Wikipedia:Pending changes/Feedback/Archive 3

Unpending suggestion
How does one suggest a page be removed from the Pending changes trial? For example Total Drama World Tour was placed under pending changes but due to the frequent completely unsourced edits being made, that page appears too frequently in the log and I cannot see what benefit it receives. Meanwhile I have added Article issues but I have also proposed that page be semi-protected instead. -84user (talk) 20:19, 30 June 2010 (UTC)


 * Requests for page protection CycloneGU (talk) 03:10, 1 July 2010 (UTC)

Should start ONLY with biographies
There's just too much work trying to figure out if something is kosher or not on other articles. Couldn't bring myself to even look at the two I listed (Rome and Roman Empire). But lots of BLP stuff doesn't pass the smell test. I'd be much more likely to watch a couple dozen bios than any thing else except articles I have too strong a POV on to be a watcher. Too few editors with too little time for all other topics - except perhaps most controversial, and then you often have to watch out for watcher's POV. Baby steps! CarolMooreDC (talk) 00:32, 1 July 2010 (UTC)
 * I agree with the pending changes and see your point. A recent vandalism I reverted here on a biography really drove me to agree with this. I didn't have the article on my watch list and was just casually reading it. The vandalism therefore surprised me as a reader and I had to look further. For almost six hours that was on the article! I think the process may be a pain for less-visited/monitored articles on Wikipedia though.--NortyNort (talk) 04:59, 4 July 2010 (UTC)
 * As I say below, it's also good for problem edit warring pages in general, many which may include BLP among other policy issues. It worked initially for the BLP I was having most problems with, but the new editors may have gotten worn down by the same old tactics of the same old POV warriors, so Don't know if it is permanent solution for the problem... CarolMooreDC (talk) 04:22, 4 August 2010 (UTC)

Usability poor
The interaction and usability of the process is poorly thought through and reaction time of the server is lacking as I have had more conflicts than successful approvals.BGinOC (talk) 03:15, 1 July 2010 (UTC)
 * Yes, the server has been VERY slow lately. However, while there may be a lot of conflicts now with many more (sample) articles than editors, it will be just the opposite once this was imposed on all articles. And we can bet people will drift towards articles they already are watching or editing. Just human nature :-) CarolMooreDC (talk) 19:39, 1 July 2010 (UTC)
 * I have the same issues. I haven't figured out exactly what is happening, but I notice slow response times and sometimes loading pages that show changes as accepted, non existent, and then pending, all within short periods of time and in random orders. Shadowjams (talk) 01:29, 4 July 2010 (UTC)
 * Same slowness noticed by me.  Anna Lincoln  09:41, 6 July 2010 (UTC)
 * The retrieval time is slow and most of the times I've made an effort to examine a pending edit another Reviewer has in the mean time completed the review process while I was looking at the article. So far it has wasted my time. -  Steve3849 talk 21:26, 6 July 2010 (UTC)

Make it possible to hide the watchilst header notice when there are pending changes
I don't think the notice ontop watchlist should be mandatory visible always. → Aza Toth 14:45, 1 July 2010 (UTC)

Comment while accept/rejecting pending revision
I left a comment (in the comment box besides Accept option) while accepting a revision of India, but my comment is not seen in the history of the article. -- Redtigerxyz Talk 17:43, 1 July 2010 (UTC)


 * The awnser to your question is further up on the page at: Pending changes/Feedback Sumsum2010  · &#32; Talk  · &#32; Contributions  17:47, 1 July 2010 (UTC)


 * My point is: shouldn't it be seen in the history? -- Redtigerxyz Talk 06:36, 2 July 2010 (UTC)


 * Oh. That would be much more helpful than in the page they're in. Sumsum2010 · &#32; Talk  · &#32; Contributions  19:56, 2 July 2010 (UTC)


 * Agreed, most people probably do not know where their comments went, and most people do not look at article logs. — MC10 ( T • C • GB •L)  16:49, 3 July 2010 (UTC)

Thoughts you should take a look at
I'd just like to say I think it generally works well. There are concerns that it will lead to moe biting and a more closed community, but I think it's the opposite-- with some exceptions:


 * Highly popular searches/vandal targets should use the standard semi prot. Examples: Barack Obama, evolution, Lady Gaga, Chocolate, Justin Bieber, Miley Cyrus, marijuana, etc


 * Some other pages are more ideal for Pending changes (PC) whereas semi-prot is too 'wikipedia exclusive'. Examples: food, pizza, all BLPs (except highly popular ones) and I think all high school pages as they are frequent vandal tagets (debatable).

I feel that if implemented this way, anons will feel more encouraged to participate where appropriate, while still discouraging vandalism. Look at it this way: if you want to look at the rate of vandalism as a measure of how effective page protection and blocking are, it's failed. There is vandalism all the time, every minute. I feel PC will help ruin the "look I change anything I want and it's effective immediately therefore I should keep doing it" vandal mentality. Thanks and comments are welcome! – Tommy  [ message ] 18:20, 1 July 2010 (UTC)


 * I agree that some pages are better left simi-protected and others are better left with PC. This would keep vandals away yet not discourage people with legitimate contributions. The vandals will know they cannot vandalize some pages but also know that vandalism on others will be reverted quickly. This would make the vandals think "These pages I cannot vandalize and on the pages that I can vandalize it won't be there long. So why should I vandalize anything?!" and then hopefully they will be come good editors. Sumsum2010  · &#32; Talk  · &#32; Contributions  02:44, 2 July 2010 (UTC)

Missing Accept in Diffs?
I've just used Pending changes on Total Drama World Tour - the interface is not very intuitive. When I clicked on "review pending changes", it gave me a diff. But what I really wanted to do was to "reject" a series of changes that were broken and never going to work (see Pending_changes/Feedback). (These changes break tables and it looks like it would just be easier to start again). So, two questions:
 * Is it possible to improve the interface so that a list is presented rather than a diff? Or can the diff have an accept/unapprove/reject button?
 * Is there an easy way of rejecting or undoing an edit? (it looks like the Total Drama World Tour article is stuck with its broken edits until someone accepts them all then undoes them). twilsonb (talk) 11:59, 2 July 2010


 * Well, if you are using Twinkle, then you can revert back to the last good edit. And you don't have to have them accepted before you can undo them, you can just undo them when you see it, and someone will accept your edit. I hope this helps.  Hi 8 7 8   (Come shout at me!) 02:03, 2 July 2010 (UTC)


 * Well, undoing was a problem due to intermediate revisions, and the complexity of the changes (I didn't want to pick them out one by one). And in this case my edits will be automatically accepted! twilsonb (talk) 02:12, 2 July 2010 (UTC)


 * What actually happened soon after I posted was that the broken revisions were accepted by User:Jcrook1987 here, then fixed by me here twilsonb (talk) 02:12, 2 July 2010 (UTC)


 * I still think it would be nice to have a die die die button for edits that should never see the light of day. But I imagine that's covered by UNDO and reject - and no diff'ing process is perfect. twilsonb (talk) 02:17, 2 July 2010 (UTC)


 * And WP:ROLLBACK. :) Probably the best way to.  Hi 8 7 8   (Come shout at me!) 02:55, 2 July 2010 (UTC)


 * But you can only roll back one edit, I think. Or a series by one user.  Not over several users.  Correct?  CycloneGU (talk) 03:00, 2 July 2010 (UTC)


 * Correct. But then you can use Twinkle to revert back to x revision, regardless of how many different editors, if there have been multiple editors.  Hi 8 7 8   (Come shout at me!) 03:04, 2 July 2010 (UTC)


 * Thank you. I was just about to do that. :)  Hi 8 7 8   (Come shout at me!) 03:10, 2 July 2010 (UTC)


 * You can also use the "revert" functionality in popups. Just highlight over the link of the revision you wish to revert to, and then press the "rv" link. It will revert directly back to that revision. — MC10 ( T • C • GB •L)  05:04, 3 July 2010 (UTC)

Why this is a bad idea
Ok, I've watched this abortion that is pending changes for the last two weeks I have finally formulated an opinion on why this is bad idea. Actually, I have formulated several opinions about why this is a bad idea. Allow me to share them with you: Its things like this pending changes trial that are driving off potential contributors to the site in droves; its the reason why Wikipedia is sinking so to speak. I for one remain firmly convinces that this is unnecessary, and urge us to pass on the pending changes trail. From where I sit, it just is not needed. TomStar81 (Talk) 06:27, 4 July 2010 (UTC)
 * For starters, no one has any idea when to implement this new protection system. We have no policy or guideline page in place to recommend when these changes should be applied. In the absence of concrete policy or even a few basic guidelines I have seen this protection applied everywhere. Its becoming an epidemic since admins seem to by shifting to this option rather than semi protection or protection; when in reality a number of pages this is deployed on probably do not even protecting the in first place.
 * These was never any training offered for this. Not that I think that admins or bureaucrats or any of the other groups here should be have to be trained on each and every tool, but for a trail it would have been nice to require us to read a walk though or to go to a designated place and try the tool out a certain number of times to get a feel for the new system. I have no desire to use something that no one has bothered to set guidelines for or to train us on; that's a quick way to find myself hung out to dry for a bad call even if it is made in good faith.
 * I recall no invitation to vote on the matter of moving forward with this trail. For that matter, I do not even remember being notified that this was being worked on. I may not be interested in using the tool, but asking if I would like to opt in for a trail run would have been nice so that we could sift out those who want to try the tool from those who would rather not have such a privilege.
 * We have enough bureaucracy here as is. We don't need another tool to limit/prohibit people from contributing, what we need now is the same thing we needed originally: the community. It should be that the community watches the articles for vandalism and reverts such instances of it on sight. Is it not we who watch the watchmen? Tools could never do for Wikipedia what a brain can: think and articulate a new revision in the terms by which the article provides for the inclusion or exclusion of certain material. The day we give up the power of policing the articles to these tools is the day we declare to the vandals that we are defeated.
 * This defeats the purpose of our encyclopedia. Our motto is "The free encyclopedia that anyone can edit." Jimmy's quote, as it frequently appears, is "Imagine a world in which every single human being can freely share in the sum of all knowledge." (or some variation thereof). How does blocking certain edits from the public eye until they are approved fulfill this mandate? The simple answer is that the two are incompatible with one another. Either we as a community believe in our statement that anyone can edit and do so without having to be cleared to do so or we believe that no one is allowed to edit unless they receive permission to do so first. Effectively, this has the disturbing potential to turn Wikipedia into Citizendum.
 * I'm not going to comment on most of what you've written, but I'd just like to point out that pending changes was not intended as an additional layer of beauracracy to be laden on Wikipedia editors. Rather, it was meant to make it easier to edit. The idea was for the bulk of pending-changes-protection to be used on pages that had been under long semi-protections already. This would allow IPs and new users with no understanding of Wikipedia to submit an edit in the same way they would make one, rather than learning how to use a talk page and the editsemiprotected template. So yes, pending changes blocks or delays edits; but the alternative is blocking them all without review. There was never support on the English Wikipedia for anything like the German system, where the entire encyclopedia is essentially under pending-changes (flagged revisions). Someguy1221 (talk) 09:37, 4 July 2010 (UTC)
 * TomStar81, I totally agree with you, especially your points about bureaucracy, guidelines, and our mission.  Them From  Space  18:09, 8 July 2010 (UTC)
 * This is a trial only. It doesn't require policy or training because neither the rules nor the lessons have been revealed yet.
 * Wikipedia is a tool. The community you rightly credit is built upon javascript and wiki-markup and Twinkle and Special page logs and a million efficient blessings of code.  The two are not in conflict.  Ocaasi (talk) 18:04, 9 July 2010 (UTC)


 * ... I don't see how any of these points are justified as it was the hope of this project to create guidelines for all your concerns OUT of this trial process. Writing strict guidelines before you can even try something is like volunteering to be retarded. You'd have no idea what needs to be controlled. Mkdw talk 23:36, 11 July 2010 (UTC)

Naming of OldReviewedPages
Special:OldReviewedPages is a lousy name for the page, IMO (what's "old" about the pages themselves exactly?). Something like Special:UnreviewedChanges would be better. --Cyber cobra (talk) 07:25, 4 July 2010 (UTC)


 * Mmm, yes, I was equally confused by this choice of name. Orphan Wiki  13:57, 4 July 2010 (UTC)


 * I figure this will be revisited upon full implementation, if that happens. CycloneGU (talk) 17:10, 4 July 2010 (UTC)


 * I'm not he only one confused by the naming! Sumsum2010 · &#32; Talk  · &#32; Contributions  18:58, 4 July 2010 (UTC)


 * Definitely not the only one. I agree that the name is not logical, I expected Special:PendingChanges myself.  --Joshua Scott (formerly LiberalFascist) 04:17, 5 July 2010 (UTC)


 * I also agree with this one. Note that mediawiki special pages can have several aliases, so all these proposals (and the current name) can be used. --Waldir talk 09:48, 7 July 2010 (UTC)


 * I agree. Why don't they call it what it's actually labeled visually? Pages with pending changes or like Joshua says, "Pending Changes". Mkdw talk 23:34, 11 July 2010 (UTC)

Accept pending changes
A warning to reviewers - It seems that if you tick "accept pending changes" when you edit a page, all previous versions are tagged as accepted by you. This could include vandalism that was added and then removed. This isn't a problem in terms of what is shown to IP users but it does mean that someone can be staring at an old page full of swear words and be told that it has been accepted by you! I have discovered this to my cost. Yaris678 (talk) 20:17, 4 July 2010 (UTC)
 * This shouldn't be the case, nor have I seen this anywhere. When you accept a revision, you're only accepting one revision.  Can you show me an example where what you described happened? -- RobLa (talk) 21:06, 4 July 2010 (UTC)


 * I see one case where you approved and subsequently unapproved an edit that shouldn't have been accepted. It was just a single edit, and it seems pretty clear when I look at. Perhaps you simply slipped and and approved it accidentally. Reach Out to the Truth 04:34, 5 July 2010 (UTC)


 * Well I have had a few problems that could have been down to a fat finger so I was especially careful this time. If you look at the revision history for Arsenal F.C. you can see a vandalised page that I supposedly approved and now can't unapproved.  My recollection was that I
 * clicked on review
 * saw that it was vandalism
 * clicked on undo
 * checked that the change was the one I wanted
 * ticked the "approve pending changes" box
 * pressed save
 * I think the undo by SuperHamster must have happened before I saved my own undo and now it just shows that I have approved the vandalised page!
 * Yaris678 (talk) 11:57, 5 July 2010 (UTC)
 * I have just unaccepted this edit. Presumably you could have done this as well. &mdash; Martin (MSGJ · talk) 11:20, 8 July 2010 (UTC)
 * Possibly I could have done. I didn't realise at the time that this was done by doing a diff on the revision you want to unaccept.  I was trying various other things to no avail.  Maybe "how to unaccept" should be documented somewhere like help:Pending changes.  Yaris678 (talk) 11:54, 8 July 2010 (UTC)

Barriers to entry
I want to voice my serious concern that this policy will alienate new IP editors, and deter them from registering. Does anyone remember that we encourage people to be bold in updating pages? I think this is a barrier to that boldness. I think it is also a barrier to the learning curve of Wikipedia, because I am also concerned that this function will lead to some reviewers and admins rejecting IP edits based on 'poor grammar' and unsourced edits. Has anyone thought about notifying these IPs of why their edit was rejected, and the burden that will come with that? Who can guarantee that we are going to be able to do that when, in my early editing days, I saw alot of IPs who didn't even get warning messages on their talk pages when they committed blatant vandalism?

While vandalism is a chronic problem and I'm not about to deny that, I think this proposal would be a greater evil when it is Wikipedia's goal to be bold, and teach editors of how to constructively edit here with neutral and cited material. And the only way we can continue creating a reliable, free encyclopedia is making the editing experience more welcoming so as to encourage more users and more editing ,and addressing these faults in quality directly at the source, not erecting new barriers around our encyclopedia to prevent some vandalism and trying to filter the best edits.

13 million registered users is quite an accomplishment, but its not even half of the population of my country, and this is open to the world! I can see the growth of that userbase start to decline if you implement this, and that will mean less editors, more burden and burearacy for our existing users, less content contributed, and therefore a worse encyclopedia that anyone*††‡ can edit. -- Natural  R X 19:14, 27 June 2010 (UTC)


 * I don't see how this is a barrier to editing when the alternative is semi-protection, which doesn't allow editing at all. Semi-protection is what discourages IPs and non-AC accounts from editing—I know I skipped a couple of possible edits when I was an IP years ago.  I believe that the pending changes will—given time—discourage vandalism once it becomes more widely known that the changes don't show up immediately.  If they have a positive contribution, I don't think a 5 minute delay in displaying it would discourage the IP from making the edit.  --Joshua Scott (formerly LiberalFascist) 04:15, 5 July 2010 (UTC)


 * To edit at level one, all people have to do is register. If someone wanted to be bold like you said why wouldn't they register. A vandal on the other hand wouldn't register just to vandalize (unless they're a sockpuppet). Besides, is 13 million the users on the English Wiki or the users on every wiki? There's a difference. Derild  49  21  ☼  16:03, 5 July 2010 (UTC)


 * Agreed with the comments above. If this is an alternative to protection, then surely this is less of a barrier to editing. At least they get a chance to submit an edit this way, unlike with protection. Orphan Wiki  09:04, 6 July 2010 (UTC)


 * I just tried making a random edit as an IP to a page covered by the PC trial, and I must say my perception was exactly the opposite. Weren't I familiar with Wikipedia's URL scheme, I wouldn't have noticed that I was actually presented a different view of the page, rather than the one everyone would see. I mean, there's a warning on top, but who reads warnings anyway? Most people will TL;DR them even if they are two- or three-liners as this one was. Thus, this greatly eased my concerns about the instant gratification aspect of editing the wiki (And though the same effect would apply to vandals, meaning they could take a while to realize their edits were not visible by default to the world at large, it also helps in that they typically won't insist if they think the vandalism worked. And even if they do, the harm is greatly reduced since 99% of the readers won't see it anyway). I'm thus very optimistic about the PC trial and hope to see it widely used soon. --Waldir talk 09:42, 7 July 2010 (UTC)


 * I concur with Natural RX that the PC proposal would hamper entry and alienate many good editors and contributors. Forcing registration will not stop vandalism. Anyone intent on causing chaos is more than willing to complete the registration process -- hundreds of times, if needed. Forcing registration will also exclude some editors that are against or fearful of any registration process as I have noted in other sections. Additionally, with the rate that new users are joining the Internet, even in the US, the encyclopedia will alienate more of these types of people. The original intent of Wikipedia was that it was an encyclopedia that anyone can contribute to, for the greater good of the end user. Generally, I am against greater restrictions. It is through freedom that Wikipedia became the prevailing Internet encyclopedia.--There is nothing civil about Civil War.Let's Talk! 19:25, 11 July 2010 (UTC)

Auto-accept
I think that a pending change should be autoaccepted by a bot after a fixed time delay (say 48 hrs), as a form of Garbage_collection_(computer_science) for pages who have pending changes activated with no reviewers keeping an eye on them. It's a corner case, but still important. User A1 (talk) 12:58, 6 July 2010 (UTC)


 * I've yet to see a backlog of pending changes, meaning if one pending change is lingering longer than others, there's usually enough of a collection of reviewers who will sort it out together. Maybe this is where we could use the as-yet pretty much dormant noticeboard or just contact the Wikiproject the article falls within? Orphan Wiki  13:04, 6 July 2010 (UTC)


 * While it's difficult to say how this effort will scale and pan out over time, there's one way in which I think pending changes is less likely to build a long backlog then we might expect, in addition to folks noticing pending changes on their own watch list, I and several other editors I'm sure make use of the list of pending changes at Special:OldReviewedPages. At this point I'd certainly pounce hard on anything marked "over an hour", but I've never seen it happen yet. So long as there's enough "hands on deck" (something made enormously easier by the fact that potentially thousands of editors, rather than a few dozen admins, can actually give a thumbs up to a protected page edit), and so long as there's some feedback as articles get old (as seems to be the case there), I think we have a good chance of doing ok.  I have no problem with there being some sort of action if an article sits for 48 hours (number TBD, but that may even be high), but it is my real hope that it won't be a problem.  In fact, I'll be a lot less satisfied unless this system can hold most reviews under 5-10 minutes.  --  j &#9883; e decker  talk  00:10, 8 July 2010 (UTC)
 * I completely disagree with this suggestion of bot auto accept. The majority of editors are not monitoring articles individually for pending changes but using Pages with pending changes to which I've never seen more than 10 pages in the log and never for longer than 30mins. Mkdw talk 23:33, 11 July 2010 (UTC)
 * Bot auto edit would accept hoaxes on low edited/watched articles. So bot approval would be a very poor thing. Regards, SunCreator (talk) 04:31, 16 July 2010 (UTC)

Ratio of vandalism/junk to good edits
The ratio of vandalism/junk to good edits should be used to judge whether to use pending changes on an article. Moving from semi-protection to pending changes when all the IP edits are schoolchildren messing about (e.g. The Game (mind game)) seems like make work. We should go back to semi-protection and Editsemiprotected in such cases. Fences &amp;  Windows  15:36, 6 July 2010 (UTC)


 * If all the IP edits are people messing around then protecting the page to some degree is definitely better. That would cut way back on the amount of vandalism to be patrolled. Sumsum2010 · &#32; Talk  · &#32; Contributions  16:39, 6 July 2010 (UTC)


 * If anybody is interested, I can make a list of pages with a ratio of reverted edits higher than a certain number (based on the analysis of the recent en-wiki dump). Unfortunately that's not exactly vandalism/junk to good edits ratio (it is very difficult to calculate this one unbiased, for obvious reasons), but still it is a good approximation. --Dc987 (talk) 00:41, 7 July 2010 (UTC)


 * Dc987, this would be wonderful. I'd be very interested in this. -- RobLa (talk) 18:19, 7 July 2010 (UTC)


 * This would greatly help to determine if PC is actually helping. Sumsum2010  · &#32; Talk  · &#32; Contributions  22:41, 7 July 2010 (UTC)

Can you be slightly more specific onto what data do you need? What I can calculate fairly easily are lists of pages with reverts ratios. These reverts ratios can be calculated based on the analysis of the wikipedia dump (the latest one was completed on April/2010). Or alternatively I can download some specific pages and do similar analysis for these individual pages (this obviously would include most recent revisions). I can also specify filters (e.g. do analysis between certain dates, only for pages with a certain number of revisions, etc). In any way the result would be a list like that: (reverts ratio, (total revisions, reverts, self-reverted, revert war, reverted questionable, reverted, regular), title) (0.42941176470588233, (1870, 803, 23, 138, 33, 737, 136), 'Chinese people') (0.42516069788797062, (1089, 463, 29, 27, 16, 470, 84), 'Inner core') (0.42328042328042326, (1512, 640, 27, 54, 17, 565, 209), 'Fudge') (0.41783216783216781, (1144, 478, 12, 31, 17, 428, 178), 'Homo (genus)') (0.41662283008942663, (1901, 792, 25, 77, 22, 713, 272), 'Old age') (0.41607956900124327, (2413, 1004, 46, 123, 33, 1018, 189), 'Ryan') (0.41368680641183725, (3244, 1342, 43, 169, 71, 1357, 262), 'Sloth') (0.41176470588235292, (2210, 910, 31, 83, 47, 906, 233), 'Hydropower') (0.41015330485046492, (3979, 1632, 56, 89, 39, 1421, 742), 'Silver') (0.40968899521531099, (1672, 685, 24, 55, 31, 614, 263), 'Roman mythology') (0.4096774193548387, (1240, 508, 30, 49, 9, 453, 191), 'Architecture of ancient Rome') (0.40959119496855345, (2544, 1042, 35, 103, 42, 1007, 315), 'Pen') (0.4079844206426485, (2054, 838, 18, 40, 17, 652, 489), 'March 10') (0.40516767454645408, (5457, 2211, 71, 202, 39, 1975, 959), 'Biology') (0.405148095909732, (2836, 1149, 34, 151, 44, 1094, 364), 'Bomb') (0.40265957446808509, (3760, 1514, 54, 122, 36, 1316, 718), 'Teacher') (0.40230621396540678, (3122, 1256, 37, 129, 45, 1160, 495), 'Ocean') (0.40182648401826482, (1314, 528, 13, 53, 7, 503, 210), 'Jelly bean') (0.4011764705882353, (2550, 1023, 42, 133, 64, 1000, 288), 'Alex') (0.4011054086063956, (2533, 1016, 39, 110, 63, 861, 444), 'Disease') (0.40098374679213, (4676, 1875, 54, 235, 90, 1748, 674), 'Moose')

This list (given as an example) was calculated using following criteria: --Dc987 (talk) 02:16, 8 July 2010 (UTC)
 * pages subset from the PAN 10 Lab training set;
 * revisions number > 1000;
 * complete page revision history (no filtering by date);
 * revert ratio > 0.4;


 * I think something like: total edits, total reverts and undos, total reverts, total undos, and percents of each except total edits. It seems to me like that is the most important information. Sumsum2010  · &#32; Talk  · &#32; Contributions  02:45, 8 July 2010 (UTC)


 * And, er..., how would you use it? Anyway, I've run the current code (filter: > 1000 revisions, reverts ratio > 0.3) against the complete Wikipedia dump and here is the resulting list http://wpcvn.com/enwiki-20100130.most.reverted.tar.bz --Dc987 (talk) 05:58, 8 July 2010 (UTC)
 * Curiously in the main namespace page 69 with 1140 revisions and 519 reverts (45% reverts) was the 'winner'. Only 87 revisions were 'regular' (not reverts or reverted). It looks like all these unnecessary edits and reverts are caused by the wrong redirect not leading to the most popular page associated with the number 69. --Dc987 (talk) 06:12, 8 July 2010 (UTC)


 * I think the information to have is the ratio of reverted IP edits (or reverted non-autoconfirmed edits, if that can be extracted from the dump) versus total IP edits. If we're still operating on the assumptiong that pending protection is only applied in the place of semi-protection, then what we're looking for are pages that get a mix of good faith IP edits in a sea of vandalism. Most of these should already be protected in some manner, but there might be more out there, like the ones you showed above. Someguy1221 (talk) 06:13, 8 July 2010 (UTC)


 * Reverted revisions made by IP edits / Total revisions made by IP edits > 0.?? AND    Total revisions > 1000 ? --Dc987 (talk) 06:25, 8 July 2010 (UTC)


 * I think it would have to be less than 1, greater than 0. Preferably close to 1. 0 means no edits were reverted; 1 means all IP edits were reverted (and thus presumed vandalism). I have to ask, though: if there is a series of IP edits to an article, and then a reversion takes place to the last edit by a registered user, what that read as all IP edits reverted or only the most recent one? Someguy1221 (talk) 06:50, 8 July 2010 (UTC)


 * Which way do you want? Currently every edit from the series is being counted, but it is very easy to ignore consequent edits made by the same user. I'm not sure it is of great importance though. IMHO what can make a big difference - is filtering by date. Rates of reverts vary with time: some pages are getting protected and rates of reverts drop virtually to zero, some have only been vandalized at a certain timeframe, etc. --Dc987 (talk) 08:13, 8 July 2010 (UTC)


 * Wow, that's really useful. I was suggesting the general principle, not the details of where to draw the line - we'll get to know what ratios determine the best protection levels over time. It'd be interesting to look at the revert levels in the month before semi-protection was used on an article vs average revert levels, that might give us an idea of what the community finds is unacceptable levels of vandalism. Fences  &amp;  Windows  12:05, 8 July 2010 (UTC)
 * Task 1: 'look at the revert levels in the month before semi-protection was used on an article vs average revert levels'. For that please give me a list of page titles and specific time frames that you are interested in (I can only do ~20 pages, that's because these recent changes are not in the last Wikipedia dump and I'll have to download pages history). Please be very precise and specific in your request. --Dc987 (talk) 19:55, 8 July 2010 (UTC)


 * Here's the stat I'm really hoping to see: during the period when a page is placed under pending changes, what ratio of IP edits for that page are reverted versus accepted? That will help us develop some ideas about which pages benefit significantly from the feature versus those pages for which the feature is creating a lot of cleanup sans benefit. -- RobLa (talk) 17:50, 8 July 2010 (UTC)
 * Task 2: 'a page is placed under pending changes, what ratio of IP edits for that page are reverted versus accepted?' Shouldn't that be available as some internal stat generated by the 'pending changes' mediawiki code?  My tool (its open source BTW) only allows me to analyze reverts, if 'rejected' edits do show up as reverts in the page history, it's the same thing as Task 1. --Dc987 (talk) 19:55, 8 July 2010 (UTC)


 * It could be. We have developers to work on this, but they're currently focused on the operational aspects in the time that they have (both of them are part time).  More to the point, working out the exact metrics we want the developers to put in is going to be tricky.  One thing that would be really helpful would be a specification/design doc laying out exactly what the developers should do.  It sounds like this may be your strong suit, so a contribution in this area would be really greatly appreciated.  -- RobLa (talk) 20:37, 8 July 2010 (UTC)


 * I'm thinking something to see which pages get the most vandalism to determine which pages should be semi-protected and which should be using pending changes. Sumsum2010 · &#32; Talk  · &#32; Contributions  19:06, 8 July 2010 (UTC)
 * Task 3: 'which pages get the most vandalism to determine which pages should be semi-protected and which should be using pending changes'. That's more of the database analysis task. I've already said, it is very difficult to do detect vandalism itself in an unbiased way. The only good proxy that I know are reverts ratios. I've already generated these for the english wikipedia, you can get the resulting list here: http://wpcvn.com/enwiki-20100130.most.reverted.tar.bz . If that list is not what you were requesting, please tell and please be very precise and specific in your request. --Dc987 (talk) 19:55, 8 July 2010 (UTC)
 * Thanks for making the list! The only problem is I do not know how to view it. I need some help with that. Thanks Sumsum2010  · &#32; Talk  · &#32; Contributions  23:06, 8 July 2010 (UTC)
 * A .tar.bz is just a compression format. It is quite widespread. Anyway, here is the uncompressed version of the same file: http://wpcvn.com/enwiki-20100130.most.reverted.txt --Dc987 (talk) 05:31, 9 July 2010 (UTC)
 * Thanks! Wow thats alot of information! Sumsum2010 · &#32; Talk  · &#32; Contributions  17:22, 9 July 2010 (UTC)
 * This is a great thread, but I'd like to throw out a preemptive, precautionary comment. Any metric which is identified as the golden ratio for a vandalized article will be prone to intentional tweaking to serve editor biases. I know it's way too early (and cynical) to be thinking about this, but there are heated views out there about pending changes, and some editors would love nothing more than a numerical rationalization for slapping protection on a page they don't want to review.  Otherwise, data is great.Ocaasi (talk) 18:49, 9 July 2010 (UTC)
 * Fundamentally, the concept of having some articles still semi-protected is a grand idea. It's definitely created more work for some editors using pending changes on articles with significant amounts of vandalism. Any sort of rules are prone to abuse but that's where the human element of admins and the community come into play. Mkdw talk 23:25, 11 July 2010 (UTC)
 * Yes, that's what I meant. We should be careful to make statistics our tools, subject to continual interpretation and context, rather than embedding them in the system and expecting that not to result in workarounds (see Campbell's Law). Ocaasi (talk) 07:23, 12 July 2010 (UTC)


 * IP editing should be completely eliminated IMHO. If a potential user does not want to register, or does not have the time, then Wiki can survive without them. Since most vandalism is committed by IPs, it makes merfect sense to withdraw their editing abilities. --GabeMc (talk) 20:15, 21 July 2010 (UTC)
 * You ARE aware that most constructive contributions also come from IPs? —  Jeremy  ( v^_^v Dittobori ) 20:26, 21 July 2010 (UTC)
 * Are you aware that the statistics in your link don't support your statement? Cresix (talk) 00:15, 22 July 2010 (UTC)
 * Depends. Are you actively defending GabeMc's statement, which is indeed a Foundation issue? —  Jeremy  ( v^_^v Dittobori ) 00:27, 22 July 2010 (UTC)
 * I'm not defending any position; I'm criticizing your statement as not being supported by the link you provided. And that does not "depend". Cresix (talk) 00:41, 22 July 2010 (UTC)
 * AFAIK there is no correlation between 'edit is anonymous' and 'edit is vandalism' features. I've seen some data supporting that claim while participating in a PAN10 lab. And here are a couple of references to relevant papers:, . --Dc987 (talk) 01:49, 25 July 2010 (UTC)
 * And lest someone leaps to an incorrect conclusion, let me clarify. The fact that there is no correlation is a direct result of the fact that there are no data. No one knows the relative ratio of vandalism to good edits for anons vs. registered users. So let me emphasize that "no correlation" could be misleading if not interpreted correctly. It does not mean there is no relationship between anon editing and vandalism. It simply means no data have been collected to demonstrate the matter one way or another. For our purposes, I think the most noteworthy statement from both of those articles is: "Vandalism is currently not documented in Wikipedia". Cresix (talk) 23:03, 25 July 2010 (UTC)
 * Yes. I agree. Perhaps 'no data' is a better statement. If you skim other the articles you notice that both researchers were unable to use 'edit is anonymous' feature in their classifiers. Same goes for the classifier that I've developed. That's why I've said - 'no correlation'. You are entirely correct - there is no reliable data. And unfortunately all the published annotated revisions datasets have inherent biases (article choice and unnoticed vandalism primarily). Still, I've just tried calculating these stats using a couple of datasets and here are the results:
 * {| class="wikitable"

! Contributor\Edits ! Regular ! Vandalism ! Regular (Revert) ! Vandalism (Revert)
 * + Dataset PAN WVC 10 (15000 annotated revisions)
 * Anonymous
 * 3692 (75%)
 * 791 (16%)
 * 375 (7%)
 * 11 (0%)
 * Registered
 * 8307 (81%)
 * 115 (1%)
 * 1705 (16%)
 * 4 (0%)
 * }
 * 4 (0%)
 * }


 * {| class="wikitable"

! Contributor\Edits ! Regular ! Vandalism ! Regular (Revert) ! Vandalism (Revert)
 * + Dataset WRDESE 10 "Rocket article" (3842 annotated revisions)
 * Anonymous
 * 228 (15%)
 * 1093 (76%)
 * 109 (7%)
 * 8 (0%)
 * Registered
 * 1198 (49%)
 * 127 (5%)
 * 1076 (44%)
 * 3 (0%)
 * }
 * 3 (0%)
 * }


 * Note of caution, it's very easy to look (for example) at the "Dataset WRDESE 10 Rocket" dataset and scream - "Anonymous contributors are all vandals, lets ban them". But note, that there are vast differences in percentages between two datasets. And you can never tell if some anonymous edits (even vandalism edits, because they bring attention to the article!) were in fact crucial in the overall article quality. So my input here would be: "Don't ban them. Work on more effective methods of detecting/patrolling/reverting vandalism.".
 * --Dc987 (talk) 23:12, 26 July 2010 (UTC)
 * And I also very much agree with Jimbo: "what is commonly called 'anonymous' editing is not particularly anonymous ... and there are good reasons to want vandals on ip numbers instead of accounts". --Dc987 (talk) 23:21, 26 July 2010 (UTC)
 * I think you understand this Dc987, but I'll state it for the other readers. The phrase "no correlation" is misleading. That implies that data were examined and the conclusion was no correlation. Thus, a better phrase is "no data" about correlation. And I happen to disagree with Jimbo. It's better for vandalism to be done by registered users. They can be blocked more easily, and if they change usernames, there's some chance a checkuser will reveal sockpuppetry. I've seen it happen many times. An anon can get away with vandalizing repeatedly for weeks or even months as long as he/she waits for the warnings to get stale. Then begin the process again after the block expires. Then move on to another IP address after a long-term block. But, like all of us, I speak without much data because we really have very little data. Thanks for you response. Cresix (talk) 00:58, 27 July 2010 (UTC)
 * Actually, IPs can be reported to their ISPs for constant and sustained abuse. Registering masks one's IP, making it very difficult to get an egregious abuser off our backs without involving the (currently severely understaffed) Checkusers, due to the Wikimedia Foundation's priv-pol. A sockmaster whom spends all his time with petty vandalism on named accounts is harder to wipe out than an IP-hopper doing the same. —  Jeremy  ( v^_^v Carl Johnson ) 03:59, 27 July 2010 (UTC)
 * As always, of course, all of us (myself included) are speculating because we have no real data except our informal impressions. My impression is that ISPs generally have no interest whatsoever in doing anything about IPs who vandalize on Wikipedia unless it strains their bandwidth or violates the law; and virtually no vandalism falls into that category. If that's the case, then the ISP solution is a moot point. And I'm not convinced that registered sockmasters are harder to deal with than IP hoppers. Checkusers understaffed? How about we (carefully) allow more admins to be checkusers, then require registration to edit. But ... we need some data. Cresix (talk) 02:16, 28 July 2010 (UTC)

Not useful to me as an editor
Every edit on my watchlist that I haven't already checked is pending review as far as I'm concerned, so I don't find it helpful to have a few edits flagged up. The few that I've looked at, I've been more inclined to simply revert if I think it needs it. It's more of a hindrance/clutter to me, but casual readers may have a different perspective.--Michig (talk) 20:19, 6 July 2010 (UTC)
 * I don't have your watchlist and you're not at the computer all the time, and if you are, you aren't on wikipedia all the time.... are you? If you aren't, than consider it some extra help from the community patrolling your pages. Beam 00:53, 7 July 2010 (UTC)
 * Sorry, I don't understand what you're saying. I have a large watchlist of nearly 9,000 articles, and those changes not flagged as 'pending' are just as likely to be vandalism as those that are from recent experience. If other editors find this feature helpful, fair enough, but I don't. This is my feedback, like it or not.--Michig (talk) 05:44, 7 July 2010 (UTC)
 * I also have quite a large watchlist and Michig's comments are correct in one way but from the viewpoint of the contributors, the percentage of uncomfirmed additions that they were able to make because of pending changes being used in preference to semi protection is the benefit to those users and to the project. Off2riorob (talk) 09:13, 7 July 2010 (UTC)
 * I can see that this is useful to unregistered readers who won't immediately see vandal edits by anonymous editors on articles - this is what I alluded to above. I can also see the argument that this allows anonymous editors to make changes which they would otherwise not be able to make. As I said though, I review all (or at least most) edits to articles on my watchlist, so having edits flagged there isn't really useful to me in terms of dealing with vandalism (you'll forgive me, I hope, for looking at this from my perspective). The key question, I think, is how much extra work this might create in reviewing and approving edits if rolled out more widely (i.e. beyond articles identified as 'problematic' or should the number of such articles increase considerably), and whether it may have a net negative effect on encouraging contributions from new editors (by their changes not immediately being visible).--Michig (talk) 13:49, 7 July 2010 (UTC)
 * I think a contributor that makes a good edit will be encouraged to see that it has been accepted by the reviewers. I also have these doubts you mention and have said that if the numbers can be provided we will see more from them. How much work has been created for how many accepted contributions from unconfirmed users? If for example 100 000 edits have been made to pending changes testing articles and the accepted percentage is low then a discussion could be required to assert net gain to the project. Other longer term issues are involved also, such as, there is a theory that if we use the system widely and vandal type contributors get the realization that their edits will not be visible they may move along as there is little gain for them, thus in the long term greatly reducing the vandal reverting workload. Off2riorob (talk) 14:01, 7 July 2010 (UTC)

Loss of power? Hatred of non-wikipedians? Simple arrogance?
I have noticed (made assumptions), after reading this thread, that some admins are actually kind of insulted at the whole idea of non-admins being able to review and accept someone's edits. I thought that this whole trial was going awesome, there is never a backlog at the page, the average is something like 3 minutes for a "review." Then I came to these talk pages! A set of caring eyes is actually reading articles for once, and it's thought as a bad thing? I bet that good edits are resulting from the simple fact that another set of eyes was drawn to the article. I just can't understand the subversive attitude towards this. Maybe I'm just a peasant regular user, but I don't see a downside at all. If you don't like pending changes, then perhaps you should advocate for the complete removal of any anons/ips from editing Wikipedia. Why even let them think they can edit the articles? I mean if you can't even give them the HOPE that their edits would DO ANYTHING WHAT SOEVER than why lie about it? Beam 00:38, 7 July 2010 (UTC)
 * I agree wholeheartedly. Not only do I think this is a great idea in general, but I think it should be implemented globally on all articles. I've always thought that mandatory registration would be the best thing that could happen to Wikipedia, but this is the next best thing, and it actually encourages *gasp* peer review! –Cosmopolitan (talk) 19:58, 9 July 2010 (UTC)
 * Beam, I'm against the pending changes precisely because it will advocate for the complete removal of any anons/ips from editing Wikipedia. Why would an ISP or anon edit a site if every edit they make must be approved by someone working in the system? It is my belief that if we adopt this freedom to edit will be granted only to those with an account. The isp editors will be permanently at a lost for anything. Why would you want to edit a sight where you edits are held to such low standard that they must be greenlit by someone with the authority to do so? Doesn't that violate the very essence of assume good faith? TomStar81 (Talk) 00:43, 11 July 2010 (UTC)


 * I agree with Tom Star. I found a user-generated music lyrics website a few weeks ago, while looking for the words to a song that were hard to decipher. The lyrics that were there were clearly mistaken, so I fixed a few of the lines, and saved to check they were taking before doing more. A message appeared telling me that my edits would be reviewed before being accepted. I decided not to waste time waiting for that, so I didn't bother to fix the rest. The reason Wikipedia succeeded is that it didn't throw up those barriers. People got the instant gratification of making a contribution. It means it's hellish patrolling it for vandalism, and I definitely want to see more protection for BLPs, but if we remove that instant feedback, we're much less likely to attract a new generation of Wikipedians&mdash;and if we don't do that, we're toast.  SlimVirgin  talk| contribs 00:52, 11 July 2010 (UTC)


 * You both seem ignorant of what goes on here. Ever heard of Page Protection? Go crusade against that while we use Pending Changes in the mean time. Otherwise you're both just bullshitting, and seem to fit in one of the categories I suggested. Love, 71.184.96.178 (talk) 01:10, 11 July 2010 (UTC)
 * Page protection doesn't give vandals a chance to dump a syringe-tide into a page history. I agree with Tom and Slim for different reasons, mainly that (1) PC will actually be a hindrance in the event of short-lived, but sustained, vandalism (such as that /b/ and /v/ normally do) because the history will get clogged; (2) most Reviewers, admins included, will, at some point, become psychologically unable to distinguish good IPs from bad, and will reject all anon edits (and, though less likely, those from new accounts that are already throwing up false red flags) offhand; and (3) gives someone new targets to chase off of Wikipedia. —  Jeremy  ( v^_^v Carl Johnson ) 18:58, 24 July 2010 (UTC)

I've heard of it. To address you crusading comment though, let me propose a scheme by which we may judge your comments in relation to the position put forth by both Beam and myself. Two wikis - identical in all respects - adopt two different positions on the matter quality control. The first is open to everyone to edit, and slow restricts editing through various means to address problems related to vandalism and other issues that a wiki must deal with. The other takes a different approach in which editing is closed to everyone, then slowly opened to community based on the needs of the site. Given a choice between the two systems, which would you choose? I pick the first one because I like the idea of openness, but my experiences will different than the experiences of others, and therefore others will select one of the two options based on what they feel will work best. My concern here is that the second option is where wikipedia is heading, and that disturbs me greatly. Your opinion will of necessity differ. TomStar81 (Talk) 02:01, 11 July 2010 (UTC)


 * Among IPs I've seen both vandals and some I've found very helpful. I suggest much tougher measures for clear vandalism - earlier, longer blocks; no re-setting the clock; "3 strikes and you're out". Then we spent much of our time doing useful work, e.g. on articles. --Philcha (talk) 03:59, 18 July 2010 (UTC)


 * But the IP addresses are usually dynamic, so you can end up hurting the wrong people. If my neighbour starts vandalising on an IP account and gets blocked for three years, then later everytime I get that IP issued to me, I have to suffer for someone else's vandalism. HumphreyW (talk) 04:07, 18 July 2010 (UTC)
 * And you could avoid that by taking 20 seconds to register. Cresix (talk) 04:10, 18 July 2010 (UTC)
 * And many IPs, baneful and beneficial, are unwilling to register, whether because they are paranoid, unwilling to reveal personal information about themselves, or what have you. —  Jeremy  ( v^_^v Carl Johnson ) 19:20, 24 July 2010 (UTC)
 * Comment re: blocks: if this occurred I'd be extremely concerned, and would want to speak to the blocking admin. IP addresses are not normally blocked for anything other than short periods - exceptions being made when it's clear that the IP address is static (i.e. your neighbour's IP address is likely to remain with your neighbour). This is for precisely the reason you imply: you might end up being affected by a block caused by your neighbour. TFOWR 19:29, 24 July 2010 (UTC)

Some input from a researcher's viewpoint
Just a general advice, try to base this decision on rigorous scientific evidence and metrics, not intuitions. Intuitions might easily be wrong. How wrong? For example if the proposed change would only stop 'silly vandalism' that is getting reverted by a ClueBot anyway, this wouldn't gain anything. Or intuitively one might think that 'most of the vandalism comes from anonymous IP editors', well, sorry, but AFAIK there is no scientific evidence for that claim. Closest claim supported by evidence that I've seen is: 'editors with 500+ edits _and_ low level of reverted edits rarely engage into vandalism activities'. Note that this _includes_ anonymous IP editors.

Anonymous IP editors certainly bring more diversity of viewpoints. They also somewhat dilute connectivity of the Wikipedia social network structure. It is hard to tell, but potentially this can be very important.

Another point that is rarely made. Vandalism, tests edits, NPOV edits, etc are maybe playing a role of random input. As a result even these edits are possibly beneficial for the Wikipedia overall. Counterintuitive. But it is well known that in large stochastic systems dynamic equilibrium can sometimes be much improved by random input.

All and all I would recommend to be very careful playing with the parameters that could affect Wikipedia popularity and availability in the anonymous editors community. --Dc987 (talk) 01:53, 7 July 2010 (UTC)


 * Lots of really interesting points to be made here, thanks Dc987. Any help you can offer in developing the specification for the metrics for this trial would be greatly appreciated.  Thanks! -- RobLa (talk) 20:31, 8 July 2010 (UTC)

Wonderful idea
Long awaited move by wiki! Most sites demand you login, if you have something to say! We can follow the same here, may be make registering imperative after two edits by an IP user, so if one really wants to add something of importance will take time to register, otherwise leave a note, which is a good idea too! Further why encourage vandalism by giving free access to text which has been created over months, by so many users! Sometimes valuable text goes missing or is replaced by unusable text. --Ekabhishektalk 06:53, 8 July 2010 (UTC)
 * Er... the idea of Wikipedia is to allow anyone to contribute. I think that's why we don't enforce registration. In fact there are quite a few IP editors who do wonderful work, like submitting lots of good stuff to Articles for Creation, vandal-fighting, etc.  Bramble  claw  x   18:10, 8 July 2010 (UTC)
 * Registration doesn't stop anyone from editing, except those who don't want to take 20 seconds to do it. It does not even require an email address. And someone can register with as many usernames as he/she wishes (unless the goal is to avoid a block by using sockpuppets, of course), so that adds even more anonymity. I think the main reason most anons don't want to take a few seconds to register is that they have no intention of making constructive edits. Yes, some anon IPs make good edits, but there is no evidence that those same people would not register and then make the edits. Cresix (talk) 19:29, 9 July 2010 (UTC)
 * I agree wholeheartedly. Mandatory registration does not stifle the "Anyone Can Edit" philosophy, since anyone can register, even without an email address, and it takes less time than making edits to wikis does. –Cosmopolitan (talk) 20:01, 9 July 2010 (UTC)
 * Its not very hard to regester which everybody should have to do. Most sites need an email to regester but wikipedia doesn't even need that. Red Flag on   the Right Side  02:51, 10 July 2010 (UTC)
 * Agreed. It would also make it easier to block users, as they have to contribute with a username and cannot edit by IP address. However, as Brambleclawx stated, there are IPs that write wonderful articles well-sourced and submit them to articles for creation, and some prefer not registering. Oh well, all choices have pros and cons. — MC10 ( T • C • GB •L)  00:32, 11 July 2010 (UTC)
 * I must dissent regarding mandatory registration, for the simple reason it takes such a very short time to register, therefore forcing registration will not stop vandalism. Anyone intent on causing chaos is more than willing to complete the registration process -- hundreds of times, if needed. IP blocking, while not perfect, does deter most vandals, and the current protection system should get the others. Additionally, forcing registration will exclude some editors that are against or fearful of any registration process (even at such a benign site at Wikipedia) because of their concerns regarding ID theft through hacking of databases, or a general misunderstanding of how registration works, etc.; yet they are extremely knowledgeable regarding very specialized subject matter & submit numerous Articles for Creation, as noted by Brambleclawx.--There is nothing civil about Civil War.Let's Talk! 19:07, 11 July 2010 (UTC)

Table change
A protection level table appears on several pages associated with the pending changes trial. I've changed the way the protection level table looks, and commented at Wikipedia talk:Pending changes/Protection level table. BrainMarble (talk) 02:16, 9 July 2010 (UTC)

What's the point?
I personally wouldn't bother with this. People shouldn't be able to edit unless they register anyway Cls14 (talk) 10:56, 10 July 2010 (UTC)


 * You don't understand Wikipedia at all. Beam 14:29, 10 July 2010 (UTC)


 * Concur with Beam. A quote directly from the article on Wikipedia itself:


 * "In departure from the style of traditional encyclopedias, Wikipedia employs an open, 'wiki' editing model. Except for a few particularly vandalism-prone pages, every article may be edited anonymously or with a user account, while only registered users may create a new article (only in the English edition)."


 * So part of what you say is already true; IPs cannot create articles. However, ANYONE (except the damne-...er, banned) is able to edit any article that isn't marked as a highly vandalised article at any time.  CycloneGU (talk) 16:11, 10 July 2010 (UTC)

Yes, of course that's the policy. If I understand things here, this is a page for opinions, and I believe Cls14 expressed an opinion. I happen to agree that registration should be required for editing, but even if I didn't, I think a little more courtesy in responding to Cls14 would be in order. Cresix (talk) 16:50, 10 July 2010 (UTC)

User:Cls14 is right. There should be a much higher bar for editing Wikipedia articles going forward, especially WP:BLP articles. I think s/he likely understands it just fine. The project is changing folks, get over it. JBsupreme ( talk ) ✄ ✄ ✄	 23:16, 10 July 2010 (UTC)
 * See Wikipedia%3APEREN Gerardw (talk) 03:37, 11 July 2010 (UTC)


 * Whoops...I should clarify. I wouldn't go so far as to say Cls14 does not understand Wikipedia at all, I was merely trying to provide support to the fact that Wikipedia is designed for anonymous users to be able to contribute. =)  CycloneGU (talk) 22:05, 11 July 2010 (UTC)


 * It shouldn't be forgotten that a large number of us are opposed to this new feature because we believe so strongly in the editing rights of unregistered editors. We fear that this feature will discourage them and have a long-term detrimental effect on Wikipedia by lowering the number of new editors who become and remain involved.  And some of us believe in the core principle that this should remain the "encyclopedia that anyone can edit" and this new feature contravenes that principle. ElKevbo (talk) 17:55, 13 July 2010 (UTC)

Interesting discussion. There was a time when I agreed with the sentiment expressed by Cls14 that only registered users should be allowed to edit. I reluctantly came to that point of view in a time of despair over IP vandals and socks, particularly with reference to BLPs. However, I am being won over by this pending changes trial. It definitely deters garden variety vandals, since they don't get the positive feedback of seeing their inanities immortalized in print. It seems like an elegant way of upholding first principles. Anyone can edit—and "edit" is defined as making a positive contribution to the encyclopedia. Sunray (talk) 23:26, 11 July 2010 (UTC)