Wikipedia:Village pump (proposals)/Archive 84

Waiting edit requests to the global block notice
There are two waiting edit requests at MediaWiki talk:Globalblocking-blocked.Jasper Deng (talk) 05:18, 16 January 2012 (UTC)

Call talk pages "Talk"

 * ''The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion.

''Consensus is in favor of changing the discussion link to Talk. Those in favor cite many reasons why this change will stimulate new editor participation, while most who oppose only state personal preference. They 'ayes' have it. — E''dokter  ( talk ) — 16:30, 9 January 2012 (UTC)
 * The discussion consensus was enacted in this edit to MediaWiki:Talk. Cunard (talk) 08:55, 16 January 2012 (UTC)

RfC tag added on 09:01, 4 December 2011 (UTC).

Hi. A lot of Wikipedia instructions tell people to post "on the talk page". However, it isn't obvious where this means, since the page is not labelled "talk page" but is labelled "discussion". This is very confusing, especially for someone unfamiliar with all the Wikipedia rules. Wny not change the text on the tab from "Discussion" to "Talk page" to match all the instructions? (By instructions I am referring to everything you get when you click Help, as well as so many of the templates used to tell posters why their edit was not accepted and how to fix it. (184.147.120.119 (talk) 23:07, 2 December 2011 (UTC))

Just a note: the page which will need to be changed is MediaWiki:Talk. sonia ♫ 09:48, 4 December 2011 (UTC)

 At 17:14, 8 December 2011 (UTC) I changed the title of the post from 'Call talk pages "Talk Page"' to 'Call talk pages "Talk"', because there is a massive consensus among people supporting this change for using 'Talk' in specific and some opposition to 'Talk Page' in specific.  S ven M anguard   Wha?  17:14, 8 December 2011 (UTC)


 * Support this one per 2011 "State of the Wiki" address. Rich Farmbrough, 23:39, 2 December 2011 (UTC).


 * Support continuity I don't much mind which is prefered or even if another name like "chatter" or "gobledigook" (well maybe not) is chosen, but I do think we would be better off with one name for all "talk pages" across the whole site.  fredgandt  00:42, 3 December 2011 (UTC)
 * There is a 2007 discussion at Village pump (proposals)/Archive AN. PrimeHunter (talk) 02:20, 3 December 2011 (UTC)
 * How about just calling it Talk. That one is even more heavily implied than discussion, although to me, discussion is so obvious that I feel this change is unneeded.  S ven M anguard   Wha?  02:53, 3 December 2011 (UTC)
 * Uh, yes I didn't read properly. Talk is what I support. Rich Farmbrough, 13:04, 3 December 2011 (UTC).


 * I support discussion. This is primarily because of the popularly brought up rule "This is not a forum for general talk about the article's subject". Georgia guy (talk) 14:31, 3 December 2011 (UTC)
 * I support Talk because that's what most people call it anyway. @Georgia guy, talk and discussion are basically synonymous, so that rule could be changed to "This is not a forum for general discussion about the article's subject". There's not really any difference. Bazonka (talk) 17:28, 3 December 2011 (UTC)
 * How about suggested improvements pages?? Georgia guy (talk) 18:25, 3 December 2011 (UTC)
 * That is a little too specific. It would also only make sense for article talk pages. The proposal (I agree with in principal) is to name all talk pages the same way so that wherever users are guided to talk pages they are not confused by finding a "discussion" page but no "talk" page.  fredgandt  18:57, 3 December 2011 (UTC)
 * P.s. Also it would be a massive tab! Face-wink.svg  fredgandt 
 * Support. I've idly wondered about why this isn't labelled "Talk" on and off for years. Thryduulf (talk) 01:59, 4 December 2011 (UTC)
 * Strong support for Talk page. I have also  wondered about  this many  times myself. It  may  also  be confusing  for non native speakers. Kudpung กุดผึ้ง (talk) 02:28, 4 December 2011 (UTC) - or just  simply  'Talk'. Kudpung กุดผึ้ง (talk) 06:16, 4 December 2011 (UTC)
 * Support "Talk". The name "discussion" is simply never used for talk pages, so it is simply counterproductive to use it in the most prominent place of all. 124.168.87.221 (talk) 05:29, 4 December 2011 (UTC)
 * Support. Improving clarity is rarely a bad thing, and I can't remember ever seeing "discussion" used except in the tab. Alzarian16 (talk) 06:14, 4 December 2011 (UTC)
 * Support "Talk". As pointed out previously, it's the most commonly used term for the page, so it would be the most user-friendly.--JayJasper (talk) 06:20, 4 December 2011 (UTC)
 * Minor Oppose I don't want to be the lone dissenter here....and my reasons are mainly aesthetic. On places where it DOES say 'talk' it just kinda looks...off. It's probably just a matter of getting used to one over the other, but also the word 'discussion' seems to carry a better connotation, IMO, than 'talk'. ♫ Melodia Chaconne ♫ (talk) 07:05, 4 December 2011 (UTC)
 * Don't worry Melodia, you're not alone. I've already stated my feelings but not my personal preference, since my preference goes against what I see as the best course of action (if any). I prefer "Discussion" too, but as is being mentioned time and time again here, "Talk" is a colloquialism we can't escape. If this proposal has any value (and I believe it does), it will be to set a standard, continuous use of one word across the site. If we were to choose "Discussion", we would have to change so many policies and templates etc. we would get nothing else done for years. Quite simply, "Talk" is too engrained to oust now, so supporting continuity may mean not getting the word we prefer, but we do get less confusion.  fredgandt</i>  07:37, 4 December 2011 (UTC)
 * Support Many templates use "see the talk page" or similar for direction, and it's more common to see "talk page" in discourse than "discussion page". Do international language versions of Wikipedia use the same word? In any case, it feels more natural and intuitive to use "talk" and I fully support the motion to change it doktorb wordsdeeds 07:12, 4 December 2011 (UTC)
 * Strong support for "talk". It should be labelled as what we call it to avoid completely unnecessary confusion (I'm thinking especially of non-native English speakers, as well as newbies who didn't get enough sleep last night) - and it's always called "the talk page".  Pesky  ( talk  …stalk!) 07:15, 4 December 2011 (UTC)
 * Strong Support Talk- been waiting for this for a long time. New people only realizes it's the "talk page" after they click the tab and see the page is actually called the "talk page". Plus all our "informative"  pages call it "Talk page" like with Talk page guidelines - Help:Using talk pages - Talk page layout - Talk page templates. So lets not confuse our editors right off the bat when there reading introductory pages. Lets make it all match. Moxy (talk) 07:27, 4 December 2011 (UTC)
 * I have listed this proposal at Template:Centralized discussion. Cunard (talk) 09:01, 4 December 2011 (UTC)
 * Support. "Talk page" is a jargon term, which we should normally avoid in the UI. The name "discussion" is more accessible. But considering how likely it is that we'll change all the help pages, policy pages, and user comments to stop saying "talk pages" (read: will never happen), this is the more prudent alternative. Dcoetzee 09:36, 4 December 2011 (UTC)
 * Weak support "Discussion" makes a bit more sense especially to the newcomer than the (relatively vague) "talk", but "talk pages" is simpler and easier for most purposes. sonia ♫ 09:48, 4 December 2011 (UTC)
 * Weak oppose. If I remember correctly, the tabs for the talk pages were labelled "Talk" when I joined the project. I think it was changed along with the switch from Monobook to Vector, and I think this particular change was based on the findings of the usability project. Apparently, we used to have a lot of readers who had no idea that Wikipedia articles have associated discussion pages and had no desire to try out the cryptic "Talk" tab. Hans Adler 13:57, 4 December 2011 (UTC)
 * This was not as you describe. Some versions of Twinkle and Friendly used to include a script which changed the title of the "discussion" tab to "talk". This script was disabled a while back (around the time of the change to Vector). — This, that, and the other (talk) 04:10, 5 December 2011 (UTC)
 * Strong support Talk for the sake of consistency with namespace: Talk:Example suggests the link to Talk page. Calling things by their names is an important aspect of usability. &mdash; Dmitrij D. Czarkoff (talk) 16:05, 4 December 2011 (UTC)
 * Support. Paolo  Napolitano  16:36, 4 December 2011 (UTC)
 * Oppose On the contrary, we should change the help, docs, etc. to speak of "discussion pages" rather than "talk pages". "Discussion" seems to make more sense as a descriptor, particularly for outsiders. --<b style="color:#3773A5;">Cyber</b> cobra (talk) 21:37, 4 December 2011 (UTC)
 * Discussion name has already failed replacing the Talk name in general use. No need to revive dead. &mdash; Dmitrij D. Czarkoff (talk) 10:29, 5 December 2011 (UTC)
 * Tentative support, but also question for clarification: are we talking only about the link in the tab row at the top of each page, where it now says "Article | Discussion" or "Project page | Discussion", or is "Discussion" also found somewhere else? If it's only that one link, I'd go for changing it, for consistence with the actual namespace labels, but perhaps go for "Talk page" rather than plain "Talk". Fut.Perf. ☼ 21:47, 4 December 2011 (UTC)
 * Just that tab link I believe, and the original proposer did say "talk page", although it's gotten a bit garbled along the way. sonia ♫ 22:06, 4 December 2011 (UTC)
 * Yes, I am the proposer and that is what I meant. "Talk" would be just as clear. I am glad the idea has support. What happens next? Do I have to request anyone in particular anywhere in particular to make the change? Or do I wait more time and then request? Or something else? Thank you everyone for understanding. (184.147.120.119 (talk) 02:11, 5 December 2011 (UTC))
 * Here's what happens next. This change appears to have significant support, so this discussion has become a request for comment to get input from the broader community. It will run for 30 days gathering input. After the 30 days, somebody that hasn't participated in the dicussion will judge the consensus and close the discussion with a summary of what the community has decided. If there is consensus for the change, somebody will file a Bugzilla request to have the software changed to make the pages show talk instead of discussion. That may seem like a long time, but once started the process runs until it's completed. So you aren't required to do anything else, but you are certainly welcome to if you'd like to. Best regards. - Hydroxonium (T•C• [//en.wikipedia.org/w/index.php?title=Special:ListUsers&limit=1&amp;username=Hydroxonium V] ) 04:30, 5 December 2011 (UTC)
 * This does not require a bugzilla request or a software change. An admin can just modify Mediawiki:Talk to make the change. --Yair rand (talk) 04:35, 5 December 2011 (UTC)
 * Cool, thanks for the info. (184.147.120.119 (talk) 22:16, 5 December 2011 (UTC))
 * Oppose The argument is flawed because in many pages, we tell users to look at the "discussion page". "Discussion" is much much more easy to understand than "talk" for a new user&mdash;what's a "talk page"? Do we call conference rooms "talk rooms" or forums "talk boards"? Discussion is more newbie-friendly; if we need to reword help pages, then we should do that to make it more sensible in the long run. / ƒETCH COMMS  /  02:37, 5 December 2011 (UTC)
 * In which pages do we tell users to look at the "discussion page"? Thanks, &mdash;{&#124;Retro00064&#124;&#9742;talk&#124;&#x270D;contribs&#124;} 05:26, 5 December 2011 (UTC).
 * There are some that refer to the "discussion page", that is true. But there are far, far more that refer to the "talk page". Take a look around. — This, that, and the other (talk) 06:30, 5 December 2011 (UTC)
 * "Discussion page" results, etc. Just because more pages use "Talk" over "Discussion" is no reason to change to "Talk". "Discussion" is far more understandable to new users&mdash;and that's what counts, no matter how many changes we have to make on our end. We must make Wikipedia simpler to understand and navigate for newbies. / ƒETCH COMMS  /  15:23, 5 December 2011 (UTC)
 * Hey, come on! Both talk and discussion describe the same process. The difference is too small. &mdash; Dmitrij D. Czarkoff (talk) 10:29, 5 December 2011 (UTC)
 * Support. We should name it what we call it. Kaldari (talk) 03:10, 5 December 2011 (UTC)
 * Comment - I would absolutely love for this to happen. Absolutely.--Jorm (WMF) (talk) 05:02, 5 December 2011 (UTC)
 * Support: It would make sense to have consistency between what the tab's text is and what users refer to the page as. It seems to me that "talk page" is the common term, so if we are going to continue to use "talk page" in our comments, help and policy pages, templates, etc., then we should change the tab to say "talk" or "talk page" to achieve that consistency. If we are going to continue to have the tab say "discussion", then we should actually start calling the page to which it links the "discussion page" instead of the "talk page" in our comments, help and policy pages, templates, etc. Regards, &mdash;{&#124;Retro00064&#124;&#9742;talk&#124;&#x270D;contribs&#124;} 05:26, 5 December 2011 (UTC).
 * Support "Talk" for consistency and clarity reasons. --TorriTorri(talk/contribs) 06:37, 5 December 2011 (UTC)
 * Oppose The word talk is misleading because we don't talk there; we write. Readers might confuse this with an audio interface such as Siri. Warden (talk) 09:30, 5 December 2011 (UTC)
 * This is actually a good argument for renaming the whole talk namespace to discussion. Good luck getting consensus on that though. Yoenit (talk) 09:33, 5 December 2011 (UTC)
 * Actually I hadn't considered that: "talk" could give the misleading impression that it is possible to hear the article read out to you by clicking that tab. Make of that what you will, though. — This, that, and the other (talk) 10:33, 5 December 2011 (UTC)
 * Support But "Talk", not "Talk page". The latter is ghastly. If you prefer "Discussion", I'm sure we can write some CSS/JS to give you the old style. —Tom Morris (talk) 14:48, 5 December 2011 (UTC)
 * Neutral If you are concerned about 'Talk' being an audio interface, then you could change the tab to Discuss, though it's not as friendly as Talk. Whatever is the consensus is fine, but I do agree that consistency is best either way. &#x0298; alaney2k  &#x0298; ( talk ) 17:05, 5 December 2011 (UTC)
 * Strong Support for "Talk". Click on "Discussion", and what does it take you to? A page that has the title "Talk:Page" right there at the top. The proposal makes very good sense, per the principle of least surprise. I've also noticed when I talk to non-editors in real life that when I explain to them how we use talk pages, they ask me how they can find the talk page for a given article. I find that I have to tell them to click on "Discussion", and that's the way to get to "Talk". Admittedly, it's not that hard to figure out, but it really is more user-friendly to make this change. --Tryptofish (talk) 20:09, 5 December 2011 (UTC)
 * Support with comments - I agree that if someone wants to call their talk page something else its fine and that the default would be better as Talk page. However, it would be fairly easy to create a script to rename the page so an individual can call it whatever they want (i.e. talk, Talk page, discuss, etc). --Kumioko (talk) 21:49, 5 December 2011 (UTC)
 * Support new, and especially foreign language editors are often confused about our nomenclature. We need to be consistent, and it's typically easier to change a label than to change what many people call them. Buddy431 (talk) 00:05, 6 December 2011 (UTC)
 * I'm in favor for consistency, but then we should go with Discussion. It describes the function of the page much better than the vague 'Talk' . —Th e DJ (talk • contribs) 00:18, 6 December 2011 (UTC)
 * Oppose - Fetchcomms is right, 'talk page' is not user friendly, actually it's such a weird name that if I were not a regular Wikipedia I would probably be confused as to what it means. In fact just a few days ago I changed my signature and talk page header from 'talk' to 'contact' for this reason. If the problem is the lack of consistency referring to the namespace, then the change is needed at the help page level, not in the display tabs. For what it's worth, I would support moving the namespace to ; in the meantime I will oppose changing the displayed tab to "talk page", as 'discussion' is much more straightforward.  CharlieEchoTango  ( contact ) 08:14, 6 December 2011 (UTC)
 * Many people at first do not know what "wiki" means, and they also learn what "talk page" means. -Wikid77 13:32, 6 December 2011 (UTC)
 * So you're saying that new users should just learn random things about how Wikipedia works before contributing? Just because people learn what "talk page" means later does not mean we should expect people to take the time to figure that out before they get impatient and just give up. / ƒETCH COMMS  /  01:42, 7 December 2011 (UTC)


 * Strong support "Talk". The menu option "Talk" is shorter than "Discussion" and the term "talk-page" provides a distinctive, while still short, name for such wiki-pages. A Google search for "talk-page" reveals it is closely linked (80%) to wiki websites, whereas a search for "discussion page" reveals wide, rambling use in many other websites, with no specific meaning as to what "discussion" entails. For over 6 years, many Wikipedia users have understood that a "talk-page" is not an open forum to blog about ideas; however, "discussion" might be imagined as a different type of page, perhaps open to blog-forum posts or other general discussions. Name the menu tab "Talk" to be more precise and avoid confusion. -Wikid77 13:32, 6 December 2011 (UTC)
 * Support for the purpose of consistency, I support "talk". "Discussion" is overly long and the page is called a talk page, not a discussion page. We shouldn't rename the talk namespace to discussion though. Talk and Discussion have the same meaning, the only difference being "Discussion" is much longer lengthwise. Alpha_Quadrant    (talk)  17:48, 6 December 2011 (UTC)
 * Support. - Dank (push to talk) 18:50, 6 December 2011 (UTC)
 * Comment - Classic skin (which doesn't have tabs) links to talk pages in the sidebar with the link Discuss this page. Would there be any change to this? &mdash; An  optimist  on the run! 21:38, 6 December 2011 (UTC)
 * Question How would this affect Monobook? I like the idea for the default skin, since it's simpler for people not already familiar with Wikipedia, but I prefer "discussion" for my own purposes simply because I'm accustomed to it.  Nyttend (talk) 03:40, 7 December 2011 (UTC)
 * Support "Talk" - as a volunteer in, I've found that many new users find it disorienting when we direct them to the "talk page". Often, the reaction is "Wait, where is that?" and we have to explain how "discussion" == "talk".  ~  Matthewrbowker   <sup style="color:#0000aa;">Say hi!  05:42, 7 December 2011 (UTC)
 * Support Talk Per common name, common sense, it's shorter, it's more inviting, it's consistent with the namespace. A small step towards UI sanity. Ocaasit 07:29, 7 December 2011 (UTC)
 * Support this obvious and long-overdue change. I don't know what caused it, but I'm impressed by the sudden outbreak of common sense in this thread. Robofish (talk) 12:42, 7 December 2011 (UTC)
 * Support talk An obviously good idea--why was this ever changed to "discussion?" —Justin (koavf)❤T☮C☺M☯ 10:18, 8 December 2011 (UTC)
 * Support, although I also think its neither here nor there. Even if English is not your first language, if you can't find the "talkpage" by clicking on the word "discussion", you should probably think about whether editing Wikipedia is the best use of your time. --FormerIP (talk) 18:49, 8 December 2011 (UTC)
 * Oppose, though I would support a broad effort to update references to "talk pages" to "discussion pages". Not sure on the namespace; I don't think keeping it "Talk" is necessarily confusing.  Powers T 19:13, 8 December 2011 (UTC)
 * Oppose I don't mind informal mentions and references to talk pages, but changing the actual tab seems to imply (intentionally or otherwise) a morphing of the pages' purpose, to general debate about the article rather than for content development as it is now. That's an erosion of wp:NOTFORUM. 66.127.55.52 (talk) 17:42, 9 December 2011 (UTC)
 * Support Talk - It's a common name, and it's more consistent with the namespace and what everyone calls it. Very rarely have I ever seen someone say "discussion page".--Unionhawk Talk E-mail 19:34, 9 December 2011 (UTC)
 * Support–It's not only far more commonly used, it's also the name of the namespace. "Discussion" does sound a bit more formal, but the consistency with most usage and with the namespace system makes it preferable to me. oknazevad (talk) 14:22, 10 December 2011 (UTC)
 * Oppose I prefer keeping it "Discussion" especially for new readers and editors. In many dialects of English talk and discussion are not the same thing.  Discussion implies for many of us civil, rational analysis and discourse.  It is not surprising that in shorthand discussions talk is used more often than discussion; it is shorter.  That does not mean we should formally accept "Talk" as the default tab. There should be no problem in providing an optional display for those editors who wish to use a "Talk" tab.  But for new users "Discussion" is clearer and more appropriate. --Bejnar (talk) 22:18, 10 December 2011 (UTC)
 * On the other hand, when we say "talk page" all the time, when several of the templates we use (such as Expert-subject) say "talk page", and when all the namespace names use the word "talk", it probably makes it confusing to new editors when the tab doesn't also say "talk". and while the second and third could be fixed, the first can't - both because we would need to go through all the pages in the Discussion: namespace, the User discussion: namespace, etc. to fix them, and since users are so used to using this term that they will keep doing so, even to new users. עוד מישהו Od Mishehu 09:01, 13 December 2011 (UTC)
 * Support - Neither term is precisely correct, and arguments could be made either way. "Discussion" is not precisely correct, because a user talk is not for discussing the user or the user page. It's important to keep things simple, and consistent. Any potential benefit of "discussion" being more clear than "talk" is lost through the constant confusion (new) users have, wondering what we mean by a "talk page". Especially when their own 'user talk' is called what it is. The word "talk" is short, snappy, matches the internals, and isn't all that mystifying. Keep it simple. I think this is a classic case where a 'usability study' did not truly consider all ramifications of a change - if we were designing a new mediawiki from scratch, we might go for 'discussion' - but we're not; "talk" is ingrained into our culture; it is unrealistic to change "talk" to "discussion" in every place (indeed, for historical discussion, it's effectively impossible). Changing "discussion" (tab) to "talk" is simple.  Chzz  ► 00:30, 11 December 2011 (UTC)
 * Support - For consistency with the mainspace and with what we actually call it, and to be less confusing to newcomers. I think their nature won't change regardless of what is actually written on the tab, so no possible harm done. Zidanie5 (talk) 13:02, 11 December 2011 (UTC)
 * Support unless we are going to change the talk namespace to be a discussion name space. The current setup is unclear. -- Eraserhead1 &lt;talk&gt; 22:03, 11 December 2011 (UTC)
 * Support, since the page in question is always Talk: (or NamespaceName talk:), not Discussion:. עוד מישהו Od Mishehu 09:11, 12 December 2011 (UTC)
 * Support - greater consistency; less likely to cause confusion among newer users. Chris (talk) 00:22, 13 December 2011 (UTC)
 * Oppose. In simple, it makes sense. Here, no. theMONO 03:11, 13 December 2011 (UTC)
 * Here, we talk about "talk pages", not "discussion pages". And the namespaces of these pages all use the word "talk", never "discussion". עוד מישהו Od Mishehu 05:31, 13 December 2011 (UTC)
 * Oppose Instead of changing the tab labels, we should rename the namespace from Talk to Discussion. Why? Because I think it is more natural to advice someone to participate in an article discussion by editing the discussion page and talk gives a stronger impression that the page is for general talk about the article topic. Should I make a counter proposal to rename the namespace? Toshio Yamaguchi (talk) 16:21, 13 December 2011 (UTC)
 * I, personally, could only support that if you can also have all past discussions changed from "talk page" to "discussion page", and make sure that all Wikipedians, including those who happen to be on Wikibreak in paralell to the discussion, will use the word "Discussion page" in stead of "talk page" - and I don't think that you could do that. עוד מישהו Od Mishehu 16:25, 13 December 2011 (UTC)
 * Support. Strongly needed.  Rcsprinter  (gas)  16:42, 13 December 2011 (UTC)
 * Oppose per most of the other oppose reasons. That and I just like the way it is now. - Purplewowies (talk) 17:20, 13 December 2011 (UTC)
 * Ovbious Support. Well, yes. It seems obvious that some people who are unfamiliar with Wikipedia, might say, "Talk page? Where's the talk page?" and all they would see is "Discussion". All I'm saying is that it seems obvious. <font face="century gothic"> <span style="text-shadow: 0 0 0.2em #8F7, 0 0 0.2em #8F7, 0 0 0.2em #8F7; color:#000">yrtneg <span style="text-shadow: 0 0 0.2em #F87, 0 0 0.2em #F87, 0 0 0.2EM #f87;  color:#000">talk  <span style="text-shadow: 0 0 0.2em #87F, 0 0 0.2em #87F, 0 0 0.2em #87F; color:#ooo">contr   23:19, 13 December 2011 (UTC)
 * P.S. I also oppose this in some matter, such as the word "discussion" being more, eh, mature than "talk". My vote is now neutral. <font face="century gothic"> <span style="text-shadow: 0 0 0.2em #8F7, 0 0 0.2em #8F7, 0 0 0.2em #8F7; color:#000">yrtneg <span style="text-shadow: 0 0 0.2em #F87, 0 0 0.2em #F87, 0 0 0.2EM #f87;  color:#000">talk  <span style="text-shadow: 0 0 0.2em #87F, 0 0 0.2em #87F, 0 0 0.2em #87F; color:#ooo">contr   23:19, 13 December 2011 (UTC)


 * Comment. Did nobody read two week ago's signpost, where it becomes clear that such inconsistencies are exactly (one of) the kind of difficulties newcomers are experiencing, or why is it that this is not mentioned a single time? Unfortunately, topics like these always result in wikiarguing, so I prefer abstaining from further input on this. Nageh (talk) 23:32, 13 December 2011 (UTC)
 * support duh. It is the talk page. Protonk (talk) 05:10, 14 December 2011 (UTC)
 * To people seriously think that the tab being called discussion somehow elevates the level of discourse on the page? And that the solution is to change the namespace to "discussion"?  Even worse, are people serious in renaming both to "collaboration" or something likewise obfuscatory and multi-syllabic?!  That's pretty surprising, so I apologize for my incredulity.  It seems to be much easier to change the site CSS to say "talk" than change every single link and mention of the talk namespace in order to satisfy our sense of superiority.  The only cogent argument I can see against using "talk" is that it mischaracterized what actually goes on in talk pages--we don't "talk", we write.  But that seems needlessly pedantic.  I guess we can chalk this up to the community's willingness to accept the status quo because they are used to it, consequences for new users be damned. Protonk (talk) 23:44, 16 December 2011 (UTC)
 * Support That was one of my first questions when I discovered non-Mainspace Wikipedia. BTW, it's already done over at the Simple English Wiki. Zlqchn (talk) 12:44, 15 December 2011 (UTC)
 * Oppose; discussion was never a good alternativ to talk - same problems. I suggest Collaboration - and change the Namespaces accordingly (I think we can set up a namespace redirect from Collaboration->Talk and change relevant links). Then make an effort to stop saying "talk page" :) --Errant (chat!) 13:06, 15 December 2011 (UTC)
 * Oppose. I oppose for reasons of dignity and the value of description. "Discussion" is a nicely dignified word, appropriate for the prominent position given to it on each article in this encyclopedia.  "Talk", in the sense of discussion, as opposed to exposition or demonstration, seems quite the opposite. And "discussion" is a better descriptor, in part because "talk" can only clearly have a sense which is general for the page.  Click on the "discussion" link and you are brought to a page which hosts a discussion, a particular discussion.  You are not brought to a page which hosts a talk, unless one takes "talk" in an obscure sense for such a usage. Even in the case of their general senses, "discussion" seems to be a better descriptor. What happens, or is supposed to happen, on the page, "discussion" accurately describes that; and that is a discursive conversation in any medium focused on a particular matter.   "Talk" can have a sense which can refer to such a thing by subsuming it under a more general heading, but as such it does not pick out it so specifically.  Lastly, "talk" having many other senses, if it is taken in these, it is prevented from referring to such as "discussion" does.  This last point has been partly illustrated above by others.--Atethnekos (talk) 03:43, 16 December 2011 (UTC)
 * Support—aside from consistency, it's more inviting to newbies to see a tab that says "talk". Tony   (talk)  10:55, 18 December 2011 (UTC)
 * Support This would make things easier for new editors, and would have some consistency when linking to the talk page. And changing the name to something different than "disucssion" or "talk" and trying to get editors to change their habits is untenable. Let's call a spade and spade. Angryapathy (talk) 15:06, 19 December 2011 (UTC)
 * Support: I found this slightly confusing when I first started editing, and I have seen cases of new users not understanding what the "Talk" page was or how to find it because it was labelled "Discussion". The pages are actually named "Talk:" in the urls, links, and search box. We could easily change any mentions of "Discussion" pages in the Help section etc to "Talk"; I think that a majority already refer to "Talk". If some people think that "Discussion" is clearer or sounds "better" I don't see that as rationale to leave the current fractured terminology the way it is. And it would be a whole lot easier to change everything to "Talk" than change everything to 'Discussion". And "Talk" is already the term most users use in posts. -MsBatfish (talk) 01:02, 20 December 2011 (UTC)
 * Support the use of "Talk:". It's confusing for new editors when they try to search for the talk page and all they find is "Discussion". In addition, FWIW, it has already been changed to "Talk" for User talk pages. — mc10 ( t / c ) 17:25, 20 December 2011 (UTC)
 * Support - There is no word that can make what happens on those pages dignified, we might as well be consistent. ▫  Johnny Mr Nin ja  18:21, 20 December 2011 (UTC)
 * Support It's a total usability howler to refer to the "Talk page" in policies, templates and communication from other users and not have a button marked Talk that the new users can click on. If we were to hire an expensive usability consultant, this is one of the first things they would tear us a new orifice over. Let's save the money and just change the label. MartinPoulter (talk) 16:30, 22 December 2011 (UTC)
 * Weak Oppose - I don't disagree with any of the arguments made by either side, but find myself ultimately preferring the present "Discussion". I will only say that *if* the tab is changed to Talk, a notice will need to be posted at the top of each talkpage clarifying that the purpose of the page is for discussion of the article and not "talk" in the sense of random back-and-forth chatter. Milkunderwood (talk) 00:35, 23 December 2011 (UTC)
 * Strong support. One of the most absurd issues with UI is that it labels elements by its own jargon distinct from the wiki parlance. But this was a WMF decision, so I'm not holding my breath any change in this area will happen. ASCIIn2Bme (talk) 22:54, 23 December 2011 (UTC)
 * Support "Talk". Continuity is a good thing. –Fredddie™ 06:24, 26 December 2011 (UTC)
 * Oppose Talk is an informal version of Discussion. Talk also has multiple meanings, like lecture or speak, while discussion refers more to debate. and, we arent talking, we are writing. i know talk is now commonly used to refer to debate/discussion, but i prefer to keep the language precise. I do note that we have chosen "talk" as the term on Simple English. I wonder if other WP's use a word that is more informal or formal for their discussion pages. Oh, and i would prefer that we refer to Discussion rather than Talk in our guidelines (maybe do some global changes, to refer back to the original posters concern. I note most people prefer Talk: maybe im just old fashioned. Id tell you all to get off my lawn, but i noticed its not mine...:)Mercurywoodrose (talk) 05:56, 27 December 2011 (UTC)
 * Strong Support The tab should be the same as the name space. Talk is also more commonly used. WP:COMMONNAME applies to articles, surely it should extend to everything. Oddbodz (talk) 19:41, 28 December 2011 (UTC)
 * Support This seems an obvious proposal, as people refer to 'the talk page', and also the pages themselves are headed Talk. Eldumpo (talk) 10:43, 30 December 2011 (UTC)
 * Support because it reflects language usage. The word "Talk" also takes up less screen space, reducing bunching in narrow browser windows. -- J N  466  14:33, 30 December 2011 (UTC)
 * Support It is rather ridiculous there is no consistency between instructions and the tag. Everyone naturally uses 'talk', as demonstrated by the instructions they have written.Sandpiper (talk) 23:51, 30 December 2011 (UTC)
 * Support per WP:USE --Neo139 (talk) 20:43, 1 January 2012 (UTC)
 * Oppose - I find 'discussion' to be a much more professional and accurate name, but I certainly concede that in common speak we most often use 'talk' anyways. 'Discussion' simply seems friendlier to new users, and I would certainly support assimilating it into regular use. Alas, cheers!    A rbitrarily 0    ( talk ) 03:39, 5 January 2012 (UTC)
 * Support Talk is more intuitive, given the existing convention of using it in templates, namespaces, guidelines, and in conversation. As an added bonus, talk is both a noun and a verb—click talk to “talk” on the talk page. To those who oppose because talk seems less "dignified" or "professional": read Protonk's comment. Braincricket (talk) 05:00, 5 January 2012 (UTC)
 * Support: Provides consistency, and is what the majority of users call the page anyway. — Huntster (t @ c) 05:37, 5 January 2012 (UTC)
 * Support per nom, that it's shorter, and also because they do it on the Simple English Wikipedia and it works just fine without confusion or loss of professionalism. Selery (talk) 07:45, 5 January 2012 (UTC)
 * Support 'Talk'. Clearer, matches current usage, and removes the erroneous implication that talk pages are for discussing the subject of the article (rather than the article itself). <b style="font-family:Times New Roman; color:maroon;">Modest Genius</b> talk 13:44, 6 January 2012 (UTC)
 * Support Talk: As per Oddbodz WP:COMMONNAME seems relevant here. Mark Hurd (talk) 13:55, 9 January 2012 (UTC)
 * Talk - this is their more common name anyway and easier for people to understand if their English is not native. This should be an easy simplification, though we might need a bot to make it happen.  Ϣere Spiel  Chequers  14:50, 9 January 2012 (UTC)
 * The above discussion is preserved as an archive of the debate. Please do not modify it. No further edits should be made to this discussion. 

Coordinated SOPA reaction in early 2012 RfC
<div class="boilerplate metadata discussion-archived" style="background-color: #f5f3ef; margin: 2em 0 0 0; padding: 0 10px 0 10px; border: 1px solid #AAAAAA;">
 * The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
 * SOPA initiative/Action was closed by, , and . Cunard (talk) 01:17, 17 January 2012 (UTC)


 * Please see also SOPA initiative/Action, where the WMF requests clarity to help it set up systems to support community consensus.

It was announced on December 16, 2011 that a floor vote on SOPA is delayed, likely until early 2012. While the threat of this legislation still looms, the brief reprieve gives our community time to reach a meaningful consensus about whether to take action, and what action to take.

Background information
The Stop Online Piracy Act ("SOPA", H.R.3261) is a piece of proposed federal legislation in the United States. The bill would expand the ability of law enforcement and copyright holders to fight online trafficking in copyrighted intellectual property and counterfeit goods. It has seen widespread opposition from all corners of the Internet, and poses a unique threat to Wikipedia's continued operation.

Two proposals for a response from the Wikipedia community have been advanced so far (first proposal, second proposal), but neither was conducted with enough lead-time to reach a meaningful consensus.

Proposal
Here is a proposal that we believe strikes a reasonable compromise, chosen as a moderate sampling of the ideas posted on Wikipedia:SOPA initiative:


 * Triggering event: When SOPA has passed committee and is scheduled for a floor vote in either the House or Senate. The banner runs for the week before the vote, and switches to the blackout on the day before.
 * Scope: Response is geotargeted to United States IP addresses only
 * Duration: Maximum of 7 days for banner component, maximum of 24 hours for blackout component. Blackout is triggered on the business day before the vote.  If the vote is on a Monday, blackout runs for 24 hours starting Friday.
 * Action (banner): Banners encourage people to contact their Senators and Representatives (priority given to whichever is urgent, House or Senate).
 * To the maximum extent possible, readers are given instant information on how they can take action. Campaign is designed to mobilize the public maximally.
 * The focus is on generating high-value congressional contacts (phone calls and in-person contacts vs letters or emails)
 * A VOIP-based callback system (such as the one used recently by tumblr) is an option if we can find one that fits our needs and allows us to remain acceptably independent.
 * Banners operate like the fundraising banners (served via CentralNotice, can be closed per-user, etc).
 * Action (blackout): All requests are answered with a black page. The page is semi-protected Wikitext.  Once the page is displayed, a cookie is set which prevents its display again.  Exact wording to be decided, but it hits the following points:
 * SOPA puts Wikipedia, and the rest of the free Internet, at risk
 * You can help by contacting your representative and senators (with maximally easy help with ways to do that)
 * A "Learn more about SOPA" link which points to the relevant article on the English Wikipedia
 * A "Why am I seeing this" link which points to a page detailing the process for reaching this consensus
 * A link to click through to the originally requested page
 * "You will only see this page once"

This gets the message across clearly, explains how to help, is targeted to people who have the ability to effect change, shows that the protest was a community decision, and doesn't reduce the utility of Wikipedia for readers.

To set this proposal in the context of similar actions in the past, see this summary.

Need
While most of us understand the power that Wikipedia has in this situation and the size of our audience, some are still skeptical that a citizen response can change the course of this legislation. Rep. Zoe Lofgren, member of the House Judiciary Committee addressed this concern on Reddit:

"My best assessment is that most members of the House who do not serve on the Judiciary Committee have not yet focused on SOPA. People should realize that incredible power they have to impact the thinking of their own Representative on the subject. For example, a very intelligent colleague who is not on the Committee approached me today asking about the bill. Why? He had received an urgent and forthright telephone call from a small business person in his district who is tremendously opposed. He wanted to know more about our Open Act Alternative. This is the power that each of you have with your own Representative. --Rep. Zoe Lofgren"

Statement in opposition
(replacement oppose statement by User: Wehwalt)

I have replaced the original oppose statement, by the proponent, as it omitted a number of points made by opposers. Accordingly, while nominally having both sides presented, it actually failed to do so.

No one has shown that SOPA is a threat to Wikipedia. That is, events after its passage would cause the site to shut down or significantly impair its functioning. When pressed, Geoff Bingham, corporate counsel to WMF (that is, they are the client, to whom his duty runs, not us) said that we might be deemed an internet search engine This seems to contradict the plain meaning of the statute, which refers to internet search engine returning a list of sites elsewhere on the internet in return to a user query. We do not send people elsewhere on the internet. For the position Geoff expresses to prevail (he qualifies his position by many a “could” and “might”), the language in the statute would have to be considered meaningless--and by the rules of statutory construction, courts dislike concluding that Congress inserted language, and meant nothing by it. All that is perfectly proper; it is routine for those arguing that a statute is wrong to present what is called in the law the “parade of horribles”. Suffice it to say that for most statutes, the parade of horribles presented by opponents resembles your favorite post-nuclear scenario.

Let’s say I’m wrong and SOPA is applicable to Wikipedia. What does that mean? Well, it means if a rights holder sued a foreign infringing site (that is, a site that neither has an office nor a designated person who can be served with legal process in the US, who commits various infringing acts), and won the Federal Court lawsuit, we might be asked to take down links to that site. In most cases, these are torrent farms which give the latest episode of HBO’s hits or new movies. Few are reliable sites. In practice, what I suspect is that the site would be placed on the spam blacklist at meta and people would search to ensure we weren’t relying on the content. Unlikely, that.

If we sourced to such a site, we would find another site for the citation. I find it very unlikely, despite the parade of horribles presented by advocates, that we would lose any content. There is no reliable source from some obscure African nation which “just happens” to be a foreign infringing site.

Absent the existential threat to Wikipedia, which many have claimed but none survived the cold light of day, we should not do this thing. Keep in mind that having the 5th-most trafficked website in the world advocate for a cause is a tempting platform, and many have already flooded in to make the decision for us. Keep in mind that this is something we can do at most once with any conviction (we may already have shot our bolt with the coverage of Jimbo’s call). The second time, even for a real threat, it’s just “There goes Wikipedia again.” (and soon “There goes Wikipedia”)  To say that this will end the Internet as we know it, as some supporters have, is dramatic, and not supported by the facts. However, it may be a step towards making Wikipedia less-well-regarded. Editors should feel free to advocate in the manner they deem best as individuals. However, a strike is a bad idea.

Let’s keep in mind that SOPA is widely supported because there is a very real problem with anything copyrighted showing up on the internet, free, and being downloaded. Many of these sites are offshore, and rightsholders cannot effectively get in touch with them to get the content taken down. They are, at best, playing whack-a-mole. That is a problem that is going to be addressed legislatively in some way.

But let’s say we went ahead and did this crazy thing. What then? Will people searching for information go and drop everything and call their congressman. No. They will say a word the civility hawks around here wouldn’t like, and go on to find the information, either from a mirror site or elsewhere. And they will remember they don’t need Wikipedia to find information. In an earlier post, I recalled James Hogan’s science fiction novel, The Two Faces of Tomorrow. In it, a self-aware computer was tested by having its power shut off. Eventually, it managed to wire around the switch. So will our public, and as a political player we will be trusted less.

Rebuttal to opposition
Because in 2009 I wrote strategy:Proposal:Wikipedia is a Web search engine, I disagree with Wehwalt. Any serious educational or academic user does not cite Wikipedia directly, but only its sources; thus, to us, Wikipedia is a resource that delivers HTML links to references in response to a search term. WMF lawyers might dispute this, perhaps even win, because the law doesn't always make sense, but let's not kid ourselves - how often is a law interpreted more narrowly than what it looks like?

A court order does not simply mean that we "are asked to take down links to the site"; it means that there must be consequences of some sort for failing to do so. Wikipedia was unable to prevent postings of the AACS key when that controversy was current, despite some admins believing them to be illegal and taking action to block them. They turned up all over the place, in articles, discussions, as screenshots of T-shirt sites, in RGB color values for a "free speech flag" (see article) - it was simply impracticable. So we're not being set up to experience some "minor" act of distasteful censorship, but rather, for the legal sanctions of failure. And there's no way to stop that in an encyclopedia that anyone can edit.

Response to Geoff
(added post, response to Geoff) I’d first like to apologize to the community for a number of grammar and style errors in my above post. I would have changed them, but with a reply on table, that doesn’t seem appropriate.

I’m grateful to Geoff for responding, as his time here has value, unlike mine. He also is putting his name and reputation on the line, I’m just posting (or being a troll, in some eyes). As a lawyer who has concentrated in certain areas (not intellectual property!) I know that the hardest person to persuade is the lawyer from another field who drops in, makes elementary mistakes, and just doesn’t get “the big picture”. If that suits me, I apologize for it. However, I don’t think I would pound the drum like this without having some reason.

A poster below disparages Geoff’s reply on the ground he has ignored technical definitions of the word “indexed” with respect to search engines, which operate to exclude Wikipedia. If Geoff has more to say on this (I will understand if he does not, his time is valuable and taken up with matters which would benefit the project), I’d be grateful for a response. If “indexed” is clearly defined or understood to have a meaning, which excludes Wikipedia, let’s wrap this up and get back to wrapping presents for Jimbo.

At the present time, if I read Geoff’s response correctly, we are called upon to take action. And why? Out of fear of sloppy legislative drafting. I grant his point that it is possible to make an argument that we are a search engine, that we do provide external links in many articles, and so some well-paid and articulate attorney can make a case that we are a search engine. The lawyer might not actually believe it (the practice of law involves advocating for clients, although a lawyer’s personal views might differ) but the case could be made.

But come on, is the search engine argument a winning one? Really? Not just one that is plausible, one that can win, or at least have a solid chance thereof, in court? That Wikipedia is a search engine? Really?

I think the difficulty with Geoff’s post, leaving aside the question of “indexed” which may dispose of the whole thing, is in his argument that we may fit the part of the definition of search engine in that it may be claimed that our primary purpose is “gathering and reporting, in response to a user query, indexed information or Web sites available elsewhere on the Internet.”

This seems dubious. Many of our articles, especially at the higher grades, have few online sources, working from scholarly books and other print reliable sources. My latest project, Cross of Gold speech has nearly no online references, but is almost entirely sourced to well-regarded books on the Gilded Age. We give the reader information, not distinguishing between offline and online information. Yes, we do contain information available elsewhere (i.e., off Wiki) on the internet, but really, is a judge going to rule that it is our one and only “primary purpose”? That this is the raison d'être of Wikipedia, the reason it has become the 5th most trafficked website in the world? Because we are (arguably) an external search engine? I would respectfully suggest that this is unlikely to be a winning argument, not in a court of law. Really, do you think any judge is going to rule that Wikipedia’s single primary purpose is being a search engine for sites outside our domain? I’ve seen judges do many ill-advised things in my time, so I never say never. However, if we are to live in fear of what seems an unnatural interpretation of the statute--we might as well hide in the basement until they come for us.

I do not say Geoff is wrong in calling SOPA a threat. Indeed, I am certain that we are in agreement on much of this. I recall a saying that everyone’s in danger when the legislature’s in session. But there doesn’t seem to be much here. Yes, SOPA bears watching, unexpected things can come out of manager’s amendments and conference committees. But right now the threat is not existential. It’s at best--or, if you prefer, at worst--foggy.

@Geoff--regarding last statement, well said. Have a good flight and best wishes of the season.--Wehwalt (talk) 19:45, 23 December 2011 (UTC)

"Internet search engine"
I want to thank Wehwalt for his thoughtful essay. As I noted in my blog, I think the biggest flaw in SOPA for Wikimedia is the loose definition of "internet search engine." Because this has been discussed above and elsewhere, let me explain in a little more detail why, in my opinion, rights owners will argue that that statutory definition covers us.

Under the new SOPA version (before the markup), the full definition of "internet search engine" (Sec. 101(15)) read as follows:


 * "The term 'Internet search engine'--
 * (A) means a service made available via the Internet whose primary function is gathering and reporting, in response to a user query, indexed information or Web sites available elsewhere on the Internet; and
 * (B) does not include a service that retains a third party that is subject to service of process in the United States to gather, index, or report information available elsewhere on the Internet."

Section (A) arguably applies to Wikipedia because Wikipedia is:


 * (1) "a service made available via the Internet" [There should be no dispute here.]


 * (2) "whose primary function is gathering and reporting ... indexed information ... available elsewhere on the Internet" [Rights owners may argue that the primary function on Wikipedia is gathering and reporting information in articles (which constitute "indexed information") that are primarily sourced through information available elsewhere on the Internet as evidenced by our reference links at the bottom of our articles.]


 * (3) "in response to a user query" [Users employ our own search function, i.e., user query, to find Wikipedia articles (or "indexed information")]

(Section (B) does not appear applicable to Wikipedia.)

This above parsing of the legislative language demonstrates at least the sloppiness in the legislative drafting, and, given the ambiguities such drafting give rise to, we can expect adverse rights owners to seek to take advantage. If legislators wished to exclude Wikipedia-type sites, they could have done so using more specific language. They have not, and that worries me. Geoffbrigham (talk) 00:18, 22 December 2011 (UTC)
 * — AlexSm 18:18, 13 January 2012 (UTC)
 * Thanks for the js Alex, I'll try it out. —Bruce1eetalk 06:31, 14 January 2012 (UTC)
 * I added the above to my vector.js and cleared my browser cache, but it's not working for me. Does this work with the vector skin? You say it doesn't work with Standard – what is Standard? —Bruce1eetalk 06:18, 16 January 2012 (UTC)
 * The code above works only if use normal [rollback] links and then taken to the page with Rollback-success message. It will not work with some other "rollback" types (Twinkle? Popups?). In particular, it's not compatible with the gadget "after rolling back an edit, automatically open the contributions of the user rolled back" (mw:MediaWiki:Gadget-modrollback.js). — AlexSm 17:03, 16 January 2012 (UTC)
 * "Standard" is the "internal" name for obsolete "Classic" skin: in preferences the "preview" link for "Classic" is ; sorry for the confusion. — AlexSm 17:03, 16 January 2012 (UTC)
 * I am using the standard rollback button and I get the "Action complete" screen, but the article is not watchlisted. I have Twinkle enabled but I don't use the Twinkle rollback because it is too slow for me. And I don't have the "After rolling back an edit, automatically open the contributions of the user rolled back" gadget ticked. I don't know why this js is not working for me. —Bruce1eetalk 06:10, 17 January 2012 (UTC)
 * In case that script is executing too early you can try this instead:
 * In case that script is executing too early you can try this instead:


 * If this doesn't help then on the rollback page you could try to open JavaScript Console in Chrome (Ctrl-J) and then on "console" tab excute some statements to find what the issue is:
 * — AlexSm 18:37, 17 January 2012 (UTC)
 * Your revised script works – it's doing exactly what I want. Thank you. —Bruce1eetalk 05:23, 19 January 2012 (UTC)
 * — AlexSm 18:37, 17 January 2012 (UTC)
 * Your revised script works – it's doing exactly what I want. Thank you. —Bruce1eetalk 05:23, 19 January 2012 (UTC)
 * — AlexSm 18:37, 17 January 2012 (UTC)
 * Your revised script works – it's doing exactly what I want. Thank you. —Bruce1eetalk 05:23, 19 January 2012 (UTC)

Alternate accounts editing user script pages
Hi I think that alternate accounts of users should be able to edit their main user's user script pages.--Breawycker public (talk) main account (talk) 20:18, 19 January 2012 (UTC)
 * The system doesn't know that it's an alternative account, for all intents and purposes it's a completely different person & account, so there would have to be a system of linking accounts created that was literal and technical as opposed to just a userpage note, so as to prevent impersonation. Doesn't seem worth it for such a minor gain that will effect very few people-- Jac 16888 Talk 20:29, 19 January 2012 (UTC)


 * This convenience is obviously not worth added complexity and potential risks. — AlexSm 20:33, 19 January 2012 (UTC)

H.J. De Blij
Hi. I noticed the page H.J. De Blij has not been rated by the geography department. Can someone rate it? Considering he is one of the top Geographers in his field, it should recieve a high rating. Hoyle Casino Man (talk) 18:31, 20 January 2012 (UTC)
 * You're more likely to get a response from editors that can help you with this at WT:GEOGRAPHY, the talk page for the Geography WikiProject; they take care of the assessment of articles on explorers and geographers if I'm not mistaken. -  ʄɭoʏɗiaɲ  <sup style="color:#3AAA3A;">τ <sub style="color:#3AAA3A;">¢  19:30, 20 January 2012 (UTC)
 * Ok, thank you! Hoyle Casino Man (talk) 22:02, 20 January 2012 (UTC)

"Quantify" Wikipedia traffic data
Wikipedia should "quantify" its traffic data on Quantcast like so many other sites have done. Quantcast is considered the "de facto standard of web audience measurement" and is much more accurate than Alexa if sites "quantify", or opt for exact statistics and numbers by utilizing the company's direct measurement tool. It's very easy to do and would take the WMF less than a minute to implement. Benefits would be, of course, more reliable and accurate traffic data, as well as page view data not available under current estimated data methods. I realize this proposal may seem silly and/or pointless but I would very much like to see it done. Here's Wikipedia's page on the site:.  Kinaro (say hello) (what's been done) 00:34, 20 January 2012 (UTC)
 * Not likely to happen, since it requires external JS to be served to visitors. As I understand it, that both introduces a reliance on a third-party service (which IIRC those who decide such things try to avoid) and is considered a leak of private information contrary to Wikimedia's privacy policy. Anomie⚔ 03:16, 20 January 2012 (UTC)
 * I suppose I understand what you're saying, though I fail to see how this minor reliance on a third-party service would cause any issues. Also, what private information? Why shouldn't we be able to see an accurate count of Wikipedia's visitors?  Kinaro (say hello) (what's been done) 04:39, 20 January 2012 (UTC)
 * The issue is enabling a third party to track every individual's use of Wikipedia. This would break the bond of trust with the readers. If the quantcast server isn't run by Wikimedia, there's no guarantee that the information will only be used in line with Wikimedia's goals. The Quantcast page you link to says "Estimate only" (and their measurements can't include users without Javascript), yet the stats server gives raw hit numbers, so on the face of it your explanation is the wrong way round. MartinPoulter (talk) 13:38, 20 January 2012 (UTC)
 * "more reliable and accurate traffic data, as well as page view data not available under current estimated data methods" data is available, but it is not published Bulwersator (talk) 14:07, 20 January 2012 (UTC)
 * I think knowing how often one article is linked from or to another would be very useful. I can see the amount of data involved would be quite large but I believe it is manageable. I don't believe Javascript would be needed for any of this. Dmcq (talk) 14:22, 20 January 2012 (UTC)
 * Eeeee... What? This data is available, see http://en.wikipedia.org/wiki/Special:WhatLinksHere/Wikipedia:Village_pump_%28proposals%29 Bulwersator (talk) 14:36, 20 January 2012 (UTC)
 * I'm sure Dmcq actually means how often one article is clicked through to or from another. Mark Hurd (talk) 12:13, 21 January 2012 (UTC)

User preference to automaticlly use https
I think that there should be a user preference to automatically use the secure server (https) when logged in. I like to use it but I find it a pain to always type in https and reload the page. It would really help a lot of people if there was user preference like that. P.S. Is this the appropriate place or is technical the appropriate place? Ramaksoud2000 (talk to me) 02:08, 13 January 2012 (UTC)
 * Please do not use images or templates in your signature; see WP:SIG for details. Please also go back and remove the template from any pages you've already "signed" with it. Thanks.
 * As for your question, there is no preference, and it would be somewhat insecure as you would have to be logged in to the insecure http site for the preference to take effect. If you're using Firefox, look into HTTPS Everywhere. Anomie⚔ 03:56, 13 January 2012 (UTC)


 * I think it would be a good idea to have a preference or some capability such as this. While I primarily use Firefox and will look into this HTTPS Everywhere extension, there are times I have to use computers that only can be used with Internet Explorer.  On a similar note, with the https site up and running at the same URL as the http site, is there a reason to keep the http site at all? –Grondemar 04:08, 13 January 2012 (UTC)
 * If you are referring to the secure server it should be kept to prevent link rot. – Allen4names 05:18, 13 January 2012 (UTC)
 * Actually I was referring to http://en.wikipedia.org, and all subpages still available through http. Why not simply redirect them to https? –Grondemar 02:45, 16 January 2012 (UTC)


 * PHP provides a rapid and simple solution to check for and redirect to https for any page visited without its use (i.e. if visiting a page at http... that can be visited at https..., the redirect kicks in). Would be a trivial fix.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  02:56, 16 January 2012 (UTC)
 * You're not thinking of the readers. They typically don't log in or even have accounts.  They (mostly) don't benefit from HTTPS, but it does slow them down at least a little; if they lack broadband it could be a deal-breaker.  The unsecure server will be around for a while, I'm afraid.  -- N  Y  Kevin  @210, i.e. 04:02, 20 January 2012 (UTC)
 * No please you don't know what all pain is behind the ssl protocol. Feature which point login page as default to https, then in case users needed, it would have option to switch back to port 80 would make sense. But redirecting all pages to https is really to much load for our poor squid and apaches. What's wrong on http? Most of people who use wikipedia only read it, so it makes a lot of sense that http should be a default type of connection, most of users should use. Re secure.wikimedia I think that best thing we could do with that, is disabling it permanently, by redirecting it to https://sitename.wiki[pm]edia.org Petrb (talk) 13:11, 23 January 2012 (UTC)


 * I use HTTPS Everywhere. Except, it's not really everywhere. It's HTTPS in one browser that I use on one machine. It'd be nice if we could allow users to opt-in to HTTPS at the user account level. It's particularly important if you happen to be on public networks and have an admin account. And ought to be bloody mandatory for bureaucrats, CheckUsers and Oversighters. (My security paranoia might be due to the fact that I recently worked for a security consultant, and I also asked my clients to run their passwords through How Secure Is My Password? in an attempt to scare them shitless. Only it scared me shitless, and they carried on non-chalantly as if having a dictionary word that can be cracked in under six seconds as the only thing between your data and the world is no big deal.) —Tom Morris (talk) 17:05, 25 January 2012 (UTC)
 * Unless we make HTTPS mandatory for all logins, there's no way to require it for any particular user. Even if the login fails when they try it via HTTP, they've still exposed their password to sniffers.
 * At any rate, I'm not clear what you mean by "public network". If you mean "my personal laptop at Starbucks", then HTTPS Everywhere would cover you. If you mean "a public terminal at the library", anyone with advanced permissions (or anyone who cares about their main account's security) should never log in at all on such a machine, as keyloggers are a greater risk than network sniffing there. If you must log in on such a machine, create yourself a throwaway sock instead.
 * If we trust users not to try to log in to the HTTP site, it would be possible to set up something like HTTPS Everywhere for the English Wikipedia: have a gadget that just sets a generic long-lived non-secure cookie, and a script in MediaWiki:Common.js that redirects anyone with this cookie set to the HTTPS site. Anomie⚔ 22:09, 25 January 2012 (UTC)


 * I think the original poster is suggesting that this is an option for users with an account. This will not affect most people who visit Wikipedia, but for the few who do log-in, it will provide them with an easy way to ensure they use a secure connection, if they wish. --Iantresman (talk) 17:08, 25 January 2012 (UTC)

English Wikipedia fork
Hi all. Looks like a part of our community is going to force a full and global blackout of the entire English Wikipedia on Wednesday, ignoring all those that think that it is against Wikipedia main goals, that compromises neutrality, all those who voted for a soft blackout or against the full blackout. I think that we need to create a Wikipedia alternative to protect us against the seize of free knowledge under false community consensus. Best regards. emijrp (talk) 12:25, 17 January 2012 (UTC)


 * You're calling the poll which had the most supporters of any single question in our entire history, a "false consensus"? You're welcome to that opinion, obviously, but I doubt it's one that's widely held.  Of course, your Right to Fork is as inviolate as it ever has been.  Good luck with that. <b style="color:forestgreen">Happy</b>‑<b style="color:darkorange">melon</b> 12:59, 17 January 2012 (UTC)


 * You have to read Consensus (specially "This means that decision-making involves an effort to incorporate all editors' legitimate concerns, while respecting Wikipedia's norms.") and Polling is not a substitute for discussion. Several issues have been raised in the SOPA discussion against a full blackout. A soft blackout would be a consensus solution (to do something). emijrp (talk) 13:29, 17 January 2012 (UTC)


 * We long ago grew too large to expect a near-unanimous consensus on large issues like this, although our written policies still describe the site as it was in 2005 instead of describing the site as it is today. I think that the options in the vote did give all voters a chance to express their concerns. The vote for a full blackout was 763 to 104; the vote for a soft blackout was 94 to 100. I think those numbers speak for themselves in terms of which option is preferred by the majority of voters. &mdash; Carl (CBM · talk) 13:45, 17 January 2012 (UTC)


 * Polling may not be a substitute for discussion, but thankfully, it's a substitute for endless discussions, which is what the SOPA thing became. —Tom Morris (talk) 16:47, 25 January 2012 (UTC)


 * You can use the time constructively improving http://simple.wikipedia.org which needs help more than enwiki. 16:55, 17 January 2012 (UTC)
 * It looks like Distributed Proofreaders will still be up. I know where I'll be... Regards, RJH (talk) 21:03, 17 January 2012 (UTC)

I'm moving this to a subpage of my own. I think we need a stable and reliable mirror of English Wikipedia 24/7. emijrp (talk) 19:19, 20 January 2012 (UTC)


 * See Mirrors and forks, we've already got a thousand useless forks that make it a pain in the ass to research anything on the internet anymore. Don't take your marbles and go home just because the outcome wasn't what you wanted. -  ʄɭoʏɗiaɲ  <sup style="color:#3AAA3A;">τ <sub style="color:#3AAA3A;">¢  19:27, 20 January 2012 (UTC)


 * If there is a stable and reliable Wikipedia mirror, paste a link please. Most of them are spammy and closed every day. emijrp (talk) 19:37, 20 January 2012 (UTC)
 * There's a reason for that: Wikipedia is constantly changing, and that involves pulling a ton of data to the mirror site every day. It's expensive to host, and not technically easy. &mdash;  The Hand That Feeds You :Bite 18:49, 23 January 2012 (UTC)
 * What are you describing is a mirror, but not a fork. He/she is looking for a fork - something to do (mostly) a clear cut of the original and then starting that as a basis. <b style="font-family:Courier New; display:inline; border:#009 1px dashed; padding:1px 6px 2px 7px; white-space:nowrap; color:#000000; font-size:smaller;">mabdul</b> 23:53, 25 January 2012 (UTC)

SOPA IS IMPORTANT - have a continual pronouncement on pages re internet restriction (vs. one shot)
we must deal with this internet liberty issue in favor of freedom, which is what separates us from China, & other intolerable living situations - SOPA/PIPA will obviously be reformed, to deal with the piracy issue in a huge way - we should *STAY VIGILANT* until it is resolved in favor of The People, by adding the SOPA/free speech awareness banner (even a 'bar': thin & black) to the top of ALL pages keeping EVERYONE abreast of the internet free speech / piracy issue until it is resolved.

thanks.

lakitu (talk) 02:30, 22 January 2012 (UTC)


 * CHOLERA IS IMPORTANT... should we display a permanent banner informing the world about that issue too? How about another banner warning visitors about Global warming? I can imagine there are hundreds of seriously pressing issue we can protest and advertise. How far down the page should the top of the articles be before we consider them too far down?  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  02:42, 22 January 2012 (UTC)
 * ...and would we need a banner about that?  Nolelover   Talk · Contribs  02:52, 22 January 2012 (UTC)
 * Preferably two.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  02:54, 22 January 2012 (UTC)
 * Support running a banner about cholera awareness. I'm glad the original poster suggested it, it's about time. Franamax (talk) 06:22, 22 January 2012 (UTC)


 * While I find the proposal unworkable and unwise, it seems to have been made in good faith and perhaps deserves a more serious response than the above. No one can reasonably suggest that either cholera or global warming pose an immediate threat to the continued existence of Wikipedia, so it's really a bit silly to make such analogies. @Factotum/lakitu: the limits of English Wikipedia lie in language, not geographical boundaries, so it's a false premise to suggest that "we" are separated from China or any other country. It's a global project. There are reasons why an "awareness banner" might be unhelpful or even counterproductive, not the least of which is that editors here are deeply divided over the notion that SOPA is a threat and, even among those who agree that it is, over what to do about it. But thanks for making the suggestion. You obviously care about Wikipedia, and that's a good thing. Rivertorch (talk) 19:16, 25 January 2012 (UTC)

Suggestion for how to stimulate improved content
Don't think this has been suggested before - couldn't find evidence anyhow - apologies if it's been dealt with already.

A common experience when looking up a topic in Wikipedia is the feeling that one has learned something but not found all the information one would have wanted on a particular topic. Often, this is largely because the editors of that page have come at the topic from a particular angle (no criticism, that's just how humans work!). If only the editors of that topic knew what it was I (and other users, obviously) needed to know, they could soon update the content to reflect the broader viewpoint required for that topic.

Why not have, at the bottom of every article, an area where users can a) add questions they'd like to see answered in the text of that article, and b) see questions other users have already posed? Each time an edit is made that answers one of the questions in this list, that question could be removed from the list by admin. Admin could also remove questions that are not pertinent to or appropriate for that article.

Perhaps there are technical issues about adding user-modifiable text to the displayed webpage? Just thought I'd see what people think, anyway. — Preceding unsigned comment added by ‎Mesbailey (talk • contribs) 03:23, January 23, 2012


 * We already have talk pages and the feedback tool thing. Talk pages do exactly what you propose, just not at the bottom of the article.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  03:53, 23 January 2012 (UTC)


 * There's a new version of the Article Feedback tool rolling out slowly across the pedia (more useful than the ratings version), so something like this is coming to fruition. --Izno (talk) 04:45, 23 January 2012 (UTC)
 * A "Make a suggestion" link next to the feedback tool, which would send readers to a feedback box which would then add to the talk page, would work quite well. Pesky  ( talk  …stalk!) 09:43, 23 January 2012 (UTC)


 * I've seen the feedback users are giving using the new Article Feedback Tool, and it is indeed quite useful. Questions they'd like answered, suggestions for improvement. But these comments do need filtering and prioritization, since some of them are random insults and nonsense - which is why they're going to be presented through some alternate interface and not on the main article page or the talk page. Dcoetzee 01:41, 25 January 2012 (UTC)

My first response to this proposal was that such questions might be better fitted on the talk pages than at the end of the articles per se. However, after a little thought, I did appreciate that the problem here is that the talk pages often have a tag stating that the purposes of them is to discuss the article as it exists in Wikipedia, not the topic in general. Perhaps we could sub-divide the talk pages into two sections - one to discuss the articles as they exist in Wikipedia, the other to ask the type of questions which the initiator of this proposal appeared to have in mind. ACEOREVIVED (talk) 16:10, 25 January 2012 (UTC)

Contents summaries for all ref desks on one page
May I suggest we create a page that gathers just the question lists (ie. the "Contents" boxes) at the top of all of the ref desks on one page? That way people who like multiple ref desks, and people who mostly want to edit mainspace, but like to have a quick check of the desks, can browse them all in one go. So in other words, the page would consist of 7 boxes full of question titles, and possibly (if there's room) of the help desk's question headings as well. IBE (talk) 10:28, 23 January 2012 (UTC)


 * If I understand you correctly: You want the section headings for all the ref desks presented as a set of (linked) lists?
 * This could be done by user script quite simply. I'm not suggesting that it shouldn't be done another way, just presenting that option.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  23:44, 23 January 2012 (UTC)
 * Just wp:transclude all of the reference desks on a single page, just like Wp:XfD today does for deletion discussions. Yoenit (talk) 09:25, 24 January 2012 (UTC)
 * I almost suggested the same thing, but seriously, that's going to be a massive page.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  09:29, 24 January 2012 (UTC)
 * A bot maintains Dashboard/Help noticeboards. I guess it could maintain Dashboard/Reference desks without too much rewriting. -- John of Reading (talk) 09:56, 24 January 2012 (UTC)
 * Transcluding with the actual transcluded material hidden behind collapse top? Unfortunately (and this may be a more general bug) clicking on a TOC link for a section that is inside a collapsed block does not cause the block to expand automatically, but rather does nothing (see my experiment at User:DMacks/transclude-test). DMacks (talk) 16:39, 24 January 2012 (UTC)
 * Yup, I've noticed that. Annoying! Pesky  ( talk  …stalk!) 21:31, 24 January 2012 (UTC)
 * Filed as 33937 if you want to track it. DMacks (talk) 22:33, 24 January 2012 (UTC)

Reference desk/all ---— Gadget850 (Ed)  talk 22:38, 24 January 2012 (UTC)

Create a category for CORRELATIVE / SIMILAR CONCEPTS
To my limited experience and knowledge, I found there are quite a lot of correlated concepts among different stream of knowledge. Like “ nothing comes from nothing” is correlated to some ideas from “Modern physics”. Another example is - matter is merely a vacuum fluctuation (seems not updated in wikipedia yet, assume it is part of the content under Matter), it correlated with the Buddha’s concept – form is void, void is form (it means anything with a physical state is void). Add a “correlative concepts”/"similar concepts"(hyper link) next to the paragraph or the name of those concepts with the correlated idea.(now we have similar things like "see also") What’s more to do is to provide a category to collect all the "correlative concepts"/"similar concepts"(or the“see also”) and arrange them according to the alphabet headings of the titles of the correlative articles/paragraphs/sentences/key words.
 * Original research is not allowed on wikipedia. Yoenit (talk) 19:05, 22 January 2012 (UTC)


 * So, even the action is linking two existing articles is counted as original research?? ohh, i just think it could be an inspiring thing if everyone suggest what they see similar concepts cross-disciplinarily...--42.2.43.151 (talk) 03:16, 23 January 2012 (UTC)


 * I'm not sure I understand your examples. But currently, if two topics are related, there are many ways of linking them. For example, they might be discussed within the article prose, using links to provide pathways for more information. If there's substantial material to be covered, new sections can be created, as in at Water, or Entropy. Your Nothing comes from nothing example also shows this. If you're thinking of something that has the effect of marginal side-notes in a textbook cross-referencing other topics, consider footnotes, which aren't just for references. See Logarithm for an example usage. Leonxlin (talk) 04:17, 23 January 2012 (UTC)

yea, what if there is a page which list all these interdisciplinary concepts. then we can easy and all-rounded to know interdisciplinary and cross disciplinary knowledge. seems concepts / knowledge have no borders to meet their partners (similar concepts), but the current classification of knowledge cannot show these concepts found their relatives in a whole picture. It's important to show it in a whole picture i think, as knowledge can actually grow horizontally, not only vertically (vertically is meant going more detailed and more detailed). but the two direction is important too i know. --42.2.43.151 (talk) 16:40, 23 January 2012 (UTC)

Wikiversity exists. And I recently created uw-wikiversity to encourage original research-focussed editors to import their articles to Wikiversity and edit.

What if the there's an article talking about matter is merely a vacuum fluctuation, and put a "see also" hyper link to another article about Buddhism - Form is void, void if form - anything has a physical state is considered as void. Assume this connection had no one created before, no reference can be found. That means it's the first hand discovery by users/editors. Can it put in Wikiversity,? it's just new discovery of 2 concepts having similar qualities, and it is partly quite subjective, that's why diversification of linkages is very welcome under this proposal.--42.2.43.151 (talk) 04:50, 27 January 2012 (UTC)

WikiProject History: time for an end?
I am not sure that WikiProject History is on the road to revitalization at anytime soon. I would like to suggest that the project either be redirected to something of a mini-WikiProject Council or just outright deleted, as it is no longer active or incredibly useful. My first choice would be to tag it as "Historical," but would that just make it an unreachable part of the past that's still obvious to viewers of the website? A mini-WikiProject Council of History-inclined Wikipedians could help increase collaborative efforts between History-related WikiProjects, but would deletion be the most efficient alternative now?  DCI  talk  22:25, 22 January 2012 (UTC)
 * ehh what? Why on earth would you want to delete it? Tag it as inactive or semi-active if want (done in WikiProject status), but deletion is out of the question. Yoenit (talk) 23:54, 22 January 2012 (UTC)
 * Deletion is sometimes an option given the general amount of activity a WikiProject has had previously.
 * However, DCI, given the large amount of content discussed on the talk pages, I would be inclined to mark it inactive, or possibly repurpose it to act as a host to history minded Wikipedians to discuss various issues, as you suggest. --Izno (talk) 04:42, 23 January 2012 (UTC)
 * Okay. Thanks for your feedback; I will certainly not mark it for deletion!   DCI  talk  00:55, 25 January 2012 (UTC)
 * I have to say I'm fairly shocked that WPHist would have so little activity that this course of action could suggest itself. Where are all the historians?  Maybe any of us who know a historian should get on their case about it? --Trovatore (talk) 01:09, 25 January 2012 (UTC)
 * I would assume historians are like any other profession: Some will focus in World War II, some on socio-economic differences during the 60s, and some on historic China. These people will naturally seek out the WikiProjects which are most pertinent to them. I think it is safe to assume that nearly all historians are like this, as most people are... --Izno (talk) 13:47, 26 January 2012 (UTC)
 * Where is there any evidence that the WikiProject is inactive? Please see Revision history of Wikipedia talk:WikiProject History and WikiProject History/Outreach/Members.  Maybe things are organized so well that only a few points need to be discussed.
 * —Wavelength (talk) 01:24, 25 January 2012 (UTC)
 * Most of these are notifications on the talk page, many of which have not gotten replies. Others are proposals/responses to proposals that I've written up in relation to the inactivity issue.  A-class reviews had been sitting around since 2009 (in comparison to very active ones like WikiProject Military history), but I cleared them out a few weeks ago, after another editor suggested that they are too old.  A newer nomination had no reviewers.  Projects with slightly smaller scopes, such as Military history, have been far more active.  This is why I am suggesting that a new course of action be taken in regard to this inactive WikiProject.   DCI  talk  03:09, 25 January 2012 (UTC)
 * There are two courses of action we follow with inactive projects: we delete the ones that never got going, and we tag formerly-active ones with historical. Ever since I first encountered it, this project has been quiet, far quieter than ones such as Military History or the project where I'm active.  Seems to me that there's a good middle ground amount of specificity for project scope.  Make it too precise, like the Weird Al wikiproject, and it's not going to attract enough people with knowledge about it.  Make it too broad, like this project, and people won't really have anything in common.  Make it somewhere in the middle, like Military History, and you're able to draw people together without excluding them.  Nyttend (talk) 01:30, 27 January 2012 (UTC)
 * WikiProject Mathematics is pretty bloody broad in scope (mathematicians from different fields can barely understand one another, a fact that may not be intuitive to outsiders) but somehow manages to chug along. No one to my knowledge has suggested breaking it up into, say, algebra, analysis, geometry/topology, and logic projects (there is a WikiProject Logic but it's not limited to mathematical logic).  So I don't really see why history should be any different.  But of course there's no point in me moaning about it; historians either find the project useful or they don't, and if they don't there's not much I can do about it. --Trovatore (talk) 01:46, 27 January 2012 (UTC)

wikimail
Hi

I think that there is a genuine demand for an email provider that promises NOT to use your private emails for marketing purposes. Can you provide it? — Preceding unsigned comment added by 94.169.34.5 (talk • contribs)

While that's a nice idea, it doesn't really fall within the purpose of Wikipedia. Our purpose here is to build and encyclopedia, and so everything we do should contribute to that. Resources are stretched as they are - having an email provider would be an unnecessary strain. There is a feature available for emailing users, here. ItsZippy (talk • contributions) 21:56, 26 January 2012 (UTC)

sort articles by " NUMBER OF WORDS"
I would like to give a suggestion to your website. Could we sort the articles by " NUMBER OF WORDS"? For example, by 100 words, 300 words, 1000 words, 1500 words, etc. Most of the time for some user, they just want to know the general information or the subject of an article only. There isn't necessary to read the whole article to get the little information. Sorting by number of words is classified articles into different categories, for lesser words -  e.g. 100 words of an article which is talking about Taoism, so readers may know what they need are just some main / key ideas ( without redundant history backgrounds). for more words, it could include more evident or findings for the subject. For even more words, it could include origins, history etc. So, all in all, just sort by different ways of summarization of knowledge.

Thank you!! — Preceding unsigned comment added by 42.2.43.151 (talk) 02:26, 20 January 2012 (UTC)


 * You don't have to read past the lead at the top. Normally people do try and put the more relevant stuff first. Are you thinking perhaps of a facility for phones where you want to restrict the amount downloaded? Anyway going further this idea could be expanded further - have a joystick pushing forward gets you deeper down with lots more detail, back and you zoom up for an overview, perhaps turn left for simpler language and less assumptions whereas turning right assumes the reader knows more of the background and can use more jargon. For editors pressing the fire button will get rid of vandalisms by identifying the edit that last changed the bit pointed at. Zap zap zap, yeah that would be satisfying for dealing with them. Dmcq (talk) 05:11, 20 January 2012 (UTC)
 * A bit sarky, Dmcq. Essentially is what we do by our choice of articles. For example, English law merely mentions murder; homicide in English law would provide (once complete) a few hundred words; Murder in English law a whole article, but whose lead might be about the summary in the previously named article. Grandiose (me, talk, contribs) 13:54, 20 January 2012 (UTC)
 * Saying what you want is the best way of getting it. Perhaps in the future computers will be smart enough to do what I said. Already we have chat rooms where messages are automatically translated into the language of the person reading the discussion so people don't even have to speak the same language. Your example doesn't satisfy the business about less words, I think what they really are asking for is a cut off which will still display something without eating into their account, that's something I believe some mobile phone service provide already by processing the pages before sending them down but I'm not into that area. Dmcq (talk) 14:29, 20 January 2012 (UTC)--42.2.43.151 (talk) 01:18, 22 January 2012 (UTC)

The best way of getting it, and also the best information we want as well. for example, when we search Albert Einstein, in the first beginning paragraphs, i don't even can see his major achievement like Theory of relativity, E = mc square. But i know that is difficult to force other's to think what i think which is more important. i just try to give an suggestion to ask users to make their beginning passage as essence as possible. essence is slightly different to general information. when articles are limited by words, i thought it can let users to think what is important to put their information. And like Dmcq said, mobile phone can apply " lesser word scheme " quite well, but the reason is not quite related the download limitation,i think, it's the phone screen's limitation, it's not that user friendly to read and move around such a big picture in a such a small screen (compared to desktop computer)all in all, sorting by words categories seems a bit far from my original purpose - essence your information. i don't know, i need to think more of that.--42.2.43.151 (talk) 01:18, 22 January 2012 (UTC)


 * The theory of relativity is mentioned in the very first sentence. Have another look. Also you can normally adjust the format of pages for mobiles so the text just goes down the screen instead of needing to pan over a large screen, there will be some option of the browser to do that and Wikipedia behaves quite well in Opera mobile for instance. There's still room for improvement but it's mainly problems in the content because of the editor generated content rather than the site itself, probably there should be a bit of a drive to deal with such things, also some extra work could be done to make tables behave better. Dmcq (talk) 13:07, 21 January 2012 (UTC)


 * Grandiose's example is a complete one to show the result of what this proposal would be archived. it shows different versions of articles would come out. That's another direction from mine, but that's ok. All this inspired me to know that readers can choose what they want more effectively in some cases. --42.2.43.151 (talk) 01:18, 22 January 2012 (UTC)


 * but this indicated another situation is that not only numbers of words should be sort, it should also include key words searching option next to / under numbers of words searching option.--42.2.43.151 (talk) 04:35, 22 January 2012 (UTC)

Dmcq's suggestions about " lesser words policy" on mobile phone is almost like that, i'm not that capable with technology thing. As mobile phone is getting online everywhere, what we absorb is not huge amount of info, like Einstein said, "Everything should be made as simple as possible, but not simpler."

An interesting proposal, but I fear it may be reinventing the wheel. We already have a system of sorting articles into stubs, start class and articles beyond start class  - would this not give the type of information for which the proposer is seeking? ACEOREVIVED (talk) 09:13, 26 January 2012 (UTC)

But that's only available in articles with stub added. right? But under my proposal, longer articles is also can be sorted as well. Depends on what readers want.--42.2.43.151 (talk) 05:54, 27 January 2012 (UTC)

It is not just available on articles that are stubs - start class articles (the next category up from stubs) are also inidicated this way. Experienced Wikipedians may remind me whether we sort articles into categories beyond that (I do seem to recall we also indicate good articles and featured articles in this way). ACEOREVIVED (talk) 09:11, 27 January 2012 (UTC)

Automatic warning when creating section heading exactly matching an existing section heading
When creating a section heading, I feel it would be beneficial if the parser would direct us to a "are you sure?" page that we must agree to before continuing the save. Since the table of contents and section links are basically useless when on whatever page there are more than one section with the same heading, this might help to stop some of the doppelgängers ever happening and thus make navigation simpler.

Further to this, it might be even nicerer to have automatic anchoring, for any precisely similar section headings that are created even after the warning/alert (alert is perhaps a better word).

I realise that anchors exist to solve parts of this issue, but they must be added by hand and known of by any editor who wishes to take advantage of them. To know where all the anchors are would be a logistical nightmare (what links here?, have fun with that). For ease and common simplicity, surely the parser should take care of all this stuff for us.

So I propose that:
 * On saving a page we are alerted to the fact we are creating a section heading exactly matching an existing section heading.
 * We must either change the section name or agree to go ahead anyway, before the page can be saved.
 * Preferably, if any page has any exactly matching section headings, the parser (whenever it spots them (every page edit)) automatically adds an anchor to the sections in question that should be labelled simply as perhaps Example heading 1, Example heading 2 etc., so that we (editors) know where to find them, that they are certainly there, and what they'll be called.

I expect there are many technical issues and editorial considerations here, but Rome wasn't built in a day (gotta start somewhere).  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  22:10, 22 January 2012 (UTC)


 * You've fallen into this trap before I think: when you start a new section (using the "+" link) you get a URL with "&section=new". This lets the parser handle just the new section text when you preview. Asking the parser to look at other parts of the page requires parsing the entire page, regardless of how large and funky it is. This fundamentally breaks the whole concept of section-by-section editing, so I don't see how can it be built in as an automatic feature of the software. Franamax (talk) 08:53, 23 January 2012 (UTC)
 * Sorry, I missed your "on saving" bit. I could see duplicate-header detection being an optional gadget, same as the thing that prompts you for an edit summary, so it probably is doable on a technical basis. Why not code up such a gadget yourself? It's easy enough to set up your own test wiki with the exact same software we run here. Franamax (talk) 08:59, 23 January 2012 (UTC)


 * That would be a good idea; here's a corollary for it, too: Don't give Wikilove messages all the same blinkin' name! Make the sender choose a section header. By the time you've been given a few kittens / cups of tea / wossnames, it's a real PITA! You type a nice reply to the most recent kitten-sender, hit the save button, and your page displays the first kittie-message on reloading, not the one you've just replied to! @Franamax: I'm sure there must be a way of achieving this without the parser having to parse the entire page, it just may not be immediately apparent. Maybe parse a "sub-page" kinda hidden thingie which has only section headers in it?  Pesky  ( talk  …stalk!) 08:57, 23 January 2012 (UTC)


 * I think a very short and simple php script could check the page for matching section headings while parsing any save made to that page. It would be the sort of script that could be chucked into the mix without (I think) any major confusions. A more complex method (from a development aspect) could be to have a database store the section headings present in all pages, then if a new section heading is created, it is compared with the DB entry for that page. Potentially faster (not that either would be slow) and cleaner (no http requests for a copy of the page in order to check it), but would require almost an entire new extension to be built. One practical upshot of the DB method would be having the DB of section headings for every page (constantly updated (every page edit)) to be used in whatever other way could be imagined. Applied to searching, that extra knowledge could make a huge difference. Just thinking out loud now.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  23:36, 23 January 2012 (UTC)
 * Definitely agree with this idea. A script would only need to check text between = characters to see whether the exact same string of text could already be found between equals signs on that specific page.  Nyttend (talk) 01:22, 27 January 2012 (UTC)
 * +1 good idea! Would also be useful for article talk-pages that are susceptible to repeated controversies (rename or merge requests, the N+1'th edit-request today, etc.). DMacks (talk) 01:32, 27 January 2012 (UTC)


 * I agree. This seems to happen a lot of user talk pages too.   Will Beback    talk    01:37, 27 January 2012 (UTC)


 * Support; I think this belongs at Edit filter/Requested. Jasper Deng (talk) 06:18, 28 January 2012 (UTC)

page full of images
Would like there to be a page that has lots of images of the same thing on it, eg: for an owl covered in images of owls, or different for different objects Trellis Reserve (talk) 14:03, 27 January 2012 (UTC)
 * Is this a proposal for gallery-pages? That exists on Commons. Choyoołʼįįhí:Seb az86556 > haneʼ 14:10, 27 January 2012 (UTC)

Proposal: Articles about extant corporations
There's been some discussion recently over paid editing, the creation of a Wikiproject addressing this and the proposal of another, and so forth. Jimbo recently talked to Bell Pottinger (described here). It's a complicated and contentious issue, and if I'm understanding the debate correctly, PR firms are offering the following types of cases where their intervention is needed or useful: It'd be silly to take this entirely at face value (because for one thing "neutral and fair" depends on your point of view, and it's only human for one's point of view may be influenced by who is cutting one's paycheck). BUT, these are valid concerns and, when they do occur, serious problems (the first two anyway). Because they are valid concerns and serious problems, these are good reasons (or excuses if you prefer) for PR firms and paid agents to claim a moral right to edit the Wikipedia and a practical need to do so.
 * There's derogatory and false (or at any rate unsourced) information about their client in their article. Their client doesn't know how to engage Wikipedia effectively (e.g. OTRS, edit within our rules, etc.) and so they need professionals to fix this.
 * There may not exactly be false information, but the article is slanted and looks rather like a hatchet job. Our client simply wants a neutral and fair article (which is what Wikipedians should want also). Again, a professional is best suited to fixing this.
 * And some clients would like to have a Wikipedia article, and we believe that they are sufficiently notable, but there's no article; and they don't want to wait years (or forever) for some random person to create the article. And since they are sufficiently notable (we believe) then an article would enhance the Wikipedia, which should meet the desires both of our client and Wikipedians generally.

For my part, I'm against paid agents being allowed to edit the Wikipedia. (There is the question of whether as practical matter it's better, tactically, to allow this as opposed to driving it all underground; that's a different issue and outside the scope of this thread.) So, is there another way, rather than allowing or welcoming paid agents, to address these concerns?

Yes, possibly, and I have some concrete suggestions. This is not going to happen right away but it's something worth talking about, maybe. What I'm proposing is: Details below.
 * As the main proposal, creation of an "Articles about Extant Corporations" policy similar to Biographies of living persons (BLP).
 * As a secondary proposal, perhaps looser notability requirements for WP:CORP.
 * As a secondary proposal, the deployment of a template which is essentially the converse of advert.

Articles about Extant Corporations WP:AEC
Articles about Extant Corporations. (This would include non-profit organizations and almost all businesses, even single stores and restaurants, since those are almost always incorporated. But some or many single-person businesses aren't incorporated. It could be "Articles about Extant Organizations" instead, which would be similar but not embracing exactly the same sets.)

Various details to be worked out but the basic thrust would be similar to WP:BLP. Corporations aren't exactly like people so there'd have to be some changes from WP:BLP, but it could be expressed with a similar summary:

With a corresponding tag for article talk pages:

This implies the creation, manning, and efficient operation of a "biographies of extant corporations noticeboard", which seems doable. The Foundation would possibly (maybe) take a hand in promoting and perhaps even monitoring this effort if it gains any traction.

Reform of WP:CORP
While this remains on the table, it's secondary and peripheral, and is a distraction from the main point, so I'm making it less visible. Discussion remains open though. Herostratus (talk) 15:42, 21 January 2012 (UTC)

WP:CORP could be made less stringent. Perhaps something along the lines of requiring just one reliable independent ref, and the requirement only proving that the entity exists, and maybe that other material from the article could come from non-independent sources -- the company's web site, for instance. Or something like that.

This would be helpful to corporations, especially corporations whose Google profile is not so good, since the Wikipedia article would likely rise to the top or near and per WP:AEC it would probably be reasonably positive, usually.

Since proof of existence is a simple bright-line test, this would also obviate a lot of contentious discussions about whether a particular entity is or is not notable, which discussions probably sometimes draw in in covert or overt paid agents, which is what we're trying to avoid.

Granted "being helpful to corporations" isn't really part of our core mission, but remember the point here is to get the PR industry off our case and out of our Wikipedia, and this helps this by removing both a philosophical argument for their involvement and a practical reason for same, to some extent.

Hatchetjob
While this remains on the table, it's secondary and peripheral, and is a distraction from the main point, so I'm making it less visible. Discussion remains open though. Herostratus (talk) 15:42, 21 January 2012 (UTC)

We have advert, which says

But we don't have the converse, something like this:

With the matching category Category:Articles with a derogatory tone (or something) as the converse of the existing Category:Articles with a promotional tone.

A step beyond this but arguably necessary would be the deployment of corresponding warning templates on the order of
 * Welcome to Wikipedia. Everyone is welcome to contribute constructively to the encyclopedia. However, please do not add derogatory material to articles or other Wikipedia pages. Scandal-mongering and using Wikipedia as investigative journalism are against Wikipedia policy and not permitted. Take a look at the welcome page to learn more about Wikipedia. Thank you.

on up to
 * This is your last warning. The next time you use Wikipedia for unwarranted vilification of entities, you may be blocked from editing without further notice.

and beyond.

General discussion
For my part, I don't especially like these proposals on the merits. But I'm a social democrat and while I appreciate the cool things that corporations make and do, for-profit corporations are amoral entities and their social impact is mixed and they need plenty of oversight. That's my opinion, and a more pro-business person might feel that these are good proposals on the merits. It comes down to a philosophical opinion on what an article about a corporate entity should be: more of a listing of their vital statistics and description of their products and so forth, or more a description of their role in society, or whatever. We're not Frontline but we're not the Chamber of Commerce either, and threading that needle is difficult and contentious.

However, I'm not not suggesting this on the merits, but as I said for two reasons: If this proposal doesn't gain traction, it doesn't mean that these concerns won't be addressed. It just means that they'll be addressed by agents of the corporations themselves, directly. This is problematic as it threatens our reputation, the morale of the volunteers, and our actual neutrality, in my opinion.
 * To address the concerns (or professed concerns if you prefer) of the PR industry.
 * To remove some of the practical reasons for the PR industry to be involved with the Wikipedia.

There's no force on earth that will stop paid agents from editing the Wikipedia, of course. The point is to strip it of its raison d'etre and reduce the need for it. Herostratus (talk) 17:46, 17 January 2012 (UTC)
 * I find the proposal repugnant and impractical. This privileges corporation and analogous entities in a very Citizens United way, not for any noble purpose, but merely to keep PR professionals from having to act like responsible Wikipedia editors. Given the tens of millions of corporations in the U.S. alone, the change to WP:CORP by itself could lead to the creation of an entire industry of "put YOUR company into Wikipedia" spamming specialists who would technically be acting within the rules. I see no burning need to whore Wikipedia out to the paid intellectual <insulting five-letter word to be found in the King James version of the Bible removed>s of the PR industry, just because these highly-paid alleged professionals are too damned lazy or stupid to figure out our interface. As an occasional journalist, I also greatly resent the false, even slanderous use of "investigative journalism" as a synonym for "hatchet job"! -- Orange Mike  &#x007C;  Talk  18:16, 17 January 2012 (UTC)


 * An absolute disgrace. This is a license to turn Wikipedia ino a censored marketing tool, and is far more damaging than SOPA could ever be.Nigel Ish (talk) 20:31, 17 January 2012 (UTC)


 * To allow paid editing or advocacy through means that are legitimized, puts Wikipedia at risk. While not all paid editing is with evil intent, it opens the door for abuse. Phearson (talk) 23:14, 17 January 2012 (UTC)
 * Your solution seems analogous to keeping your front door unlocked so burglars won't force the lock when they come to rob your stuff. Yoenit (talk) 23:47, 17 January 2012 (UTC)
 * OK, but look. It says here, for instance, that a corporation was described as "wanting to kill you" (and the citation was an extremely unreliable source). And this was there for quite a while. And nobody noticed it, or cared. But the corporation noticed it. And they cared. But they couldn't change it (because they don't know how to edit or engage with the Wikipedia properly). So they hired paid agents. I don't like paid agents roaming the database. But if the alternative is that entities will be described as "wanting to kill you" (if it's not justified; it might be in some cases), then bring them on. I think many Wikipedians would agree: bring them on. You want that? People depend on these entities for their livelihoods, you know. It's real important. Why shouldn't they have the same consideration as provided under WP:BLP, or at least some modified version. If we can't solve our problems ourselves, they will perforce be solved by other means -- other means that bring their own problems. Herostratus (talk) 05:19, 19 January 2012 (UTC)
 * Let me get this straight: I'm not familiar with the specific case, but you're saying the corporation "cared" but didn't know either how to edit or how to engage with WP. Well, first off, corporations don't care; despite bizarre court rulings suggesting otherwise, corporations are not human or mammalian or even alive, so they're incapable of caring or indeed of having feelings of any kind. As for the people affiliated with the corporation who cared—well, I'm finding it a little hard to imagine that they were capable of using a web browser to find their corporation's article yet were utterly stymied by the links reading "edit this page", "discussion", and "Contact Wikipedia" that appeared above and alongside that article. I have sometimes described myself as an AGF extremist, but I have to tell you I don't believe that for one second. In any event, put me down as opposing this proposal in the strongest terms. You seem to be suggesting we invite the wolf into the fold because otherwise it'll just sneak in anyway. Good grief. We have policies (NOR, NPOV, V) to deal with bad content; we do not need to give corporations special consideration on top of that. "Unwarranted vilification of entities"? Do you have any idea how dystopian that sounds? Also, the likening of "investigative journalism" to advertising is absurd. Rivertorch (talk) 06:49, 19 January 2012 (UTC)


 * We have NPOV policy to deal with people slanting an article. The proposer seems to at the same time want corporations to be in Wikipedia when they have no notability, and yet for us to treat them with kid gloves like BLP. They have not read the bit in BLP about the strong need for verifiability as well which goes with the kid gloves bit. We definitely do not need loads of corporations noted when they are not notable. From my reading of that case of Bell Pottinger it seems to me their problem was they assumed bad faith so they tried to do things in an underhand way and so acted in bad faith themselves. If they'd done things in a straightforward manner in the first place there wouldn't have been a problem. It does not sound to me from what that says that they have learnt anything either except to be more careful, their attitudes seem unchanged, lets jut hope they follow the policies in future rather than trying to be more devious in 'how best they can use us'. Dmcq (talk) 11:34, 19 January 2012 (UTC)
 * You're ignoring something that applies to straight BLPs as well: BLP victims are typically not familiar with Wikipedia. Therefore, they might not know any methods of fixing the problem other than underhanded ones--that's what unfamiliarity means, they don't know.  They might not even be familiar enough with Wikipedia to know that something is underhanded.
 * A persistent problem with BLPs is that the BLP victim violates the rules to fix his BLP, and a lot of attention is given to banning or blocking him while little attention is giving to fixing his BLP or preventing BLP violations. Pointing out "oh, they assumed bad faith" or "they weren't straightforward" or other examples of misbehavior is an example of this--they don't know Wikipedia, how in the world would we expect them to know about AGF?  All they know is that someone is telling lies about them--to an outsider, that looks like reason to assume bad faith.  So they violate the rules to fix the lies and people like you jump on them because you care more about the rule violation than the fact that we are spreading lies. Ken Arromdee (talk) 16:44, 19 January 2012 (UTC)


 * We are not talking about BLP here. I fully agree that for BLP we should take extra special precautions. I even do that if they are dead never mind the living bit. But this is about organizations and in particular that complaint was about a PR organization and moreover one where the head man still doesn't see anything wrong with what they did. Dmcq (talk) 05:24, 20 January 2012 (UTC)


 * I'm a paid editor that's been involved in the linked projects and the general effort to make this dynamic better. I think the three bullets up top are very good as problem statements (I would add the desire to make pre-existing articles more complete), but not sure these are the right solutions. A few comments:


 * <ul><li>Negative POV is often less scrutinized than positive POV, but policy already addresses both equally. It's more of a cultural and motivational problem. </li>
 * <li>I don't think it makes sense to erode WP:CORP simply because it's difficult to enforce. However I will say there are a lot of very large notable companies who simply aren't in the news much. </li>
 * <li>The biggest problem is PR people don't read or even know about the existence of policies, so creating more policy for them won't change anything, since they won't read it. </li>
 * <li>I'm not sure if this was intentional but I do appreciate the language of "hiring a professional." I think PR needs to recognized Wikipedia as an expertise and there needs to be experts that can be a guardian of ethics, protect them from themselves and know policy.</li>
 * Bell Pottinger basically said they didn't know how to edit Wikipedia ethically. Why do PR people keep accepting work they have no expertise on?
 * King4057 (talk) 16:49, 19 January 2012 (UTC)

The proposal above is yet another reason why Santa Clara County v. Southern Pacific Railroad is one of the worst US Supreme Court decisions of all time, in that it created "corporate personhood" without a corresponding check on the power of a "person" which was effectively immortal and, often, richer than Croesus. We treat the biographies of living persons differently from other articles for the simple, humanistic reason that real life-and-blood people can be conceivable be harmed by irresponsible editing of those article. Corporations, on the other hand, have vast resources at their beck and call, and can counter any inadvertant inaccurcies with public relations, advertising and as much "spin" as they're willing to pay for. There's no compelling reason for us to institute a corporate equivalent of our BLP policy, and every reason to be on guard for their attempts to warp our neutral articles to their liking with paid editing. This may not be David vs. Goliath, but there's certainly no reason to give the corporations our assistance in skewing our articles in their favor. Beyond My Ken (talk) 06:35, 21 January 2012 (UTC)


 * Most corporations don't have vast resources. Upper Crust Pizzeria doesn't. They've got 20 stores, but they're not Exxon-Mobile. This is typical. Is it right and fair that half their article should consist basically of attacks? Maybe it is. But I'm just asking. (And "real life-and-blood people can be conceivably be harmed" by this sort of thing, yes. Upper Crust Pizzeria is not owned and staffed by robots.)


 * I hatted the peripheral and distracting sub-proposals, to clarify that the main proposition is:
 * A notice on the talk pages of these articles, directing people with a problem to a noticeboard where they can seek relief.
 * And the creation and manning of such a noticeboard.
 * And a policy supporting the noticeboard, to the general effect of "negative information which is unsourced or improperly sourced should be removed without discussion". It could be hedged all around with various caveats about how this doesn't mean the article has to be a puff piece, or whatever.
 * What's wrong with these three simple things? Who could be against this? Herostratus (talk) 15:42, 21 January 2012 (UTC)


 * While I support making it easier for businesses to challenge unsourced or unreliably sourced negative information, I also want to be sure that we're not allowing them to to exclude serious but unproven allegations just because the allegations have yet to be proven conclusively. Also, we'd want to prohibit the selective inclusion or exclusion of reliably sourced information in a way that violates NPOV, such as listing their product in an article as "a product that is specifically designed to clean up spilled water is the Big Mop by Mops Inc" while not specifically mentioning their competitors and alternative solutions if competitors and alternative solutions are available. So I support this proposal in the sense that it can help with NPOV and requiring reliable sources, but I want to be sure that we don't go too far in allowing the exclusion of allegations and/or competitors' products. <b style="color:#01796F;">Pine</b>talk 23:02, 21 January 2012 (UTC)

Question why is this proposal for corporations and not businesses in general? Why not also include forms of business like LLPs which are likely to be used by small businesses? <b style="color:#01796F;">Pine</b>talk 23:07, 21 January 2012 (UTC)
 * I think Pine has a good point. Businesses can use a number of different legal frameworks and these can vary a little from country to country. (And in British English, "corporation" looks like an americanism although historically a number of local public-sector bodies used to call themselves "corporation"). So, we should take care to use a more inclusive term.
 * We do have a problem with some business articles being hatchet jobs - although we might fret about paid editors making an article too positive, there's no shortage of editors out there who dislike big businesses generally, or have an axe to grind against a specific retailer or former employer, and hence collect criticism from various angles and wrap it up in decidedly non-neutral text... I think a noticeboard and a couple of templates would be very helpful but am wary of making this a bigger thing with substantial policy changes, like BLP. Simply applying NPOV &c to business articles should be sufficient, I feel, and we should concentrate on ways to get extra eyes on potentially-problematic articles to ensure they fall in line with our existing policies. bobrayner (talk) 00:14, 22 January 2012 (UTC)
 * Well, right, it could be "Organizations" instead of "Corporations" (although I think that LLPs could be shoehorned into "corporation". "Businesses" would be no good since that leaves out not-profits maybe. "Organizations" though would (I suppose) include political parties and possibly bands and so forth so I dunno about that. A minor point of semantics though. Herostratus (talk) 05:57, 22 January 2012 (UTC)
 * I'd support changing this to "organizations." Political parties and bands could have unreliable or unsourced criticism directed at them just as easily as any other type of organization. This doesn't mean that we should remove bad news or credible accusations from articles just because an organization wants us to censor the bad news when the news is backed up by reliable sources, but we also shouldn't be including every unreliable or unsourced negative news and rumor. We need to achieve balance. <b style="color:#01796F;">Pine</b>talk 06:28, 23 January 2012 (UTC)

No. Fifelfoo (talk) 02:01, 22 January 2012 (UTC)
 * Just.... no? Well, while that's succinct, I gather that you preference is for paid agents to be roaming the database instead? Herostratus (talk) 05:57, 22 January 2012 (UTC)
 * Given the choice between professional PR agents and the naïve corporate affiliated people who try to write their own articles, the PR agents at least do a more consistent and usable job, with greater potential for improvement. But very few of them ever fully internalize the basic concept that while they are automatically thinking in terms of what the subject wishes to communicate to the public,  an encyclopedia article must  think in terms of what the public might wish to know.
 * Yet, we are greatly deficient in usable content in this subject area--perhaps more so than any other broad field. I can think of several approaches. The minimum is to consistently watch what they do ,and fix it--but to do this effectively requires legalizing it,and enforcing the standard that they declare their identity. Perhaps we need to modify our policy on anonymity to the extent that anyone editing for pay or part of a job, declare their true identity and affiliation. This would at least provide a better way or tracking the articles,   The second, might be to accept articles on corporate entities in the form of infoboxes, which could then be rewritten by people who understand our rules--this would at least provide the basic information and have the side benefit of providing a channel through which we could look at them. The third, which has the advantage that we are already doing it, is to actively work with the various professional agencies on their field to raise their standardsof work here.  DGG ( talk ) 10:18, 23 January 2012 (UTC)
 * Well, I guess I just fundamentally disagree about the "PR agents at least do a more consistent and usable job, with greater potential for improvement". What PR agents know is how to slant things subtly, gladhand, offer treats (like references (but only the references they want you to use)), and so forth. Of course they know all these tricks. They're professionals! Better some hack job that can be detected and reverted. In my opinion welcoming PR agents into the fold is a dagger to the heart of the volunteer ethos. I for my part am not willing to contend with paid professionals as a hobby. But I'm getting the impression that this is distinctly minority view, so... Herostratus (talk) 13:24, 24 January 2012 (UTC)
 * I agree with Herostratus that PR agents are untrustworthy. They typically have a conflict of interest, and there have been reports of PR agents bragging about the changes that they've been able to make on WP. I don't oppose them working here, but I would want full disclosure of who's paying them and for what purpose, and I hope that their actions could be flagged for especially rigorous oversight by other editors. <b style="color:#01796F;">Pine</b>talk 09:54, 25 January 2012 (UTC)
 * Oh I think there's quite a few general editors around on Wikipedia who are pretty good at slanting things without being professionals! And some seem to put in more time on their hobby horse than any professional would. Disclosure though is what I would hope for from any professional. Dmcq (talk) 01:58, 29 January 2012 (UTC)

Potential student project
I am applying for a summer student to do a Wikipedia Medicine research project through my department. One potentially project I am looking at is having them review all the edits made to Wikiproject Medicine articles. The student will go through each edit and I am thinking of collecting a weeks worth of edits. If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be handling all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine if: The proposed student will be either between first and second year or second and third year medicine and will be working 40 hours per week for 6-8 weeks during the summer. This is still a rough draft thus appreciate comments? Would also need someone who can create a bot to apply PC to the articles in question if we get to that point. -- Doc James (talk · contribs · email) 13:44, 21 January 2012 (UTC)
 * 1) determine if the edit is okay and revert it/fix it if it is not
 * 2) determine which edits are made from IP/new users verses long term edits
 * 3) calculate the percentage of positive/negative edits from each group
 * 4) they will be going over edits more than one day old and thus we will be able to determine how good Wikipedia is at repairing itself.
 * 1) pending changes affects the numbers of IPs editing
 * 2) to what degree pending changes reduces the visibility of poor quality content.
 * Seems to me that this would require turning whole swathes of an Encyclopedia into a semi private test facility. Whatever work is done should be for the benefit of Wikipedia, not an outside research project. Asking that pending changes be applied to (what one can imagine is) a large number of articles for study purposes, is in my view an unacceptable use of page protection tools.  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  14:14, 21 January 2012 (UTC)
 * People here have asked for data regarding if pending changes works or not. This is a proposal for a trail to determine this. The number of articles in question is about 24,000. Concerns raised regarding PC in the past have been 1)does PC turn people away 2)how much time is required to manage PC 3)how much poor content does it prevent going live. We can determine all of this. We could try it with a one day trial to determine if the effects are large before looking at doing a week. Since this project primarily / only benefits Wikipedia it is going to be a hard sell to my department. But just the effort will raise awareness regarding Wikipedia. I find the comment regarding "outside research project" strange as I am trying to get funding for an inside research project. -- Doc James (talk · contribs · email) 14:24, 21 January 2012 (UTC)
 * There was some trial of pending changes a while ago, did it come up with figures and what happened about it all does somebody know?
 * Well one would certainly need the trial to last for a while to get over any transitory effects, also one would need to monitor some similar pages say on biology or sport as the numbers of vandals and good editors may vary anyway e.g. when other countries have a summer holiday. Dmcq (talk) 14:44, 21 January 2012 (UTC)
 * Yes i have seen a bit of an overview from before will try to dig it up latter if someone does not beat me too it. This trial would just be on medicine pages as that is where the founding is coming from and my only interest. Would be great to have others run trials on other topics though. Doc James (talk · contribs · email) 14:54, 21 January 2012 (UTC)


 * (Reply to User:Fred Gandt). This is pretty clearly a project intended to benefit Wikipedia, so please try to conduct discussion on the basis of an assumption of good faith. Phil Bridger (talk) 19:33, 21 January 2012 (UTC)
 * You are assuming that I was conducting it any other way. Please follow your own advice. f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  02:45, 22 January 2012 (UTC)
 * I was not assuming anything, but basing my comment on the evidence of what you wrote. Claiming that this is not being proposed for the benefit of Wikipedia is an explicit failure to assume good faith. Phil Bridger (talk) 11:06, 22 January 2012 (UTC)
 * Whether or not I misunderstood the intentions of the proposal, I am insulted by your patronizing suggestion that I in any way didn't assume good faith. As I read the proposal it seemed to be a suggestion that some medical students would privatize a subject in order to study it (etc.). I however at no point considered the proposal to be made in anything but good faith (the tone is obviously serious and considered); I just disagreed with the proposal (as I understood it). Your implication that I did not assume good faith was rude (to me) and unwarranted. Even my Mother doesn't know what I am thinking. You didn't assume good faith in my response to the proposal. I didn't assume anything. I respond and wait. It has paid off. The proposal is now clearer to me. f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  17:41, 22 January 2012 (UTC)
 * Have you run this by your IRB already, or is this at concept stage? Will the student be writing a paper, or is this a pure WP project? (We heavily frown on external experiments run on us lab rats)
 * My first concern is the expertise of your selected student, expecially if they are handling PCs, 2nd-3rd year meds sounds better. The equal ranking concern is how well they can learn that they are not in charge of anything at all. I think you should change your design so that they are also analyzing responses to their own edits.
 * Setting up PC on 24,000 articles is a tall order, is the function even enabled anymore? Can you select a smaller subset for the trial? Say 2-3,000 articles? Getting an adminbot approved to turn on PC may take the entire summer just to get through BRFA. Franamax (talk) 06:07, 22 January 2012 (UTC)
 * What the student would do would be collecting data for Wikipedia. I guess I should say they would be dealing with PC when it is on so that this would not generate extra work for the community ( a concern previously raised ). Yes PC is still enables (if you are an admin you can see it under the protect opinion). If by write a paper you mean write an article for the signpost yes. I unfortunately do not think anyone cares about the effect of PC on Wikipedia but us thus seriously doubt we could find an academic journal. Doc James (talk · contribs · email) 14:18, 22 January 2012 (UTC)


 * I see clearer now that what you are proposing is a re-evaluation of the potential usefulness of the pending changes system. I'm actually a big fan of the idea. There are issues that kinda fly in the face of open editing though. One has to wonder if Wikipedia would be what it is today if there hadn't been the opportunity for any passing Tom, Dick, or Harry to add their little bit. But that is the question isn't it? And with this study you intend to find out? If pending changes was ever going to work well, I feel the judgement of those editors proven by track record would be superior to the judgement of outsiders (however well versed in the subject they might be). Certainly though, if a re-evaluation of PC is what you're after, and a way can be found to do it without disrupting the Encyclopedia too much, I might support it. 20000 pages is far too many to play with though. A longer running test on a far smaller subcategory would (in my opinion) be far less disruptive (and possibly thus, far more fruitful).  f<i style="color:#0dd;font-size:10px;">red</i>g<i style="color:#0dd;font-size:10px;">andt</i>  17:41, 22 January 2012 (UTC)


 * (*grumble*)
 * Yes, there is data from the previous trial. You can see some of it at Pending changes/Metrics/Anonymous edit quality.  Yes, the data proves that PC works for permitting new and unregistered editors to make improvements (about a third of the edits to these articles) while preventing vandalism and other bad edits from ever seeing the light of day (about two-thirds of the edits).  NB that the articles in question were selected primarily from among semi-protected BLPs, i.e., articles known to have had problems in the past.  The ratio of good:bad edits is likely to be (much) higher if you're randomly selecting articles.
 * Doc James, the better way to run this trial is to randomly assign half the articles to PC and half to current status during the same week. This eliminates problems with unexpected media exposure, holidays, etc., and also halves the workload.  WhatamIdoing (talk) 04:52, 23 January 2012 (UTC)
 * Thanks for that. I just had a quick look around and I just don't see any metrics to answer the questions I'd have wanted to answer. In particular there is no comparable sample chosen with pending changes not used and they should count ordinary editors to see the effect on them as well and if possible I'd like to have an idea of the number of watchers for each article. I think I'd have just stuck it on a random sample of other pages too for the trial period to see the effect where there wasn't a preexisting problem. I'd have thought there would be some good statisticians around on Wikipedia who could have helped with setting up the trial and interpreting the results. Dmcq (talk) 09:10, 23 January 2012 (UTC)
 * By putting pending changes on a group of articles and reviewing all the edit made, than by not having pending changes on a group of articles and reviewing all the edits made one can determine if pending changes affects the number of edits made by IPs / new users. We would also divide the list of articles in half such that (half have pending changes the first week and half do not, than the second week they are switched so that one can hopefully take into account a change in editing volume from week to week even though there is not much of one). We are than comparing articles to themselves (each article will be its own control).-- Doc James (talk · contribs · email) 19:02, 28 January 2012 (UTC)
 * I strongly oppose running this trial. First, the subject is inappropriate, as the content of or medical articles is well-watched, and has been repeated evaluated by outside evaluators as having very high quality, If we were to do such a test, there are many areas much more susceptible to problematic edits--PC was introduced at first as a proposal for BLPs, and was used as such; BLPs in certain fields particularly, such as entertainment and politics, have a much higher frequency of problems ,
 * More generally, I think we've discussed this enough.There is already an excellent trial running: the German Wikipedia. On the one hand, their articles have a higher quality of writing--on he other, they are by our standards very often inadequately documented. Perhaps a more detailed analysis of the differences here might be the more productive approach. It does however, require a degree of fluency in German uncommon in the US.
 * Even more generally, regardless of what might have been the case two or three years ago, quality is not now our most pressing problem The overall quality of Wikipedia is well accepted--that is, the overall quality as judged by appropriate standards for a quick reference site, not the standards for an academic treatise. and the public now seems to understand that such is the appropriate standard. Our problems are rather the attraction and especially the retention of new editors and the introduction of spam articles for both companies and non-profit organizations. Patrolled Changes is irrelevant to the problem of new spam articles, and almost certainly counter-productive in terms of editor attraction and retention. What we need to solve, are the currently critical problems.
 * Overall, I well recall the  tens of thousands of hours for us all devoted to this problem: for us discussing it, for our testing it and explaining it, for the programmers attempting to meet the constraints of our high editing rate.  During the trial, the difficulties were such that I at least simply refrained from editing any article under the trial despite my admin status which meant anything I edited would be automatically approved by the system.  (The effect of the deWP system is such that I no longer attempt to do even simple error-fixing there--which I must admit is all I'm generally capable of in that language. --that's part of the basis on which i anticipate a similar discouraging effect here.)  I think the best way of distracting us from positive work on the problems of Wikipedia would be to reintroduce the subject. I'm glad the programmers made the final decision--their disgust at working so hard on what was not implemented led them to refuse to work further unless we would commit, and since we would not commit without proof that it worked better, this put an end to it.  DGG ( talk ) 10:02, 23 January 2012 (UTC)
 * Yes this was sort of the response I expected. Doc James (talk · contribs · email) 02:20, 25 January 2012 (UTC)

Binding content discussions
<div class="boilerplate metadata" style="background-color: #edeaff; padding: 0px 10px 0px 10px; border: 1px solid #8779DD;">
 * The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page.  No further edits should be made to this discussion.

I table a proposal to the Wikipedia community that I hope you will support. Since May, I have been rather active in attempts to reform the dispute resolution processes. Back in June, I proposed the creation of the dispute resolution noticeboard, which has been reasonably successful in its aims to provide an open style of addressing content disputes.

Since then, I have been working on a few other ideas. While I want to come up with a way to tackle POV pushing, my current proposal is Binding RFCs, a method for resolving intractable content disputes. The proposal explains how the process would work, but in essence, it's a two part discussion which would be closed by three users, an admin, a user experienced in the subject area, and a user experienced in dispute resolution. I envision the discussion structure would somewhat resemble the recent RFC on the verifiability policy, but with some changes, part one of the discussion would only be to present evidence in favour of X proposal or Y proposal (policies, reliable sources, past precedent etc) and the second part being an AfD styled discussion, with comments weighed depending on strength of argument.

I'm happy to answer any questions relating to my proposal and clarify any details. I feel the proposal page itself explains how the process would work, thus I have not rehashed it here. I think that this differs as opposed to other binding content proposals because it puts the power to resolve these issues in the hands of the community. I encourage comments on this and hope this is something the community will support. Regards, <font face="Verdana"> Steven  Zhang  Join the DR army! 07:14, 21 December 2011 (UTC)


 * Essentially, an editorial board staffed by the DR people? That's not putting decisions in the hand of the community, that's putting them in the hand of a cabal. Now there is nothing wrong with decisions being made by cabals (every area has its regulars)... but binding ones? Dangerous stuff. Binding decisions, if they should ever be taken, should only be taken by people vetted by the community as a whole. There is already an arbitration committee for handling "binding" decision; and no, I don't think we need an editorial committee to rule over actual content. This is not the idea I have of a wiki, and definitely not the one I have of Wikipedia. CharlieEchoTango  ( contact ) 07:34, 21 December 2011 (UTC)
 * I think you completely misread my proposal. At the moment, AfDs, RFCs and many other discussions are closed by admins. This proposal would not create a cabal at all. It would mean that instead that three independent users would close the discussions, as opposed to one. The suggestion of a user experienced in th area (say a WikiProject participant of the topic) may be able to add perspective, and a user experienced in DR would help ensure that other venues of DR were tried first. There could be a requirement for these users to be admins, though I note a few discussions that were closed by non-admins well (ie the Ireland article names RFC a month or so ago). But I want to emphasize this is not a creation of a new content committee, I agree that's a bad idea. <font face="Verdana"> Steven  Zhang  Join the DR army! 08:16, 21 December 2011 (UTC)
 * Hmm, yes I did indeed misread the proposal. Disregard, and apologies. CharlieEchoTango  ( contact ) 16:25, 21 December 2011 (UTC)
 * Random comment "To table" something in America means to remove it from discussion, which is quite the opposite in Britain. Just a friendly reminder for future discussions so we damn Americans can follow. Angryapathy (talk) 18:54, 3 January 2012 (UTC)
 * Eh, to me, it meant, well "to table" as in, "to bring a proposal to the table" but your comment is noted. <font face="Verdana"> Steven  Zhang  Join the DR army! 21:13, 3 January 2012 (UTC)
 * American and British usages are exactly opposite. On one side of the pond it means to take up and on the other side it means to lay aside. Diplomats are schooled to avoid the phrase because it has led to some embarrassing misunderstandings. The two dialects deceptively similar, with countless booby traps like this. ~ Ningauble (talk) 18:24, 19 January 2012 (UTC)
 * Steve, I still think this is a bad idea. Editors that were once active in one area of Wikipedia may have switched to another area of wikipedia.  Also, requiring 3 people to close a discussion can drag out discussions unnecessarily.Curb Chain (talk) 19:12, 19 January 2012 (UTC)


 * Comment' - While I'm very much in favor of reining in and, if necessary, eliminating POV warriors, I'm not sure that a draconian device such as this advertised on obscure noticeboards populated by certain sorts of WP volunteers is the answer. I've seen too much of the drama board lynch mob mentality around here. Ultimately this should be the function of our elected representatives, ArbCom. That they seem to have no taste for "resolving content disputes" (even though they, in practice, do exactly that) is part of their own group failing, in my view. They need to work faster, to seek less massive and often irrelevant testimony, and to be more aggressive topic-banning POV warriors off their treasured battlegrounds. I'm not saying that binding content rulings is a bad thing, I just don't trust the precise mechanism you propose to deliver fair and well-considered results. Carrite (talk) 17:09, 21 December 2011 (UTC)
 * Ya know, part of th problem with many current discussions is a lack of knowledge by the community that they exist. I would think that using watchlist notices advertising the creation of a binding discussion would attract the attention of more editors than something like an AN thread. I also do think that having three closers will deliver a more balanced result as opposed to just one closer. That said, we won't know unless we try. I've in fact been discussing this with arbitrator Casliber as an alternative to Remedy 5.1 of the Abortion case, as I feel it would be a good test case, but realise this process needs to get the support of the community first. Thus, I am asking for the support of the community. If the test case goes well, then great. If it crashes and burns, then at least we know it doesn't work, but we won't know if we don't try. <font face="Verdana"> Steven  Zhang  Join the DR army! 19:44, 21 December 2011 (UTC)
 * And if the discussion is binding people might well be keener to take part. And if the RFC is binding we could probably at least list it at WP:CENT. -- Eraserhead1 &lt;talk&gt; 22:25, 1 January 2012 (UTC)


 * Comment: Getting warriors out would be a user issue, so Arb, right? So I would support but for content disputes, with nothing about indevidual users.   Ebe 123  → report on my contribs. 21:19, 23 December 2011 (UTC)
 * Only if its clear cut, and that doesn't stop more warriors joining the party post-arbitration case. -- Eraserhead1 &lt;talk&gt; 20:10, 1 January 2012 (UTC)
 * Support seems sensible beyond getting arbcom involved to remove the worst troublemakers. -- Eraserhead1 &lt;talk&gt; 14:49, 1 January 2012 (UTC)
 * I don't understand. RFCs are binding, aren't they? You can't just ignore consensus. The only thing this proposal seems to introduce is a limit on how soon another RFC can be opened. Do we actually have a problem with people opening new RFCs after the first one doesn't get the answer they want (assuming the first one did result in a clear consensus)? --Tango (talk) 16:34, 1 January 2012 (UTC)
 * That and the fact that discussions get over-closed as no-consensus. That at Talk:China to make China about the People's Republic rather than the civilisation/history the discussion was closed as consensus even though only 51% of editors were in support was helpful to moving forward. And that seems to be an unusual step. -- Eraserhead1 &lt;talk&gt; 20:00, 1 January 2012 (UTC)
 * That strikes me as exactly the kind of reason I don't want such majority votes. I think it was the wrong close, one likely made for political reasons, and one which introduced more instability by making a snap decision.  I still don't understand quite how this was arranged..... Wnt (talk) 18:56, 4 January 2012 (UTC)


 * @Tango, RFC's are not binding, no. People often ignore consensus, because it's not the consensus that they want, and often things like topic bans do not work. The dispute over the images at Muhammad is a good example of this. The other problem with continued discussion is that if one RFC closes with X result, some users will not be happy with the result, even if there was a clear consensus, and they may open another RFC(s) until they get the result they want. This often causes disruption to the wiki, and drives editors away, and this is what the process I am proposing is designed to tackle. <font face="Verdana"> Steven  Zhang  Join the DR army! 21:34, 2 January 2012 (UTC)
 * RFCs are not binding, and they are occasionally ignored, but the usual method of "ignoring" is to claim that you're asking a new question. And sometimes you actually are asking a new question, so we can't just ban that.  To give an example of a complicated situation:  There was an RFC a while ago about whether to have an art nude be the first picture at Pregnancy.  Most editors said no.  They gave two basic classes of reasons, one of which boiled down to editorial judgment (what they thought best for the article), and the other of which was a fairly technical issue about whether the nude woman had consented to have her picture spread around the internet in the first place.  So it closed as "no nude in the lead—unless there really was valid consent for posting the nude photograph to Commons, in which case who knows ("no consensus"), because we honestly have no idea what all of the editors worried about consent would think if the consent were correctly verified". Well, the consent issue was (finally) resolved (apparently at the cost of greatly irritating the photographer).  The pro-nude group said that the now-lack of consensus meant that the article was required to have an artsy nude in the lead.  The other editors opened a new RFC, to deal with the "new" question, which was "Now that nobody has to worry about that tricky consent issue, what do you want to do in this article?"  That is, IMO, a valid "new" question, but the losing side (the pro-nude group knew they were going to lose, and in fact it was ultimately about 70% against using the nude in the lead) spent weeks complaining that this was "ignoring" an RFC that should have been binding. The thing is that most disputes aren't binary.  It's not a question of "yes" or "no".  To use this same example, the "simple" question was really multiple questions:  "Shall we have any image in the lead?  If so, what image (from among hundreds already available to us) shall we put there?  Shall we use this art nude in this article at all?  If so, shall we use it in the lead or in another section?  What caption shall we give the art nude, if we use it?"  It's hard to envision a binding decision on such a discussion that we'd really want to make binding.  "Let's stop arguing over having a nude in the lead" would be helpful (having that nude in the lead had generated persistent complaints since it was first added), but "We decided on the woman in the blue shirt" isn't necessarily something that anyone would want to consider binding.  WhatamIdoing (talk) 23:13, 2 January 2012 (UTC)
 * Sounds like the first RFC was incorrectly closed. -- Eraserhead1 &lt;talk&gt; 23:16, 2 January 2012 (UTC)
 * I envision that this process would be used for things like naming disputes. Stuff like "should we use X or Y image of Julia Gillard in the lede" should not require a binding RFC, and can be resolved through normal editorial discussion. <font face="Verdana"> Steven  Zhang  Join the DR army! 23:25, 2 January 2012 (UTC)
 * And if it was subject to continued discussion it would be quite difficult to argue that the new RFC had been carried out in good faith, as appears to have been the case with Pregnancy. -- Eraserhead1 &lt;talk&gt; 22:32, 3 January 2012 (UTC)
 * Support. Great idea! Ironholds (talk) 21:06, 3 January 2012 (UTC)
 * Support. (Disclaimer - though it is mostly Steve's work, I had a small hand in creating this proposal.) It has become increasingly apparent to me that we need something like this to handle really intractable disputes. This proposal is simple, powerful, and more community-focused than the current de-facto method of leaving ArbCom to organize things. — <b style="text-shadow:0.15em 0.15em 0.1em #555; color: #194D00; font-style: oblique; font-family: Palatino, Times, serif"> Mr. Stradivarius ♫</b> 22:19, 4 January 2012 (UTC)
 * This not a well crafted proposal for the reasons explained above by Carrite and Tango, as well as a few others (myself included) on the proposal's talk page. Amusingly, since this proposal aims to impose a more structured form of DR, it kinda fails at dogfooding by having this split discussion, which was probably unintentional, but still smacks of WP:FORUMSHOP. (And as a side note, I've looked at a few closed threads on the vaunted DRN. They are a mere continuation of the bickering on talk pages, and most of them are closed inconclusively.) ASCIIn2Bme (talk) 10:13, 9 January 2012 (UTC)
 * I disagree that it is forum shopping, I am merely trying to get more input. If you feel it is not crafted well, you feel free to make changes. I don't see anyone else trying to improve DR, but if you think you can do better, than be my guest. I don't claim to have all the answers but at least I am trying. I'd welcome constructive comments and ideas but do not take kindly to comments that amount to "X and Y are crappy proposals". And DRN isn't the problem, it's a lack of people to deal with the issues (read- not enough mediators) Is it perfect? No. Could it be better? Of course. But has it helped people? You betcha. <font face="Verdana"> Steven  Zhang  Join the DR army! 10:37, 9 January 2012 (UTC)
 * Merely posting a notice here is not forum shopping. However, presenting this discussion, which is held in a rather irregular place for policy adoption (if you look at Requests for comment/Wikipedia policies and guidelines), as proof of support is testing the boundaries of forum-shopping, particularly when there have been more comments on the proposal's talk page than here. Also switching the place of the CENT-advertised RfC  is not exactly kosher either. ASCIIn2Bme (talk) 11:00, 9 January 2012 (UTC)


 * Eh, discussion is needed. If I was posting in multiple places that I wanted people to support, that'd be forum shopping. Anyways, discussion has been rather slow. What I'm after is more comments on the process, suggestions and ideas so we can make it viable, and go from there. The proposal possibly isn't 100% yet, so I'd welcome comments. <font face="Verdana"> Steven  Zhang  Join the DR army! 20:20, 13 January 2012 (UTC)


 * I'm aware that this section is not intended to be a vote, so I won't bold my support. But I think most people agree that a way of making long term, binding decisions short of Arbcom is needed. This is not the finished article because the structure of these RfCs hasn't been clearly defined, but the prerequisites to and principles of it are the right ones. —WFC— 01:34, 14 January 2012 (UTC)


 * Support Since the current system doesn't seem to be working particularly well, I think it's a good idea to try something which sounds a bit more streamlined. Miniapolis (talk) 15:46, 14 January 2012 (UTC)
 * Support, since this certainly seems a realistic attempt to solve a real problem, though I'm not convinced we need more process here, just a change of mindset - we need to see admins being (a) more willing to close (and enforce the results of) all kinds of protracted discussions, not just the ones for which we have a process for like AfD and RM; (b) willing (and permitted) to do a bit of focused chatting and informal mediating as part of the closure process, rather than just jumping in. Oh, and (c) prepared to look at the substance of the arguments when doing closures, rather than just counting. --Kotniski (talk) 16:50, 14 January 2012 (UTC)
 * I can think of some areas of Wikipedia where a binding RfC process would be useful; I'm reminded of the Ireland article name controversy. However, I would want any process such as this only being used when absolutely necessary.  My suggestion would be to require ArbCom to first certify a dispute as "good faith but intractable" before a binding RfC could be authorized.  –Grondemar 18:55, 14 January 2012 (UTC)
 * Support Some issues will never be resolved, and the health of the community would be best served by a binding arrangement, even if it's the wrong result. Who cares about the length of a dash, or which image is used in an article, or whether X is English/British/Polish-English? The answer is that lots of people care, and the interminable back-and-forth is a huge distraction. There needs to be a review process, but that has to be time limited (for example, a binding RfC might say "this article will use this image in the lead; can review in one year"). I would prefer to lose an argument if the matter were settled, and I could focus on something hopefully more productive for a few months. Johnuniq (talk) 06:59, 15 January 2012 (UTC)


 * Oppose. Here is the proposal language as seen above and I disagree in general on the grounds that this proposal is redundant. The present process used is enough. I find RFCs to be regarded with some contempt. I don't see this proposal helping with that problem. Just to cite an example from the specific proposal language under "Enforcement": -- it seems to me we have adequate ways to address this already.-- Djathinkimacowboy  what now?! 14:47, 15 January 2012 (UTC)
 * Opppose redundant to arbitration commitee committeeCurb Chain (talk) 15:05, 15 January 2012 (UTC)
 * The arbitration committee doesn't handle content. -- Eraserhead1 &lt;talk&gt; 19:32, 15 January 2012 (UTC)
 * Why doesn't it? I probably should.  In anycase, the 3 people giving the final say on an issue of how content should be?  The first problem I already see with this is how this will change editing for everyone simply because a decision had been already made.  This is the nature of wiki.  You can't assume that 3 people can be the judicial system of content on wikipedia.Curb Chain (talk) 12:03, 17 January 2012 (UTC)
 * Firstly you aren't going to get a single arbitrator to agree with you - and secondly because it would impede on their ability to tackle behavioural issues. -- Eraserhead1 &lt;talk&gt; 23:49, 19 January 2012 (UTC)
 * The problem I see with this proposal is how it is trying to allow another decision making process on wikipedia to take place, contrary to the communitywide decision making process of consensus. I see, if this was to go ahead, one decision making process used for non consensus issues, and everything else goes through WP:CONSENSUS. What is the point of that? Who decides a discussions must go through a binding RfC?Curb Chain (talk) 00:01, 23 January 2012 (UTC)
 * I really have to disagree here. Think of it this way. Pretty much all discussions on Wikipedia are closed by one or more users. RFAs are closed by a bureaucrat after 7 days of discussions, where they evaluate the consensus in the discussion. XFDs also run for 7 days and are closed by an uninvolved admin, who closes the discussion as per the consensus. This is pretty much what would happen here as well, except we increase the amount of closers from one to three. These users would not make unilateral decisions, but evaluate consensus. That's the norm, and I too would disagree with a community process where three users decide an issue unilaterally. That is not the case here. <font face="Verdana"> Steven  Zhang  Join the DR army! 21:41, 23 January 2012 (UTC)


 * Oppose What an invitation to Wiki-Lawyering this would be. People noting the dates in their calendars as to when the rumble could begin again. People make peace! Consensus can be hard to find over contentious topics. Locking any POV is going to stir trouble, not solve it. Sometimes NPOV is that there is contention. htom (talk) 17:24, 15 January 2012 (UTC)
 * So we should just discuss controversial topics forever?
 * Forcing endless discussion just allows the partisans to win as they care more. -- Eraserhead1&lt;talk&gt; 19:32, 15 January 2012 (UTC)
 * Yes, that's what you do with controversial ideas; and only if you give up the "good fight", abandoning the field to the other POVers. Your call.htom (talk) 22:52, 16 January 2012 (UTC)
 * Not everyone who gets involved in controversial topics is a POVer. Some of us want to follow the policies and principles we follow elsewhere in the project.
 * Calling it the "good fight" is implying that everyone is out to WP:BATTLEGROUND which isn't true. -- Eraserhead1 &lt;talk&gt; 23:50, 19 January 2012 (UTC)
 * Support This will take a bit more discussion but basically I think it would be a good idea to have RfCs to have a reasonable time set on their result so we don't have POV warriors grinding their way through to sticking in their ideas and removing other editors by exhaustion. Bad results might occasionally get set in stone for a while but it would free up editors to look at disputes properly and not being in eternal war mode. Dmcq (talk) 17:37, 15 January 2012 (UTC)
 * Weak Support - When mediation and arbcom have failed, this should be available. This wouldn't mean this procedure would have to be used in that case, just that it could if a request for its use was granted (think certiorari). SMP0328. (talk) 19:29, 15 January 2012 (UTC)
 * I think part of the problem here is that no-one knows what such a discussion would look like. It won't look like an RFC, that's just the title I chose (for lack of a better name). I'll create a mock-up page of how it would work so you all can get a better idea. Regards, <font face="Verdana"> Steven  Zhang  Join the DR army! 21:11, 15 January 2012 (UTC)
 * We do know what a binding RFC would look like. We have had several examples of them. Each one was different according to the needs of the community involved in the dispute, but they have shared common features. -- RA (talk) 23:46, 15 January 2012 (UTC)
 * I'll create a mock-up, then let you decide. <font face="Verdana"> Steven  Zhang  Join the DR army! 03:44, 17 January 2012 (UTC)


 * Comment. RFC's, if closed by an uninvolved admin or in some cases by an editor are already binding, that is until a new discussion is started were consensus might shift against the RFC's outcome and therefore can overrule any previous attempt of solution. I don't think we should make changes that go against established principles that, so far worked out just fine most of the time.TMCk (talk) 01:38, 16 January 2012 (UTC)
 * No, RFCs are not binding right now, even if they are closed by an uninvolved admin. The results are commonly respected, but that's largely due to the implicit threat that people not accepting the results could be blocked for disruption.  WhatamIdoing (talk) 18:30, 16 January 2012 (UTC)
 * Which is what makes them binding (or ought to). If someone keeps recreating an article that's been deleted through AfD, there's nothing physically to stop them, until they get blocked (or the title protected, though they can get round that by creating it under different titles). Same with RfCs: if you keep editing against the consensus, then the page will be protected, or (preferably) you'll get blocked. "Binding" on WP can only really mean this. And we should be a lot more vigorous in ensuring that disputes are actually settled through the discussion, i.e. through (not necessarily unanimous) consensus, as opposed to the only alternative - which is settlement through edit-warring and random page protection, which can hardly be of benefit to the encyclopedia.--Kotniski (talk) 10:02, 17 January 2012 (UTC)
 * Comment. Re "I envision the discussion structure would somewhat resemble the recent RFC on the verifiability policy..." — That didn't work out so well. There was 62%  support versus 34% oppose for 444 respondents and the 3 closing admins decided that wasn't a consensus because the support comments were not sufficiently definite in their support. This was found not to be true. See Support comments of previous large RfC re first sentence. --Bob K31416 (talk) 03:16, 16 January 2012 (UTC)
 * Oppose The Israel-Palestine conflict area shows that it does more harm than good. We actually had a "centralized discussion" on an issue. The close was contentious, and opening a new RfC or even a sub RfC to address outstanding issues has not been possible. The implementation was terrible. It simply didn't work even though it could have been a great thing. Consensus grows and shifts over time. Relying on an RfC that could be out of date is not the way to improve contentious topic areas. We should not be adding yet another level of bureaucratic hoops since it will only be used to shout down further discussion by Wikilawyers. Cptnono (talk) 03:20, 17 January 2012 (UTC)
 * Speaking as someone who mediated an Israel/Palestine case for a few months back in 2008 over one word, the I/P dispute outlines exactly why such a process is needed. Everyone agreeing on something is never going to happen. Structure is important to such discussions, and is something I intend to demonstrate (when I get a chance). As for doing more harm than good, I'd make the argument that endless arguments on such matters (many examples from I/P disputes, for example) drives away more editors than a binding discussion would. <font face="Verdana"> Steven  Zhang  Join the DR army! 03:42, 17 January 2012 (UTC)
 * Nonsense. The centralized discussion (Legality_of_Israeli_settlements) was an enormous if imperfect success. It's probably one of the most significant improvements in the topic area that has taken place since Wikipedia started. It largely put an end to pointless disruptive discussions and edit warring in articles all over the topic area about this issue, but far more importantly, policy compliance has been increased in a very large number of articles. The only people edit warring the content out now are vandals who are reverted on sight and blocked by admins. The issue was not contentious at all in reliable source world and what the sources say is not going to change. It was contentious in Wikipedia because of many editor's inability and disruptive unwillingness to simply follow policy. That is exactly why these kind of discussions with binding decisions are necessary. <small style="border: 1px solid;padding:1px 4px 1px 3px;white-space:nowrap"> Sean.hoyland  - talk 06:53, 17 January 2012 (UTC)
 * This happens without RfC's. Our many policies and guidelines cover over what what the community deems to be best practice.  Using a separate venue analogous to a judicial branch of government is contrary to WP:BUREAUCRACY and WP:CONSENSUS.Curb Chain (talk) 19:24, 19 January 2012 (UTC)
 * So why did it take 10 years for the world's largest country to have a title which satisfied WP:COMMONNAME and WP:POVTITLE? -- Eraserhead1 &lt;talk&gt; 23:55, 19 January 2012 (UTC)
 * That is entirely political. There are problems in the real world where people don't agree.  As in real life, calling the People's Repulic Of China "China" is a political move, especially in the presence of Taiwan or certain Taiwanese nationalists.  The situation on Wikipedia is the mirror of the real world.Curb Chain (talk) 20:56, 20 January 2012 (UTC)

Reflections from the Ireland-names example

 * Apologies in advance for very long post.

I was very involved in the Ireland-names example: I proposed that the issue be resolved through a binding RFC. Despite that, however, I would be very cautious about bringing binding RFCs into the fold of every-day dispute resolution mechanisms.

There seems to be a sense here that a proposal like this is an obviously good idea. I would absolutely disagree that. Binding RFCs can lead to deep rifts in editors (as it did on Wikiproject Ireland) and do not lead the the "right" answer (the dispute they resolve are those where there is no "right" answer). This is because they run contrary to normal collaborative and consensus-based practice on Wikipedia and, by their very nature, they strip whole swathes of (constructive and well-intentioned) editors of their voice either indefinitely or for a set amount of time.

Because of this, anything like a binding RFC is rarely, if ever, a good idea and, therefore, having an explicit process for them would be a bad idea IMO. In practice, if binding RFCs were commonplace, I believe they would quickly fall into something akin to what I call the "tyranny of consensus": decisions where consensus is not allowed to change because, we are told, the existing arrangement is consensus and attempts to change it are, therefore, disruptive. Situations like that are incredibly vexatious and not a good idea, either for dispute resolution or content development. Certainly, I do not believe that commonplace binding RFCs would promote either harmonious relationships or good solutions to difficult questions.

Despite this, I would suggest that there is no need for a proposal of this sort: the means for binding RFCs already exist — and bindings RFCs do have a place — but their place is under WP:IAR. Under WP:IAR, the existing precedents for binding RFCs already allow for binding solutions to inextricable disputes. However, under the existing arrangement they are entered into only when all other means are exhausted and where participants to the dispute agree mutually that a binding RFC, outside of the normal "rules" of Wikipedia, is the best, probably only, way forward. That is what gives the existing examples of bindings RFCs the strength of legitimacy they need.

Further, I don't believe it is wise that a proposal for a process that diverges so greatly from core wikipedia policy and current practice should come from the ether. The best policies and guidelines come from already existing practice. They begin as consensus already and are simply formalised into policy. Therefore, I don't think that a proposal like this is a wise idea. An essay or guideline that drew on the lessons of the examples of binding RFCs that we already have may be a good idea. But coming up with a brand new process for something like this, without basing that process on the learning we already have, would, in my opinion, be a bad idea.

Finally, from my experience on the Ireland example, I would say that the following would are important points to consider for any binding RFC:


 * Consensus for a binding resolution: The impetus for a binding resolution must come from the participants in the dispute and be agreed in consensus by them. Otherwise, there will be an enormous sense of aggrievement and a feeling that outsider are enforcing resolution to a dispute that they do not understand. This view would not only be serious, and may be long-lasting, but it would in many case be quite justified.
 * Exhaustion of other means: This impetus should be driven by the absolute exhaustion of other modes for resolution. A binding resolution should only be adopted in the most extreme of cases, where discussion and dispute resolution has been on-going for years. It should never be entered into lightly or for the sake of convenience. In all, the decision to arrive at an binding resolution should come the absolute exasperation of all sides.
 * Consensus can change: Time limits on how "binding" the outcomes are are probably best. I suggest 12, 18, or 24 months are appropriate terms. Any longer would be normally excessive IMO. In the Ireland-name case, two years was long enough for views to settle to a reasonable level.
 * Clear, objective and mutually verifiable outcome: A vote is more likely to be suitable than a discussion. If the dispute was resolvable by discussion then it would not need a "binding" solution. A clear, single and objectively measurable outcome is necessary. Otherwise, the decision will be disputed. In the Ireland case, we used the alternate-vote system. This allowed multiple options to be voted on in terms of preference.
 * A mutually-agreed process in advance: A boiler plate process may not be suitable for all disputes. Different disputes have different components and no two disputes are the same. Additionally, the need for consensus among the participants to a dispute for a binding RFC, and for the consensus of participants to recognise the outcome of the RFC, mean the participants in the dispute will need to be involved in shaping the process through which the binding resolution will be arrived at. That means that no two binding RFCs will likely be identical. However, they may share common features.
 * Recognition that the outcome is not the "answer": There needs to be recognition that the outcome is not consensus — or the "correct" answer — but rather it is a "best we can do" to resolve a long-runing dispute for the time being . There is no benefit in running roughshod over genuine and informed views of well-intentioned editors just because they happen to be on the wrong side of the views of others (often the views of people who know less about the subject than they do).
 * Agreed "facts" and a voice for opposing views: Clear, understandable and mutually-agreed background info needs to be provided to outsiders in advance of inviting them to participate in the RFC. Additionally, individual participants need to be able to express their own voice. In the Ireland case, individual participants in the dispute were also able to provide their own statements on a subpage and these were able to be endorsed by other participants in the dispute ahead of the vote being opened to "outsiders". This allowed for both a commonly agreed backgrounder as well as enabling individual editors to set out their own stalls.
 * A legitimate authority: There is a suggestion that a process for a binding RFC could by-pass ArbCom. I would completely disagree. There needs to be an authority to back up the decision. That authority needs legitimacy and absolute authority to enforce what will. in effect, be an arbitrary decision. The only body that can effectively and legitimately do that in all circumstances is, IMO, ArbCom. Regardless of whether ArbCom are involved or not, without an strong legitimate authority to enforce the decision, the decision won't be accepted by a consensus of editors into the future and will eventually come apart or turn into farce.

-- RA (talk) 22:49, 15 January 2012 (UTC)
 * Haven't read your entire comments, but I would see this as a process that is not commonly used. A few requirements would be the use of mediation and one other DR process to have been tried and been unsuccessful at resolving the issue. As for Binding RFCs being a bad idea, well, I would suggest a test case would be the best way forward. I also note that endless discussion can drive away editors, which is pretty bad as well. As for ArbCom, I think that this should be a community process. We're the ones who deal with resolving content issues, and I think we're capable to do so in this sort of situation as well. <font face="Verdana"> Steven  Zhang  Join the DR army! 22:58, 15 January 2012 (UTC)


 * I didn't say that binding RFCs were a bad idea. In fact, I said they have their place. However, I do believe that having an explicit procedure for binding RFCs is a bad idea. For one reason, I think it would invite binding RFCs where there would be no need for one or where one would not be a good idea. It would also, to some degree, "normalise" binding RFCs, whereas they are in, fact, "not normal" and should remain "not normal".
 * WRT "test cases", we already have at several "test cases". Binding RFCs already exist. I suggest that, if you are not aware of these, you should look at them and learn from them. As I wrote, the best examples for new policies and guidelines come from exiting practice. Rather than coming up with something new out of the ether, you should look at how binding RFCs work right now. The oldest example, I think, is Gdansk. We have also had Ireland and Macedonia. All three have worked successfully (in their own terms) — or as well as could be expected.
 * WTR to ArbCom, when you do take the trouble to read what I wrote (and I apologies again that it was so long), you'll see that I said that binding RFC need to be a community process. In fact, they need to be one that is initiated by the community in dispute according according to a process they can agree. This is the example set by binding RFCs to date and is one that has worked for them. However, an authority is needed to enforce the outcome. Again, in the examples of binding RFCs that we have to date, that has been ArbCom. -- RA (talk) 23:30, 15 January 2012 (UTC)
 * Steven, you claim that you want the community involved in a the decision making process when a no consensus is reached in RfC's, because of a lack of participation of the wider community or otherwise. Your proposal is to introduce voting so these disputes get resolved.  Or so is there another solution you are thinking of?  A cursory look at the proposals here and on the project page does not seem like there are any better solutions.  I see a our current unworkable solution, but I see worse proposals to solve it.Curb Chain (talk) 19:06, 19 January 2012 (UTC)
 * Not raw voting, of course not. I do apologise for not creating a mock-up page as of yet (been crazy busy lately but will try today) but turning it into a vote, not exactly. We have many processes that are an evaluation of consensus that in essence resemble votes. RfA, XfD, requested moves and even this discussion are examples. But they are in fact closed per the consensus of the community. A binding content discussion woul be no different. I'll make creating the mock-up page my first priority today so you can see what I mean. Regards, <font face="Verdana"> Steven  Zhang  Join the DR army! 21:41, 23 January 2012 (UTC)

Binding content discussions. Section break

 * Support the idea of binding RFCs. Some disputes, specifically those with strong arguments and substantial amount of supporters on both sides, can last forever and even make people quit editing. Though I would note, that the particular attention should be given to making the wording of request content-oriented, as there are just too many examples of RFCs that fail to help dispute resolution just because the question doesn't fairly represent the issue. Given the binding result, this type of RFC should be subject to a very careful choice of wording. &mdash; Dmitrij D. Czarkoff (talk) 01:12, 16 January 2012 (UTC)

Comment. On first glance this sounds too complicated. There are thousands of content disputes. I think people want quick comments first from admins and arbitrators (we need a lot more of both) on the article talk page. See my userboxes for more info:

This idea for binding content discussions is one proposal of various content dispute arbitration methods. It should be used only after other simpler feedback from admins and arbitrators. Right now admins and arbitrators don't officially comment on content. They should be summoned right away. And they should comment directly on the article talk page. See: User:Timeshifter/Unresolved content disputes. Editors are leaving because of this. --Timeshifter (talk) 11:11, 16 January 2012 (UTC)
 * Yah, this is kind of the crux. Content discussions are binding (or should be), in the sense that you shouldn't be editing against the consensus as revealed in those discussions. And if it's not clear what consensus has been revealed in those discussions, then you need an admin to come and sort it out, say what the result it, and make sure it's enforced. Unfortunately, admins are fairly unwilling to do this in the case of ordinary (non-process) page-content discussions, I think because of two strands of muddled thinking: (1) "The fact that there's still disagreement show's that there can't be consensus." Nonsense, of course, since our definition of consensus does not require unanimity; and this is clearly seen when admins close the process discussions (Afd, RM, ...) (2) "Admins are not supposed to decide content disputes." Well they're not supposed to impose their own opinions on content; but that's no reason why they shouldn't assess the result of content-related discussions - as of course they do every day at xfD and RM, which are also content discussions, just a particular type thereof.--Kotniski (talk) 12:10, 16 January 2012 (UTC)
 * ...but.... the way discussions are closed at the moment is unsatisfactory, particularly for debates about article wording, where multiple possibilities exist. I mean, it works all right for most simple everyday RMs etc., but not for long, complex or fraught debates. Here we need the admin to act not as a one-person jury ("I've looked at the discussion and have decided this"), but someone who can possibly mediate a bit so as to focus the discussion, possibly suggest compromise solutions, or in any case talk to the participants a bit just to be sure he's noted and understood all the arguments and counterarguments correctly. And there doesn't have to be just one admin doing this, either. --Kotniski (talk) 12:19, 16 January 2012 (UTC)
 * Admins do not have any authority to decide content disputes. And they should not have that authority except as part of a process we set up for settling content disputes. But their opinions and experience are valuable, and admins should be summoned much sooner concerning content disputes. And it should be quick, and above all, done on the article talk page. Oftentimes editors don't understand that WP:NPOV is not necessarily implemented by picking one description of the facts, but by pointing out all significant viewpoints. Admins can explain this quickly, and editors are less likely to believe it is wikilawyering since the admin is new to the discussion. --Timeshifter (talk) 13:21, 16 January 2012 (UTC)
 * I believe admins do have the authority to decide content disputes (as they do regularly at AfD and RM, which are particular types of content dispute; as well as at RfCs, sometimes, when asked). We don't need to set up a new process; we just have to get people into the habit of asking admins to help reach a settlement when the editors themselves can't work out what they've decided (of course, in most cases this won't be necessary), and admins into the habit of responding positively to such requests, and everyone out of the mindset that "closing" a discussion must be a single God-like act without any interaction with the participants. So what I'm suggesting is fairly similar to what you're suggesting - summon an admin for help in resolving the dispute, but the reason it's useful for it to be an admin rather than any other experienced editor is that admins have the authority (and should be given more of it) to take firm action against those who continue to disrupt the process or edit against consensus.--Kotniski (talk) 17:53, 16 January 2012 (UTC)
 * Admins should absolutely not be given that authority. It goes against a core belief in Wikipedia that goes all the way back to the dispute between Jimbo Wales and Larry Sanger concerning "experts." AfDs and RMs are not content disputes. One is about deleting an article. One is about changing the name of articles. Neither decide the content of articles. WP:Edit warring is one of the most disruptive of the guidelines because admins are using it arbitrarily to decide content disputes. It should be used first as a warning by an admin, and not used first as a block. Many editors leave Wikipedia due to such abuse by admins who give blocks without warning from an admin. --Timeshifter (talk) 23:58, 16 January 2012 (UTC)
 * Yes, sure, blocks should be used only as a last resort. But it's nonsense to suggest that having admins "decide" RMs and AfDs is fine, while having them decide "content disputes" is fundamentally wrong. This is the sort of prevalent muddled thinking I refer to above. RMs and AfDs are simply particular types of content dispute - in the first case about that part of the content of an article that goes above the top line, and in the second case about whether to include the whole content of a given article in the encyclopedia. If other content disputes can't be decided by adjudged consensus, then they'll end up being decided by edit-warring, which is contrary to our fundamental principles.--Kotniski (talk) 09:56, 17 January 2012 (UTC)
 * I am against single admins making any content decisions. AfDs aren't decided by a single admin. RMs aren't decided by a single admin. They are closing a discussion made by a group of people. If you read my proposal linked from my userbox higher up you will see that at the very end of the process a group of people make content decisions. --Timeshifter (talk) 14:14, 17 January 2012 (UTC)
 * "Single admins closing discussions made by a group of people" is exactly what I'm suggesting should happen, and sometimes does happen, for discussions about page content.--Kotniski (talk) 14:33, 17 January 2012 (UTC)
 * Arbitration currently is decided by a vote of the arbitrators. That is what should happen concerning content disputes too in the end after all else fails. The arbitrators do not consist of the editors involved in the dispute. See my proposal: User:Timeshifter/Unresolved content disputes. --Timeshifter (talk) 16:13, 17 January 2012 (UTC)
 * OK, I agree, but the arbitrators should act as a final court of appeal, not of first instance. So first an admin tries to assess the consensus, then (possibly) someone will challenge that decision and take it to (say) the admins' noticeboard, and then finally if the matter is still disputed it should go to ArbCom. I've been saying this sort of thing ever since I started commenting on dispute resolution (which I started doing when I saw first hand how broken the present system is). Basically what's wrong with the system as it is is that it encourages and foments drama and disruption and endless fruitless debate, when it should be stamping on them hard, and focus on getting real resolutions to real problems in real time.--Kotniski (talk) 17:14, 17 January 2012 (UTC)
 * An admin is no better at assessing "consensus" than anybody else. So I don't want admins assessing consensus more than anybody else. As in Animal Farm: "All animals are equal, but some animals are more equal than others". Go stamp on somebody else hard, but not on Wikipedia. --Timeshifter (talk) 00:41, 18 January 2012 (UTC)
 * So who do you think should be assessing consensus? I hardly think ArbCom will have time to consider every AfD, RM and disputed RfC, unless we elect about 100 arbitrators and they divide into subcommittees or something. (Of course we have non-admin closure too, but custom has it that this is only for clear-cut cases - and that admins are asked to review such closures if they're disputed.)--Kotniski (talk) 08:17, 19 January 2012 (UTC)

@TimeShifter, we need to work within the options we have at present. ArbCom had stated over and over that they will not intervene in clear cut content disputes, and are reluctant to comment in regular discussions due to the fact they likely would have to recuse if an issue over conduct came to ArbCom. Admins assess consensus on many issues, such as XfD, RM etc. Bureaucrats assess RfAs. Non-admins can close some clear cut discussions as well. I decided on this approach because the ideas of content committees have been shot down in the past repeatedly, and this seems the best option to address intractable content disputes by the community. Set up a structured discussion with clear possible outcomes, users add material supporting X or Y, and members of the community opine on the discussions. After a period of time, three admins evaluate the consensus and close the discussion. Having three closers is key here. It reduces the possibility for bad closes (three opinions instead of one). I'll work on that mock-up for you all to see. <font face="Verdana"> Steven  Zhang  Join the DR army! 21:41, 23 January 2012 (UTC)
 * Oppose. While certainly voting is not evil, I do think that it shouldn't be used to resolve content disputes.  As things are, we've got three stages in an attempt to reach a consensus.  1. We start on the article page, and try to talk things out.  2. We invite uninvolved editors in in an RFC, 3. As a last resort, we ask Arb Com to talk it out. Arb Com doesn't rule on the page itself, but on user behavior.  In all of these cases, there is an attempt to reach a flexible consensus.


 * Binding RFCs won't do that; you'll make winners and losers, you'd force people to take sides, you invite Wikilawyering, and unless but there's a very clear vote on wording, you're still going to have further arguments over the nature of the vote, only now with the winners being self-righteous and the losers pissed off. Further support or oppose vote cast by a drive-by-editor will have the same weight as the contributions of an editor who takes a great deal of time attempting to reach a compromise. (And the same weight as all the text posted by a single filibuster whose posts make up 95% of the entire discussion, but I'm not sure the benefit from that is worth the detriment of downplaying attempts to reach a consensus.)


 * Generally, I think most users are reasonable, although some can have Achilles' heels where there's an issue they particularly care about (this seems to be the particular case with ethnically based conflicts), but even Arbcom tries only to make sure that they remain civil and contribute fair, verifiable information. Even if it takes awhile, a consensus is reached eventually, and can be changed when it needs to be.  Wikipedia is not a race. --Quintucket (talk) 22:15, 16 January 2012 (UTC)


 * Perhaps it's just me, but I feel that many of the users voting oppose here may not to understand the proposal, so let me make a few comments here.


 * The first point is that this would be a last resort dispute resolution method, reserved for issues that have failed both mediation and other forms of low level dispute resolution, and where conduct is not a primary issue. It's not something that would be used often at all.


 * Some here have commented that the style would allow for !voting by disinterested people. Isn't that the whole point? If we only allow comments by people involved in the dispute, we get nowhere, otherwise the dispute would have been resolved without requiring a binding discussion. I also think a watchlist notice would aid in this.


 * Some have commented that it will create a win/lose situation for some. Unfortunately, it may have to come to that. The alternative is that discussion continues until the heat death of the universe, and the person who digs their heels in the hardest wins, the rest leave exasperated, and after all that's what we're trying to fix, editor retention. I would much rather a structured discussion based on policy decide outcomes of deadlocked content disputes (generally, naming disputes) rather than the former.


 * I also note that none of you have seen the proposed structure of this process. That's partly my fault, but assuming that it will look like a normal RFC is incorrect. I also note that few here actively participate in dispute resolution. It is true that most users are reasonable, and most issues can be resolved through mediation, but some cannot, and that's what this process would be for, when all else fails.


 * If there's no clear consensus as a result of a binding content discussion, then the status quo would remain. I don't see a 51% consensus being enacted. Note the closing of a discussion by three, minimising the chances for bad decisions that have possibly occured in past binding discussions. Note, bad != everyone doesn't agree.


 * Wikilawyering is always a factor in any discussion, but take the Pro-life/Pro-choice vs Abortion-rights/Anti-abortion titles. There would be a clear outline of what issue the binding RFC is to address, and users would present evidence to back up why outcome X vs Outcome Y is appropriate (policy, backed up by other info). For example, if an argument was made for common name usage, a demonstration of how common it is, and so on. After this, editors would partake in an AFD style discussion, commenting on which they prefer and why based on the information presented, eg, "I prefer X as is the common name as demonstrated in Y sources" or something like that.


 * All in all, binding discussions of content should indeed be the exception to the norm, but part of the problem in the past has been a lack of a unified method to hold such a discussion. That's what my proposal is about, bringing structure to these discussions, limiting long-term disruption in intractible disputes. Locking down content isn't a great idea, but it's better than endless heated discussions over X versus Y, which people end up leaving over. That's my opinion, and is the reason I made this proposal. Try it. I don't see any harm in a test case. If it flops, hey, at least we know it doesn't work. If it does, well that's great. But I feel it's a viable solution to a real problem. <font face="Verdana"> Steven  Zhang  Join the DR army! 04:11, 17 January 2012 (UTC)


 * Steven, I appreciate that you are well-meaning here but, rather than supposing that those who oppose the proposal do not understand it, I suggest you could consider another possibility: you may not be as conscious of the issues involved in a binding RFC as you may believe you are.
 * First, I think you need to acknowledge that binding RFCs already exist. You seem to believe that they do not - and that this is an original idea that merits "trying out" to see if it works. In fact, we have had several examples of binding RFCs. I've linked to some above. I think you should first begin by looking at those examples, see how they came about, if they resolved the dispute and the issue affecting editors involved.
 * Some of those who oppose the proposal here (myself included) do so because of our experience with binding RFCs in the past. Here is an important point: it is a policy on binding RFCs that is being opposed, not bindings RFCs. Binding RFCs have their place. The question is would anyone who has been involved in resolving a dispute through a binding RFC believe that they should be a part of the stated toolset of dispute resolution? Certainly, I do not.
 * Their place to date has been under WP:IAR and with good reason (not least of which is because every one is different). In contrast, an explicit policy ("a unified method") would run counter, I believe, to the very reasons they have worked in the few examples that we have. Let them happen on a case-by-case basis where there is a strong consensus and an over-riding reason to have them. But "legislating" for them, or prescribing how they should happen, is not a good idea IMO. -- RA (talk) 10:13, 17 January 2012 (UTC)
 * How do you distinguish "binding" RfCs from "non-binding" RfCs? What on earth is the point of having a discussion that leads to a result which is not then binding (i.e. enforceable)? If consensus is not binding, then what is? The only answers I can think of are (a) the decision is made by the best edit-warriors; (b) any sufficiently bloody-minded editor can block any change to the status quo; (c) decisions are made randomly, by admins locking a page in some version. It's not hard to see why any of these three solutions would be worse than having decisions made by (qualified) consensus. Does someone have any other alternative? (And don't say "keep discussing until you reach a solution that everyone can accept", since we know that's not always possible or even desirable, and is effectively just the same as my (b).)--Kotniski (talk) 10:45, 17 January 2012 (UTC)
 * "If consensus is not binding, then what is?" - Consensus isn't binding by it's very nature. It is consensual. That's the point. If there was consensus then there would be no need for a binding RFC. You are confusing democracy (which Wikipedia is not) with consensus.
 * In the cases of binding RFCs that I pointed to, there was no consensus. And there was no hope of consensus ever being reached. So, instead, in stark contrast to normal practice (see WP:IAR), we decided things by democracy (or at least we did in two cases). The RFCs were closed with a decision one way the another, and deemed to be absolutely binding, regardless of whether it attained consensus of not. Even regardless of whether consensus changed or was later reached! In fact, the kibosh was put on all further discussion of the matter.
 * The point that I am making is that processes of these kind fit under WP:IAR (as they are right now). Codifying them into policy would invite POV pushing and gaming and provide a disincentive to reaching agreement by consensus. People would reach for it too early, when in fact it needs to be the very last option. Something you never even imagined would be an option.
 * Even more importantly, a policy would try to fit binding RFCs to a single codified procedure. In contrast, in each of the examples that we have of binding RFCs, a different procedure was used in each case. Crucially, this was one that was agreed in consensus by the community in dispute according to the needs of resolving their particular dispute. Bindings RFCs are right at the fringe of dispute resolution and therefore straight-racketing them according to one policy, devised out of the ether, is not a good idea. The process needs enormous consensus from the community in conflict (and so needs to come from them), otherwise the result will lack legitimacy. -- RA (talk) 14:14, 17 January 2012 (UTC)
 * You seem to be confusing consensus with unanimity (as someone always does when these matters  are discussed). In the sense in which it is used on Wikipedia, consensus does not have to mean unanimity,  and therefore not everyone needs to have consented to it,  and therefore some people might still want  to thwart the decision made, and therefore someone has to stop them from doing so, otherwise the decision  would not  in fact be made, and  the whole idea that our decisions are made by "consensus"  would be a myth. Sometimes it eventually comes down to something resembling a majority vote (quite often, in fact), although we sometimes require more than 51% to change something, and we don't look only at numbers. And sometimes we really do have to stop further discussion of a matter, let it stay settled for a certain period, so that people can get on with other important things. (But I agree with you that there can't be a one-size-fits-all procedure for all types of disputes; and that we don't want people to seek enforced resolution of a disagreement too quickly - though too slowly is also bad, as some matters end up consuming vastly excessive amounts of editorial time and attention.) --Kotniski (talk) 14:30, 17 January 2012 (UTC)


 * "You seem to be confusing consensus with unanimity (as someone always does when these matters are discussed)." - No. I'm not. Indeed, I took the trouble of linking to a dictionary defintion of consensus so that there should there be no such misunderstanding.
 * I am talking about situations where there is no widespread agreement. There may be two or more very divided, evenly numbered, and equally "right" divisions in opinion. Those are the situations in which we have used binding RFCs to date. -- RA (talk) 15:58, 17 January 2012 (UTC)
 * Well, the ones you're thinking of, perhaps. But there are RfCs every day, some of which produce more or less conclusive results - it would hardly be respectful to those taking part in those discussionsg to tell them that their opinions count for nothing in the face of a determined edit-warrior or two. --Kotniski (talk) 17:08, 17 January 2012 (UTC)


 * I whole heatedly agree with that! I'm shocked at how little action comes out of some RFCs. I'm thinking of RFC/Us in particular but the same can be said for content RFCs. Often, even when there is consensus, nothing happens. Another problem is arriving at consensus in some RFC (again I'm thinking of RFC/U, in particular). The structure of some RFCs often does not lead to action.
 * However, I would very strongly contrast that problem with a new policy on "binding RFCs". I would see the current problem as a weakness in the current RFC procedures that needs to be addressed. I would wholeheartedly support a review of those weaknesses and to improve the current system. -- RA (talk) 17:29, 17 January 2012 (UTC)

Binding content discussions. Section break 2

 * Comment: Perhaps it is helpful if an extension of the Palestinian comments above could be considered. It seems what is really wanted here is something a bit different than an RFC. It seems it's like arbitration plus dispute resolution, beginning with an RFC-like process. What seems to be the main point is the cessation of edit warring and disruptions. I think what is sorely needed is admins and other active editors who care enough to meet the urgent challenges, not some weird new process. For example, something of an orphan article can flare up within hours; in that case, no one else wants to look at it because it may be two or three editors disagreeing. Do two or three editors deserve to be ignored because they are so few? No, but that is what happens and then not even an RFC is successful. It's not a party when no one shows up! Then on the other hand you have, as so many have commented, the biggest blow-hard winning a one-sided argument.-- Djathink imacowboy  12:55, 17 January 2012 (UTC)
 * ASCIIn2Bme (talk) 14:14, 23 January 2012 (UTC)
 * Oppose in its current/stated form. I see no problem with all RfCs being binding for a set period of time, but creating a discussion where a policy-level enforcement is to be binding forever is problematic. Can ArbCom overturn this? Jimbo? Who really decides? What happens if circumstances change? It seems to me that an ArbCom-level board would best sort these kinds of problems out. Buffs (talk) 22:09, 17 January 2012 (UTC)


 * You may have missed it, but I didn't say that results of binding RFCs would be binding forever. It clearly outlines that it would be for a period of time, and if real world changes occured, the outcome could be dissolved. Arbcom-style committees for deciding on content has been shot down repeatedly in the past, thus the idea for a community based process to do so. (I haven't ignored other comments, just rather busy and I want to give them decent replies, at this point I don't have the time to give well thought replies to these comments. Cheers. <font face="Verdana"> Steven  Zhang  Join the DR army! 22:31, 17 January 2012 (UTC)
 * Your essentially asking for a certain version of an article be kept. This is just another form of page protection.  And your asking editors to vote on the position they like best.  This goes against the core base of Wikipedia.Curb Chain (talk) 19:00, 19 January 2012 (UTC)
 * No, I am not. This sort of thing I see mostly utliized in naming disputes, but it could also be used in other disputes with a small scope. There would not be any X vs Y version of the whole article discussions. As for such a binding RFC being contrary to the core principles of Wikipedia, I would say that we need to consider that discussions where editors dig their heels in until their opposition gives up is not how we want things to go around here. At times, consensus = who gives up last. People leave over this. I hope this process puts a stop to that. You are entitled to your opinion, of course, but so are others. Have a good day. <font face="Verdana"> Steven  Zhang  Join the DR army! 20:28, 19 January 2012 (UTC)
 * But then you are instituting voting. RfCs have little participation.  That is a fact of life.  One person proposals a side of a position.  There's no consensus or supporters of this single person's opinion.  Too bad.Curb Chain (talk) 20:59, 20 January 2012 (UTC)


 * Support. The current system, while still delivering some good, simply doesn't live up to the show. Increasingly it is getting difficult to get participants into discussions to have an argument/conversation that leads to a visible consensus. And, now wonder discussions are being closed after long-drawn silence as no-consensus, with issues remain unresolved. In case of POVs the stalemate can become silly, even harmful. Also consider that the other option - DR is too heavy for many of those cases. Aditya (talk • contribs) 15:06, 19 January 2012 (UTC)
 * Oppose RfCs are a consensus-based mechanism. When that breaks down, things move to ArbCom.  I'd rather trust elected Arbitrators to decide contentious issues, rather than having a binding RfC, in which the quality of the closers is randowm & unknown, and there is no accountability -- especially considering how difficult it is to get someone desysoped for anything except the worst possible offense.  If Arbitrators don't perform as expected, we don't vote for them next time around, if the closer of a binding RfC abuses the position, what's the mechanism for removing them from that position in the future?  Effectively, there is none. Beyond My Ken (talk) 05:53, 21 January 2012 (UTC)
 * Eeh, but my experience of arbitrators is that they're self-satisfied know-it-alls with an exaggerated sense of their own infallibility and of the inferiority of all other forms of human life (no offence). They reach their decisions through a private gossipy mailing list, and are not interested in amending them (except through a long and tortuous process) when mistakes are pointed out. And you can't vote them out - you'll be outnumbered by those who just vote the established arbs back in every time (most voters won't have been paying much attention to the details of any decisions, any more than real-life voters pay attention to the details of the laws that their electees pass). So in practice, ArbCom is no more accountable - less, I would say - than admins (most of whom are at least prepared to talk to editors and explain and perhaps even modify their decisions). But in any case, given there isn't any perfect solution, and that ArbCom has limited capacity, and that two (or three) instances of wise but possibly fallacious judgment are better than one, it seems clear to me that it should go (1) admin decides; (2) possible appeal to other admins; (3) possible appeal to ArbCom. (And if the original admin was way out of line - I mean, really incompetently wrong - ArbCom could then consider that admin's future as well as the issue itself.)--Kotniski (talk) 14:00, 21 January 2012 (UTC)


 * Support with changes I agree that situations (Israel settlements, Ireland naming, Taiwan) were pointed in which the discussion turned or was in danger of turning intractible. In these discussions the consensus approach has a hard time yielding results. Heavily structuring these discussions might solve this by stronger involvement of neutral parties which guide the discussion. Furthermore, by ensuring all the elements are presented in a structured way this would prevent or at least hamper people repeatedly reintroducing arguments or performing other non-constructive discussions. However, I do not see why this should be in the form of lawyer-like 'binding RfC' system. If the consensus system has broken down because of user behavior, ArbCom should be sufficient. If the consensus system yields no results because of the many aspects of the discussion, heavily structuring the debate would suffice. If the consensus system yields no result because of conflicting sources, a voting system or binding RfC proces should be avoided and a way should be found that addresses the multiple truths. This might be hard in the case of naming disputes, but I am confident that a structured debate would in the end yield acceptable results (see Talk:Republic_of_China). A binding RfC process, on the other hand, would be too lawyerous. Jhschreurs (talk) 09:19, 21 January 2012 (UTC)
 * Oppose at this time There doesn't seem to be a significant demand for this type of dispute resolution, as there are alternatives such as administrators' noticeboard, third opinion, regular request for comments, mediation, etc. Based on the proposal, it appears that it may take a lengthy amount of time for the entire process: certifying, discussing then, having 3 editors make a binding decision - if a dispute is lengthy, overlapping and confusing, it will take quite awhile to decipher and filter through the problems for each viewer. I understand that a structured RfC will make it easier to read but, like any other dispute resolution form it takes some time to familiarize and understand the process.  Whenaxis  about &#124; talk 02:57, 22 January 2012 (UTC)


 * Comment As a user who was banned by the community in July 2007 in just 5 hours 11 mins for being a "POV pusher and harasser", I agree that dispute resolution needs to be improved. But I think that it needs more discussion, in order to ensure that it addresses certain issues. Based on my own experience, I feel that existing dispute resolution discussions tend to be "free-for-alls", full of acusations and rhetoric. Here is what I believe needs to be addressed.
 * 1) All statements must include diffs. In my case, a few editors noted that they were "having a hard time finding a single shred of evidence against" me. If I am a "POV-warrior", it will be trivial to include several diffs. Despite an editor requesting whether "anyone here provide a single diff", none were forthcoming.
 * 2) Some editors noted there was no "due process". Such a process must be fair and timely. One of the reasons the nominator gave for discontinuing Community Bans, was that mine was an "amazing example .. five hours and eleven minutes [..] who finds that rather unfair"
 * 3) Due process must include consideration of points made. For example, during the little time I had before being banned, I had expressed my concern that the person I was alledged to have harrassed, was actually a sockpuppet. At least one editor commented that there as some "serious explaining to do". Subsequently, it was confirmed that the editor I was supposed to have harassed, was indeed a sockpuppet (one of at least 4) that were being used abusively. At the time, my comment was effectively ignored, as has been my attempted to get any Admin to assess my concerns, subsequently.
 * I believe that my case could have been resolved more efficiently if there had been due process, that included a couple of impartial Admins (or whatever). --Iantresman (talk) 18:11, 22 January 2012 (UTC)


 * Commentary: It seems no one should be needed to assess whether consensus exists or not. Problems begin after a "well-sort-of" consensus is approached, but still has dissenters. I don't think admins are listening to the dissents, and I see too many admins simply excusing bad editorial behaviour. That is, when admins bother pay any attention to a problem at all. I had a problem with an article and got echoing silence. Why not just let it all keep failing that way! What I'd love to see is admins being taken to task for failing to help in the first place. Add to that some binding enforcement to keep editors from posing as admins, (which I have seen from time to time).-- Djathink imacowboy  22:55, 22 January 2012 (UTC)
 * WP:BRD is the mechanism when to evoke when you have a problem and get echoing silence. This is not a failure.  This is actually successful in the 8 years? it has been in existence.Curb Chain (talk) 23:56, 22 January 2012 (UTC)
 * Hear hear. I am familiar with one particular editor who, (a) it was acknowledged, was given "leeway", ie. break the rules with impunity, and then I was penalized for complaining "ie. Wikilawyering". (b) When the editor complained that they should be given more freedom to "fight", when I responded that that wasn't the Wikipedia way, I was given a ban for harassing them. (c) I've had an Arbitrator endorse the editor's incivility towards me (d) And when the editor claimed to be a professor, when in fact they were a student, there was no criticism, and another editor continued with the pretence sometime later.
 * Sometimes I think it would be useful to raise issues anonymously, so the facts can be assessed without the personalities. --Iantresman (talk) 00:00, 23 January 2012 (UTC) (Original posting date)
 * Note to (a) "you should not edit or delete the comments of other editors" (b) I assume this was in error, as you gave no explanation in your edit summary, and you did not notify me of your reason. I've restored my comment, which I feel is relevant to the ongoing discussions. --Iantresman (talk) 10:18, 23 January 2012 (UTC)

I think that this is a good proposal, and a needed part of the process. The thing is, we already sometimes have Binding RfCs, it's just that it isn't particularly obvious--in effect, many RfCs that occur after an Arbitration decision are binding RfC's, because an uninvolved admin can easily argue that raising the same topic another time in spite of the RfC is tendentious editing in violation of the sanctions (see, for example, this RfC close on Talk:Senkaku Islands regarding the naming of that article). I'd love for us to have this process available as an intermediate step between mediation and arbitration, especially in cases where the behavior of participants isn't actually the main problem. While I don't know much about the details, isn't it possible that Tree shaping could have been solved by this approach? I understand the concerns above that binding RfCs can actually increase tensions/battleground mentality in some cases. But having no means whatsoever seems to me to be actually worse. Furthermore, it the current state of affairs gives enormous power to those who have the ability to control their own tongues while maintaining an entrenched position; they may be able to outlast the more neutral editors, and sometimes even push them into outbursts that show up as "behavioral problems" and thus lead to sanctions. Qwyrxian (talk) 03:22, 23 January 2012 (UTC)
 * Oppose as written. Since it has come down to a vote, I'll register my opposition for the reasons stated here and on the proposal's talk page. ASCIIn2Bme (talk) 14:08, 23 January 2012 (UTC)
 * Support, given that I don't know what I'm talking about. - Dank (push to talk) 19:21, 23 January 2012 (UTC)

Comment. Something needs to be done. More articles and less editors will mean many lower-quality articles unless content disputes are resolved more efficiently in a less time-consuming and abusive way. Editors are leaving for various reasons. See: User:Timeshifter/Unresolved content disputes. Some editors are being driven away by the unresolved content disputes. The number of active editors might actually start rising again if we find ways to more efficiently and fairly resolve content disputes. --Timeshifter (talk) 20:13, 23 January 2012 (UTC)
 * I'm one who is currently about to leave (possibly not for good) because we just don't have a system in place for making sensible editors' consensus decisions actually happen against the opposition of the drama queens, the edit-warriors and the "nothing must ever change" brigade. Too much of a waste of my time and nerves staying around here.--Kotniski (talk) 23:26, 23 January 2012 (UTC)
 * I'm an editor who actually left. I absolutely refused to edit Wikipedia at all for a year or so, and for several years after that refused to log-in, feeling, quite correctly, that it would lead me to get more invested in things than editing as an IP.  But I left because I was quite frustrated by the proliferation of rules on Wikipedia, which made it harder to be a casual editor.  Getting back to the present, I'm already unhappy with the bright line rules we have, such as the 3RR. Having bright-line rules seems to remove the requirement to assume good faith; using the aforementioned 3RR, I've noticed that on the ANI editors and admins alike can be quite sanctimonious ("You broke the rule, there's no excuse.") which is likely to lead to more bad blood than simply letting a couple of opinionated editors have a pissing contest.  In the case of non-binding RFCs where consensus happens, the bulk of the editors can wear down a few opinionated ones.  This happened on Arab Spring a few months ago when a couple editors wanted to drag the AI conflict into it, and while they filibustered and revert-warred, the rest of us reached a consensus without them, and simply by our numbers enforced it until they got tired of it. Also it seems to me that this will give administrators power over content when closing a binding RFC where consensus isn't clear. --Quintucket (talk) 12:24, 24 January 2012 (UTC)


 * Oppose This only invites lynch-mob mentality, pedantry, wiki-lawyering and wowserism. Voting on content can quite simply mean more people with a stupid POV win. Just like in populist politics. Everton Dasent (talk) 00:47, 24 January 2012 (UTC)
 * Forgive me, but I do not see how this proposal could do any of the above, indeed it would do the opposite. We already have voting on content in many talk page discussions where users have differing opinions. This process is by no means a process to just vote, it's one to determine a consensus which often happens with various forms of votes (RfA, XFD, requested moves). This would introduce structure into intractible disputes, and perhaps be the intermediate step between mediation and arbitration. There would be no room for wikilawyering, because users would only be able to present information on option W vs option X vs option Y etc, and then members of the community (as well as the parties) opine on the discussion. The notification of the discussion being in progress (say through a watchlist notice) will help offset this too, because uninvolved users could analyse the information supporting the alternative options and weigh in. That's really how AfD should work as well, because often comments at afd resemble "Delete/Keep - Meets/fails [random policy]." But I think a demo of how a discussion would work will be of use. <font face="Verdana"> Steven  Zhang  Join the DR army! 01:47, 24 January 2012 (UTC)


 * Support: I've been in my share of content disputes in the past, and things always got pretty ugly when they went on. There were always at least a few editors who got blocked/banned, or otherwise had their reputation as Wikipedians spoiled. In my opinion, any form of contract/agreement/truce/any other term that could summarize it, that would help prevent future conflicts from getting that advanced is a good idea. Wilhelmina Will (talk) 11:14, 24 January 2012 (UTC)


 * Oppose. One problem is the quality of the commenters. On one topic (Venezuela, for example) there was a "lurker" who never edited but always commented favorable to the regime. The quality of commenters, generally is suspect. I hope for comments from article editors or try to solicit Project editors. Anyway, we have mediation for this sort of thing. Admins have been pretty good about "commenting" to single dissenters that "they appear to be in the minority" to avoid disruption and further escalation. Student7 (talk) 16:15, 24 January 2012 (UTC)


 * Oppose: This would be against WP:CCC. Also, Wikipedia is an encyclopedia that everyone can edit (or atleast use edit requests to get the content in on protected pages), this will completely lock down content for newcomers who might have better arguments and reasons than given in the RFC that is "binding" the content to the article. There's always place for fresh minds to reopen a discussion. On the contrary an option should be made in the mediation to open an RFC as a subpage of it or something which will help the mediators to get a consensus. -- <b style="color:#060">lTopGunl</b> (<b style="color:#000">talk</b>) 16:28, 24 January 2012 (UTC)


 * Oppose. If we want to remove something from the WP:BRD cycle, it should be done by people specially appointed by the community for the task, namely, ArbComm. We don't want to give the editors who show up to a discussion that kind of power. Nor do we want content discussions to be entirely about whether a binding discussion applies, what was decided, whether a given edit violates it, etc. And in order to be binding, the discussion would have to be enforceable. This means that administrators would have to decide blocks etc. on the outcome of these discussions, and so would have to be able to interpret these discussions. Right now, admins have to master policies and guidelines in order to carry out their duties. Binding discussions would require that they master policies, guidelines, and content discussions. This looks like a mess. As bad as the current situation is, this proposal would make them worse.  RJC  TalkContribs 16:00, 26 January 2012 (UTC)


 * Oppose- I don't see the need, and I'm concerned that this proposal would make the encyclopedia an even more hostile, bureaucratic and legalistic place than it already is. Besides, what ever happened to WP:CCC? Reyk  <sub style="color:blue;">YO!  22:14, 26 January 2012 (UTC)
 * WP:CCC != giving POV pushers a free pass and allowing circular discussion to continue until one "side" gives up and the other side "wins". The need is caused by disputes like Senkaku Islands and Abortion. I don't see this as an often used option, but I think it would be a necessary one. Consider, this would be a last resort before ArbCom, after mediation has failed. I invite those opposing to read over the Senkaku Islands dispute, as well as the Abortion arbitration case, and offer an alternative. Admins close discussions all the time. If y'all hate this idea so much, think of an alternative. And an alternative is not "free reign for POV pushers". <font face="Verdana"> Steven  Zhang  Join the DR army! 22:31, 26 January 2012 (UTC)
 * Sorry, but I just don't see the need to permanently lock down an article's content and presentation. That runs afoul of WP:OWN as well as WP:CCC. If you have problems with the behaviour of POV-pushers, you know the way to RFCU. You might see your proposal as an absolute last resort; I see it gradually almost immediately morphing into the first resort for people who think they currently have the numbers but want to nip any possible dissent in the bud. Reyk  <sub style="color:blue;">YO!  23:15, 26 January 2012 (UTC)
 * But it can't be the first resort. A binding content discussion would only be allowed if all other forms of dispute resolution has failed. A prolonged mediation would be a requirement for a binding discussion to even be considered. <font face="Verdana"> Steven  Zhang  Join the DR army! 23:18, 26 January 2012 (UTC)
 * And I disagree. I think it would cheapen and demean all prior attempts at dispute resolution. I say "first resort" because it will mean people become less likely to engage meaningfully in the nominal first steps, when that might require compromise. Instead, they could just hold out for a vote. Reyk  <sub style="color:blue;">YO!  23:28, 26 January 2012 (UTC)
 * Do you have an alternative idea? <font face="Verdana"> Steven  Zhang  Join the DR army! 23:31, 26 January 2012 (UTC)
 * Since I'm not as adamant as you that there's a problem that needs fixing with mechanisms that aren't already there, I guess my alternative is to leave things as they are. Reyk  <sub style="color:blue;">YO!  23:40, 26 January 2012 (UTC)


 * The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page, such as the current discussion page. No further edits should be made to this discussion.


 * Since the discussion is already closed, can someone please remove the notice that appears on our watchlists? A Quest For Knowledge (talk) 01:34, 30 January 2012 (UTC)
 * Yes check.svg Done Anomie⚔ 02:35, 30 January 2012 (UTC)