. John254 14:32, 28 October 2006 (UTC)

Yes, it was improper for him to use administrator rollback, but nonetheless it is not necessary to discuss uncontroversial changes, such as grammar and typo fixes, but I strongly object to the inclusion of this proposal. (Guess that makes it two sysops for and two sysops against if you're counting, but that's wholly irrelevant anyway, heh). I've made my comments on the proposal page, but won't edit war any further to get this removed. Instead I ask that you please remove it while this is sorted out, and seek a wider consensus. And wee, we're making a wave of threaded conversation.. I hate when that happens :/ Cowman109^Talk 14:48, 28 October 2006 (UTC)

I have posted something on the village pump to hopefully attract a wider audience to this discussion. JYolkowski // talk 18:46, 28 October 2006 (UTC)

John, I agree that policy pages should be semi-protected; in fact, I'd go even further and introduce minimum-edit qualifications or something. However, it's something that's been proposed before and while many supported, many objected too, so it needs thorough discussion before being added here. SlimVirgin ^(talk) 20:14, 28 October 2006 (UTC)

I find the manner of 'implementation' here rather unpleasant. "I propose this! I hear some objections! They are wrong to object! Thus this is policy! No, you may not say it isn't!". More substantively, it doesn't matter in the least if policy pages get some vandalism, since it doesn't actually change the policy. You don't suddenly get a barnstar for vandalism if someone changes WP:VAND to say you do; you don't suddenly get to write opinionated articles if someone changes WP:NPOV to say you can. The vandalism does no damage - revert it, and move on. Don't start using pretty blunt instruments to 'solve' a 'problem' that doesn't actually exist. Now, policies like WP:NPOV and WP:NOR are the subject of much edit-warring and disagreement, particulary from people who don't understand that even if they shoehorn a new piece of law-esque stuff in that says they can do the silly thing they can't otherwise do, they still can't do it. THis I suppose is the idea behind SlimVirgin's edit-countitis on policy pages - again, though, these people being wrong doesn't actually do any damage to the policies, since their changes never stick. -Splash - tk 23:35, 28 October 2006 (UTC)

The changes don't stick, but making sure they don't can involve a lot of hassle and time-wasting. It's not just vandalism, Splash, but people with three edits and 10 minutes experience turning up to make changes; but, because it's not vandalism, we can't just roll back, can't violate 3RR, but have to discuss, persuade etc. It is sometimes very time-consuming, especially when you get a few of them at once. SlimVirgin ^(talk) 23:41, 28 October 2006 (UTC)

I think that the solution here is not to restrict editing to experienced users, but to explicitly allow rollback and maybe 3RR exemptions for generally unhelpful policy changes, even if they aren't vandalism. I think that the criteria under which admin rollback may be used are far too narrow right now (Note: This is not an excuse for my use of rollback as discussed above). JYolkowski // talk 20:16, 29 October 2006 (UTC)