Wikipedia talk:Semi-protection policy/Archive 4

billions of accounts will be made
People will sign up billions of accounts if they are a vandal to get around this we need to make sure that they can't do this. --  A dam1213 Talk [/w/index.php?title=User_talk:Adam1213&amp;action=edit&amp;section=new +] 01:36, 16 December 2005 (UTC)
 * Umm....what? Please read the discussion above: no IP can create more than 10 accounts per day, and the percentage number will be coded as a variable in case the number of accounts gets too high. There is an idea to limit the number of accounts created per hour, but I'm not sure where that's headed. Please read the issues before commenting like this. Further, how can or will vandals get around this, and what would you propose? In the future, try to make your comments a little more coherant. Thanks. - Mys  e  kurity ( have you seen this? ) 03:31, 16 December 2005 (UTC)
 * Somehow, I am reminded of Boothy443...:-(. Voice of All T 03:49, 16 December 2005 (UTC)
 * 10 accounts a day seems so high...if the only reason we're allowing it is because of classes signing up behind a proxy at the same time, we should have a process where a teacher requests from any admin to allow their ip to create more than 3 accounts a week. --kizzle 04:01, 16 December 2005 (UTC)
 * Throttling down from 70 accounts per IP per week to 3 and raising the limit only upon request from faculty wouldn't help students who want to create their own accounts independently of a class but are limited by the residence hall's caching proxy, or for that matter users behind any other web proxy such as that of AOL, or any ISP in an IP-address-restricted less-developed country. --Damian Yerrick 15:21, 17 December 2005 (UTC)
 * I'm thinking we add an option to Special:Userlogin about creating more accounts if the user is from a University/school/legitimate group behind a single IP. (Maybe an e-mail/contact link that would get sent to any admin who signed up for the task?) I think that would probably be more useful, plus a 3 account creation limit, or maybe just two accounts. - Mys  e  kurity ( have you seen this? ) 15:37, 17 December 2005 (UTC)

The "newest X%" could suffer from someone creating a constant stream of accounts and just waiting until they become active. Once the first is active, the pipeline is full and will keep pumping out active accounts. A better heuristic might be account activity. Count the number of unreverted changes made by the user (and maybe look at the time distribution of those changes). Yes, this is more complex and may require more computing power, but I think it is much more accurate and still transparent.

It should be always many ways to vandalize if u want to, but at least this method could put some obstacles for that , and this method would apply in few sensitive pages which are under excessive monitoring and control. so this method is originally against the intended Vandalism made by untolerant ppl.--Unfinishedchaos 09:40, 16 December 2005 (UTC)
 * Likewise, I think this is more to prevent the casual vandalizers who just feel like coming to mess with pages and then leaving rather than those who are seriously determined to vandalize.--Chiklit 13:51, 18 December 2005 (UTC)


 * Captchas? Dnavarro 17:11, 24 December 2005 (UTC)

Question1
Now that we seem to have a pretty solid consensus on this, where do we go from here? Jimbo has been informed of this but maybe another shout out on his talk page now that we've come to a general consensus in terms of what the policy should say? I've been working on the project for a year now but I've never done policy. Any enlightening would be helpful. :) And yes I read the How To Create Policy thingy. --Woohookitty(cat scratches) 04:15, 16 December 2005 (UTC)
 * Actually, I will leave a note on Jimbo's page. --Woohookitty(cat scratches) 04:16, 16 December 2005 (UTC)
 * I also left a note on the Village Pump page. Hopefully someone will tell us what to do next. :) 100 to 4 is a damn good consensus. Don't think anyone can argue with it. --Woohookitty(cat scratches) 05:32, 16 December 2005 (UTC)
 * Village Pump probably won't yield anything. Another call to Bugzilla should probably be the first step, as well as notifying developers in the IRC channel of the status.  Also, any developers monitoring this page should leave feedback as to what the status of implementation is. --kizzle 09:16, 16 December 2005 (UTC)
 * Bugzilla Link --kizzle 18:43, 16 December 2005 (UTC)

100 list
Splash, I guess we can add this proposal to the Times_that_100_Wikipedians_actually_agreed_and_voted_to_support_something list, now, huh. --kizzle 04:43, 16 December 2005 (UTC)
 * Indeed, which is excellent news! I'm surprised to see such large numbers visiting. I'm also pleased to observe that discussion has continued even as the poll is working; a good balance between the evility of voting and the process of discussion. Note that Kim Bruning's comment (he'd never vote) sounds like an oppose to me, although it's a pity he didn't revisit the discussion. There are a number of experienced Wikipedian's whose names I'd like to see somewhere, but if they aren't here, they aren't. However, User:Datrio is one of the Stewards, so that's encouraging, especially since he indicates that he has changed his mind over the course of the process. -Splash talk 14:50, 16 December 2005 (UTC)
 * Sweet. --kizzle 18:08, 16 December 2005 (UTC)

Woohoo!
Apparently, we have Jimbo's attention. Looks like he's going to talk to Brion about how hard it'd be for this to be implemented. Given all of this civil lawsuit nonsense, this is probably a very good time to hit him up on something like this. --Woohookitty(cat scratches) 15:33, 16 December 2005 (UTC)

Classes of Users
Being one of the newest X% of registered users is not simple or predictable.

I propose a waiting period of 14 days and 50 edits before a new user is enabled for editing the frequently vandalized pages. An admin could examine those 50 edits and see if they were merely an evasion of the requirement (i.e. adding a comma in 50 random articles, for example)

The reason a number of days is important is because it's not variable on the actions of others (i.e. the number of other new registrations) and 50 edits is indicative of a desire to contribute to the Wikipeda and not a "hit and run" on a political target. patsw 15:39, 16 December 2005 (UTC)
 * The developers have indicated that, at present, there is no means for determining account-life in software (odd as that sounds), so we'd be asking for structural changes to the database. Counting edits is apparently too computationally draining. Also, correcting 50 commas is fine: we don't have a standard of editing, other than it be not-vandalism. I'd appreciate someone fixing 50 of my mis-commaed articles, for example! On the precise values you choose: 14 days is vastly too long. I would prefer something that equated to 2-4 days and no more. 50 edits is perhaps alrihgt, although whatever the barrier is, it should be nearly trivial to overcome by any new, good-faith contributor, imo. We want all the good-faith edits we can get our hands on. -Splash talk 16:07, 16 December 2005 (UTC)
 * Yes, remember this is not to punish new users, but rather to stop "hit & run" vandals that plague certain pages. So it restricts anon & new accounts during the time needed by non-admins to fix the vandalism while still being able to contribute, and maybe a few hours more in case it's someone sitting at a workstation or school library. "New accounts" is needed to discourage "well if anon accounts can't edit, I'll just quickly create a new account & vandalise anyway". So really as long as the "new user" ban extends to accounts that are less than 1/2 a day old, it works. New accounts which vandalise already have a system for banning/dealing with repeat vandalisms. Right now I think the % suggested works out to 3-4 days, right? And the % can be tweaked if the minimum time gets too short later on. Jabrwock 16:18, 16 December 2005 (UTC)
 * Yes, apparently 3-4 days is about right. (Although a calculation based on growth of 10% per month at a constant rate per-day suggests rather shorter than that - less than a day). I suppose the devs are able to amend it if they need to, since it is currently the same restriction as on move, and I presume thinking-ahead has been done there for us. -Splash talk 19:50, 16 December 2005 (UTC)

Splash, I can accept the difficulties involved in implementing my proposal making it impractical. If something like this is to be done, it should very, very difficult to game. I thought we did have a standard of editing reflected in policies of the Wikipedia. In terms of relative needs, I think the Wikipedia needs a good solution to the vandalism problem more than it needs all the good-faith edits Splash can get his/her hands on. If not as if improving the vandalism situation will trigger a boycott by the good editors.patsw 18:36, 16 December 2005 (UTC)
 * I would hope that allowing virtually all of our editors to edit a page except from those who are wanton vandals will be a good balance. We have blocking powers over those that game the system, and that's given added effectiveness with semi-protection since any sockpuppet a blockee may create won't be able to edit a semi'd page. -Splash talk 19:50, 16 December 2005 (UTC)


 * Patsw, the possibility of including an edit count check along with a time calculation is not off the table yet, we're just going to see if we can do without it first. If it becomes apparent that these pages still become heavily vandalized through legions of undead sleeper accounts, then we'll take this back to the developers, but initially, we don't want them to do more work then they need to do. --kizzle 21:38, 16 December 2005 (UTC)


 * Avoiding the sleeper vandals -- How about auto-invalidating inactive users accounts based on number of edits (where less than 100 lifetime edits will trigger an auto-expire on the user account in 180 days after the last edit, to throw some numbers out). patsw 23:07, 17 December 2005 (UTC)

Poll closed
I've archived the poll, the discussions that went with it and the comments more than about a day old to Archive 3. By simple number counting, we have 103/4/2, but there is at least one comment that reads as an oppose and one that reads as a support; in both cases the not-voting appears deliberate. Discussion continues, naturally, and do note that Jimbo called by indirectly as mentioned above in. -Splash talk 19:20, 16 December 2005 (UTC)
 * Having an official vote now seems way too bureaucratic, with 96% support... it's time to implement and tag this with policy. Sadly, I can't reach IRC to go to #wikimedia-tech to ask a dev to look into this in detail... Tito xd (?!? - did you read this?) 19:26, 16 December 2005 (UTC)
 * Agreed. The proposal received 96.26% support and was spammed all over Wikipedia, so I believe it is fair to say there is a strong consensus behind implementing this feature without a recount.  Hall Monitor 19:36, 16 December 2005 (UTC)
 * Wow! that's the first time I've seen so many wikipedians actually agree on something. G-Man  19:40, 16 December 2005 (UTC)
 * Yes. I would, however, disagree with tagging this as policy without Jimbo actually blessing it. It's a change in the way we do things that would need his overt sanction. IMO, anyway. -Splash talk 19:42, 16 December 2005 (UTC)
 * *Sigh* Well, it seems that I was in the (vast) minority here. I would still like, if this is technically implemented, to see a policy drafted or something added to the blocking policy saying that pages shouldn't be semi-protected permanently or for extended/long periods of time, and that this kind of protection should only be applied to articles facing serious and extraordinary levels of vandalism. Thanks! Flcelloguy (A note? ) 19:46, 16 December 2005 (UTC)
 * That was part of the agreement, though. So, I don't know if it should just be made more explicit. Tito xd (?!? - did you read this?) 19:49, 16 December 2005 (UTC)
 * Will this page become the policy page? I'm assuming that either Protection policy will be modified or a new page will be created to document the policy. Flcelloguy (A note? ) 19:50, 16 December 2005 (UTC)
 * Umm, well this page is called Semi-protection policy which seems like a reasonably decent name to me. We could merge it into the full protection document upon formal acceptance. However, for now, I think we should keep the discussion here. -Splash talk 19:59, 16 December 2005 (UTC)
 * Well, it seems we now count with Jimbo's Blessing. Let's move ahead and tag this with policy Tito xd (?!? - did you read this?) 00:04, 17 December 2005 (UTC)
 * I really do think that the language on the front page is pretty clear about this. -Splash talk 19:52, 16 December 2005 (UTC)

Non english projects
I just discovered today that at least one non english pump has received a call for participation to discuss this policy. I am a bit wondering about this. Mysecurity, are you implying that a policy currently under work (well, apparently now closed) on the english wikipedia, will be applied to all projects ?

Anthere 22:02, 16 December 2005 (UTC)
 * Not really, Anthere. It would be an extension to MediaWiki, which could then be applied to other wikis, after local discussion in each wiki. So, the proposal could affect other projects, but won't unless they want it. But we wanted to genererate as much discussion as possible (and still want to), so we wanted to adjust it to the need of other Wikimedia projects too by asking for external opinions. Tito xd (?!? - did you read this?) 22:06, 16 December 2005 (UTC)
 * After seeing the message, I doubt many understood that.... you might have mentionned a change in the software... By the way, I am curious. How much feedback did you get from other languages ? Anthere 22:20, 16 December 2005 (UTC)
 * Probably it should have been reworded to make that clearer, but no, it is not mandatory for other wikis to adopt... but we did get Datrio from Polish Wikipedia comment and think it is a good idea, as well as Unfinishedchaos from the Arabic Wikipedia comment on the issue. I'm active on the Spanish Wikipedia, so I tried to give a few interwiki thoughts as well. Tito xd (?!? - did you read this?) 22:38, 16 December 2005 (UTC)


 * A couple of wikicities-hosted sites (wikicities:c:furry:talk:furry and uncyclopedia:uncyclopedia:Village Dump) have already expressed some interest, but is the not-so-minor detail that Wikicities is still MediaWiki 1.4 going to break things? I understand that semi-protection is a new MediaWiki extension created for Wikipedia, and the main en.wikipedia is on 1.6alpha? --carlb 20:24, 17 December 2005 (UTC)
 * Special:Version says we're running MediaWiki 1.6devel. Since, so far as is apparent to those of us that don't understand Bugzilla and the like, semi-protect has not been implemented at all: probably there are 0 lines of code written. So it's not possible to predict in what form SP will manifest in MediaWiki. -Splash talk 21:16, 17 December 2005 (UTC)


 * Also to note, is that even if changes the softwre and makes this 'feature' available to all wiki's they would not have to implement it. xaosflux  Talk  / CVU  23:36, 18 December 2005 (UTC)

Following up on proposed policy
The recently discussed/voted on policy seems reasonable, but we should be thinking of how the vandals will get around it. The simple answer is that they will create numerous accounts and let them sit for however long is required to bypass semi-protection before launching attacks. That requires sufficient 'dedication' that it should deter most 'impulse vandals', but the zealots will immediately go for that loophole. Is there any way to prevent/discourage or at least identify that sort of activity? Also, what about the growing list of indefinitely blocked usernames which this policy will spur? Should we have a process for clearing these out so the names can be claimed by legitimate users in the future? Overall this 'semi-protection' idea seems better than the current alternatives, but we should anticipate issues which will arise from it. --CBD ☎ ✉ 00:13, 17 December 2005 (UTC)
 * These are questions that have been asked about issues other than SP. The username list thing is easy: the devs could clear out accounts with no edits if they wanted to, but there's no technical reason to do so. Any account with >1 edit can't be removed, for obvious reasons.
 * We regularly block large numbers of accounts indefinitely seconds after their creation as it is. The determined vandals have been doing this forever. I wouldn't think we will create (m)any more determined vandals with this protection since there is already enough motivation to create throwaway accounts as it is e.g. pagemove vandalism, vandalbots etc. We've handled them all before, and will handle them again (sometimes with dev help, but that's life).
 * How to discourage it is worth thinking about, in the larger sense (it isn't a problem peculiar to SP, imo). I don't honestly think that implementing the double hurdle of including an edit count threshold will make much difference to determined vandals: they're determined after all! We couldn't make it a very high numerical threshold for obvious reasons and so your average vandalbot could deal with it easily. I don't have any suggestions off the top of my head; I'm just giving some context. -Splash talk 00:24, 17 December 2005 (UTC)
 * I think an edit count + time check will indeed make a significant change should our initial sp of only a time check fail. This will remove impulse vandalism, but its not to say that any kid can still write down on a piece of paper the 10 accounts they created today, and wait a few days and then use every single one.  Utilizing both checks requires each fake account created to have 50 edits, which is a daunting task unless one uses a vandalbot, which will be in the minority. I think we do need to seriously consider the proposal of deleting fake accounts, as you're going to see account creation substantially increase, and there's going to be a lot of indefinetely banned accounts with 1-10 vandalism edits.  There's no reason to keep them around. Is it truly technically impossible to remove these accounts? --kizzle 01:05, 17 December 2005 (UTC)
 * We'd be breaking the attribution requirement of the GFDL if we removed the accounts, unless we also removed their edits and any edits that derived those edits from the articles' histories in which they appear. -Splash talk 01:35, 17 December 2005 (UTC)
 * Yikes, that seems like that's going to bite us in the ass years down the road. --kizzle 01:37, 17 December 2005 (UTC)
 * I have wondered about that. The counter argument is that, if the edits are all vandalism that was immediately reverted, then there are no derivative revisions in the later history of the article that would be affected by e.g. a history deletion. We occasionally do remove edits from articles' histories when they are of this nature, althuogh some have questioned whether that is permissible or not. The vandals are authors too, after all, even if they are unwelcome ones. -Splash talk 01:40, 17 December 2005 (UTC)
 * Hmm, well I guess 20 years down the road if people are using User:JohnSmith30948230948203984, they can just change their sig. --kizzle 01:42, 17 December 2005 (UTC)
 * .....if we don't usernameblock them for being too similar to the one before! -Splash talk 01:45, 17 December 2005 (UTC)
 * Yeah, User:JohnSmith30948230948203983 makes us all look bad. -User:JohnSmith30948230948203985
 * Would it be simple to rename the accounts User:JohnSmith22 becomes User:vandal200512171800... 82.12.106.84 18:10, 17 December 2005 (UTC)
 * In fact, yes, it would, as long as the vandal has fewer than 6800 edits. We can already change usernames so that might be a way of freeing up desirable names in future, as long as the vandal never became established or anything. -Splash talk 18:14, 17 December 2005 (UTC)
 * But I'm not sure if that would violate GFDL. Since that user in a way requested to be identified as User:Johnsmith22, making them User:vandal200512171800 might violate their "right" to be identified as authors.  Ral315 (talk) 20:49, 18 December 2005 (UTC)

co-operation with Tim Starling's new feature
I saw on GWB that Tim is going to implement a "last good version" feature in January for pages like GWB... will this be used in addition to, instead of, as a primary solution to semi-protection? --kizzle 02:22, 17 December 2005 (UTC)
 * Hrm. Is this getting confused with article validation that is supposed to be coming in Jan? That was a Brion Vibber project, I thought, so perhaps not. Difficult to say without reading something about the feature. It reeks of content review by admins, though. -Splash talk 02:34, 17 December 2005 (UTC)
 * Amen, brotha. I wonder if the Wikipedia community will actually have a say in this implementation. --kizzle 03:07, 17 December 2005 (UTC)
 * Of course not. Users telling us what to do? Ridiculous idea. ;-) 86.133.53.111 17:21, 17 December 2005 (UTC)

Now that we have approval
Are we going to modify the existing protection policy page to incorporate this or do we need to rename that page full protection policy? Seems like a minor thing, but it's not actually. I think we should rename the other page full protection and have this exist as 2 pages, since they are 2 different policies to a large degree. --Woohookitty(cat scratches) 02:57, 17 December 2005 (UTC)
 * I would agree with the seperateness, with obvious links between the two. Perhaps in time, they will naturally become One. I would, though avoid impinging upon other, already functional policies, until this one actually switches on. We have no reason, at present, to suppose it will be any time soon. (I am willing to be proved wrong, and to consume my second hat of the day.) -Splash talk 03:02, 17 December 2005 (UTC)
 * Yeah we need to wait until it goes live before actually making any changes. But I wanted to get this debate going because it is important. If it's 2 pages, we need to change PP so it clearly states that it is just for full protection and that there's another page for SP. --Woohookitty(cat scratches) 03:14, 17 December 2005 (UTC)
 * Are users going to be able to request semi-protection or will this be applied at the whim of an admin? --kizzle 03:39, 17 December 2005 (UTC)
 * "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". -Splash talk 03:45, 17 December 2005 (UTC)
 * Sweet, although I have a feeling there's going to be a lot of semi-protection requests. --kizzle 03:46, 17 December 2005 (UTC)
 * So? :-D We can handle it. We currently have about 5-6 new requests a day on RfP. We have 4-5 admins patrolling it. Another 20-30 a day could be handled and I doubt we'll get to those levels. We might get a flood to begin with (and honestly, I think GWB and Daniel Brandt might be SP without specific requests as tests for the policy), but it'll be something we can handle I think. --Woohookitty(cat scratches) 04:01, 17 December 2005 (UTC)
 * Ok, ok, just commenting, wasn't trying to argue! :) --kizzle 04:58, 17 December 2005 (UTC)
 * I know. :) As I said, I'm proud of the discussions we had as a group. I wish all discussions on Wikipedia were that intelligent and civil. --Woohookitty(cat scratches) 05:05, 17 December 2005 (UTC)
 * Ya too bad we all don't live near each other so we can't celebrate and get drunk at a bar and sing karaoke. --kizzle 05:11, 17 December 2005 (UTC)
 * I drink to that! ≈ jossi ≈ t • @ 03:40, 19 December 2005 (UTC)
 * Well, don't forget that the final vote was 95% in favour. It was a discussion without argument, because everyone was already basically in favour. No argument=no heat. Stevage 20:53, 19 December 2005 (UTC)

Great idea
Fantastic idea... I wonder if they can implement this in the MediaWiki software? - Ta bu shi da yu 04:00, 17 December 2005 (UTC)
 * Glad you like it! In Archive 3 (see link up top), Aevar and Brion have both commented without saying it can't be done. They have pointed out that certain things would mean more work or server load than others, and the version on the front page, and subjected to the poll, deliberately avoids asking for those things. -Splash talk 04:10, 17 December 2005 (UTC)
 * Yep. And Jimbo gave pretty damn fast approval. :) He must think it'd be doable or else he wouldn't have. This debate hasn't even been going on for a month yet and Jimbo said yes. --Woohookitty(cat scratches) 04:16, 17 December 2005 (UTC)
 * This makes me wonder if it's a good time to bring up another MediaWiki idea I had - locking sections rather than entire pages. This would probably be well beyond the scope of what's being discussed, though. --Starkruzr 11:41, 17 December 2005 (UTC)
 * I highly doubt this is possible- the main problem being that even if it were possible, vandals could likely remove the section, or just vandalize other areas of the page. Ral315 (talk) 20:46, 18 December 2005 (UTC)
 * I think Starkruzr is probably talking about normal protection for controversial topics (rather than frequently vandalised ones). In that case if it's possible to lock a section, presumably the people would have the common sense not to go around the lock to edit it again. Stevage 20:51, 19 December 2005 (UTC)

Btw
I added a note to the top stating that this has yet to be implemented. Hopefully that'll stave off early requests. Edit it if you feel like it could be worded better. --Woohookitty(cat scratches) 04:08, 17 December 2005 (UTC)
 * I also changed the tense on some of the words and spelled out the acronyms since this is official policy now. :) Needs to be readable for John Q. Wiki. :) --Woohookitty(cat scratches) 05:23, 17 December 2005 (UTC)

WebDSign
There is a way to implement such protection in a non-intrusive way: WebDSign application to Wikipedia.

Nasty side-effect
There seems to be a nasty side-effect to this change. Consider the pages that a new user is most likely to visit and edit. It will commonly be one of the most popular articles, which are probably also the most vandalised articles. In a way, this change prevents a new user from making an edit to a page as they see necessary.

I would have thought that just disabling anonymous edits on semi-protected pages would have been a good place to start, before denying new users edits on popular pages. - James Foster 11:38, 17 December 2005 (UTC)


 * We've discussed and debated all of that. Here is a quick rehash. And btw, popular doesn't always equal most vandalized. And we're talking 20-30 articles at most. Anyway, you can read the rest for yourself. :) --Woohookitty(cat scratches) 12:27, 17 December 2005 (UTC)
 * There does remain the possibility of letting X = 0. The implicit supposition at present is, so far as I understand it, that it would actually be the sam X as currently restricts access to the "move" button. -Splash talk 16:59, 17 December 2005 (UTC)


 * From the users point of view, vandalisation of popular pages has more impact on the users perception. At some point, user perception is the main point of this kind of mechanism (since the history permits seeing all earlier, unvandalised edits.)WolfKeeper 01:48, 21 December 2005 (UTC)

New topics immune to Semi-protection
I like the idea in general, and while Semi-protection is designed for oft-vandalised pages, I would like to recommend that it not be possible for anyone to semi-protect a new article for ndays after its creation.. even if it's being vandalised. This allows the maximum number of people to participate in the article's 'birth', even if it's a painful birth. My primary reason for this recommendation is that it ensures the Wikipedia is able to be immediately up-to-date with breaking news from people who've never edited before.--RickMeasham 12:22, 17 December 2005 (UTC)


 * I just don't see the need for that. For one thing, articles are rarely protected now that are only a few days old. Secondly, as we've stated before, we're trying to start with as few restrictions as possible. I don't see admins semi-protecting pages that are 2-3 days old. They don't fully protect new articles, so I'm not sure why they would semi-protect new articles when it's clearly for prolonged vandalism. Admins are pretty self-policing. Trust me. I've been yelled at enough to know. :) --Woohookitty(cat scratches) 12:30, 17 December 2005 (UTC)


 * heh .. thanks for assuaging my concerns :) --RickMeasham 12:39, 17 December 2005 (UTC)
 * Indeed. One might reasonably suppose that brand new pages would usually be free-ish from vandalism at first anyway. THey might well face an edit-war over any number of things, but as the front page says, semi-protection would definitely not be appropriate then. -Splash talk 16:41, 17 December 2005 (UTC)
 * See 2005 Sydney race riots for an example. A new and controversial issue which brought in a lot of people who really have no idea how wikipedia works.  It would be very tempting to semi-protect that page.  It would be wrong, but still very tempting.  :-)  Regards, Ben Aveling 20:56, 18 December 2005 (UTC)

Good work
And here I was thinking that Wikipedia would be able to avoid the classic decay and decline of other internet communities, the steady erosion of the basic principles that made the community and project work in the first place. Ah, innocence. Doesn't anyone remember that VotingIsEvil? The Cunctator 13:53, 17 December 2005 (UTC)


 * If properly implemented, I think the gains in quality and professional tone will attract more than enough new editors to make up for the ones turned off by the policy. If improperly implemented, then you're right, Wikipedia will end up turning in on itself like other internet communities.  We'll just have to see to it that it's properly implemented, eh?  Or I suppose we can give up, declare the whole thing a loss, and go home.  We all have to make that choice individually.  Stephen Aquila 15:37, 17 December 2005 (UTC)
 * This is, at very worst, a minor 'erosion' of principles. It will be applied no more widely, I hope, than full-protection is at present. So there will be no greater number of pages protected than there already are, and for no longer. The community will see to that. It is, at best, a de-erosion of principles since, rather than having 0.11% of editors able to edit a page, and forbidden by long-standing convention from doing so, something like 96-98% of contributors will be able to edit the page. There is no suggestion that we apply this to random pages just 'because'. Do read carefully the wording on the front page that refers to pages facing serious vandalism. Voting is evil. Yes, we know. The immense amount of discussion that went with it (you did read the whole of the archive, yes?) and the allaying of most fears that it resulted in is a good thing. The straw poll is just an optional extra to clarify the process. -Splash talk 16:41, 17 December 2005 (UTC)

Should Semi-protection policy be semi-protected?
;-) The policy page just got slashdotted 4 ½ hours ago.  Can&#39;t sleep, clown will eat me 15:56, 17 December 2005 (UTC)
 * No, since Semiprotection is not a proposal to pre-emptively protect pages that might face vandalism. :) It's a well-written policy, you see! -Splash talk 16:41, 17 December 2005 (UTC)
 * Although it was a lighthearted joke when I first wrote that, it looks like we're starting to see some of the residual effects of the slashdotting now, in the form of vandalism. Can&#39;t sleep, clown will eat me 17:01, 17 December 2005 (UTC)
 * Now . Flcelloguy (A note? ) 17:05, 17 December 2005 (UTC)

Smooth move to whoever protected it
You protected the wrong version. Notice the nice "HA HA" @ the bottom! 68.39.174.238 17:02, 17 December 2005 (UTC)


 * Now reverted. Cut it out, or you will be blocked. Flcelloguy (A note? ) 17:05, 17 December 2005 (UTC)


 * Then why is it still there? Do you people know what it's like for a pedant when he/she can't edit a syntactically incorrect page? I may have to print out this article and either apply some white-out or some judicious red-pen markings just to get some sleep tonight!! ThePedanticPrick 18:42, 17 December 2005 (UTC)


 * I don't think it was him. &mdash;Bunchofgrapes (talk) 17:12, 17 December 2005 (UTC)


 * It wasn't. 68.39.174.238 is the infamous "RVC" anti-vandal.  Can&#39;t sleep, clown will eat me 17:13, 17 December 2005 (UTC)

Support from Jimbo
How is that one sentence relevant? I get the feeling this is used to round up votes from those who admire him, a sort of appeal to authority or respect. Anyway, I do not think a statement like that should be put in bold fonts on a project page, it's just silly. --Friðrik Bragi Dýrfjörð 17:16, 17 December 2005 (UTC)
 * It was Jimbo who wrote it there, and if Jimbo says green is red, then green is red. → Aza Toth 17:19, 17 December 2005 (UTC)
 * I think he's talking about the note at the very bottom of the page, "Jimbo has also expressed his support.", which Splash added, not the press disclaimer at the top. &mdash;Bunchofgrapes (talk) 17:21, 17 December 2005 (UTC)
 * Ah, ok, that should be removed yes → Aza Toth 17:22, 17 December 2005 (UTC)


 * Well said, that man. Jimbo might call the shots at the upper level, but he is not a God. Statement such as, "if Jimbo says X is Y, then X is Y" are completely ridiculous. If Jimbo told you to jump off a bridge, would you do it?


 * As to whether or not Jimbo's support for something should be allowed to influence whether or not people support it as well; well, if they do, that's their lookout. At the end of the day, are you a dog, a sheep, or one of the pigs? 86.133.53.111 17:26, 17 December 2005 (UTC)
 * I just feel like it goes against the NPOV policy, even though it isn't an article. If a person has an opinion and supports it with valid and good argument, should that opinion be on the project page in bold fonts?  If I had used this as an argument in support of this policy, those who argued against it would have said that such an argument is invalid, the fact that Jimbo supports something or not doesn't make it right or wrong.  I don't have anything against Jimbo, I just don't like to think of him as something more (or less) than an equal even though he might be thought of as a first among equals. --Friðrik Bragi Dýrfjörð 17:41, 17 December 2005 (UTC)

Jimbo expressed his support only after the close of the straw poll and the {policy} tag was added only after Jimbo had expressed his support. It was not used for recruitment purposes, and is there merely as a justification for the {policy} status of the page. And anyway, if Jimbo says we do something, we do it. -Splash talk 17:54, 17 December 2005 (UTC)
 * I wasn't accusing you (or anyone) of rigging the votes, but using this as a justification is silly. The votes justify the policy (and are sufficient), Jimbos comments are completely irrelevant as a justification in this example. If Angela had noted somewhere she didn't like the policy, should that be in bold fonts on the policy page? --Friðrik Bragi Dýrfjörð 18:31, 17 December 2005 (UTC)
 * Jimbo's comment is not irrelevant, because this does represent a change in the way we do things and, if he had said "no" the policy would have failed irrespective of the straw poll. If he had said nothing, we'd not yet be calling it policy. Angela isn't Jimbo. -Splash talk 18:32, 17 December 2005 (UTC)
 * Please. Citing the endorsement of the man who runs this place, while being sufficient but not necessary, is entirely relevant. --kizzle 18:39, 17 December 2005 (UTC)
 * Yep. I don't remember where the list is, but there is a list of policies that have been passed in the last couple of years and 2/3rds got passed because Jimbo pushed them through. You bet it's relevant. Ah, here is the list. --Woohookitty(cat scratches) 21:50, 17 December 2005 (UTC)

"temporarily" protected from editing
The template message states that the page is "temporarily" protected from editing by anonymous users. However, the policy is then described as a solution to the constant vandalism on, for example, George W. Bush. This seems like a rather disingenous template message -- how often would George W. Bush be unprotected? -- Creidieki 18:40, 17 December 2005 (UTC)


 * Yeah, we should perhaps change that rationale part. However, the current vprotected uses the self-same wording and GWB is usually unprotected, so there's nothing disingenuous there. There is no intention that pages be permanently semi-protected, and permanent protection of even GWB found little support and significant opposition when discussed on WP:AN/I. -Splash talk 18:42, 17 December 2005 (UTC)
 * I think you're making a bit of a joke here, but I don't get it, and it's not funny. ThePedanticPrick 18:44, 17 December 2005 (UTC)
 * I was attempting to point out that the phrasing used to gain support for this policy didn't seem to mesh with the described usage of the policy, and that the policy was very unclear in describing its usage. I find that disturbing, especially since I hadn't seen this discussion earlier.  It's rather the opposite of a "joke".  -- Creidieki 19:32, 17 December 2005 (UTC)
 * Part of this stems from the fact that someone has brutalised the paragraph dealing with un-semi-protection into what it is now. I will change it back when the page is unprotected, which should be soon. We don't protect against mere Slashdotting.-Splash talk 19:45, 17 December 2005 (UTC)
 * I don't think it's a joke. Creidieki, initially I had proposed a system of incorporating time limits into the process of semi-protection with an increasing minimum level of admins required to consent to extended semi-protection, but Splash has assured me that admins will keep themselves in check without having to resort to fixed time limits specified within policy.  I think (and hope) what will happen on GWB is that it becomes semi-protected for a week or so, then lifted so we can analyze the results.  Even if there is 100 vandalism edits a day, then none under semi-protection, SP should still be lifted on a weekly basis at the very least, if not sooner. --kizzle 19:15, 17 December 2005 (UTC)
 * I think that would be a good balance. Based on the way things work at present, in a fairly wide variety of admin actions, admins as a whole are reasonably good at self-regulating by a mixture of shouting at one another, reaching compromises and respecting the spirit as much as the letter of policies and guidelines. Then, there's How to create policy which is good advice we've followed in the construction of this policy. In this particular context, points 2 and 3 are especially salient. -Splash talk 22:55, 17 December 2005 (UTC)

I fail to see the problem. GWB would be temporarily protected for 3-4 years. Once he is no longer president, I imagine his article will be much less liable to vandalism. (Oh wait, I forgot: 4 years is forever.) —Steven G. Johnson 05:38, 18 December 2005 (UTC)
 * If you're living in the U.S., it is. --kizzle 22:22, 18 December 2005 (UTC)

What happens now?
Now that it has (as far as I can tell) been agreed that this should become policy. What happens now?. How long might it be before we see this in action? G-Man 22:07, 17 December 2005 (UTC)


 * Hello, anyone there??. G-Man  17:23, 18 December 2005 (UTC)


 * As far as I know, we wait for a developer to implement semi-protection. Other levels of protection have been alluded to by Jimbo, such as "the ability to put pages into a 'validated' state", but I do not know where those discussions took place.  Anyone have a link?  Can&#39;t sleep, clown will eat me 17:47, 18 December 2005 (UTC)


 * The 'validated' talk as been going aroudn for quite some time, frequently on the project to make a print/cd/dvd version. Once the dev's are complete they will need to publish the proceedure for admins to work with the new protection level.  xaosflux  Talk  / CVU  23:40, 18 December 2005 (UTC)


 * This all sounds very interesting, where can I read about it in greater detail? Can&#39;t sleep, clown will eat me 00:40, 19 December 2005 (UTC)


 * Try Google, ya can't go wrong :). - Mys  e  kurity ( have you seen this? ) 00:43, 19 December 2005 (UTC)
 * Well....Article validation (I'd been watching this for sometime now, and I JUST forgot about it...) - Mys  e  kurity ( have you seen this? ) 00:44, 19 December 2005 (UTC)


 * Pushing to 1.0 has all of the info you'd ever want on it all. I dislike it myself. I just dislike this idea that people are going to decide what is "correct". --Woohookitty(cat scratches) 00:43, 19 December 2005 (UTC)

Getting back to the original question. I surpose what I really want to know is what sort of timescale are we talking before we see this in action?. Will it be days, weeks, months or what, or doesn't anyone know. G-Man 20:15, 19 December 2005 (UTC)
 * Days or weeks. As far as I know, semi-protection has now been coded, and it is just undergoing testing in the developers' test wiki. It won't be too much before it is actually deployed on Wikimedia. Tito xd (?!? - help us) 05:47, 22 December 2005 (UTC)
 * Tito, do you have the right Bugzilla link there? I don't quite understand what that has to do with semi-protection. Blackcap (talk) (vandalfighters, take a look) 05:53, 22 December 2005 (UTC)
 * Scroll down to the very bottom, it says it is a fix for Bug 675 too. :) Tito xd (?!? - help us) 22:18, 22 December 2005 (UTC)
 * Oh, got it. Thanks! Blackcap (talk) 22:21, 22 December 2005 (UTC)

GWB stats
Just thought I'd post something interesting I found:

According to statistics, George W. Bush is the most edited article (by far), with 21507 edits by 1762 unique registered users and 4902 unregistered users or 59% registered. The total size of the article is by far the largest: 1323.8 Mb, which is ginormous. Just thought people might like to know, Mys  e  kurity ( have you seen this? ) 23:54, 17 December 2005 (UTC)
 * Why do they let soooo many versions get saved, that is just a horrible waste of server space. There should be a limit, like 2000 versions. A separate list can be made for the GFDL under a "contributors" tab for old versions. This site must use sooo much server space. Voice of All T 18:08, 18 December 2005 (UTC)
 * According to Wiki is not paper, it costs one (USD) penny per 2 million 7-letter words, and the statistics page says that all the revisions of George W. Bush uses up something like only 1 GB (1323.8 Mb), which equates to something like a dollar worth of hard drive space. Probably, the main cost is time of all the volunteer editors having to clean up the vandalism. --AySz88 ^ -  ^  18:30, 18 December 2005 (UTC)
 * Ever heard of the GFDL and its requirement that authors be credited for their work? Vandals are authors too, just unwelcome ones. -Splash talk 18:43, 18 December 2005 (UTC)
 * If a vandal's edits are reverted, the vandal is not considered an author. Nothing contributed by a vandal is kept. —Guanaco 19:10, 18 December 2005 (UTC)
 * IANAL, but I think that's a matter of opinion. -Splash talk 20:05, 18 December 2005 (UTC)

IAAL (indeed, IAAIPL), and here's my analysis:

The purpose of the GFDL is for editors to license their contributions. If a vandal blanks a page, or randomly adds some vulgarity or other nonsense, and that vandal is swiftly reverted, what is his contribution? There is none, nothing to give credit for, nothing to license. In fact, the question of licensing is probably already irrelevant in such circumstances, because there is rarely anything that the "author" could claim a copyright interest in. The mere removal of a block of text is not copyrightable under any circumstance, and neither are individual words or short phrases (such as "pelican s**t" or "wikipedia is communism"). The phrases are so short that the idea has merged into the expression (there is simply no more concise way of saying it), and therefore no copyright protection is possible (ergo, nothing is licensed). However, even if something copyrightable were added by a vandal, it violates no rights of the copyright holder to remove both the information and the history of the information having been added, because nothing that was added remains in the article - which is really what the GFDL applies to.

Similarly, although the reversion of vandalism is an important service, there is no contribution, no original thought or information added by the reverter, that would ever be protectable by copyright. In sum, the edit history of any vandalised page would probably benefit greatly by the removal of all the vandalism and reversion of same, and no rights would be violated in so doing. bd2412 T 21:31, 18 December 2005 (UTC)
 * Thanks for saying what I didn't feel like bothering to. And as I said, a list can do for the non-vandals of super old page versions. Also, Hojaboob45687(some vandal at a library) is not a very trackable, legal, name to attach to the libelous GFDL "contribution" "Bush rapes 14 year olds at fdfn-93454wew04".


 * Just saying...applying GFDL to vandals is not very legal or ethical; this is just bureaucracy for the sake of annoying bureaucracy. Voice of All T 21:41, 18 December 2005 (UTC)
 * Agreed. Thankya, VoA. - Mys  e  kurity ( have you seen this? ) 23:53, 18 December 2005 (UTC)
 * There are two benefits I can thing of to keeping such edits in the edit history. One is that it allows us to identify a returning vandal - if an individual has conducted similar vandalism from the same account and then been dormant a long time, they can be blocked without warning. The other is for pure sociological curiosity, to keep vandalism statistics like the ones that have been raised here already. Both can be accomplished with more economical means.  bd2412  T 00:56, 19 December 2005 (UTC)
 * Just curious - how is vandalism currently stored? Does use of the admin rollback feature allow the system to avoid saving a new copy of the same exact material? --AySz88 ^  -  ^  02:11, 19 December 2005 (UTC)

Here are some more stats I found using the Wikimedia.de tools server. Tony's tool counts the number of vandalism reverts, which looks like it's around 35%, but I'm not sure how the tool calculates it, and I'm too tired to fully read the script. Is this perhaps only Admin reverts? In any event, 35% is pretty damn high. What are others thinking? - Mys  e  kurity ( have you seen this? ) 05:41, 22 December 2005 (UTC)

Important question
I have not made up my mind whether I think this is a good idea or not.

My concern is that as Wikipedia (and other mediawiki sites, which I presume will have access to the proposed semi-protection feature by implementing new updates to the mediawiki software) grows, and more people sign up, the amount of time/effort/edits it will take for a brand new user to get out of the bottom X% of the food chain will increase in direct preportion to the total number of users on the site who signed up before them, making it gradually more and more difficult to become a new user. It's like inflation. And as I understand it, the principles of the place so far as I've understood them imply that there aren't supposed to be any elites. What is the limit on this percentage? Can it be set to 50%? Will it be? Perhaps we need some kind of formula to determine the required percentage rate based on the total number of people on the site.

So I guess basically my simplified question is, "What is going to determine X?" --Nerd42 03:05, 19 December 2005 (UTC)


 * I believe this will be worked out as semi-protected is tested, but remember that x% is a hack to accomplish what ideally should be a fixed time limit of somewhere around a few days. As the total population of Wikipedia users changes, the x% will be modified so the time limit in terms of days will stay the same. --kizzle 03:48, 19 December 2005 (UTC)
 * That omits the mathematics of the English Wikipedia (for which the policy was writte). As the rate of account registration rises, the time it takes to leave the newest X% falls. In recent months, we have gained new users at the rate of some 10% per month &mdash; an exponential rise. At present, it might be reasonable to assume 1% on the English WIkipedia, this being the same percentage barrier for access to the move feature. -Splash talk 05:09, 19 December 2005 (UTC)


 * That's only assuming this kind of substancial growth continues unabated. If, for example, if a wiki (not nessicarily en.WP) gets 1000 new users signing up in it's first month, and they're pretty acitve but there's no new users for six months, then five join, the original 1,000 will continue to be elite relative to the five until more people sign up. Not every wiki is wikipedia - with a feature like this it kinda makes it even easier for admins to discriminate on other wikis, probably without even meaning to. --  NERD42 [mailto:nerd42@gmail.com EMAIL ]  TALK  H2G2 UNCYC NEWS   22:09, 5 January 2006 (UTC)

A Great Idea! Here are some more ideas
Semi Protection is indeed a great idea. Jus a few days back, I thought of what we could do to make Wikipedia better. I came up with a lot, much similar to Semi-protection Policy. May be some of them can be implemented...


 * Bring in email-verification of users while registration.
 * Discourages editing/registration, for what benefit? Stevage 03:42, 19 December 2005 (UTC)


 * Anonymous users shall be allowed to provide information to be added, in the corresponding talk page of any article. Pages therefore shall be edited in bulk by the other X% of registered users compiling the information from talk page, and locked for further edits, and then edited again after some time interval, adding more information.
 * Discourages editing, increases workload of logged ins, for what benefit? Stevage 03:42, 19 December 2005 (UTC)


 * The semi-protection can be extented in an effective way by creating a cached, protected, offical version and a editable unofficial version.


 * A legal time-marked version of articles in Wikipedia shall be delineated, so that it can be used as a reputable research source. Any information later than the official version of the article should bear the stub mentioning 'use of the source later than offical version should be at users risk and unofficial'. So there need to be two versions of the same article. One official locked version and one unofficial 'being-compiled' version. Every 'being-updated' stub version should provide a official version tab along with the edit this page, discussion and history tabs.
 * Yes, I agree. I describe it as, published version is version X, editable version is X+n.Stevage 03:42, 19 December 2005 (UTC)


 * Often we link to an article in Wikipedia and find the information changed. Every official version should provide information as to how to cite that page along with the date and time stamp.


 * To cite an article, everyone shall be advised to provide the date/timestamp beside the link (This is not even necessary if we have regulated the editing process as said above) to official wikipedia article from which they took the information, which means that any information till that date is accountable by Wikimedia Foundation and it should be held responsible for official versions of all articles.
 * Even better, the link should take you back to the original article that the research was based on. Stevage 03:42, 19 December 2005 (UTC)
 * Even better, click on "Cite this article"! Tito xd (?!? - did you read this?) 03:45, 19 December 2005 (UTC)

- By making this happen, we can still have a fully editable wiki version of wikipedia which is contantly compiled and a protected version which can be used as great research piece!

-- Magnus astrum 03:36, 19 December 2005 (UTC)

Most of these suggestions have been brought up several times before. The Village Pump is a place where they will get a wider hearing, and where there are people who can mroe easily direct you to the existing discussions. This talk page is about the semi-protection policy. -Splash talk 05:09, 19 December 2005 (UTC)

Protecting the policy page
It seems like a good number of recent edits are vandalism and reverts on the policy page, likely due to media coverage. Should we put protection on it until it's out of the limelight? Oberiko 14:25, 19 December 2005 (UTC)
 * Nah, it's getting a few 10s of edits a day which is way below what even moderately popular articles get. It'll be ok. -Splash talk 15:56, 19 December 2005 (UTC)

BTW
We've had 4 requests for semi protection since this became official 3 days ago. I don't know if the note on the top of the policy page needs to blink and make noise or what for people to notice it. --Woohookitty(cat scratches) 16:45, 19 December 2005 (UTC)
 * That's interesting. Would they have qualified? Might be nice to make a subpage /Early requests or somesuch and list them there, along with their actual request text and a theoretical decision you might have made. Or would that be too precedent-setting? I think probably it could be in green red and orange and offer to visit you for dinner and there'd still be some people who didn't notice it...-Splash talk 18:38, 19 December 2005 (UTC)
 * Well the big blue tick saying "Hey this is in operation" might have something to do with people thinking it's, well, in operation. Stevage 20:49, 19 December 2005 (UTC)
 * The word "operation" doesn't appear anywhere on the page. -Splash talk 22:02, 19 December 2005 (UTC)
 * I think the problem is that the page says that the policy is "yet to be implemented", but it doesn't give any indication as to when it will be implemented. Is this going to be implemented tomorrow or 2 years from now? Kaldari 23:15, 19 December 2005 (UTC)
 * Developers are actually working on it right now. There have to be some changes to the protection user interface, which takes a while to do, but there is work being done. It probably will be done at around the same time as article validation. Tito xd (?!? - help us) 23:18, 19 December 2005 (UTC)
 * To answer the question from up above, I would've turned down all 4 requests because they were seeing this as a first resort. I mean even for vandalism, it's not a first resort. It should only be used for heavily vandalised articles and they were asking for S-P on articles where the vandalism was light (3-4 times a day). --Woohookitty(cat scratches) 00:09, 20 December 2005 (UTC)
 * I wish I could use this policy on the John Seigenthaler Sr. article. What a pain the ass! Too bad I can't do it yet. Kaldari 05:08, 20 December 2005 (UTC)
 * That's getting about 5-10 edits per day. It doesn't need any level of protection. Semiprotection isn't for protecting long-term against minor vandalism any more than full protection is. To quote, "Semi-protection is only to be applied as a response to serious vandalism", which 10 edits a day doesn't really reach the level of. Basically SP is appropriate if full-protection would be at present, otherwise it's probably not. -Splash talk 10:57, 20 December 2005 (UTC)
 * I think if I was the one reverting 10 edits a day, I would disagree with you that that's not "serious vandalism". If each vandalism survives 10 minutes, that's nearly two hours a day that the page is vandalised. Stevage 12:59, 20 December 2005 (UTC)
 * Maybe not all pages should be compared to whether its vandalism approaches GWB level. Something like a qualitative calculation of how many editors there are working on a page vs. the amount of vandalism... what do we do if a page has only one editor with it watchlisted and it gets vandalized 10 times a day? --kizzle 17:46, 20 December 2005 (UTC)
 * Stevage, it isn't quite THAT bad. It looks like most of the vandalism on that page is reverted after 2 minutes, at most. If the Seigenthaler Sr. case was up for semi-protection, I'd probably say yes just because right now, he's in the news and we are seeing alot of vandalism but we're also seeing alot of real edits. Good candidate for it. Kizzle, it'll be on a case by case basis. I don't think the admins who patrol the page are going to compare everything to GWB. Generally, we take things case by case. We rarely compare requests to previous pages we're protected. That's partially why I didn't want time limits and such. The reasoning for protection just isn't uniform enough for uniform rules. And as for just 1 person watchlisting a page, that doesn't happen with RC Patrol and the vandalism room on IRC and the Vandal Fighter. Some vandalism gets through but not if it's that sustained. --Woohookitty(cat scratches) 18:22, 20 December 2005 (UTC)
 * Obviosuly the nature and frequency of vandalism depends a lot on the page. There's an anon the frequently makes exactly the same apparently incorrect change to Paris. There's an editor who keeps changing it back.  Assuming the editor is right, this vandalism is extremely annoying, but will never be caught by any vandalism detectors, because you have to know the subject to know that it's wrong. SP would be great, but the vandalism level is well below the threshold (3-4 times a week). Stevage 21:45, 21 December 2005 (UTC)

question
do users have a right to request their user page/talk page to be semi-protected? --kizzle 23:41, 19 December 2005 (UTC)
 * Yep. As far as I know, anyone can request protection and I don't see why that would have to change for SP. On RfP, we often get requests from users to have their talk and user pages protected. Sometimes it's yes, sometimes no. --Woohookitty(cat scratches) 23:55, 19 December 2005 (UTC)
 * But that would violate the "pre-emptive" principle. Sorry, I meant more in the lines of if users could request semi-protection without having their pages actually vandalized.  Which do you think should trump in this case, a user's ownership of his user page or the "pre-emptive" clause in SP? --kizzle 23:59, 19 December 2005 (UTC)
 * No. We don't allow that now for protection. We do *not* do pre-emptive protection. I mean the thing is, even if we protect a page that's been linked from the main page, we do it only after it's been heavily vandalized. --Woohookitty(cat scratches) 00:01, 20 December 2005 (UTC)

Full protection
I'm the only one here who has contributed heavily to this discussion and who also patrols RfP alot. Generally, we only full protect pages if the edit warring or vandalism is heavy and long running. We do have some exceptions. If it's something that is getting nailed by dynamic IPs, we are more likely to protect after only 3-4 days of abuse, but then it's only protected for a short time. As for edit wars, we do protect for slow edit warring at times, but only if it's sustained. We reject probably 2/3rds of the protection requests, because protection is a last resort. And honestly, I see semi-protection the same way. It can't be pre-emptive and it has to be only for cases where everything else (including full protection) has been tried. And us admins are pretty uniform on that. There might be exceptions for pages "big in the news" if it's for a sustained period, such as the president of Iran. --Woohookitty(cat scratches) 00:07, 20 December 2005 (UTC)
 * RC patrollers will protect pages at no notice if there's a dynamic IP being a real pain. Typically for a few hours, while the ICT lesson at school ends, or something. I would suggests (quite strongly) that we need not 'try' full protection in the case of simple vandalism unless semi doesn't work. If RC patrol can't nuke an obscene vandal, there's no need to wait 4 days (!) before applying full protection, but now they could simply apply semi for the same few hours. Since semi is never a solution to an editorial conflict it shouldn't have to go through RfPP any more than current full-protection does. -Splash talk 00:16, 20 December 2005 (UTC)
 * That makes sense to me. It'll save a great deal of work from these pages. - Mys  e  kurity ( have you seen this? ) 12:34, 20 December 2005 (UTC)
 * Reducing workload is probably the real goal here? The less time wikipedians spend on crap tasks like verting vandalism, the more they have to spend on actually writing WP. Stevage 12:58, 20 December 2005 (UTC)
 * There is a more noble goal too: to stem a temporary case of serious vandalism that cannot be stemmed any other way whilst allowing good editors to continue editing an article. It isn't really a time-saver first and foremost, since that would lead to the conclusion that George W. Bush should be indefinitely semi-protected, which it shouldn't be. -Splash talk 21:26, 20 December 2005 (UTC)

One line summary
I've added a one-line summary here as on other guideline pages. Please feel free to tweak the wording on the display of the Template:Guideline one liner template. Stevage 21:42, 21 December 2005 (UTC)
 * I know I'm nitpicking, but.... I think perhaps it'd be better to say "Pages being frequently vandalized can...." instead of "Frequently vandalised pages can..." to emphasize that the current-ness (i.e. it is not pre-emptive, as temporary as full protection, and is being frequently vandalized right now and not over time)? --AySz88 ^ -  ^  22:05, 21 December 2005 (UTC)
 * That's a good point. I mildly disagree with that aspect of the policy, but that's definitely the consensus. Don't wait for me to change it ;) Maybe even "Pages suffering frequent vandalism..."?  Hmm...Stevage 22:09, 21 December 2005 (UTC)
 * I agree with that wording, but might we possibly have too many boxes on the page? I will admit it is nice to eliminate the ugly templates/change them, but is there a better place for them? I don't think that pre-emptive semi'ing will be helpful or good, but I think it should not need as much vandalism as v-protection does now to be implimented. - Mys  e  kurity ( have you seen this? ) 22:49, 21 December 2005 (UTC)

This box repeats the first sentence of the document. It is hard to imagine a more redundant tag on a page. If people aren't going to read the very first sentence, then they aren't going to understand it at all. I do not see the need to try to boil every policy and guideline down into 10 words. -Splash talk 23:39, 21 December 2005 (UTC)
 * Agreed. Blackcap (talk) (vandalfighters, take a look) 23:43, 21 December 2005 (UTC)
 * It definitely shouldn't repeat the first sentence, so by all means get rid of the first sentence or change it. The goal here is to encapsulate every policy into a single line that can be then copied to WP:RULES. I suspect that a box that says "This is basically what you have to know" is much more inviting than a first sentence which in most cases does *not* summarise the rest of the page (from the dozen or so examples I've worked on so far). I'll have another go. Stevage 23:47, 21 December 2005 (UTC)
 * No, get rid of the template. It does not achieve a useful purpose. It encourages a simple-minded approach to things that can and should be approached with a degree of subtlety. Make WP:RULES if you like, but don't templatecruftise every policy and guideline page in the process. It's more than a little silly to accuse the first sentence &mdash; the first sentence of a page of being redundant! -Splash talk 23:54, 21 December 2005 (UTC)
 * Let's discuss it at template talk:Guideline one liner. Fwiw, I'm not accusing a sentence of anything :) Anyway rather than get into an edit war on this page, I'll attempt to gather some more opinions on the worthiness of the project as a whole. Stevage 23:58, 21 December 2005 (UTC)
 * It seems to me that the problem isn't that we need a box to sum things up but that we need better first lines. Just take the information which you would put in the box and instead put it in the first line. Then, you don't have a massive ugly box, you have all of the same information, and it removes the simplistic summary of policies (i.e. from WP:BAN, the useless and PowerPoint-eque "Users may be banned for varying times. They may appeal, but must not circumvent the ban, so don't bait them or help them to try."). I heartily agree with Splash on this one. Blackcap (talk) (vandalfighters, take a look) 00:00, 22 December 2005 (UTC)

Semi-protection online
Well, looking at my Protection log, you'll see that semi-protection is now online. I'll dig up the diff where Brion said this was possible. Tito xd (?!? - help us) 07:16, 22 December 2005 (UTC)
 * Here's the diff: . Tito xd (?!? - help us) 07:17, 22 December 2005 (UTC)
 * Oh, this is so brilliant. Blackcap (talk) (vandalfighters, take a look) 07:27, 22 December 2005 (UTC)
 * Geez! Protect GWB and not put the SP tag on it. Jeez! :) --Woohookitty(cat scratches) 08:05, 22 December 2005 (UTC)
 * I put the tag on. Look at the history of the page. Amazing what making it S-P can do. Looks like Ral put the correct tag on it but a user who didn't know about S-P removed it. :) --Woohookitty(cat scratches) 08:26, 22 December 2005 (UTC)


 * Wow, that was fast. It seems the feature doesn't work quite as described in the diff, though, since it presently only restrict anons, and even a brand new account can edit through SP. Deliberate or bug? -Splash talk 10:46, 22 December 2005 (UTC)


 * I'm guessing bug. --Woohookitty(cat scratches) 10:50, 22 December 2005 (UTC)
 * I left a note on Brion's page so he can take a look. --Woohookitty(cat scratches) 10:58, 22 December 2005 (UTC)
 * If it was a bug, then it was a bit hasty to update this page to reflect the fact. The page should reflect the policy (ie, newest X% and anons can't edit), rather than the (buggy) implementation. Stevage 13:40, 22 December 2005 (UTC)

Wrong way round - policy reflects reality. Dan100 (Talk) 15:03, 22 December 2005 (UTC)
 * What or who are you talking about or to? The policy reflects what was agreed during much discussion. You didn't participate, but that doesn't reduce its validity at all. -Splash talk 15:08, 22 December 2005 (UTC)


 * Btw, the protection scheme was changed a tad since last night (it now says "Allow all" instead of "default"), so I assume it's still being worked on. By the end of the day, it might be 100% working and this will be moot. Btw, the policy is unregistered and 4 days or less. At the moment, that's not how it's working out, but it will be soon. --Woohookitty(cat scratches) 15:35, 22 December 2005 (UTC)

Requests
Where are we going to put requests? In a subpage? We ought to also have a list of currently semi-protected pages somewhere. We could do both of these on the main page, as does WP:PP; that seems like a good idea to me. Blackcap (talk) (vandalfighters, take a look) 07:30, 22 December 2005 (UTC)


 * The requests are going on the requests for page protection page, just like any other protection request. If we need a separate page for it, we can do that, though I don't see that happening. I will create a separate subsection for semi protection on PP just like we have for moves, redirects, etc. --Woohookitty(cat scratches) 08:00, 22 December 2005 (UTC)
 * Sounds fine. Sorry, I missed that on the main page somehow. I'm going to add a wee note saying to list semi-protected pages on PP, just like fully protected pages. Blackcap (talk) (vandalfighters, take a look) 08:06, 22 December 2005 (UTC)
 * Okey doke. Just for everyone's reference, here is the place to put semi protected pages. I split the real articles section into 2. Should work for now. If we need to do more later, fine. --Woohookitty(cat scratches) 08:25, 22 December 2005 (UTC)

2 pages semi protected thus far
GWB and Kerry. I'm treating them as tests. I just really want to see how much it decreases vandalism. It's *not* permanent. I think everything beyond this needs to be request only. But I just wanted to see how this would affect it. I'm thinking just 24/48 hours to see. --Woohookitty(cat scratches) 08:56, 22 December 2005 (UTC)
 * Please add the Hitler article too, it seems to be more vandalized than Kerry. Can&#39;t sleep, clown will eat me 13:47, 22 December 2005 (UTC)
 * OK, but requests should really go to requests for page protection. But I'll semi protect. And actually, someone beat me to it. :) It's semi protected already. --Woohookitty(cat scratches) 13:53, 22 December 2005 (UTC)


 * Might be too early to tell - but I am naturally impatient... Is it working??? novacatz 16:57, 22 December 2005 (UTC)

Consequential changes
Could some admin who know which Mediawiki pages it is, fix the description that turns up when a page is semi'd, since at present it still says "editable by admins only", or words to that effect. -Splash talk 11:00, 22 December 2005 (UTC)


 * I'd leave a note on Brion's page for that. Btw, I had the same question. --Woohookitty(cat scratches) 11:03, 22 December 2005 (UTC)
 * Perhaps the template has been modified to reflect the current reality of the semi-protection. Stevage 13:51, 22 December 2005 (UTC)
 * I changed the MediaWiki text earlier in the day. You can go to Special:Allmessages and search for the text of the message to find out which MediaWiki text it is (MediaWiki:Protectedtext, in this case.) I also updated This page is protected to reflect the changes. Thanks! Flcelloguy (A note? ) 23:18, 22 December 2005 (UTC)

The warning to admins language
I thought your language was a bit too punative sounding, Splash, so I changed it to a modification of the caution that's on the RfP page. I think it works better. If you don't think so, let me know. --Woohookitty(cat scratches) 13:46, 22 December 2005 (UTC)

Template inaccurate
The template should be updated to indicate that "unregistered and recently registered users" are blocked. As it is currently worded ('unregistered' only) a well meaning new user would try to edit and be blocked, click the link in the template itself to register, try to edit again and still be blocked... despite the template seeming to indicate that they won't. This will cause unneccessary confusion, questions, and frustration. I'd update it myself, but the template is protected. --CBD ☎ ✉ 13:57, 22 December 2005 (UTC)
 * BTW, this brings up another point. Right now we are considering which open pages should be semi-protected. We should also consider which PROTECTED pages can be downgraded to SEMI-protected. Since the semi-protection template isn't going to be on a vast number of pages it might make more sense for it to be semi-protected itself rather than protected. Ditto various meta-templates which have been identified as server hogging, but which can't be fixed by many users because they are protected. --CBD ☎ ✉ 14:00, 22 December 2005 (UTC)
 * The template was changed last night because as of right now, there is a bug that allows brand new accounts to edit S-P pages. It's being worked on. Brion has been alerted. As for your second point, don't worry. As we've been saying, admins are good at self policing. I go through the *entire* list on the protected pages list every few days to make sure that articles that are protected should be protected. Once this bug is worked out, I'll go through the list. Btw, I don't like the term "downgrading". It's not really downgrading. S-P and full protection are 2 different things really. --Woohookitty(cat scratches) 14:04, 22 December 2005 (UTC)
 * And btw, the policy doesn't really allow for templates to be semi protected unless they are being heavily vandalised. So unless the S-P template starts getting vandalized a few times a day, semi protection isn't appropriate for it or any other template. --Woohookitty(cat scratches) 14:07, 22 December 2005 (UTC)
 * Thanks for the update on the bug. As to the other, how can permanent full protection be more 'appropriate' than semi-protection? I understand that it wasn't contemplated for such a use, but once the dust settles I think we should revisit other potential uses of semi-protection. There are lots of pages which are permanently protected right now to prevent the possibility of casual vandalism which could fulfill that function just as well with permanent semi-protection while allowing established users to make improvements. --CBD ☎ ✉ 14:12, 22 December 2005 (UTC)
 * Patience my friend. :) Gotta start simple. I agree with you. I think there are other uses for semi-protection, but we want to start simple. We can always add to it either. We have alot of interested parties in this, so if changes are needed, it'll happen. Patience. :) But I understand ya. Like I said, I'll go through the list. The redirects and pages that aren't to be recreated can't really be used for this. As for the others, it'll depend on the original reasoning and all of that. By the way, "permanent semi protection" scares alot of people. Let's just put that down for a future discussion. Like you said, until the dust settles. --Woohookitty(cat scratches) 14:20, 22 December 2005 (UTC)
 * Protection templates should always be fully protected to avoid vandal mockery. We have everything to gain and nothing to lose by protecting a simple template. Its not like it needs constant updating. The only update it will get is when the new user bug is fixed. Voice of All T 18:08, 22 December 2005 (UTC)
 * If it's not being edited or vandalised, why protect it? -Splash talk 18:09, 22 December 2005 (UTC)