Wikipedia talk:Semi-protection policy/Archive 3

From Wikipedia, the free encyclopedia

This 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.

Contents 1 Rehashing 2 By the way.... 3 An alternate proposal 4 Idea from Talk:Wikipedia 5 How much is too much? 6 Articles for creation experiment 7 great idea 8 Concern: Vandals make more usernames 9 Semi-protection proposal v.02 9.1 Semi-protection proposal v.02 straw poll 9.1.1 Support 9.1.2 Oppose (please explain why) 9.1.3 Neutral (please explain why) 10 Question about the KISS principle 11 Why base it on how long a user has been registered? 12 Back to the nut 13 A note to those that oppose this proposal 14 Banner should read "...has been semi-protected..." 15 Straw poll deadline 16 Ads 17 Not a good idea 18 Can a semi-protected article be edited as normal... 19 Improving the downsides 20 An issue for later 21 Inferior to delayed gratification 22 A textbook case for semi-projection 23 Another text-book case: Stanley Williams 24 question for developers 25 My personal $0,02 26 Concern from LV 27 My opinion 28 My 2 Eurocents ( 0,024$) 29 My 2 (African) American cents 30 Support

Rehashing

I've invited several users who have recently reverted vandalism on GWB, and a couple have responded at my talk page that they think there's simply too much text here to be useful to them, and were wondering if we could briefly rehash the stuff we've discussed. Anyone chiming in (NOT DISCUSSING) would be helpful. Thanks in advance, Mysekurity^{(have you seen this?)} 04:14, 5 December 2005 (UTC)

1. There's a proposal on the front page that is quick to read. It institutes a level of protection that prevents anons and very new accounts from editing an article.

   Note that anons already can't edit protected pages, so this doesn't amend their status in that regard.

2. It is intended for use only as a vandalism reducing tool, so will have the same usage patterns as {{vprotected}} does at present.
3. Some people would prefer more levels of semi-protection based on whatever criterion we are using as the threshold.
4. Some people would like automatic expiry of semi-protection.
5. Significant concerns include:
   1. Protection creep across articles
   2. Protection creep up the scale
   3. Vandalism being displaced to other pagesBen Aveling

-Splash^talk 04:22, 5 December 2005 (UTC)

Thanks Splash, that was a very good and even-handed rehashing. Are there any questions for our newly-enfranchized users? Tip: When away from the page, use the history to see what you've missed. I personally think MediaWiki needs a better system for talk pages, but at present, I guess this works okay. -Mysekurity^{(have you seen this?)} 04:32, 5 December 2005 (UTC)

Here's my take: (very briefly)

1. Simple is good - I've been in the software business, and you rarely implement things "perfectly" the first time around. You do things incrementally -- especially when you are making changes to established software. There may be a need for increased stratification of protection levels, but one extra one seems okay as a first step. You need somewhere to start, and you need something as a proof of concept to begin with.
2. I would prefer to see something based on the newest x% of users. Simpler to implement, simpler to change.
3. I think the community is robust enough to handle adequately the sorts of consequences which may happen as a result of its implementation. It's good we are thinking about this, but it does not seem productive if we are overly concerned with all the nuances and subtleties of its usage.
4. I don't think "protection creep" will occur to the extent that we may worry about. This sort of semi-protection seems only applicable to a very small subset of articles on Wikipedia. --HappyCamper 04:40, 6 December 2005 (UTC)

I agree completely with HappyCamper, and suggest we move more and talk less. Titoxd^{(?!? - did you read this?)} 07:12, 6 December 2005 (UTC)

Agreed. What's the next step? Judging from the comments under #Semi-protection_proposal_v.01, it is time for another version addressing the concerns. Once we have consensus on an acceptable version (even if the next version isn't it), what comes next? —HorsePunchKid→龜 2005-12-06 07:26:28Z

I'd hazard that the version currently on the front page comprehensively supercedes v.01. That was why I wrote it. v.01 can be kept up our sleeves for future deployment once we understand better the sociology of SP. -Splash^talk 13:34, 6 December 2005 (UTC)

IIRC, in order to become policy, a proposal must be approved by the community with 80% support. Once we get that, we can ask a developer to code it (the move-restrict analogy will really make coding easier), run it for a while, then send it to Meta. But we have to agree on something here first. Titoxd^{(?!? - did you read this?)} 07:33, 6 December 2005 (UTC)

Yes, what's on the project page needs a rewording. What does IIRC mean? "It is reasonable to conclude"? I'm not sure where to begin though. --HappyCamper 01:21, 7 December 2005 (UTC)

Agreed; the project page needs to at least reflect what has been agreed upon so far. Or is the intent to keep the project page small to avoid intimidating people or garnering more "but it's creep!" protests? (P.S. IIRC: If I recall correctly.) —HorsePunchKid→龜 2005-12-07 02:08:42Z

The project does represent what has been agreed on so far, in my understanding. -Splash^talk 02:31, 7 December 2005 (UTC)

It is a conservative subset of what is accetable to most. There seem to be things that everyone who has commented so far agrees upon in principle that are not reflected on the main page (for example, the process of inititiating semi-protection). The majority of the disagreeable portions under the three headings for v.01 seem like they would be trivially addressed, and the supporters of v.01 do not seem to be voicing much objection to the points that the disagree-ers have raised. Hence the time may be ripe to explicitly codify a v.02 and seek consensus again, with the goal of updating the project page with more of the points that have been agreed upon. —HorsePunchKid→龜 2005-12-07 03:08:46Z

I don't think there's any agreement at all on any of the elements of v.01, and I would oppose each and every one of them. -Splash^talk 03:12, 7 December 2005 (UTC)

So you also oppose most of what's on the project page then? Maybe it would be most constructive for you to write a separate proposal, then. I, too, opposed each specific formulation in v.01, but as I said, it looked like quite a few people opposed for the same reasons, with no objection from the supporters (perhaps my memory or reading comprehension is being selective? :)). Thus, again, it seems it would make sense to come up with a second proposal that is closer to being agreeable. How else is this going to move forward? —HorsePunchKid→龜 2005-12-07 05:42:54Z

I wrote the current proposal. I oppose >1 level of semiprotection, I oppose hard and fast rules for instituting or releasing SP and I oppose anything other than a simple tag on articles indicating their protection. The current proposal has none of those things, and v.01 has them all. -Splash^talk 12:30, 7 December 2005 (UTC)

I agree strongly with Splash here: policy should never be made more complicated than absolutely necessary. A single level of semi-protection and flexible rules on establishing semiprotection are exactly the right way to proceed. Tim Pierce 13:40, 7 December 2005 (UTC)

I think that we should start with only one level of semi-protection, and go from there if necessary. As HappyCamper said, it's hard to foresee exactly how the feature might develop in the future, so it's best to start simple and go from there if necessary. As for how the semi-protection should be placed and removed, I think consensus at WP:AN would do a good job in figuring that out. Titoxd^{(?!? - did you read this?)} 20:37, 7 December 2005 (UTC)

Ah, I see! Thank you for clarifying. For some reason, I was starting to think you were strongly opposed to even what was on the project page (v.00 for simplicity's sake?). I see now that you object mainly to the complexity introduced between v.00 and v.01, and that indeed most of the shared objections above fall under this (extra levels, extraneous links in the template, more rules for admins to have to grapple with when applying it). For what it's worth, I would favor v.00 over v.01; an obvious reason is that because it is simpler, it is more likely to garner enough support to be enacted in a reasonable time frame! —HorsePunchKid→龜 2005-12-07 21:05:20Z

My thoughts exactly. Also, having a single level of SP inherently prevents protection creep upwards, since there is only the full-blooded full protection above it, and we already use that sparingly. The concerns about creep across articles will have to be addressed by vigilance and AN/I shouting matches. If it helps clarify, v.01 above was written before the current project page. I wrote the project page as a suggestion here under #KIS(S) (subsequent to having problems with v.01) and it was fairly quickly suggested it migrate to the front page, so it did. If anything, it is v.02. -Splash^talk 21:18, 7 December 2005 (UTC)

Write up a v.02 on the talk page then? --kizzle 21:27, 7 December 2005 (UTC)

v.02 is on the project page, so no need for a duplicate. That said, we might need to hold another poll like we did for v.01 soon. Titoxd^{(?!? - did you read this?)} 21:29, 7 December 2005 (UTC)

Am a little confused by the triplication we now have. We've the project page which appears not to be objected to (although people have sought to expand onit), it appears above in #KIS(S) and now it appears below in poll, of all things, too. Consensus can be guaged from discussion in which there is significant agreement or where those in disagreement indicate thay they can live with it. There's no need to go bean-counting. Anyway.-Splash^talk 21:53, 7 December 2005 (UTC)

I had a quick question that I haven't seen anywhere else. What exactly, currently, is the 1% limit? 15mins, 20days? I *do* know we have what, 700,000 accts? This means, that one would have to create 7,000 accounts to influence this limit. I would rather see a set time period. 1d, 1w, whatever, I dislike using a variable like a %, as this enables a user to at least think he/she/it may 'game' the system, like, for instance, hooking /usr/dict up to a RNG, in conjuncture with a long proxylist, in an attempt to influence the average age of that bottom 1% of accounts. Just a thought, but I would love to hear what y'all think of this.... --негіднийлють ^{(Reply|Spam Me!*|RfS)} 06:06, 16 December 2005 (UTC)

By the way....

Anon page creation is going to be disabled temporarily. Titoxd^{(?!? - did you read this?)} 17:48, 5 December 2005 (UTC)

Actually, that's only anon creation of new pages. The email says clearly that anons will still be able to edit. -Splash^talk 17:50, 5 December 2005 (UTC)

I know, I had seen my mistake and corrected it [1] before you posted here. Titoxd^{(?!? - did you read this?)} 17:52, 5 December 2005 (UTC)

A direct link to the letter: http://mail.wikipedia.org/pipermail/wikien-l/2005-December/033880.html xaosflux ^T/_C 01:54, 6 December 2005 (UTC)

An alternate proposal

What if we did something like this.

1. Track unproven users in some way similar to what's already been discussed.
2. When an unproven user makes an edit, set a flag in the DB on that edit, but otherwise treat it normally.
3. The next time the page is edited, insert:

• the usual heading (eg Editing Wikipedia talk:Semi-protection policy (comment) )
• a message at the top of the page saying something like "Wikipedia cannot be sure if the previous editor(s) are experienced or not. You might like to take a moment to review their changes."
• a diff of the current version and the last edit by a proven editor
• the usual edit buttons and subject/headline and edit box and everything, exactly the same as usual

Nobody is prevented from doing anything. Nobody can do anything they couldn't normally do. Nobody has to do anything more than they do now. It would just help experienced users do what they can already do, if they want to.

Regards, Ben Aveling 21:11, 5 December 2005 (UTC)

Sounds good to me. +sj + 00:06, 14 December 2005 (UTC)

It's too close to a content review, which has been continuously shot down by the community. Besides, article validation is supposed to go online soon, which may address some of those problems too. Titoxd^{(?!? - did you read this?)} 07:14, 6 December 2005 (UTC)

Where has article validation as a concept been shot down? People have not agreed on any particular suggested implementation, but this sounds like a very mild and constructive one, without the downsides of other suggestions that would require (rather than suggest) validation. +sj + 00:06, 14 December 2005 (UTC)

That's not what I said. Article validation is going ahead; content review is what has been shot down several times. An interesting example was Alan Dershowitz, where Jimbo just dumped a whole new role to admins by making the page admin-only; see all the protest in his talk page about that. Titoxd^{(?!? - did you read this?)} 01:04, 14 December 2005 (UTC)

Idea from Talk:Wikipedia

Ahoerstemeier (aka andy) presents an interesting concern at Talk:Wikipedia#Vandalism. In a nutshell, he is concerned that protecting (or semi-protecting) frequently vandalized pages might cause vandals to spread out and vandalize a wider variety of pages, where their vandalism might not be caught as quickly as it would be on pages that admins are watching closely, such as GWB and Wikipedia. --TantalumTelluride 00:24, 6 December 2005 (UTC)

On the other hand, it's much less interesting to vandalize pages that aren't currently cool or in the news. +sj + 00:06, 14 December 2005 (UTC)

That's exactly the point of this policy. Determined vandals have been here for a while, and have been enabled by the ability to hit big articles. If we can stem vandals from their source (if only for a little bit), the world would be a better place, I think. High-profile pages, while more likely to be watched by a greater number of users, are much more likely to be vandalized than lower-profile pages. -Mysekurity^{(have you seen this?)} 05:22, 14 December 2005 (UTC)

How much is too much?

Assume we try this out. How many pages would have to be semi-protected before we decide it's not working and turn it off? 30? 300? 3000? Regards, Ben Aveling 07:05, 6 December 2005 (UTC)

My guess would be less than 30, more like a dozen. We should be able to see a difference one way or the other and track it over a period of a week or two. Lately, the vandalism to GWB is down slightly, but that may coincide with the probablity that these talk pages are being read by vandals....--MONGO 08:17, 6 December 2005 (UTC)

A good answer to a question I didn't mean to ask.  :-) What I meant to ask is, if the number of semi-protected pages creeps to 300, is that a problem? Is 3000? I guess so long as we can turn it off without causing a big fight, it can't become a problem? Regards, Ben Aveling 10:44, 6 December 2005 (UTC)

Articles for creation experiment

Personally, I think this is a good idea, but I'm wondering what everyone else thinks. Does anyone believe that this "experiment" [2] is really temporary, or feel that its implementation is un-wiki? Can't sleep, clown will eat me 08:58, 6 December 2005 (UTC)

I can't believe its temporary. I probably would have voted against it if asked, but looking at AFC it does seem to be working. There certainly feels like less cruft appearing in recent changes, and there are still plenty of anon's editing pages. IMHO, of course. Regards, Ben Aveling 11:16, 6 December 2005 (UTC)

I don't think anyone believes it to be temporary, and I don't believe it to be anti-wiki either, whatever that means. Although this definitely will not prevent the types of mistakes that were made in the Seigenthaler case, it is a positive step in the right direction towards much-needed quality control; we now have page moves limited to N% of registered users and new page creation restricted to registered users as well. It's obvious to where we are heading, but I do sincerely wish that Mr. Wales would comment on this proposed policy we've invested so much time discussing. Hall Monitor 18:04, 6 December 2005 (UTC)

Heading towards 'restriction' in this fashion is a quick hack to reduce vandalism. Better, more subtle solutions would not feel like restriction; but rather like community-assisted guidance -- which is where I hope this proposal will direct itself as well. +sj + 00:06, 14 December 2005 (UTC)

For the curious, anon page creation is being discussed here and some of the effects so far can be seen at WP:AFC. Regards, Ben Aveling 11:12, 6 December 2005 (UTC)

AFC is a fine idea, in general. We should surely offer new users a place to make such requests, since they are often full of good ideas but scared of creating a new article on their own -- what if I get it wrong? &c. I don't believe restricting new-page creation is a good idea; it is one of the simplest ways to get new users hooked on the system. And the restriction seems to have cut back on the creation of useful new articles by 25%. +sj +

great idea

But it misses a big bit!!! Users that are vandals should not be able to edit pages that are protected like this. I suggest a way of marking vandals as vandals, as well.
Also that certaint users that remove vandalism a lot and are not yet admins can still use this. --Adam1213 Talk + 09:15, 6 December 2005 (UTC)

Semi-protected pages would be editable by anyone with an account older than 1% of all current accounts. That would essentially mean accounts older than 3-4 days. So yes, any non-admin who has been here for more than a week could edit the pages. IPs could not edit semi protected pages, which would stop most of the vandalism, if not all. --Woohookitty^{(cat scratches)} 10:49, 6 December 2005 (UTC)

It is impossible to discern which users are vandals until they vandalise. The proposed semi-protection is a much more liberal version of the current protection scheme. If a page is being vandalised, I would much rather have it open to registered editors with an account older than 1% of all current accounts rather than lock it out completely. Hall Monitor 18:05, 6 December 2005 (UTC)

Established vandals are blocked indefinitely. That's a better solution than any kind of protection. -Splash^talk 18:21, 6 December 2005 (UTC)

Indefinite blocks are not always an option, such as the case with AOL users due to collateral damage concerns. The same applies to school proxies; it generally takes at least 10 to 20 temporary blocks due to severe vandalism from a school IP address before something more permanent is discussed. Unfortunately, from what I've observed, the bulk of our vandalism originates from either an AOL IP or a school proxy. Hall Monitor 18:30, 6 December 2005 (UTC)

I figured Adam1213 was referring to accounts >1% old that were known to have vandalistic propensities. These are usually killed off as "vandalism only" accounts in fairly short order, or find themselves without an admin to unblock or whatever. I agree that most vandalism originates from IPs, and hope that this version of semiprotect will nobble quite a bit of that on articles where it is a serious problem. (I'm a he, btw.)-Splash^talk 19:08, 6 December 2005 (UTC)

Aye, my apologies. Hall Monitor 19:27, 6 December 2005 (UTC)

Concern: Vandals make more usernames

First of all, I think this is a great idea. However, I am a bit concerned that this may lead to would-be vandals simply creating many, many more user names. It is relatively trivial to register two usernames each day, and that would mean that begining 4 days later, you would have 2 names per day to vandalize with. My real concern here is that, for instance, in my war with vandals on the Chuck Norris article, seeing contributions from IP addresses makes it a bit easier to spot potential vandals. If all vandals determined to add nonsense started using multiple new accounts (because they cant just use multiple IPs anymore), this could get extremely difficult for everyone. -Lanoitarus 00:15, 7 December 2005 (UTC)

I suspect the reduced volume of edits would offset the loss of the convenient red flag that IP addresses raise. I'd rather check a few edits from registered users who have made dozens of edits than dozens of edits from anonymous users who have made few edits. I doubt we'll know for sure whether the tradeoff is a good one before we can actually test semi-protection. —HorsePunchKid→龜 2005-12-07 03:10:08Z

I certainly agree it is worth a shot-- I simply think this is a potential issue which is worth considering and being prepared for. Perhaps some mechanism to flag multiple usernames being created from the same IP address in rapid succession? Or perhaps modify the requirement so that a certain number of edits need to have been made to non-protected pages (this admittedly may cause contamination of other articles, so it is only a rouge idea at best. -Lanoitarus 04:21, 7 December 2005 (UTC)

Would only allowing like 3 logins per ip address be anti-wiki? --kizzle 05:39, 7 December 2005 (UTC)

Certainly you cant restrict the number of accounts that can USE each IP, because of public terminals... Perhaps a slightly more lenient system where there is a time-based limit on multiple account creation from a single IP? For instance, if you create more than a handful of accounts in one week from the same computer, you go on "cool down" and cant create any new ones till next week? It seems that this would almost never affect legitimant users, and even when it did the damage would only be their having to wait a few days to create an account. -Lanoitarus 06:21, 7 December 2005 (UTC)

It already exists: 10 account creations per IP per 24 hours. And that still runs into problems when Joe Professor tells his entire class to sign up for wikipedia and yet they all reside behind the same proxy. Dragons flight 06:27, 7 December 2005 (UTC)

Hmm, didnt know that policy existed, nor had I thought of the proxied computer lab scenario. Would it be feasable to flag accounts created "excessively" from the same IP in some way, then? Maybe so that they would have a longer period where they couldnt edit semi-protected or so that they get flagged in some way on history pages for the first couple weeks? Seems a bit big-brotherish, to be sure, im just throwing out thoughts here. -Lanoitarus 06:33, 7 December 2005 (UTC)

That is a concern of mine as well. The good news is that we can try this to see if it works or not. If it doesn't we can go back to te current status=quo. Not trying is not acceptable... ≈ jossi fresco ≈ t • @ 05:56, 7 December 2005 (UTC)

As above, I agree with you that it is certainly a good idea worth trying. Before trying, however, I think it is at least worth considering this issue and brainstorming ways we can prevent it. -Lanoitarus 06:18, 7 December 2005 (UTC)

We do have Special:Contributions/newbies to watch them, though... Titoxd^{(?!? - did you read this?)} 01:12, 9 December 2005 (UTC)

Semi-protection proposal v.02

Ok, since there seems to have been progress and we're now agreeing on the core of the proposal, let's verify the consensus on this page with a poll, just like we did for v.01. If there's agreement, then we'll send the entire proposal to a community-wide vote. Titoxd^{(?!? - did you read this?)} 21:37, 7 December 2005 (UTC)

---

Semi-protection of a page prevents the newest X% of registered users and all unregistered users from editing that page.

Semi-protection will only be applied if the page in question is facing a serious vandalism problem. It is not an appropriate solution to editorial disputes of any kind since it may restrict some editors and not others. Administrators will thus apply semi-protection in the same manner as current protection against vandalism is applied — either on their own initiative or following an alert on an article's talk page, WP:RFPP, WP:AN/I or some other relevant page.

Requests to lift semi-protection should generally be unnecessary in the same way that unprotection against simple vandalism at present is generally swiftly seen to by either the protecting admin or another. Generally, a simple note to the talk page or WP:AN/I should be sufficient, but WP:RFPP can be used if necessary.

Articles that are semi-protected will be indicated with {{sprotected}} (or similar) and listed at WP:PP in the same way as protections are at present.

Suggested template (note the links):

As a result of recent vandalism, editing of this page by new or anonymous users is temporarily disabled. Changes can be discussed on the talk page, or you can request unprotection.

Note that with full protection at present, anonymous editors are prevented from editing the article in the same way as are all non-admins. This proposal does not restrict unregistered editors more than they already are in the case of protection. This is not a proposal to prohibit anonymous editing.

Semi-protection proposal v.02 straw poll

Question about the KISS principle

If we keep it purely simple right now so that it will have more of a chance to be accepted and the developers make the changes we want, how likely is it that they will want to make code changes again later on down the road for issues that should have been solved in the first place? Secondly, does v.02 specify that the tool from the software end will have a fixed time limit or a variable time limit that will be controlled by policy? --kizzle 03:40, 8 December 2005 (UTC)

My understanding (CMIIW) is that no explicit time limit is defined. The timing policy is inherited from the current protection policy, minimizing the amount of new knowledge that admins and other editors need to assimilate in order to take advantage of semi-protection. Note "Administrators will thus apply semi-protection in the same manner as current protection against vandalism is applied…" and the paragraph following that line, for example.

I don't think the developers will want to make code changes regardless of when they are asked about it. I suspect that regardless of whether we choose percents or explicit intervals, semi-protection has the same odds of being proven effective or ineffective. (As long as the percent translates to roughly the same time interval.) Hence my support of this version, even though I think an explicit time would more accurately reflect our intent. —HorsePunchKid→龜 2005-12-08 08:13:48Z

Why base it on how long a user has been registered?

Instead of allowing a user to edit a semi protected page based on how long they've been registed, why not base it off how many edits they've made? Reub2000 05:24, 8 December 2005 (UTC)

As discussed earlier (we refer to this as v.01), this would promote editcountitis, be less technologically feasable (as we already have the code for the pagemove function, and it is potentially taxing on the server), and does not show dedication to the project by being there a few days, and making productive edits. There are problems with both proposals, and hopefully a middle ground can be reached. -Mysekurity^{(have you seen this?)} 05:45, 8 December 2005 (UTC)

Agreed. Edit counts are very easy to fake, and would actually lead to more vandalism (if i cant edit this page yet, ill go make 5 edits over somewhere else and come back). Im not sure why 1% is used instead of a fixed time period, however. -Lanoitarus 07:17, 8 December 2005 (UTC)

It has been stated farther up (Concern: Vandals make more usernames) that you can easily register an acount (or even many), wait, and then make your vandalism. An edit count means that the user actually has to do something in order to be able to edit semi-protected articles. And if a user decided to use vandalism to get to the required edit count, then they should be banned before they reach the point where they are able to edit semi-protected articles. Reub2000 08:03, 8 December 2005 (UTC)

Your questions are answered here on this page and several times over in the archives. Can't sleep, clown will eat me

Primarily vandalbots which can make edits very quickly. Ideally, it should be based both upon a time limit and a minimum edit count to combat both vandalbots and someone who simply registers, waits a couple days, then starts vandalizing. --kizzle 09:13, 8 December 2005 (UTC)

Maybe new users (and maybe even all users) should have a limit on how often they can make edits. Something like 30 seconds should do it. That makes it take longer for the vandal to make a lot of edits, by which point someone will have noticed, and banned them. If a user registers an account, and doesn't do anything with it, nobody will notice. Maybe have a policy that account that don't make a certain number of edits after they get created get deleted? Reub2000 09:36, 9 December 2005 (UTC)

There are legitimate reasons for a non-editor to register an account. Off the top of my head, he/she might want to create a watchlist, e-mail other users, and avoid receiving talk page messages intended for another recent user of the same IP address. —Lifeisunfair 22:07, 9 December 2005 (UTC)

By the way, a suggestion I heard on Jimbo's page is a request for Special:Newuser to list the recently created usernames, this would help ferret out newly created usernames purely for vandalism sake once semi-protection becomes a reality. --kizzle 09:22, 8 December 2005 (UTC)

Special:Log/newusers is already an excellent tool. (I believe this was created by User:Curps.) -Splash^talk 12:20, 8 December 2005 (UTC)

No ee didn't! (see Special:Version). —Ævar Arnfjörð Bjarmason

I agree with you kizzle. I think we'll start with just age of accounts, but I bet we add edit count #s eventually since we'll see vandals try to get around it. The "experiment" that stops anons from creating pages is on iffy ground already because vandals are then just simply creating accounts and still posting copyvio and hoaxes. --Woohookitty^{(cat scratches)} 12:54, 8 December 2005 (UTC)

But on the other hand, it *would* discourage a lot of the vandals I've seen, who are the "spur of the moment" anonymous vandals. Cutting down on the "I feel mischevious so I'm going to blank a few pages & insert some swears randomly during class" vandals will allow monitors to concentrate more on the determined vandals who have the patience to create an account & wait a few days. 9 times out of 10 the vandals I see are logged in from school IPs, likely through a library terminal, and all they do is insert "penis" or "homo" into 4 or 5 articles before they get banned for 24 hours. This proposal would allow the higher traffic articles to be spared the "anon tagging". Jabrwock 17:21, 8 December 2005 (UTC)

Back to the nut

...of my philosophical question about whether this is an encyclopedia that happens to use a wiki format, or a wiki that enjoys tinkering with encyclopedia topics. Jimbo's post here seems to settle that one. [3] "Anyway, this is first and foremost an effort to write a high quality encyclopedia, not a wide open wiki" · Katefan0^(scribble)/mrp 16:55, 8 December 2005 (UTC)

I'm not sure what the answer to that question will help here, Katefan. Whether one emphasizes a wiki over an encyclopedia or vice versa, this is something that is just common sense so that a page doesn't get vandalized 30 times a day. --kizzle 17:23, 8 December 2005 (UTC)

This sounds like a good policy to me.Homey 23:35, 8 December 2005 (UTC)

You know what the frightening thing is, kizzle? You are way under the actual amount. In the last 24 hours, GWB been vandalized 49 times and I believe that's actualy low. --Woohookitty^{(cat scratches)} 07:21, 9 December 2005 (UTC)

A note to those that oppose this proposal

I would encourage you to visit the RC IRC channel #wikipedia-en-vandalism and spend 15 minutes just observing the activity. As an administrator that spends considerable time each day in RC patrol fighting back the ever-growing hordes of vandals, I can assure you that without some tools to limit the openness of the project to curt vandalism, we will not prevail. The number of vandals increases exponentially as Wikipedia becomes more popular, but the number of admins that chose to spend time on the thankless job of vandal-fighting, is not growing nearly as fast. We need better tools, and we need them fast. ≈ jossi fresco ≈ t • @ 00:45, 9 December 2005 (UTC)

You've hit the nail right on the head. I've been in that channel and it's scary. There's roughly one new occurrence of vandalism per second, and although anti-vandalism admins will burn out, the sheer horde of vandals (each of whom only contributes a small number of defacings) never gets tired. --Cyde Weys [u] [t] [c] 01:33, 9 December 2005 (UTC)

The worst part about vandal fighting in RC IRC is that you see articles like GWB, Hitler, etc over and over again, generally dozens of times an hour, sprot would greatly reduce them! xaosflux ^T/_C 04:00, 9 December 2005 (UTC)

But would it? My worst case scenario is that just as many vandal edits are still made, but instead of happening at GWB/Hitler/etc they move to the user and talk pages of people making valid contributions of GWB/Hitler/etc. I don't think that's the most likely outcome, but I think it will happen to some extent. Regards, Ben Aveling 04:32, 9 December 2005 (UTC)

I think that that's fairly unlikely, and if it is, then it's better than vandalizing articles. Note that on the sort of contentious articles that we're speaking of, e.g. Hitler, Jesus, George W. Bush, etc., the great contributors to them are not brand spanking new or anonymous users, but WP veterans, generally speaking, who have been around long enough and know what they're doing enough to be willing to work on such pages. Regardless, I don't think it's enough of a deterrent to not implement the idea; it's certainly worth a shot, even in the case that you're right. Blackcap (talk) 06:27, 9 December 2005 (UTC)

Not only that but the vandalism room doesn't catch alot of the vandalism. --Woohookitty^{(cat scratches)} 07:22, 9 December 2005 (UTC)

Yeah, I prefer WP:CDVF, myself, for that and other reasons. Blackcap (talk) 07:39, 9 December 2005 (UTC)

Banner should read "...has been semi-protected..."

Currently the average user can't distinguish between this and full protection. If they complain to an admin, they will say "the page has been protected", which isn't quite true. So it should clearly distinguish itself in the banner. Stevage 17:34, 9 December 2005 (UTC)

Maybe add a smaller line "Note: this page is still editable by *registered* users."? Jabrwock 19:19, 9 December 2005 (UTC)

Newly created accounts would be blocked too, though. --TantalumTelluride 00:46, 10 December 2005 (UTC)

Straw poll deadline

To keep things rolling toward sending this proposal to a community-wide vote, I have established a December 16th "deadline" for the v.02 poll. If support for this is ~75% or greater we should push this to the top for greater consideration. Hall Monitor 18:02, 9 December 2005 (UTC)

Ok, although I am concerned that this has not been publicised widely enough given the fairly low participation. We need either to restart this from scratch now we know that we have to use %age rather than no. of days, or we need to advertise it much more widely. I see it's already on RfC,and Current Surveys. Perhaps it needs highlighting on VP and AN, too. I would observe that, although we are approximating a policy-making proposal, this kind of thing hinges on Jimbo's approval. He can overrule any of the 3 possible outcomes. -Splash^talk 18:11, 9 December 2005 (UTC)

I would, however, urge those currently with concerns about %-v-days to review the comment by Aevar and study the proposal again to see if they can support it. -Splash^talk 18:11, 9 December 2005 (UTC)

I agree with restarting the vote now that we know that %age will be used rather than no. of days, mainly because some were basing their vote off the assumption that one or the other was used. So we should make it clear which version the v.02 will be using, and then vote on v.02. I would also suggest maybe advertising this proposal vote, is there a place to let people know about it? Jabrwock 18:38, 9 December 2005 (UTC)

I don't think it is necessary to restart this vote, this is just a straw poll. Hall Monitor 18:41, 9 December 2005 (UTC)

Actually, only 3 people appear to have that basis, so presumably it's not so serious a problem. -Splash^talk 18:43, 9 December 2005 (UTC)

Kizzle and Mysekurity come here quite often, so there's probably not a need to restart the poll on those grounds. We can always ask Lanoitarus to comment on Aevar's point. Titoxd^{(?!? - did you read this?)} 18:47, 9 December 2005 (UTC)

Thank you for trivializing my concerns because I'm the minority, Splash. Since we have actual feedback from a developer saying this is the way to go, of course I give my support. Just make the project page reflect what we currently have agreed. However, I think we need to think of the upcoming vote as more of a petition than a pass-vote. There really hasn't been a lot of advertising of this page, and I think for any serious consideration by anybody who could spearhead such a change in both policy and software, we need to aim for a goal of at least a 100 support votes. Basically, when you vote support, act like its a chain letter and post the link to this page on at least 10 people you know on your watchlist. If we get a simple majority of only like 20-30 people, I doubt that's even enough to be considered, as it will be said this proposal wasn't advertised enough. We also have to take into consideration that we have been operating this debate under the principle that semi-protection is justified in itself. We're going to get a lot of flak from people who dogmatically consider it "anti-wiki" and who will furiously hold to the notion of anon-editing no matter what as a fundamental aspect of not only the concept of a wiki but of Wikipedia itself. We might want to consider adding a section of replies to common objections about the justification of using semi-protection itself, as this is the primary reason why this proposal has previously gotten no attention on bugzilla. --kizzle 21:13, 9 December 2005 (UTC)

I didn't trivialise; I simply said that I don't think we need to restart the whole thing because of three opposes on a particular point that has been now been resolved so far as it is likely to be. Disagreeing with you on the detail of this doens't mean I'm out to get you.

To my understanding, the project page does indicate what we have agreed, it being identical to the v.02 proposal that people have been discussing in the section(s) above.

We shouldn't chain-letter it, since Talk page spam (such it would seem) is generally very unwelcome. We simply advertise via the usual channels, some of which we are already utilizing and let it run for a little longer. We shouldn't set hard numerical limits, because they're too easy to miss while still having a clear support. Also, 100 Wikipedians hardly ever agree with one another. See Wikipedia:Times that 100 Wikipedians actually agreed and voted to support something. We just sound out the community as widely as we can, after tidying this talk page up some and see whether it sinks or swims. -Splash^talk 21:45, 9 December 2005 (UTC)

Of course disagreeing with me doesn't mean you're out to get me, it was the part where you said "only 3 people" disagreed so it wasn't "so serious of a problem". Of course I'm not setting hard numerical limits, I was just setting a goal, basically as many votes as we can muster, and I don't think "talk page spam" is beneath us, people post on my talk page all the time about polls they think I'm interested in, I don't take offense. --kizzle 22:00, 9 December 2005 (UTC)

Kizzle, please take a deep breath and relax, I don't think Splash meant anything by his comments. As for spamming the Semi-protection proposal, I don't see that as a problem either, especially in light of some of the surprising and unilateral experimental changes which have been made recently. Hall Monitor 22:03, 9 December 2005 (UTC)

How dare you sir! I will not calm down. The grievances which you have commmited against me will not be forgotten for a long time, I can guarantee you that! Nah, I'm calm, I just didn't like Splash's treatment of the minority opinion, it's not like its eating inside at me. Call me ultra-sensitive, I guess. --kizzle 22:09, 9 December 2005 (UTC)

Ads

What's funny, Kizzle, is that your comment got picked up by Lupin's vandalfighter-tasticalprogrammythingy, and thought you were a vandal. Bad, bad. ;). Have a great day, everyone! -Mysekurity^{(have you seen this?)} 22:13, 9 December 2005 (UTC)

It's cause I'm thinking like the enemy. --kizzle 22:16, 9 December 2005 (UTC)

Suuure! I've created a new userbox template at {{User Semi}}, enjoy! -Mysekurity^{(have you seen this?)}

I've made posts at the Village Pump now, both in Policy and in Proposals. Now, what we should do is announce it on Wikien-l too. Titoxd^{(?!? - did you read this?)} 03:41, 10 December 2005 (UTC)

I'll be sure to check it out. I created a template (always subst'd) at {{User:Mysekurity/Semi}} for people who have reverted vandalism at GWB to get them involved in this page. As they are related to this issue, I don't think it's spamming, but if this practice is uncooth, feel free to state so. You may use and/or modify/copy/make it awesome, in any way (yes, it's released into the Public Domain...not that it has to be), just as long as we can get a bunch of users here to help us out and give thoughts before we post something site-wide and have an (even) big(er) talk page. -Mysekurity^{(have you seen this?)} 05:45, 10 December 2005 (UTC)

Jossi Fresco already put something out on WikiEN-l here: [4]. Do you want to put out more than that? Blackcap (talk) ^{(vandalfighters, take a look)} 02:37, 11 December 2005 (UTC)

Yes, because not everyone reads the mailing list, the Village Pump, and others (they are too often flooded with unimportant things, and hard discern what's relevant to them). The nice idea of putting a notice on the talk page is that this policy is not a small deal, and it relates to them, as they have recently reverted vandalism on that page (targeted ads, without urging them to buy anything, and without a bot). I started it out by copy-and-pasting the message, or writing something origional, but I had not interracted with many of the users, and I realized that it would not be as impersonal if I subst:'d the template, and would deliver the same message. Feel free to edit it as you see fit, but I thought I'd keep it light and friendly, while not overly dire or unfocused. This way, we don't have to have massive amounts of people coming here via a sitenotice, and on the same token, we don't have to keep this a secret. Posting to talk pages gets that user directly involved, and anyone who is reading that page or keeps watch on it. -Mysekurity^{(have you seen this?)} 03:42, 11 December 2005 (UTC)

Oh yes, I completely agree. I actually meant to ask, "Do you want to put out an additional post to WikiEN-l?" as Jossi's post wasn't responded to and it had been posted before Titoxd had posted his comment which I was asking about, above (and so I thought maybe he meant an additional post). Sorry about that. I now realize that my post had been pretty ambiguous. Blackcap (talk) ^{(vandalfighters, take a look)} 03:49, 11 December 2005 (UTC)

If possible, yes, place another post on WikiEN-l, please... and be sure to mention that it is gathering strong support already. :) Titoxd^{(?!? - did you read this?)} 06:23, 11 December 2005 (UTC)

O.K., I just have. As I've just joined the list, the email I sent won't go out for a while (12 hours, I understand), as it needs to get moderator approval first. Once I see it go out on the list I'll post a link to it here. Blackcap (talk) ^{(vandalfighters, take a look)} 07:33, 11 December 2005 (UTC)

All right lads, here's what I sent out: http://mail.wikimedia.org/pipermail/wikien-l/2005-December/034724.html. Blackcap (talk) ^{(vandalfighters, take a look)} 08:29, 13 December 2005 (UTC)

Not a good idea

This is another discrimination against anons, right?

I joined here on the basis of the foundation issues.

I understand jimbos position wrt temporary experiments, he's under extreme pressure. I'm not sure the same can be said about this measure, however. Kim Bruning 06:07, 10 December 2005 (UTC)

I think you are quite correct, on the surface: It is a discrimination against anons. However, I dont think it is unwarranted, as long as it is only used in certain cases. I view it less as a step down from pure wiki as a step UP from the standard protection. -Lanoitarus 06:16, 10 December 2005 (UTC)

Hmmmmm Kim Bruning 06:20, 10 December 2005 (UTC)

This is only tangentially related, I suppose, but I stumbled across this in a Google search. Compare point #3 with the paragraph (from #2) directly above it. They're not really contradictory, but it seems like you could read the end of #2 to support or reject this proposal, while #3 clearly rejects it. —HorsePunchKid→龜 2005-12-10 06:17:15Z

However, the very existance of a admin-only protection violates both points just as much. I see this new policy as a way to use that level of protection less, which I think it a good thing. (note: I copied your signature to the above, since i was responding between your two bullets and you only signed the bottom one - if you meant them to be together feel free to rearrange my comment.) - Lanoitarus 06:27, 10 December 2005 (UTC)

For what it's worth, previous brainstorms of semi-protection (lost in the massive GWB talk archives) involved an immediately editable version of the article separate from the real article, leaving the admins (or perhaps point #2's "regulars") to merge the changes into the real article. This system would not violate sacred point #3; I do not recall why the idea was abandoned. —HorsePunchKid→龜 2005-12-10 06:17:15Z

It sounds fairly impractical to me, which might be why it was discarded. Blackcap (talk) ^{(vandalfighters, take a look)} 06:24, 10 December 2005 (UTC)

Not at all impractical. Opinion was divided between what HorsePunchKid describes and the version I favored, in which the real article would remain editable. It was intended only for George W. Bush and perhaps a handful of other articles. A simple procedure:

1. At the top of the George W. Bush article, we add a note warning that it's frequently vandalized, and that a scrutinized version free of vandalism is available here (wikilink to George W. Bush (scrutinized) or whatever it's called).
2. Once a week or so, an admin who has pledged never to edit the article for substance comes along and finds the most recent unvandalized version of George W. Bush. The admin copies that version.
3. The admin then goes to George W. Bush (scrutinized), which is a protected article. It begins with a note that it reflects the content of George W. Bush as of a specified date and time.
4. The admin changes the date and time to reflect the new updating, deletes all the previous text of the scrutinized article, and substitutes the text that the admin has just copied from the main article.

That's it. The job's done. It takes one admin about five minutes per week. I don't see why it should be impractical. There's a fuller discussion at Wikipedia talk:Experimental vandalism protection. JamesMLane 03:05, 14 December 2005 (UTC)

Is this handier? : m:Foundation issues Kim Bruning 06:25, 10 December 2005 (UTC)

Yup, that's exactly what I was looking for. Bookmarked; thanks! :) —HorsePunchKid→龜 2005-12-10 06:42:24Z

The vast majority of anons that edited GWB vandalized, I had two deal with WoW and IPs making the article into 200 pictures of the same two penises again and again. Most good IP edits there would have been made my users anway. That is the point, there, we would only use it on articles were IPs and Brand "New" Users = Trolls 99% of the time. To be blunt, lets stop defending thousands trolls and a handful of good IPs and make a good encyclopedia.Voice of All^T|@|ESP 06:25, 10 December 2005 (UTC)

Hear, hear. Blackcap (talk) ^{(vandalfighters, take a look)} 06:30, 10 December 2005 (UTC)

This is actually designed to make editing high-profile articles easier. Just look at the protection log for George W. Bush, and you'll see that the article is actually uneditable most of the time (it is currently protected, by the way). With this policy, we can actually return editing priviledges to most non-admins, and those non-admins that will be affected by the proposal are already restricted some privileges, like moving pages. As Jimbo himself said, we're trying to build an encyclopedia here, not a wide-open wiki. If an article is less protected, it is easier for both a) RC Patrollers to contribute more content to the actual encyclopedia, and b) regular editors who haven't vandalized to edit a page that is usually non-editable by most. Titoxd^{(?!? - did you read this?)} 06:37, 10 December 2005 (UTC)

Also, when you have penises, vaginas and POV getting mixed in, it is hard to make any good changes, as they may accidentelly get reverted.Voice of All^T|@|ESP 06:43, 10 December 2005 (UTC)

Tito, once again, you prove your brilliance ;). This will definitely help, as, yes, we're trying to build an encyclopedia, our main objective, and a wiki second. (okay, I know that was bad grammar, but still). It's too difficult to edit GWB as-is, and this would help in more ways than three. -Mysekurity^{(have you seen this?)} 06:45, 10 December 2005 (UTC)

Corollary: When I've been on RC patrol and I see an anon edit a controversial article doing something such as changing dates, I'll often revert it on the assumption that it is vandalism, as I just can't know. So I guess maybe I'm already anon/newuser prejudiced a bit, but I just don't want to have vandalism sit around for ages because I was worried about biting a newbie. I alway apologize later if I'm wrong... Blackcap (talk) ^{(vandalfighters, take a look)} 06:50, 10 December 2005 (UTC)

No, it's not another discrimination against anons, since they can't edit articles that are protected anyway. Semi should only be used on articles that would have been protected anyway. It doesn't reduce their editing ability at all, as long as it is applied with good sense. Semi should be used to replace full protection in the case of short-term mindless vandalism that won't yield to blocks. At least that way, good editors can still edit. -Splash^talk 16:47, 10 December 2005 (UTC)

Can a semi-protected article be edited as normal...

...or does it face the same restrictions that protected pages currently do for admins (i.e. they can edit them, but aren't allowed to until it's unprotected)? In other words, on a semi-protected article, can normal editors continue to edit the page or are they expected to discuss all changes on the talk page? (Sorry if this is a dense question, but I didn't see it outlined anywhere). Blackcap (talk) ^{(vandalfighters, take a look)} 01:47, 11 December 2005 (UTC)

I believe that semi-protection only applies to a page that is facing vandalism, not one that is simply controversial. So normal editors should continue to work as normal on the page.Stevage 02:04, 11 December 2005 (UTC)

Yes, as the proposal says, Semi would emphatically never be the solution to an edit war, since it could treat some parties differently to others. There should be a presumption that editing semi-protected pages is fine. There needn't even be the big red message that admins get when they go to edit a protected page. It should be treated with as little ceremony as a move is at present, since it is basically the same level of restriction on established editors. -Splash^talk 03:09, 11 December 2005 (UTC)

O.K., thanks. That's what I thought, but I wasn't sure. Blackcap (talk) ^{(vandalfighters, take a look)} 03:30, 11 December 2005 (UTC)

Improving the downsides

I thought of several objections and how to address them:

• dedicated vandals will create series of accounts for 'the future' - consider implementing a block in one IP creating other accounts. Besides, sooner or later some vandal will devise a spambot which will spam us with scores of accounts.
• Make sure that the template mentions that you can edit the talk page - so the postive contributions by new users can be recorded at talk.--Piotr Konieczny aka Prokonsul Piotrus ^Talk 02:22, 11 December 2005 (UTC)

Sounds good, I think we should add a common objection/reply section below the actual proposal once we start advertising other places, as there are a lot of people who still think semi-protection is "anti-wiki". --kizzle 02:27, 11 December 2005 (UTC)

To the second poiint: the template does already say to discuss changes on the talk page. That message seems to work ok for the current protection. To the first: the devs have already implemented just such a block after the recent vandalbot attack which did, as you predict, spam us with several accounts per second for a lengthy period. Now, my understanding is that a blocked IP address (resulting either from a checkuser or an autoblock) cannot create accounts any more than it can edit. -Splash^talk 03:13, 11 December 2005 (UTC)

Also, there is an account creation throttle, so any given IP cannot create more than 10 (is that the right number? I can't ever remember it...) unique accounts per day. Titoxd^{(?!? - did you read this?)} 05:48, 11 December 2005 (UTC)

10 accounts per day is good enough for most vandals. There should be a limit like 1 account per hour, perhaps. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 15:40, 11 December 2005 (UTC)

You mean, 24 per day?Stevage 18:31, 11 December 2005 (UTC)

One per hour stops the recurring vandal. I would say 1 per hour, 3 per day. 10 is just stupidly high.Voice of All^T|@|ESP 23:56, 11 December 2005 (UTC)

3/day would likely hit major issues in school computer labs... especially if a teacher is making kids create accounts. Although the multiple limit (hourly and daily) sounds pretty good. -Lanoitarus _^(talk) .:. 04:13, 12 December 2005 (UTC)

I'm very concerned about someone creating "sleeper" accounts to do later damage. Why not just set a minimum number of edits? Almost no one will make 25 or 50 good edits or talk page comments just to get to do vandalism later, so we'd have a chance to catch the vandal before they get to these sensitive articles. -- Jmabel | Talk 01:15, 12 December 2005 (UTC)

Ideally, the real way to do this is to implement both an time check through either account life or % of users and an edit check, however currently it has been advised that this would not be technically feasible. If we're going to pick one out of the two, a time check is better, as a vandalbot could easily overcome 25-50 edits within a matter of minutes. --kizzle 01:20, 12 December 2005 (UTC)

Vandalism is, by and large, an act of convenience and boredom. There will be a few people who are so focused and maniacal about committing vandalism that they may consider creating sleeper accounts, but they will be far and away the exception. A system based on edit counts has the problem that there is no way of checking whether the edits are "good" or meaningful - a vandal could manually rack up 25 trivial edits as fast as a bot. Tim Pierce 01:32, 12 December 2005 (UTC)

I agree with Tims statement 100%. -Lanoitarus _^(talk) .:. 04:14, 12 December 2005 (UTC)

I think it would be helpful for everyone to keep in mind that this proposal will not and was never intended to stop all vandalism. Vandalism will always be a problem on Wikipedia, and "dedicated vandals" will always find a way around any protections that we put into place. That's a fact of life; semiprotection won't stop it and isn't going to try to stop it. I hope that any criticism of the plan takes that into account.

What semiprotection hopes to do is to reduce the amount of time that admins and editors spend on the most often vandalized pages. It will not prevent those pages from being vandalized at all. I think that it will, however, mean that they spend less time in a vandalized state, and I think that it will allow admins to spend their valuable time improving other parts of Wikipedia. Tim Pierce 01:44, 12 December 2005 (UTC)

GWB acts to a degree as flypaper. Vandals discover wikipedia can be edited and hit a random article. They then think ah yes GWB and vandalise that. The first vandlism is not picked up. the second is and that persons contributions are cheacked.Geni

Most of the vandalisms I've seen were spur of the moment blankings or "tagging" (adding in LOLZ or douchebag, etc) to high profile pages. The pages I monitor mostly get vandalised through school accounts, so IP banning is very counter-productive. This semi-protection would allow regular users to continue to edit the pages protected, but would stop those who do a quick "hit & run" edit. Jack Thompson is a good example. It was vandalised 40 times over the weekend, and every edit by a registered user over 3 days was to revert the vandalism, except for one, which was a new account. So SP would have helped keep the page under control. Determined vandals will still find ways to get around this (sleeper accounts, etc), but like Tim said this will reduce the volume of "quick" vandalisms that occur on a regular basis to high-traffic pages, allowing admins/page-monitors to keep up. Jabrwock 17:05, 12 December 2005 (UTC)

An issue for later

What if the account creation rate slows down. Wont the newest editors get locked out of articles, even though they are not really that new anymore? Why can't Wikipedia record account creation date, then we could have 3-4 days instead of "newest 2-3 percent". Just a though for the devs in the future. For now, X% should work very well.Voice of All^T|@|ESP 21:44, 12 December 2005 (UTC)

• I see no reason at all why account creation rates would slow down. "X%" or "X weeks old" are merely two separate ways to calculate how "trusted" an account is. I think whichever puts less load on the database should be used. --Cyde Weys _talk^contribs 21:49, 12 December 2005 (UTC)

Inferior to delayed gratification

The whole idea behind allowing anon edits is that it maximizes the ease of making small changes. Don't remove that ease. Find other ways to reduce vandalism... like delaying gratification. +sj + 00:13, 13 December 2005 (UTC)

This doesn't restrict anon editing any more than full-scale protection does at present. There is no proposal to apply this to all articles, or even to a significant number. It is intended that it will only be used to combat vandalism in the same way as full-protection is at present. So, for say, 24 hours at a time on a particular article that has a hard-to-block anon vandal. Permanent semi-protection would, I hope, be as distasteful to the community as permanent full-protection. -Splash^talk 01:32, 13 December 2005 (UTC)

Please read Wikipedia:Protection policy carefully. Protection is intended to restrict all edits to a page; even admins should generally avoid editing these pages. Protection is intentionally an extreme measure, to help guarantee that it is applied sparingly (some 0.02% of articles : [5] [6]). What you are suggesting here is a dangerously 'mild' variant which, while favoring certain classes of uers (as you note in the proposal) nevertheless encourages users to edit the 'semi-protected' page if they are favored. This is significantly different from standard protection. +sj + 23:42, 13 December 2005 (UTC)

As for "hard-to-block anon vandals" -- this doesn't seem to me the best way to deal with such vandalism. But if you want such a policy to apply only to those cases, please make its scope much more narrowly defined. +sj +

No, this wouldn't be the best way to deal with that, and perhaps vprotection might, but I highly doubt many users would create sleeper accounts to be used on proxy servers over and over again. This is good for pages that have been vandalized because they've been in the news, or are high-profile. This would allow trusted—non-admin—editors to contribute without worrying about vandals as much. It would only be used as a preventive measure. -Mysekurity^{(have you seen this?)} 05:26, 14 December 2005 (UTC)

I think you're going to find little distaste of extended semi-protection on pages within the non-admin community. --kizzle 02:31, 13 December 2005 (UTC)

What, like having all anon/new acc edits screened by trusted users first? And yes, I agree completely with Splash, in that this would be the best, if not perfect, way of dealing with vandalism usually dealt with by page protection. We're talking about a very small number of articles, for a short period of time, which would greatly reduce wikistress overall, and keep people smiling, happy, and able to fix other things. -Mysekurity^{(have you seen this?)} 01:44, 13 December 2005 (UTC)

A textbook case for semi-projection

Mysekurity and talked about this on the RfP page. I think that the Daniel Brandt article is a textbook example as to why we need this policy. In the last 24 hours, it has been hit by 16 vandalizing edits, but it's also had many, many constructive edits. One area that we haven't discussed where semi protection would be very helpful would be with cases like Daniel. He's big in the news right now because of the JFK assassination controversy. Instead of blocking constructive edits by protecting, semi-protection is a wonderful alternative. Same with the stuff like when celebs die. The Richard Pryor article got nailed after it was announced that he died. Well we had to protect, which isn't the best option since again, you are freezing out good edits by doing that. We could easily semi-protect for a day or so until things calm down. Anyway... --Woohookitty^{(cat scratches)} 01:43, 13 December 2005 (UTC)

I think we leave it unprotected for the first little while so an anon can add in information from a newswire (I read somewhere that Wikipedia got the information faster than MSNBC, CNN, and the other guys), and then semi-it-up (provided there's vandalism, no pre-emtive stuff here). But yes, in the beginning, it is most assuredly helpful to get a few new editors adding things, but after a few vandal edits (16 seems a bit high), we go ahead and semi- it. A day or two seems perfect. -Mysekurity^{(have you seen this?)} 01:49, 13 December 2005 (UTC)

Ya, we should probably stress that semi-protection should only be used in response to vandalism and not in preparation for vandalism. --kizzle 02:30, 13 December 2005 (UTC)

Good point. Let me state, in large, bolded letters:

Semi-protection will not be used as a pre-emptive measure against vandalism.
It will be used as a response, just as regular protection for vandalism is now.

Just so it is clear; this will basically replace full protection for vandalism, which is undoubtably a good thing. -Mysekurity^{(have you seen this?)} 02:34, 13 December 2005 (UTC)

Um, no. There still will be full protection for vandalism, but it would be applied as a "next level" measure: if semi-protection fails for some reason to stop the vandals, then full-protection would be applied. Titoxd^{(?!? - did you read this?)} 02:39, 13 December 2005 (UTC)

That's why I'm baffled at keeping only one level of semi-protection. Protection should almost never be used for simply combating vandalism if we have the availability of a tool to only allow "trusted" users with a variable amount of editing life. Having the ability of a graded response to various levels of vandalism threats has so far been rejected solely because of simplicity's sake, and not for any reasons in itself. Ideally, if semi-protection initially fails, we should lower the amount of accounts who can edit from 99% to 95% rather than shutting it down to only the 0.11%. If we do this right, semi-protection should replace protection in 99.9% cases of vandalism, whether its a minor vandal, a significant number of vandals, or a full-scale attack against Wikipedia. In any of these cases, semi-protection becomes much more "wiki" as it enables the largest amount of editors possible to edit given circumstances that would normally warrant page protection. --kizzle 02:49, 13 December 2005 (UTC)

Let's KIS(S) for now (no, not like that), and worry about other things later. Tito says it's going to be coded as a variable, so other foreign-language wikis can participate (And your point is duly noted, thanks for clarifying). Does anyone know what should happen next publicity-wise? -Mysekurity^{(have you seen this?)} 02:53, 13 December 2005 (UTC)

(After edit conf): Yes, Kizzle, I agree: we need this as a more wiki solution, an anti-anti-wiki, if you will. -Mysekurity^{(have you seen this?)} 02:53, 13 December 2005 (UTC)

I already warned about that, when I said "easing prophylaxis". God, I'm good. Ahem. We should re-read that grilling to see if we've fallen foul of any of it. -Splash^talk 02:58, 13 December 2005 (UTC)

Clearly, we may still need to fully protect articles from time-to-time on vandalism grounds, such as when an established editor goes loopy. But, simplicity and vertical creep aside, I can't see the reason for needing a higher level of protection. If an editor is going to be a determined vandal, we block them and that's the job done. Their new account is caught by semi, as are they by an anon. There is no phenomenon like dyanmic IPs for accounts. Anyone having sleeper accounts will get them successively blocked, or nuked all at once after a quick Checkuser. -Splash^talk

Chagrin...uh...agreed. And yes, you are damn good, but that's no reason to gloat ;). Any thoughts as to what to do next? I keep spamming the reverters on the GWB history with User:Mysekurity/Semi (please subst:), and feel free to add to/change any bit of, and use it! Please, drop it on talk pages to get people involved. But outside of that, I see there's been discussion on the mailing list. Maybe an IRC discussion? Ideas? -Mysekurity^{(have you seen this?)} 03:03, 13 December 2005 (UTC)

(after three edit conflicts): Well, since there is a "formal" closing date for the poll (December 16), right now, all we have to do is wait. But then, the level of support that the proposal has gotten (60+ votes, 94% approval), it tempts me to actually tag this as "policy", since it would have passed the 80+% threshold we usually hold for policy proposals. I probably won't, but the next thing would be asking a dev to code it, and asking Jimbo, Angela, Anthere and Tim for their input (they're Foundation members, although Tim might not reply, he is not very active on en.Wiki). That is what I would do, but then, there may be other opinions on what to do next. Titoxd^{(?!? - did you read this?)} 03:09, 13 December 2005 (UTC)

We should probably re-post or re-locate the straw poll to a more visible section, maybe place on the project page or archive everything but the poll? --kizzle 03:11, 13 December 2005 (UTC)

We can't call this kind of change policy without a Jimbovian blessing, imo. I think it worth adding a mention of Mysekurity's bolded message to the proposal page or somesuch since this is an obvious point to challenge it on. -Splash^talk 03:14, 13 December 2005 (UTC)

After the USA Today editorial, the chances of Jimbo looking at this approach zero, unless it actually passes, so it might be easier to throw it to Angela or Anthere for approval. And I think that adding the banner is a good idea. Titoxd^{(?!? - did you read this?)} 03:17, 13 December 2005 (UTC)

Keeping with the spirit, it might be a good idea to create a "replies to common objections" section, like proposed above. Titoxd^{(?!? - did you read this?)} 03:26, 13 December 2005 (UTC)

Uh, if an established editor goes "loopy", I think a block is much more warranted than entire page protection. A higher level of sprotection would be needed given Tito's hypothetical of semi-protection not being sufficient. If there is still vandalism occuring (of which the possibility will be remote) even with sp set to 99%, the answer should be to reduce this number gradually to 97-95% (or somewhere around there), rather than reduce the number drastically to 0.11%. Given the availability of applying semi-protection at variable levels, protecting a page in response to vandalism becomes inherently "anti-wiki". Of course, responsibility, caution, and discretion must govern the process of semi-protection's increase, but as you have so far advocated, admins will keep themselves in check. --kizzle 03:07, 13 December 2005 (UTC)

If vandalism is still ocurring, then blocks fix it on the spot. A blocked editor without a sockpuppet has had it as soon as they are blocked since any new account they make can't edit through SP. A blocked editor with socks isn't much more of a problem once you have an admin on the scene. Or am I misunderstanding the point? -Splash^talk 03:14, 13 December 2005 (UTC)

(double edit conflict again... sigh) To be honest, I'm not sure how the semi-protect+block sequence would fail to address almost all the vandalism within six sigmas. In a recent case of an editor going loopy, Wonderfool, a former admin at wikt: got banned from Wikipedia by the ArbCom for sneaky vandalism, and newly-created socks of him wouldn't work on high-profile articles. Titoxd^{(?!? - did you read this?)} 03:15, 13 December 2005 (UTC)

My point is that if we find out that the 99% check is insufficient to prevent vandalism, the next step, contrary to what Tito suggested earlier, should be to conservatively increase the level of semi-protection rather than protect the article. --kizzle 06:58, 13 December 2005 (UTC)

As I've said from the start, let's start it simple. If we have to add other levels later, we can. Btw, Stanley Williams is another textbook example. it's been protected multiple times today and it's pointless. --Woohookitty^{(cat scratches)} 17:21, 13 December 2005 (UTC)

As long as we leave the option open for later. --kizzle 19:28, 13 December 2005 (UTC)

Another text-book case: Stanley Williams

First off, I may be giving credit to Kizzle's "non-admins might like semi-protection" (paraphrased) statement as I'm not an admin.

Over at the Stanley Williams article today we saw a text-book case of a high-profile article getting vandalised by anons repeatedly. Now the page has been protected. That means that I can no longer edit the article, even though I've been editing/reverting vandals there all day. Bring on semi-protection, it would have solved the problem much nicer than the brick-wall-solid protection status that exists now. CanadianGuy 04:23, 13 December 2005 (UTC)

I was more enunciating that non-admins might like extended semi-protection. --kizzle 05:46, 13 December 2005 (UTC)

The Stanley Williams article really is a perfect example of a page which could have and should have been temporarily semi-protected rather than protected. In cases such as this one, we have two options; block everyone from editing (and protect the "wrong version"), or fight the uphill battle of trying to revert vandalism while struggling to make positive contributions and not have them edit-conflicted or lost in the shuffle of confusion while vandalism is being simultaneously removed. These are not productive options for building an encyclopedia. Hall Monitor 18:02, 13 December 2005 (UTC)

question for developers

In a paragraph or so, can you briefly describe what it would take to implement an edit count check in addition to the time limit associated with x% of users? What is the most significant roadblock for this to be implemented? In what circumstances could this be implemented? --kizzle 21:41, 13 December 2005 (UTC)

To be really feasible, this would require adding an edit counter on the user table. Counts on revision would be a touch expensive to do all the time. Possible if needed. --Brion 03:21, 14 December 2005 (UTC)

We've been talking about it tho' - Tim and Avar were on about a separate table for it at one point. 86.133.53.111 17:37, 17 December 2005 (UTC)

Ok, well lets see how solely using time limits derived from %x of users works out, if for some reason we still get excessive vandalism due to hordes of sleeper accounts even with setting the time limit to 95%, we may want to consider implementation at that time to get this right. --kizzle 07:59, 14 December 2005 (UTC)

Perhaps a new database only containing editors could be devised. If during an edit, the editor is not in the list a relatively time consuming check could create the count for that entry from scratch. After that, they are in the database, it would be only necessary to update the value and this would be much faster.WolfKeeper 05:50, 17 December 2005 (UTC)

My personal $0,02

Okay, this will be an outside comment, since I'm not active on the English Wikipedia.

At first, I didn't like the idea. I'm a pretty strict conservatist in these matters - if Wikipedia is a free encyclopedia, then let it be. Just as with the recent move to block anonymous users from creating new articles, I was oppose to the idea.

But now, I think it might by pretty useful. If all anonymous users and new users will get blocked from editing the article, vandalism rate should drop by... Hell, I don't know how much, but it will drop - that I'm sure off. I don't think I have any further ideas here. If this idea will get approved, I'll be more then happy to promote it on the other Wikimedia projects. Datrio 21:47, 13 December 2005 (UTC)

If you support the idea, you might want to help by putting your vote here. --kizzle 21:55, 13 December 2005 (UTC)

What I like about it is that it would probably even mean less "normal" blocks - G. Gearloose (?!) 03:55, 14 December 2005 (UTC)

Concern from LV

Okay, I get the fact that current event articles like Stanley Williams are vandalised, but isn't this policy effective for just one article? I see this already has a lot of support, but can someone convince me that this isn't just being established for protection on this specific article? Thank you. --LV ^{(Dark Mark)} 22:17, 13 December 2005 (UTC)

You can see the history of Edward George Honey, for example: this page was hit by a vandalbot editing through proxy servers, so blocks didn't work, and rangeblocks wouldn't have done a thing either. I went mano-a-mano with the bot (one revert every 15 seconds or so), to try to maintain the integrity of the visible article (by the way, it was linked from the Main Page), while I could find another admin to v-protect. That could have easily been stopped by a s-protect, which would block throwaway accounts and allow good users to contribute to the article. Titoxd^{(?!? - did you read this?)} 22:26, 13 December 2005 (UTC)

Except it was just a short time thing. What would be worng with protecting it for a short time, until they went away? Were other people really desperately trying to edit that article? It still seems aimed at one specific article. Also, what about the new great influx in registered users? Won't they hit the top n% quickly? --LV ^{(Dark Mark)} 22:30, 13 December 2005 (UTC)

Well, the new influx of registered users will certainly affect the percentage, just like it is happening with page moves now; that is why it should be coded as a variable that can be changed from one wiki to the next and from time to time. As for temporarily protecting the article against vandalism, that was what ended up happening; however, in this case, it didn't need full protection, as in many cases.

But no, it is not aimed at one particular article. Several articles, particularily those listed at Wikistats, would benefit from this kind of protection, as would all the current events articles. Our current events articles are where we really shine, due to the collaborative nature of wikis; therefore, there is no need to restrict editing rights to 0.11% of registered users, just because of the actions of a few (or in many cases, just one). Titoxd^{(?!? - did you read this?)} 22:42, 13 December 2005 (UTC)

The x% of new users is a hack, as apparently there is no way to calculate account life to my understanding. As for your estimation that this policy is for one article, I think this semi-protection would have much more use than that; not only pages like Adolph Hitler, and the others in the 10 most highly vandalized pages on Wikipedia, but pages like Stanley Williams that suddently become high profile due to current events. Given a vandalism situation that would warrant page protection, why shut these articles off to only 0.11% of the editors when we can effectively eliminate vandalism and still keep the article open to 99% of the editors? --kizzle 01:03, 14 December 2005 (UTC)

I generally support this proposal, but I must say, I wouldn't like it if this was put on a page from the Main Article (like the Featured Aritcle). I was watching the Newton article today, and I could really see it get better during the day. Quite a few of the small-but-improving edits came from anon ips (and also quite a lot of vandal... c'est la vie). I think that s-protect is useful for something like GWB / Abortion / Hilter -- the obvious vandalism targets. novacatz 15:55, 14 December 2005 (UTC)

Exactly: SP would only be used on an article that is suffering heavy vandalism of the sort that would currently result in protection. Vandalism of the main page article does not currently result in its protection, and so neither would a well-applied semi-protection policy. -Splash^talk 16:49, 14 December 2005 (UTC)

My opinion

I think something to this effect should be implemented soon. The page move delay was a good idea and I think this is, too. I urge users to consider this as a viable method to keeping the wiki safe. --King of All the Franks 04:44, 14 December 2005 (UTC)

I assume you are supporting then? Titoxd^{(?!? - did you read this?)} 04:51, 14 December 2005 (UTC)

Yes. --King of All the Franks 05:28, 14 December 2005 (UTC)

My 2 Eurocents ( 0,024$)

Basically this discriminates between registered users and anons. It will probably eliminate some vandalism, but I would like to seek ways to discriminate between vandals and good editors. --Snowdog 13:44, 14 December 2005 (UTC)

That'd be great. What do you propose? -Splash^talk 14:01, 14 December 2005 (UTC)

Maybe a tougher blocking policy? Stop with all this {{test}}, {{test2}}, etc... Someone vandalises twice, they get blocked for 12 hours. No questions asked. If they keep coming back to vandalise, block for 48 hours. Much tougher than the current policy. Something like that? --LV ^{(Dark Mark)} 17:43, 14 December 2005 (UTC)

Given that a substantial majority of vandalism is drive-by impulse vandalism, your solution wouldn't fix this problem at all. It also vastly increases the workload on the RC community, which is one of the reasons this proposal is receiving such heavy support, so that people can go back to building a better encyclopedia rather than spending all their time reverting some kid who puts in "GWB sucks".--kizzle 18:04, 14 December 2005 (UTC)

Except, this would "only be used on an article that is suffering heavy vandalism of the sort that would currently result in protection." (per above thread). This would mean that drive-by vandalism would not really qualify under this policy. It is supposed to be used as a response and not a proactive restriction. I'm still not convinced this wouldn't just be slapped on the GWB article and would never be removed. --LV ^{(Dark Mark)} 19:40, 14 December 2005 (UTC)

Of course drive-by vandalism would qualify under this policy. It is possible for a page to suffer heavy vandalism from a significant amount of drive-by vandalism. You don't honestly think GWB or Stanley Williams got vandalized by the same 10 people? Heavy vandalism can, and most of the time is, made up of drive-by vandalism to high profile pages, thus the page gets protected. As for GWB, if it solves the problem, than sp should still be lifted every once in a while to see if the vandalism continues. It's not perfect, but its the best solution we have, including the status quo and simply blocking drive-by vandals who won't stick around long enough to feel the effect of the block. --kizzle 19:54, 14 December 2005 (UTC)

Occasional vandalisms can be handled by people watching those pages, or watching edit lists. It's when it's obvious that it's a problem keeping up, that this would get applied. For example, one of the pages I monitor gets vandalised by "drive-by"s once or twice a day during the week, so SP isn't really needed. On the weekend though, the number climbs up to 40 vandal edits a day. At that point, SP would be requested (maybe lasting up to 24 hours), just long enough to discourage the vandals who do 3-4 "repeat" vandalisms in a row, and discourage any other vandals until the registered users can clean up the mess of having to repeat the same revert command over and over 20 times. Jabrwock 21:07, 14 December 2005 (UTC)

It is a small, yet essential tool that we need to stop these vandals. Much can be said against it, but for those who have spent even five minutes on RC patrol, this will work wonders. -Mysekurity^{(have you seen this?)} 23:09, 14 December 2005 (UTC)

Agreed. The main thing to keep in mind is that this will only be used on very high profile, well developed pages. New users can still make edits, just not to the Hitler, Bush etc.. It basically comes down to math: Do we experience more flash vandalism or flash contribution? For low profile pages, the latter far outweighs the former; for high profile pages, it's the opposite. I believe this will allow Wikipedia to grow at virtually the same rate and free up a great deal of resources from our anti-vandal patrols. Oberiko 14:09, 15 December 2005 (UTC)

My 2 (African) American cents

I support this idea. ....εγκυκλοπαίδεια* (talk) 18:39, 15 December 2005 (UTC)

Support

Hi ... my edits here in English Wikipedia are somehow little but i,m active Sysop in Ar_Wikipedia and i have to admit that vandalism from anonymous users and newly users are so bothering and anoying , and mostly unresponsible , they tend to express their fanatic thoughts and to put unneutral viewpoint . offcourse this issue is serious especially in particular issues like religious and political issues , so i totally supports this new policy of semi-protection and i think it is important step to improve the quality of Wikipedia articles in all languages . And as it is just semi-protection , it provides better idea than full protection cause these pages will still editable for trustfull and known users to edit these sensitive subjucts , so this policy make the Wiki more Wiki and could increase the Production in such articles , i have to admit that religous pages are always brone for excessive vandalism and we have to find better way for protection rather than full-protection . --Unfinishedchaos 18:51, 15 December 2005 (UTC)

Please add your name to the semi-protection v.02 propsal vote.Voice of All^T|@|ESP 01:31, 16 December 2005 (UTC)

already done --Unfinishedchaos 09:46, 16 December 2005 (UTC)