Wikipedia talk:Reviewing pending changes/Historical/Archive 2

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 1 Archive 2 Archive 3

Is protection better?

@G.G. and Dr.K: As I pointed out below, we currently use page protection to screen out good-faith but unwanted editorial content from pages such as Liancourt Rocks. It may be that you're proposing we should continue using protection for those pages, rather than PC. If that is what you're saying, could you explain why you feel protection is better? — Carl (CBM · talk) 15:40, 16 June 2010 (UTC)

If the good faith content isn't sourced, it's not a worry, it can be skived. If it is sourced, it should be in the article along with other PoVs, under sway only of WP:UNDUE and WP:NPOV. Protection is but a flag to show that open editing has hit a glitch. I don't worry too much about high traffic articles staying under semi-protection for very long stretches and full protection nudges everyone towards working through disagreements much more quickly. PC is not meant as a means of handling content disputes, at all, but as a way to allow editing throughout a content dispute. Alas, as I keep saying, many editors already have groked that this design is a handy means for getting rid of unwanted PoVs and sources.Gwen Gale (talk) 15:51, 16 June 2010 (UTC)
Clearly if the content is not sourced, as Gwen says, it can be declined. But a sourced POV should not be screened, at least not until an exhaustive discussion has taken place at the talkpage. As Carl mentions the current state of affairs at Liancourt Rocks is not good either. Because who decides what POV is allowed in or not. Neither system appears to be ideal, but given a choice I would choose protection. At least it is something reversible and that gives an incentive to use the talkpage to reach consensus on an equal basis, unlike the FR system which creates permanent over and under classes of editors. Dr.K. λogosπraxis 16:03, 16 June 2010 (UTC)
It's amazing isn't it? what was sold as a process to eliminate libels and vandalsim of BLP is now going to become little more than censorship. One personal opinion against another, with the winnner being the one most tooled up.  Giacomo  16:06, 16 June 2010 (UTC)
If Wikipedia is compared to a MMPORG it looks more and more like editing has entered a phase like an arms race where the guy with the most weapons wins. Something like Halo 3. Dr.K. λogosπraxis 16:11, 16 June 2010 (UTC)
I expect almost every active editor will have reviewer rights, so there won't be any disparity in that respect.
The issue of "sold" is complicated, because the proposal wen through so many steps. It has been sold as:
  • A way to review every edit for quality
  • A way to protect BLP articles from vandalism
  • A way to replace the protection system with something more accomodating to IP editors
So I think that everyone probably has a different idea about the system depending on we followed the discussion and which announcements we saw.
But I don't see a concrete answer to my question: is protection better? — Carl (CBM · talk) 16:15, 16 June 2010 (UTC)
I think I answered positively just above and also gave my reasons. Dr.K. λogosπraxis 16:18, 16 June 2010 (UTC)
I think that the real issue here is the creation of a permanent overclass of editors (the reviewers) who control the edits of the underprivileged lower classes (everybody else), something that under the protection system did not exist. Dr.K. λogosπraxis 16:29, 16 June 2010 (UTC)
I think that we are experimenting with the very DNA of Wikipedia, which thus far has made it so successful and dynamic. The new editing regime is not as dynamic and free-wheeling as in the old model and may even lead to loss of editing energy and editing momentum while creating editing bottlenecks which will cost dearly in terms of new editor recruitment and current editor satisfaction. Dr.K. λogosπraxis 17:01, 16 June 2010 (UTC)
Protection stops people from editing. PC does not. Given that I can't see how you can honestly make the claim of a "permanent overclass of editors who control the edits of the underprivileged lower classes", as the claim would apply far more strongly to protection than to PC as PC is strictly less controlling. --Gmaxwell (talk) 17:38, 16 June 2010 (UTC)
It depends on what type of protection. If it is semi then it excludes IPs but autocofirmed users get to edit normally. If it is full protection then nobody edits but at least everyone is equal in not being able to edit the article including the IPs. Of course the admin class is exempted from all this anyway but it doesn't really matter because regular editors are still equal to each other in not being able to edit under full protection. With this new PC system we have autoconfirmed editors who are not reviewers or who do not want to be for one reason or another such as conscientious objectors, technically challenged or simply gun-shy of the new responsibility and therefore are the new underclass, the IPs are relegated to permanent underclass status because their edits have to be reviewed to be accepted. What do you call this PC system if not an overclass/underclass generator? Dr.K. λogosπraxis 02:42, 17 June 2010 (UTC)
Any honest person could only come to the same conclusion. Malleus Fatuorum 02:48, 17 June 2010 (UTC)
Of course Malleus, honesty is the operational word here, the keyword. How could I forget? Thank you :) Dr.K. λogosπraxis 02:51, 17 June 2010 (UTC)
I just call it a wiki. I see little difference between somebody just quickly removing the IPs edit in the normal way, and somebody using this feature. And if the reviewer is up to no good, it's quite easy for an IP to bring it to an admins attention, and the evidence is still in the history list. Compared to some of the things that a rouge admin can do, this is almost entirely benign.- Wolfkeeper 02:57, 17 June 2010 (UTC)
The difference is that only the new reviewer elite can accept or reject the IP edit. The non-reviewer editors are disenfranchised from this action even though they were able to participate before PC on an equal footing with all the other autoconfirmed editors. So yes it is still a wiki but a divided one. There are two classes of editors now. They are separated, they are apart and developing separately no longer being able to share similar experiences at least on a significant part of their hitherto common editing activities. Dr.K. λogosπraxis 03:07, 17 June 2010 (UTC)
The above is factually incorrect: Anyone can reject an edit on a PC page. Anons are equal to the power of reviewers in that regard. They get the ability to do this because they can edit at all, something they lack on a semi or protected page. Calling the approval process "accepting the edit" is a little misleading too: once an edit is made it's made, the edit is there— an (almost) indelible mark in the article history instantly available to the whole world. What PC does is it controls if themost current version of an article or a past version is displayed by default to anons on the main article page. So if you make an edit and no one reverts you... but no one accepts your changes then the page will continue to only show the old version, even if a reviewer (or an administrator, for that matter) subsequently edits the page. When that admin approves his own edits your pending edits in the article will be necessarily included along with it. The only way that someone can reject your changes is to revert them, like always. This is an active and public act for which complaint can be raised if performed inappropriately, and it's not a new power, nor is it limited to a small set of privileged users.
If the fact that other people can revert you with the full public records, 3rr, and other processes that we have in place is cause for claiming a dire risk of suppression by a powerful elite then we've been doomed since day one.
Of course, PC _is_ a restriction. But it is arguably the most minimal restriction possible to achieve its particular effect... The structure of PC ensures a high degree of equality: No user has a special privilege to reject changes, changes can't be passively rejected by the "elite" through inaction without obstructing their own editing, and pending changes are approved as a side effect of the self-interests of the "elite" in approving their own edits. Not to mention the reviewer right is granted pretty liberally, and if you think making it more liberal would help then you should advocate that.
In most cases where equality is of concern the most important privilege is the privilege of being able to speak. Often we can not judge the evil of a protection action fairly because the editing that it prohibits is a silent harm. PC removes most of the harms, the ones that it potentially leaves are not unique to PC— we've always had the problem that some good edits are reverted— and they are not silent, nor are they "passive" harms which people can excuse on the basis of inattention.
I sometimes wish that we had deployed the first crap review system that came into someone's mind— just so we'd have a basis for comparison and could see all the things that were carefully done right by the collection of hundreds of people and thousands of man hours that went into shaping the concepts behind PC and flagged revisions. I'm sure there is plenty of room to refine it further, but all the black helicopter talk isn't helpful. A lot of people who cared deeply about openness, inclusiveness, and transparency poured their hearts and hours into this system. Perhaps they screwed it all up, time and experience will tell. But to pretend that people were indifferent to these issues is an insult to the entire Wikipedia community. --Gmaxwell (talk) 04:31, 17 June 2010 (UTC)
(edit conflict) You gave a lot of technical info which is interesting enough and needs further examination to see if it pans out the way you described it. Time will tell; on this we both agree. But I don't understand your concluding comments about black helicopters, pretending that people were indifferent to issues and insults to the entire community. Unless I didn't follow the entire discussion, I don't see how these points fit to the discussion we are having. Unless you imply that people who disagree with you are insulting the devs and by extension the whole community, even though I would think that the people who disagree with you are also part of the community and thus they cannot insult themselves. As far as conspiracy theories this is also a new angle. I don't know where this came from either. Dr.K. λogosπraxis 04:55, 17 June 2010 (UTC)
Apart from the tab, I can't see much difference. There are now the following grades of user:
  1. admin
  2. reviewer
  3. autoconfirmed
  4. other
There are the following grades of article:
  1. protected
  2. PC
  3. semi
  4. other
Each type of article can be changed by a user whose grade number is no larger than that of the article. Peter jackson (talk) 10:16, 17 June 2010 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── That's not entirely accurate, IMHO. PCed articles can still be edited, but those changes are not shown by default to non-logged in readers — who, by the way, can see those changes, if they so wish —. As said before, this tool doesn't restrict editing privileges; it merely limits the number of people who by default see them. Salvio ( Let's talk 'bout it!) 10:46, 17 June 2010 (UTC)

I did start off by saying "Apart from the tab". Peter jackson (talk) 17:07, 17 June 2010 (UTC)
Oops, I didn't read accurately enough; sorry. Salvio ( Let's talk 'bout it!) 22:59, 17 June 2010 (UTC)

Sea of Japan and Liancourt Rocks

I want to start a new thread about reviewing, based on the specific example of Sea of Japan, which is scheduled to go under PC as part of the trial. Another example of an NPOV protected page is Liancourt Rocks; this one may be more well known because it was the subject of an arbitration case. The problem with these pages is not vandalism. Many of the edits that we are trying to avoid through protection are misguided but made in essentially good faith.

As I understand it, the goal of PC is to be able to loosen our restrictions on these pages by allowing more IP edits through, while still filtering out the types of edits that lead to protection. So when we review edits for these pages, we have to keep in mind the reason for protection, which is not juvenile vandalism.

The point of PC is that, because new IP edits don't go live immediately, there's no hurry to revert them. So instead of preventing all IP edits, as in the previous system, we now have time to look at the edits on an individual basis and allow at least some of them through. But we can't review for articles like this by approving every edit that is not juvenile vandalism. The reason behind the page protection has to be taken into account. — Carl (CBM · talk) 14:19, 16 June 2010 (UTC)

Agreed 100%. We don't want vandalism to be taking up the entire history; if a page is going to receive heavy vandalism, protection is still appropriate. However, for issues where we may have to protect just to maintain an article's integrity, pending changes seems to excel as a much better alternative. --Shirik (Questions or Comments?) 14:23, 16 June 2010 (UTC)
I actually quite disagree with the idea of using pending changes for articles that are protected in some form because of ongoing (or in the case of these examples, long term) content disputes; in fact, I think that pending changes would be likely to either reignite or exacerbate such content disputes rather than to ameliorate them. In these two particular cases, the consensus community decision on appropriate content, after long and difficult dispute resolution, was being regularly challenged through editing of the articles. While consensus can change, sticking a reviewer unfamiliar with the long and complex background into the middle of the situation and expecting them to make a good decision is setting the reviewer up to fail, and exposing the article to renewed edit warring (even if the edit war is outside of the view of the reading public). Risker (talk) 15:43, 16 June 2010 (UTC)
Why would a reviewer be "stuck" in a situation? I think that, like protected edit requests, reviewers can always let someone else handle a situation that they are not comfortable with. The usual editing of most articles should take care of keeping the reviewed versions current, as long as there are plenty of people with reviewer rights. — Carl (CBM · talk) 15:50, 16 June 2010 (UTC)
Let's see. We have these articles where good-faith, quality edits have been made on both sides of a POV, to the point that dispute resolution was required to come up with an internal decision on which of the two paths to follow. Despite that internal decision, good-faith anonymous/newly registered editors continued that same editing pattern, which was disruptive enough that longterm semi-protection was needed. Now, we superimpose an editing review system that explicitly presses for rapid review and decision-making, to be carried out by editors who are not subject matter experts, whose "rules of engagement" are to accept good-faith, reasonable edits that do not appear to be vandalism or some form of BLP violation. There is no way, without a huge amount of reading, to determine whether a specific, good quality edit in pending changes is going too far in expressing one point of view or the other. This is a much more intensive qualitative analysis than is expected for the review process, and is outside of the scope of this particular protection feature, which expects all good faith, quality edits to be accepted, and the regular editorial process to take over from there. Risker (talk) 16:04, 16 June 2010 (UTC)
Those are exactly the articles I am thinking of here, and exactly the problem of new editors who don't know the current consensus.
Are you arguing that those articles should just remain protected, rather than being in pending changes? I thought that the point was to allow at least some edits through while being able to filter out the ones that caused protection. So rather than telling new editors "you can't edit", we can point out the internal decision to them instead, because we don't have to worry about their changes being cached on google for hours. — Carl (CBM · talk) 16:10, 16 June 2010 (UTC)

Risker, if some users are involved in a long term content dispute then they should be topic banned or otherwise blocked from editing. We shouldn't stop people from making good edits simply because we're too lazy to kick out the people making bad ones. Perhaps some of our problems with ejecting persistent PoV pushers have arisen because of the ease of papering over the misbehaviour with protection while the openness and inclusiveness of Wikipedia silently suffers for no good reason. There are cases where we can't eject the bad eggs— because they are a mob being sent by an outside party, coming in on a hundred accounts, for example — and we need protection for that. But that should really be the extent of where protection is strictly required as opposed to PC. --Gmaxwell (talk) 17:34, 16 June 2010 (UTC)

  • You are justifying the use of this wonderous new tool using reasons for which it was not introduced. Have we all been conned?  Giacomo  19:02, 16 June 2010 (UTC)
Maybe so, Giano, though I'd hope it's only a heedless design. For the nth time, the consensus for PC and even the policy page have nothing to do with handling content disputes. Gwen Gale (talk) 19:07, 16 June 2010 (UTC)
I haven't a clue what either of you are talking about. What specific use are you discussing? The purpose of the tool, for which is was discussed, designed, and introduced was to allow us to reduce the amount of edit prohibiting protection by offering a less harmful measure against obviously bad edits. I'm not suggesting handling "content disputes", I'm talking about handling users with behavioural issues so severe that they can't refrain from revert warring unless the page is locked so that no one can edit it, which is really just a form of vandalism. Content disputes should just be handled by the regular mechanisms as always. ... Persistent protection isn't a tool for content disputes either, Wikipedia:Protection#Creation_protection. --Gmaxwell (talk) 19:23, 16 June 2010 (UTC)
No, it was sold as a means to cut down BLP violations from showing up live, along with vandalism, legal/PA/libel woes and the odd copyvio. Meanwhile, editors who revert war are blocked, edit warring isn't allowed, PC has aught to do with edit warring. Gwen Gale (talk) 19:33, 16 June 2010 (UTC)
Thank you Gmaxwell - it was like drawing teeth, but at last we have part of the truth - nothing at all to do with BLPs.  Giacomo  20:41, 16 June 2010 (UTC)
Whats to draw teeth about,— what was proposed was explicitly not limited to BLPs. Many people have used BLPs as a primary example of why it is useful, but the idea that it was ever limited to them is an unfortunate hallucination. --Gmaxwell (talk) 21:19, 16 June 2010 (UTC)
The idea is that no one is obligated to do careful fact checking, but there is also not an extreme urgency to approve (or undo) an edit as soon as possible. The lack of urgency exists because PC greatly reduces the primary harm from bad edits. Even ignoring the enormous inclusiveness benefits that removing protection offers, PC can also provide the benefit that users are less likely to be reverted by careless video-gaming RC patrolers— thus the good edits that are made are more likely to stick. But this benefit goes away if you continue to play PC out as just another MMORPG feature enabling people to act as review-farmers cranking out acceptances or reverts as fast as possible— on the road to the next big level-up (adminship)— heedless of the quality impact on the encyclopaedia and the enfranchisement of our editors. Just as was described in the text where this feature was sold: quality, and enfranchisement both matter. Common sense and due care is required in order to achieve both outcomes. --Gmaxwell (talk) 20:19, 16 June 2010 (UTC)
What would you call a "bad edit"? Gwen Gale (talk) 20:21, 16 June 2010 (UTC)
oops. I accidentally cut off part of my own message. First let me set the record straight about how this feature was sold:
"The aim of marking an article revision as patrolled is to indicate that the revision has no major problem requiring immediate attention. Reviewers are expected to apply common sense when marking revisions as patrolled. Because there is no urgency to marking revisions as patrolled, reviewers should consider editing the article to fix problems, reverting unconstructive edits, or bringing issues to the talk page or the appropriate noticeboards.", on [1], which was later also nutshelled to "If edits haven't made an article worse, they ought to be confirmed."
A bad edit in this context is whatever you believe to be bad. One of your few obligations on the site is to not willfully make the site worse. Wilfully and knowingly making the site worse is pretty much the essential defining characteristic of vandalism. If, in your casual opinion— which is all PC requires, not research—, an edit is probably not harmful/malicious/etc then you should confirm it. But if you _know_ that it is bad, at least as far as you can tell in your quick analysis, you should take no action (because nothing obligates you to fix anything) and leave it to someone else to confirm it or fix it, or you should take the time to fix it (improving the project). The latter is best if you are willing to put the time in to do it well but it isn't required, and if you're not willing to do it well you should leave it to someone who is.
If you instead confirm an edit which you personally believe to be bad you will have directly and actively contributed to reducing the quality of Wikipedia (at least as measured by your own personal criteria). Moreover, you will have actively denied other reviewers an opportunity to look at the edit and decide to fix it as once reviewed the edit and the prior revisions are not included in any review span diffs. Haste would be justified if we had many edits sitting around for a long time, but that doesn't currently look like a risk. --Gmaxwell (talk) 20:35, 16 June 2010 (UTC)
That's not at all what the policy says but yes, PC does at least seem to be socially engineered into fooling editors towards carrying forward the site's already very heavy systemic bias in its core high traffic articles. Gwen Gale (talk) 20:42, 16 June 2010 (UTC)
I quoted and linked to how the feature was sold. The policy would say different things if you didn't revert things which were consistent with how the feature was sold, as well as everyone's obligations on the site, but which you find personally disagreeable. I think your claim about engineered for furthering systemic bias demands a supporting argument rather than just an assertion. I think you're full of it. PC reduces systemic bias by making it so that people outside of the club-of-regulars are able to edit _at all_.--Gmaxwell (talk) 20:50, 16 June 2010 (UTC)
This? It doesn't support your take on PC at all (and be aware, I've never edited that project page). The policy, oops, "trial guideline" is quite straightforward, PC, as put forth, is meant to screen only for vandalism, BLP worries, legal threats/PAs/libel and copyvios. As you've said, though, it can be wielded for so much more. Gwen Gale (talk) 20:58, 16 June 2010 (UTC)
You edited the page which you cited above [2] as justification for your policy argument. You appear to be ignoring the argument that I'm making that Wikipedians are obligated to not intentionally make the project worse, and that people who do so are normally called vandals. --Gmaxwell (talk) 21:17, 16 June 2010 (UTC)
We both know I only undid your edit to take the guideline back to where it had been all along. Nobody has undone what I did, which should tell you something. Gwen Gale (talk) 21:20, 16 June 2010 (UTC)
I asked people not to revert you while we discussed this, FWIW. If you promise to not edit war over it I'll withdraw the request. I see that you are continuing to ignore the issue wrt Wikipedians obligation to not take actions which they know will make the project worse. I am beginning to get the feeling that you are not participating in this discussion in good faith. --Gmaxwell (talk) 21:24, 16 June 2010 (UTC)
Promise not to edit war? That's a smear. We both also know I haven't edit warred. As for any lack of good faith, you brought it up, now live with it. Gwen Gale (talk) 21:27, 16 June 2010 (UTC)
I apologise for smearing you, it certainly wasn't my intention. So you're not going to state any opinion on a Wikipedian's obligation to not intentionally lower quality of the content of the articles and how that intersects with the PC approval process, or otherwise refute the argument that I've presented? Are we in agreement? --Gmaxwell (talk) 21:58, 16 June 2010 (UTC)
I think PC was engineered for you. Gwen Gale (talk) 22:00, 16 June 2010 (UTC)
I prefer to describe it as 'created in My image' Cheers! --Gmaxwell (talk) 22:42, 16 June 2010 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────These two articles have been semi-protected (note: not fully protected, autoconfirmed editors can edit them) because anonymous and newly registered editors, most from the geographic area of the real-world dispute attached to these articles, have repeatedly attempted in good faith to reinsert and/or remove the information that was covered by a difficult and complex Wikipedia dispute resolution. I will say this again: the articles are semi-protected from good-faith but POV edits, not from vandalism, because they have repeatedly caused edit wars with newly registered and anonymous editors who seek to import a real world dispute into this encyclopedia. (I don't disagree with the idea of sanctioning autoconfirmed editors who reignite this edit war.) Under the rules of this trial in particular, reviewers are obliged to accept these edits. Why anyone would want to reignite a very significant content dispute is beyond me; it's not like this longstanding international real-world dispute has gone away.

And Gmaxwell, please don't try to game the system by editing the rules of the trial to your own preferred version; I know you have invested a lot of your time and energy into the both the software and the principles of pending changes, but you have to let it go now and see where the community takes it. You need to understand that when you try to force changing this trial to follow your own vision of what pending changes should be used for, you increasingly risk losing the support of people who want to give this a fair shake, and you give those who are skeptical to start with valid grounds to distrust the entire process and everyone connected with it. Risker (talk) 05:35, 17 June 2010 (UTC)

Risker, it is clear to me, as a reviewer, that "Reviewers should take special consideration of the reason given for protection, and attempt to uphold it. For example, if the article is protected because of repeated inappropriate edits by a sockpuppeter, and if the same inappropriate edits are made by a new or anonymous account, they should not usually be accepted." Wikipedia:Reviewing#Reviewing_process Anyone reviewing ought to not accept the (good faith) POV edits you describe. Is your concern that reviewers are not reading these guidelines? ErikHaugen (talk) 06:03, 17 June 2010 (UTC)
Reviewers should protect articles from bad faith edits, not from good faith albeit POV-pusing ones. PC is not meant as an alternative to other dispute resolution methods. Reviewers, in my opinion, should not reject a pending change to prevent an edit war or to preserve the WP:TRUTH in an article. Salvio ( Let's talk 'bout it!) 09:36, 17 June 2010 (UTC)
In the case of articles like Liancourt Rocks, the POV good-faith edits are the "bad" edits. Also, reviewers cannot reject an edit, that is not an option in the PC system. — Carl (CBM · talk) 11:00, 17 June 2010 (UTC)
Of course "good" and "bad" are not the right terms. But the POV good-faith edits are the edits that he protection was intended to prevent, so these are exactly the edits that should be filtered out under PC. — Carl (CBM · talk) 11:04, 17 June 2010 (UTC)
I'll rephrase what I meant to say: reviewers should not refuse to accept or unaccept (or undo/revert without accepting) edits that are not a BLP-vios, blatant vandalism or copyvios. Neither preventing edit wars nor stopping what one perceives as wrong bits of info from being included in an article is within the province of a reviewer, IMHO, and I see it as a policy violation. Salvio ( Let's talk 'bout it!) 12:54, 17 June 2010 (UTC)
Of course it's not a policy violation (we don't even have a policy yet, that's what we are developing). Moreover, for example, if a page is under PC because of persistent sockpuppeting, edits that continue the patterns of the sockpuppets should not be accepted even if they are not blatant vandalism, BLP violations, or copyright violations. It would completely ignore the purpose of PC in that situation if we accepted the edits of the sockpuppets. — Carl (CBM · talk) 13:05, 17 June 2010 (UTC)
In case of sockpuppetry I agree: I think I've never said otherwise. It's in case of edit warring or plainly "wrong" edits that I disagree. Salvio ( Let's talk 'bout it!) 13:42, 17 June 2010 (UTC)

Risker: I think you're excessively limiting the "rules of the trial" to fit your opinion. Since there is no history of using pending changes, there is no established set of rules for its use. The point of the trial, and this conversation, is to discuss these things. But with that said,I see nothing on Help:Pending changes that excludes articles protected due to edit warring. PC was sold, most recently, as an alternative to page protection. It seems to me that you're saying you prefer to use page protection, so that no IPs can edit the articles, rather than use PC to filter out only the undesirable edits. That seems less than ideal to me, if the point of PC is actually to loosen editing restrictions on articles. — Carl (CBM · talk) 11:00, 17 June 2010 (UTC)

Carl, above you identify one of the significant concerns that comes with pending changes: that there are some situations where an article practically needs a "don't review this article unless you understand the past dispute and know what content has been determined inappropriate to the article" notice when what constitutes a "bad" edit has nothing at all to do with vandalism or BLP violations, and looks like a perfectly reasonable edit. Reviewers aren't supposed to require special knowledge to review articles, and that includes not having to know that Edit XX is okay anywhere but Article YY. They're square pegs being shoved into round holes; pending changes has never, in any of its configurations, been intended to cover every article, only to be one tool in the drawer.

Having said that, maybe it is time to lift all protection from these two articles and see what happens. I think that would be a better course of action for articles like this. Risker (talk) 14:19, 17 June 2010 (UTC)

PC/flagged revs has been intended for many things. I only learned recently about how it is currently configured.
At the moment, when I read WP:PENDING I see: "The conditions for level 1 pending changes protection should be the same as to what the current semi-protection policy allows.". I didn't write any of the proposed policy documents, so there must be at least one other person who thought that PC was appropriate for replacing semi-protection.
I think it's a bad idea to encourage "drive-by" reviewing. Ideally, reviewers should be familiar with the area of the article and should look at why the article is protected. The whole point of the PC system is that there is not a rush to deal with the edits, so we can take a little time to do it right. If someone is reviewing edits on several articles per minute, they need to slow down. — Carl (CBM · talk) 14:27, 17 June 2010 (UTC)
What would you think of adjusting Special:OldReviewedPages so that it only shows pending pages which have gone without review... say, 5 minutes or more, on the normal view— but with watchlisted pages subject to no such delay. This would give preferential review access to people with past experience on an article without depriving us of the benefit of the entire reviewer base for the purpose of keeping the delay reasonable. It also keeps us out of the business of telling people what they should and shouldn't review. I think the people who do bulk patrolling activities perform an _essential_ service for the site, but if we have both an subject area experienced editor and a high-speed patroller available to review a change, I think that it's obvious that we'd prefer the experienced editor and save the high-speed bulk patroller for the pages where there aren't experienced people currently available. Sadly, the way that its currently setup the system rewards haste. If you're not the fastest on the draw you'll have nothing to do. --Gmaxwell (talk) 15:06, 17 June 2010 (UTC)
Actually, I think that this solution risks giving the impression of reviewers' ownership over given articles; with the exception of sockpuppetry, which requires a little bit of research, all other cases when an edit shouldn't be accepted are rather clear-cut and do not require particular knowledge. Ideally, all the edits I would revert without accepting would be reverted without being accepted by you too, because if there can be any doubts, then, IMHO, it's not proper for a reviewer to use his tools. Salvio ( Let's talk 'bout it!) 23:12, 17 June 2010 (UTC)
Yuck. That wouldn't be a desirable outcome. ... My initial though was to propose that we adjust the software so that people who spend very little time reviewing don't see the most recent revisions pending... but I thought people would like that less. With your description in mind, perhaps it would be the other way around. Review-time is a 'neutral' metric that wouldn't create any consistent ownership. Of course these should be obvious things, but all things equal more care will mean better decisions. If someone changed a DOB in an article is that a good correction, an honest error, or a piece of nasty semi-subtle vandalism? Someone who is simply trying to review as fast as possible won't, on average, make as good of a decision as someone taking a little more time, but if we don't change the way the queue is handled the reviews work disproportionately go to the fastest, even while we have more careful users eager to contribute. --Gmaxwell (talk) 06:08, 18 June 2010 (UTC)
You've picked a very good example! When I'm huggling and I see that kind of changes (a DOB, a chart position, the length of a song or something like that), assuming good faith I usually leave it alone, unless it is clear that I'm dealing with vandalism (e.g. John Doe was born on 31.02.2058). As far as PC are concerned, anyway, I understand your point and to some degree I agree with you: vandalism should never be allowed in articles — and sneaky vandalism is the worst kind thereof —; however, if we begin to check the DOB of someone, before accepting the edit in which it is changed, we're taking the first step towards editorial control over edits, which is what, quite frankly, I'd be trying to avoid here. In that case, I would accept the new DOB, even though I know it to be incorrect and, then, I would correct it (ideally, providing a reliable source). If the editor should change it again, then I'd approve it and engage said editor on the talk page. Salvio ( Let's talk 'bout it!) 13:11, 18 June 2010 (UTC)

One proposal I heard from Risker, which I think is very good, is to see if we can get a "review notice" for each page, like the "edit notice". This could be used by admins to give a short, neutral description of the purpose of protection. It would have more room than the protection summary. If this was shown on the "accept revision" screen, it would make it easier for reviewers to know what to look for. — Carl (CBM · talk) 15:10, 17 June 2010 (UTC)

This is a very good proposal! Salvio ( Let's talk 'bout it!) 23:12, 17 June 2010 (UTC)
Fantastic. --Gmaxwell (talk) 06:08, 18 June 2010 (UTC)

Automatic granting of "reviewer" bit

Just curious... could someone point me to the discussions regarding who would get the bit? I'm curious as to the standards, as there seems to be an auto-generated list of editors with a certain number of edits/recent activity that are getting the bit automatically. For the most part, I'm sure this isn't an issue, but I've noticed at least one case where an editor with past issues (blocks/edit warring) was given the bit without these factors being considered. When I removed the bit (noting the editor in question that he could of course apply for the bit if he felt he needed it) it was restored by another admin, and I was informed that it could only be removed through an AN discussion. (Note that I don't have any desire to stir up trouble for the admins who did the above; I just want to review the policies in play for granting the bit.) Thanks in advance. --Ckatzchatspy 19:48, 18 June 2010 (UTC)

These editors had the ability to edit semi-protected articles with immediate effect prior to pending changes going live. Why should that be any different now? –xenotalk 19:52, 18 June 2010 (UTC)
Xeno, I'll try to look this up myself as well, but could you please clarify - does the "reviewer" bit mean your own edits are auto-approved, or that you have the ability to approve other edits? --Ckatzchatspy 19:58, 18 June 2010 (UTC)
Both, but prior to Pending changes going live they had the ability to tacitly approve other edits as well - by not undoing them. –xenotalk 20:00, 18 June 2010 (UTC)
So what happens when an autoconfirmed editor without the "reviewer" right edits a pending changes protected page? Does it have to be approved or is it accepted automatically? HJ Mitchell | Penny for your thoughts? 20:03, 18 June 2010 (UTC)
It is accepted automatically unless there are pending changes from non-autoconfirmed users. –xenotalk 20:04, 18 June 2010 (UTC)
OK, if you'd humour one my question, what would happen if a reviewer or admin edited the page with pending changes (that had yet to be accepted) but didn't review the prior edits? HJ Mitchell | Penny for your thoughts? 20:07, 18 June 2010 (UTC)
The reviewer/admin's edits wont go live until they review the pending changes. –xenotalk 20:14, 18 June 2010 (UTC)
Basically, it doesn't matter all that much. The first time (or so) that they abuse the Reviewer bit they'll get it removed, and their edit will be undone, and that will be that. You don't want to give it to non autoreviewed editors, or even recently autoreviewed editors (to avoid account farming issues), but apart from that, it's of very limited harm. Really, people are making a big song and dance about this feature, but it's still just a wiki, and nothing much has changed, if you're up to no good, people will notice and stop you.- Wolfkeeper 00:40, 19 June 2010 (UTC)

Pending changes patrol

Is there a page where all unreviewed changes are listed , similar to that for patrolling new pages? Mjroots (talk) 14:26, 18 June 2010 (UTC)

Yeah, Special:OldReviewedPages, linked at the top as "Pages with pending edits". VernoWhitney (talk) 14:40, 18 June 2010 (UTC)
Couldn't see it for looking! Mjroots (talk) 14:49, 18 June 2010 (UTC)
Complete with Special:Log/review. Should this be linked to from Special:OldReviewedPages? ErikHaugen (talk) 14:52, 18 June 2010 (UTC)
When you click on the top left link (when it displays "Pages with pending edits"), so that it goes to the Special:OldReviewedPages page, it also seems to fill in the "Category" fill-in form with the current pg that I'm on, which narrows down the search, of course. Is anyone else having this problem, or is this the way it's supposed to work? Thanks, --Funandtrvl (talk) 21:12, 18 June 2010 (UTC)
1. How can one tell if someone is already reviewing the edit? Is that the "under review" note? If no, what method is to be used. If yes, how is it inserted on the page?
2.How does one watchlist the list of edits awaiting review? Bielle (talk) 16:00, 20 June 2010 (UTC)

On denying requests for reviewer permission

When should a user be denied the reviewer permission? Clear guidance needs to be provided on this.

For reference, requests have been denied 7 times at WP:PERM/RW (5 +2). One of these users brought the matter to ANI (perm).

I'm of the opinion that the reviewer permission simply explicitly (re-)enables an ability that all autoconfirmed users had previously: the ability to approve (tacitly - by not undoing) non-autoconfirmed edits. If an established user requests the ability to do this, it should only be denied in exceedingly rare cases. –xenotalk 15:01, 20 June 2010 (UTC)

My thoughts on this are the same. I've wondered why this can't be given automatically, by the software, to autoconfirmed users (perhaps the threshold could be set a bit higher, say 50 edits and 5 days, whatever). Gwen Gale (talk) 15:29, 20 June 2010 (UTC)
Because anyone could create an army of sockpuppets and get them over that threshold and because nobody with a mere 50 edits and 5 days experience would have sufficient knowledge of policies like vandalism, BLP etc to really know what they're doing. No comment on this specific case. HJ Mitchell | Penny for your thoughts? 15:35, 20 June 2010 (UTC)
(Non-admin here) I concur with HJ Mitchell. If there has to be a threshold, I'd say at least 500-1000 and 3 months (and no blocks if possible, except on counts of 3-rr). Salvio ( Let's talk 'bout it!) 15:44, 20 June 2010 (UTC)
To avoid conflating issues, any suggestions regarding auto-granting of the right should be discussed above at #Automatic granting of "reviewer" bit. –xenotalk 15:46, 20 June 2010 (UTC)
  • Interesting to see the granting of this new bauble (and no doubt in due course its removal) developing in exactly the direction that I predicted that it would. Malleus Fatuorum 15:49, 20 June 2010 (UTC)
Yep. Either way, I think reviewer rights should be handed out like Chiclets to autoconfirmed users (which is almost already happening). Gwen Gale (talk) 15:56, 20 June 2010 (UTC)
So what do we do when someone with 12 edits starts allowing libel into articles, whether abusively or inadvertently? Nobody with that little experience can have sufficient understanding of our policies (vandalism, BLP, etc). HJ Mitchell | Penny for your thoughts? 16:00, 20 June 2010 (UTC)
Block them or at the very least, take the PC bit away until they acknowledge the worry. Gwen Gale (talk) 16:02, 20 June 2010 (UTC)
Any editor who has been around for a reasonable amount of time (3mo?) or with a reasonable number of edits (1,000?) without any major problems ought to be granted the right on request. Any problems and the facility can easily be removed. Mjroots (talk) 16:17, 20 June 2010 (UTC)
Like Chiclets, at a grammar school picnic, if that hasn't been made illegal somewhere, yet. Gwen Gale (talk) 16:21, 20 June 2010 (UTC)
  • Since I advocate automatic granting of the tool at 1 month/100 edits, I think refusing it at any less level than that is improper; the only exceptions would be (1) community- or Arbcom-banned users and (2) indefinitely blocked users. The latter could request it if their indefinite block is lifted. I do not believe that block logs should be used as a criterion for refusing to grant the tool. And I think that 1000 edits is far too high; people who don't use edit-accelerating tools like Twinkle or Huggle, or who don't do repetitive tasks like adding categories (all of which are regularly used to beef up edit counts) don't usually accumulate edits at that rate. Risker (talk) 16:34, 20 June 2010 (UTC)
Auto at 1 month and 1000 edits, or before that happens, to anyone (not banned or indef blocked) who asks sounds fit enough to me. Gwen Gale (talk) 16:38, 20 June 2010 (UTC)
Well before I got it, I thought it should be handed out like confetti, but now I have it I can clearly see it is a Badge of Merit for Very Important Editors wait! 100 edits and 1 month sounds about right but could the granting of it be accompanied by an automatic little instruction box stating precisely what it is for, given the confusion about that point on these pages - ie obvious nonsense/vandalism/BLP/copyvio only - not edits you happen to disagree with or things better dealt with by tags like [citation needed] ? Fainites barleyscribs 16:45, 20 June 2010 (UTC)
So long as the edit count threshold is low, I don't see a big worry. Gwen Gale (talk) 16:48, 20 June 2010 (UTC)
  • Here's a crazy idea: Since the point of the tool is to indicate the revision is free of vandalism, BLP violations, and the like, why not base it on whether the editor seems trustworthy to accurately identify whether an article is free of vandalism, BLP violations, and the like instead of some arbitrary edit count or time since registering? Anomie 23:04, 20 June 2010 (UTC)
    • How would you judge that? Bear in mind as well that this new "right" actually takes away the right that all autoconfirmed editors previously had, and turns it into an unseemly patronage. Malleus Fatuorum 23:49, 20 June 2010 (UTC)
      • Look at their edit history? Ask them questions? I also don't really buy the argument that this takes anyone's rights away, as all it does is show a slightly older revision of problem articles to readers and search engines until someone verifies that a later version is free of poop jokes and other crap. It's doesn't even prevent readers from seeing the poop-laden latest revision, they just have to notice the link in the corner. Anomie 01:03, 21 June 2010 (UTC)
        • The plain fact is that this change has taken a right away, whether your eyes are open to that or not. I'm not trying to sell you anything, just pointing out the facts. Malleus Fatuorum 01:06, 21 June 2010 (UTC)
          • Believe whatever you want. I'm done here. Anomie 01:10, 21 June 2010 (UTC)
            • Eyes still firmly blinkered then, obviously. Malleus Fatuorum 01:16, 21 June 2010 (UTC)
              • What, exactly, is this depriving you of Malleus? If you edit a page protected under this, it is automatically accepted. The only people this affects are readers who aren't logged in and brand new or unregistered editors and even then, only on the most problematic 2000 of our 3.something million articles. I don't necessarily support this move, but I'm not important enough to stop it, but you're acting as if you've been told you need permission to edit which is just not true. HJ Mitchell | Penny for your thoughts? 01:32, 21 June 2010 (UTC)
                • It's not "depriving" me of anything, as I've chosen to ignore it. What it has done though is to delay the edits of autoconfirmed users, including myself, which would otherwise have been immediately available, until they're accepted by a 14-year-old admin wannabbe. I just don't see that as a step in the right direction. Malleus Fatuorum 01:41, 21 June 2010 (UTC)
                  • That's not quite accurate. Go and make an edit to one of the pages protected with this and see what happens (I know you don't want to, but please, humour me). It'll be automatically accepted because you're autoconfirmed. The only delay your edits will have is if there are edits from new editors pending, in which case you have to wait for those edits to be accepted or reverted before yours appear in the live version, but if you'd accept the reviewer "right", you wouldn't even have to wait in those rare cases. HJ Mitchell | Penny for your thoughts? 02:05, 21 June 2010 (UTC)
                    • I have done, and so I know that it's perfectly accurate. What you ought to know is that I will never accept any of these "rights" until wikipedia's basic corruption is sorted out. Malleus Fatuorum 02:12, 21 June 2010 (UTC)
  • The one I declined wasn't (and still isn't) autoconfirmed so I wouldn't expect that anyone would grant reviewer to them. —DoRD (talk) 00:04, 21 June 2010 (UTC)

Under review?

When Special:OldReviewedPages says an article is "(under review)" does that mean that it is awaiting review and I should hop on in to review it, or does it mean that another reviewer is in the process of reviewing it and I should stay out? Sbowers3 (talk) 18:24, 20 June 2010 (UTC)

I would guess that it is a WP:BOLD-hop-on-it situation. Someone reviewed it, and now the pending version lies there awaiting for a reviewer to hop. Rursus dixit. (mbork3!) 18:35, 20 June 2010 (UTC)
It appears from a limited test on the lab site that the "(under review)" indicator shows up when someone views the diff between the stable and current revisions of the page. Looking into the code of the extension, it looks like the indicator is supposed to disappear 20 minutes later or when a new unreviewed edit is added. Anomie 23:19, 20 June 2010 (UTC)

more pages

Hi all, I have semiprotected oodles of pages - anyone is welcome to go through my logs and try them on this page. My time has been limited so have at it if y'all are looking for more pages to try. Casliber (talk · contribs) 03:37, 21 June 2010 (UTC)

No expert review

It should be stressed that this "review" has nothing to do with quality assessment of the article. This "review" doesn't even remove hoaxes nor synthesized neologisms nor other blatant WP:NOT, just WP:VANDALISM and [[WP:BLPVI O]]s. Rursus dixit. (mbork3!) 18:29, 20 June 2010 (UTC)

Keep in mind, as a guideline, straightforward hoaxes are more often than not handled as vandalism (Wikipedia:Do_not_create_hoaxes#Do_not_create_hoaxes) but, good faith mistaken content is not the same thing as a hoax, is not vandalism and should be accepted through pending changes. Editors can then deal with such content as they always have done. Gwen Gale (talk) 18:38, 20 June 2010 (UTC)
Sorry, good faith mistaken content should be allowed through for other editors to clean up? Either I've misunderstood you, or that sounds a bit ridiculous - to knowingly let through an improper (though not vandalistic) edit for someone else to fix. S.G.(GH) ping! 20:31, 20 June 2010 (UTC)
Not improper, taken as "mistaken" by an editor. Maybe the reviewing editor is mistaken that time. Either way, the edit is only being accepted as not vandalism or a BLP vio and that's the policy. Meanwhile I can already foresee some editors now and then trying to game the tool by not accepting PoV edits they don't like as "hoaxes." Gwen Gale (talk) 20:46, 20 June 2010 (UTC)
Ah, I got you. S.G.(GH) ping! 20:59, 20 June 2010 (UTC)
With a good faith mistaken edit, you accept, and then deal with it as you normally would with an edit summary or a tag and a message or whatever.Fainites barleyscribs 21:06, 20 June 2010 (UTC)
Except that the way you would normally deal with it is to just undo it – which is what you should also do under pending revisions. There is no reason to "accept" an edit that you are going to undo with your very next edit. The edit you are undoing is already in the edit history of the page, and nothing with about pending changes can remove it from there. — Carl (CBM · talk) 21:25, 20 June 2010 (UTC)
I understand the thinking, but that's not what the trial guideline says: Wikipedia:Reviewing#Reviewing_process. It also straightforwardly notes, "accepting doesn't prevent you from later editing the article to address concerns you may still have." Gwen Gale (talk) 21:31, 20 June 2010 (UTC)
Actually, it was what the trial guideline said before you removed that line [3]. But the entire "trial guideline" is up in the air right now until we actually have enough experience with the trial to see how things work; that's why it's marked as "proposed". So I see no reason to be hasty in re-inserting the line, and I have not. Nothing in the trial guideline can be viewed as binding at the moment.
The point of pending changes is to leave us time to look at changes carefully; there's no reason to set up criteria that encourage hasty reviewing. There is no difference in "editorial control" if I approve, or do not approve, an edit before reverting it. But if I approve an edit I plan to revert, the only effect is to publish the edit temporarily to a broader audience, which is a misguided thing to do if I already know I will revert the edit.
Can you explain what possible "control" I might gain by not approving an edit? — Carl (CBM · talk) 21:41, 20 June 2010 (UTC)

To follow up on the previous post: I want to emphasize that I, and many other editors, scan every edit to a large number of pages using tools such as [4]. None of those pages is under pending changes control, and I still have the ability to undo or change any edit that I wish. I do not see pending changes as changing the ability to do this in any way. — Carl (CBM · talk) 21:44, 20 June 2010 (UTC)

The point of pending changes is to screen edits for vandalism and BLP vios (the odd copyvio too if it can be spotted). See also the proposed core trial policy at Wikipedia:Pending_changes#Reviewing, it's very straightforward about this. Note the wording, It does not constitute a guarantee of quality or accuracy, or total adherence to content policies.
Meanwhile, please don't take anything I say here as a hint I think you'd do anything untowards with any tool here. Gwen Gale (talk) 21:47, 20 June 2010 (UTC)
The section you linked to says,
"The process of reviewing then accepting new edits is intended as a quick check, to be reasonably sure that they don't contain vandalism, violations of the policy on living people or other clearly identifiable inappropriate content."
Material that I am planning to immediately revert is, almost by definition, material that I feel is inappropriate. Similarly, material that I can clearly identify as wrong is inappropriate. — Carl (CBM · talk) 21:50, 20 June 2010 (UTC)
Putting it that way, you shouldn't ever be reverting GF material you think is "wrong," you should be undoing it with an edit summary. I think you might take your thoughts to the trial policy talk page, Wikipedia talk:Pending_changes for more input, because you seem to be disagreeing with the core policy behind this tool. Gwen Gale (talk) 22:00, 20 June 2010 (UTC)
Yes; I would undo the changes with an edit summary, not use rollback. I very rarely use rollback. Or, preferably, I can fix the edit to make it better without undoing it. There is no reason I need to "approve" an edit before doing that, though, as far as I can see.
Now, as far as I know, I am in agreement with the "core policy" that clearly identifiable inappropriate content should be filtered out. On the other hand, it appears to me that you are arguing that clearly identifiable inappropriate content should be let through. If that isn't correct, could you explain your position about what I should do when I identify inappropriate content? — Carl (CBM · talk) 22:49, 20 June 2010 (UTC)
There is nothing about PC that has to do with "approving" anything. Gwen Gale (talk) 23:28, 20 June 2010 (UTC)
Hmm? When I look at an unreviewed edit, I can either mark it as approved (accepted, reviewed, etc), or I can edit it and mark my edit as reviewed (or I can do nothing). — Carl (CBM · talk) 23:31, 20 June 2010 (UTC)
I would like to add that I have contributed to some extent at articles for creation. There it is necessary to grant a somewhat through review of the article, and its referenced sources. Mainly this type of focus takes time which would probably not be available should this trial develop into a functioning project of Wikipedia. Therefor if it's not vandalism, or contentious content for a BLP, It should quickly be published for the community with bots alike to find copyvios and deadlinks and these sorts of things. This is where challenges to content can be made and worked out on talkpages. My assertion is that there simply will not be the luxury of time to be as through as you might like. My76Strat (talk) 01:49, 21 June 2010 (UTC)
Bots (which are logged in) will already see the latest version whether it is reviewed or not. Similarly, all edits will appears on editors' watchlists as usual whether the edits are reviewed or not. So the community still has exactly the same ability to do these things regardless whether the edits are reviewed. — Carl (CBM · talk) 01:54, 21 June 2010 (UTC)
It would help if WP pages agreed with each other. I can't find the above mention of "inappropriate" on this page. Peter jackson (talk) 09:55, 21 June 2010 (UTC)
I think there's confusion about the policy. My interpretation is as follows: Stopping vandalism, BLP, and CopyVio issues are the MINIMUM use of pending changes. Because, those issues can't be permitted to exist under any circumstances. I do not think this means that reviewers must treat that as the MAXIMUM use of the tool. In other words, you have to at least stop V, BLP, COPY, and not to at most stop them. How much discretion each editor uses is up to them and their knowledge about the article. If an edit is an obvious violation, don't accept it. If you don't know enough about the context or substance of an edit, let someone else decide, or bring it up on talk. The stakes are frankly not that high, so long as it's not a V, BLP, or COPY violation and it can always be reverted afterwards. If you're misinformed, or waltzing into an edit war, a more experienced editor can step in, and, if necessary, seek a higher level of protection. (talk) 01:17, 22 June 2010 (UTC)

Undoing vandalism with Twinkle


I have a suggestion on dealing with vandalism. The guidelines say that if you see vandalism in an edit pending review you should roll it back. I did just that, but used the Twinkle "rollback" function, which apparently is not an actual rollback, so evidently my edit and the vandalism edit were both in the "to be reviewed" queue for a few minutes. I'd suggest clarifying the guidelines to say that Twinkle rollback should not be used in such situations. ScottyBerg (talk) 13:54, 19 June 2010 (UTC)

Just to clarify: I see that there may be an issue here. My Twinkle rollback (basically the same as an undo) was not automatically accepted, which is contrary to what it says in the guidelines. The guidelines say that an undo of a vandalism edit is automatically accepted, but it wasn't. However, a rollback that I just performed on a test vandalism edit was automatically accepted. ScottyBerg (talk) 14:30, 19 June 2010 (UTC)

I would gather that Twinkle does not (yet?) have anything built-in to handle the acceptance part of a non-native rollback. I don't think it uses the 'undo' button as such. You'll need to either use native rollback or go in and review it manually until such time as it does. –xenotalk 14:34, 19 June 2010 (UTC)
If you use rollback rollback, not Twinkle rollback, it actually takes you to a new page (something like &page=foobar&action=rollback) and the use of that method of reversion will be automatically accepted, but if you don't use that, the software doesn't register it as a revert, so your edit join the queue. Have a look at this bit of vandalism on David Cameron, followed by this revert. HJ Mitchell | Penny for your thoughts? 14:43, 19 June 2010 (UTC)
Yes, I've noticed that in the one rollback that I did on a test page. The procedures page says to either undo or rollback vandalism. I think the problem that I encountered is that Twinkle doesn't use either, but rather its own automated revert mechanism. I may test a Wiki undo just to be sure. ScottyBerg (talk) 14:48, 19 June 2010 (UTC)
My guess would be that Twinkle retrieves an old version of the page- either the one prior the one you're rolling back or the version of your choice if you click "restore this version". By making a dummy edit to an old version of the page, you revert every change since that version, but that wouldn't be recognised as a revert by the software- try it on this randomly selected version of the sandbox. HJ Mitchell | Penny for your thoughts? 15:56, 19 June 2010 (UTC)

YesYFixed. You'll need to bypass your browser cache. Amalthea 10:04, 22 June 2010 (UTC)

What gets flagged and what does not

Not sure where to ask about this, so redirect if there's a better place. This morning, the Barack Obama flagged an IP's simple change of "in" to "within", which I marked as accepted. Less than an hour later, a different IP made this lovely change ("He is gay retarted...") which went right through and had to be reverted out of the article. Why was the first edit held for approval, but not the second? Tarc (talk) 15:40, 22 June 2010 (UTC)

I don't think it went through; someone used rollback to automatically decline the edit (note the "pending revision" at the top of your diff there). –xenotalk 15:41, 22 June 2010 (UTC)
Ahh, I see. Hrm, would be nice if there was a clearer "edit rejected" tag or somesuch. Tarc (talk) 16:02, 22 June 2010 (UTC)
There is no such thing as rejecting an edit... PC doesn't change the normal editing process and it doesn't confer any special reverting privileges to reviewers. It just controls which revisions are displayed by default. The reason for this if more obvious if you consider a sequence of edits: A bad one, followed by a couple good ones, followed by one that undoes the bad one. That final one would be accepted, but the intermediate ones would not, even though only the bad one gets undone, yet they aren't "rejected" in any meaningful sense. --Gmaxwell (talk) 18:51, 22 June 2010 (UTC)
I see that now, yea. I was under the impression that it was like another layer to the editing process, where these edits that are "on hold" are either accepted or rejected. It's all clearer now though. :) Tarc (talk) 19:23, 22 June 2010 (UTC)


I would like to say that the guidance suggested in item 6 of the information box is currently over exaggerated. It makes sense that this should be a resource for interaction and information, but currently has demonstrated a bad track record of answering questions. I have asked at least 5 unanswered questions. I think people who are interested in this project need to considering logging into the IRC channel. Especially experienced editors with advanced familiarity of this trial so as to help disseminate information. If you come on in I'll talk to you there. My76Strat (talk) 15:52, 22 June 2010 (UTC)

"Edit" Button on Reviewing

I'll clarify what this is by explaining how I stumbled upon this "bug".

Per this talk page post (placed there because that admin. originally granted me the reviewer user right), I point to the Revision History of Cat at 04:01 June 23. An edit was made to the list of European words that also mean "cat" adding "chat" as the French word. After going to a translator through a Google search, I verified this information as well as the pronunciation "shah".

I go back to the edit review page. On the reviewing edit (the one on the right), I click "edit" next to "undo". I make the addition and put in my edit summary. I save changes. I am promptly told that the changes are not yet viewable because other pending changes have to be accepted. The edit it's referring to is the exact edit that I just made another edit to. So I click the Accept button to save the previous pending edit, thinking it's going to then put my edit over it. All of a sudden, my own edit at 04:01 says "Accepted by CycloneGU". The previous edit has no indication of acceptance, though I clicked to edit it instead of accepting it first. It's also silly to see a user accepting his own edit! (It should have been auto-accepted NEway.)

I consider this a bug. We need to fix that "edit" button to work a little differently with reviewers somehow. CycloneGU (talk) 04:37, 23 June 2010 (UTC)

I agree in principle to what was just said. I experienced a similar issue to an edit which was missing the closing ref tag. An easy fix was to simply add the closing tag and publish. But this easy fix is not available. I think the review should be delivered in the form of a preview so the editor can make quick corrections and publish. If there is an edit conflict, deal with the conflict as you would any other. To me that seems more simple, and prevents any such publishing which otherwise occurs prior to the review. I'd happily elaborate if my comments are confusing as presented. My76Strat (talk) 06:04, 23 June 2010 (UTC)
I do care to discuss one detail, My76Strat: a preview. Just the preview popping up is a failed attempt at reviewing, as we then have to reread the entire article to verify it's valid. If the preview appears as I presume you mean below the comparative edits, on the other hand, I'd be supportive of it. The reviewing can only be done easiest with the comparative edits as it's set up now, but having the preview below and the ability to edit the preview, automatically approving it when we go to edit and then our submission saving on top - that seems about right. CycloneGU (talk) 06:26, 23 June 2010 (UTC)
What I am suggesting is that when the pending edit is sent it should be sent and received exactly the same as if the Show Changes button had been pressed. From there the reviewer can see the changes, cancel the edit, or make changes and publish. Just seems more simple to me. Currently the edit is quasi published, delivered showing the difference. It must be undone to change and the fact that the edit history shows the edit, the garbage is somewhat visible. If it is transferred in the show changes status, canceling the edit prevents it from ever being part of the articles history. It also allows the reviewer to make spelling corrections or other copy-edit improvements quickly and efficiently, and the edit which is in order can be quickly published as well. My76Strat (talk) 06:46, 23 June 2010 (UTC)
This is an interesting idea, but insofar as we use editing history to "prove" bad behavior for the purpose of imposing sanctions in various places, wouldn't this be detrimental? Or are you suggesting keeping the editing history associated with a user or IP, but just not associated with the article itself? I wonder if that would even be technically possible. — e. ripley\talk 17:03, 23 June 2010 (UTC)
The edit being "quasi-published" is a feature, not a bug. The only thing this tool does is change from "readers and search engines see the latest revision" to "readers and search engines see the latest revision that is marked 'accepted'", the idea being to prevent even the few minutes' visibility of the vandalism. It doesn't change anything about how the actual editing process works. Anomie 18:16, 23 June 2010 (UTC)

Two points:

  1. It is intentional that if a reviewer edits a non-reviewed revision, the changes are not marked as "automatically reviewed". The idea is that we don't expect reviewers to review every time they edit. If they edit an already-reviewed version, we assume they don't add problems, so we automatically marked the change as reviewed. But if you edit a non-reviewed revision, and you want to publish your changes, you need to check the diffs and then mark you edit as reviewed.
  2. It is perfectly acceptable to mark your own edits are reviewed. Just remember that you need to check the whole content of the revision (or look at a diff to the last reviewed version) to make sure no other bad changes slip through.

— Carl (CBM · talk) 18:33, 23 June 2010 (UTC)

In my opinion "bad changes slip through" as soon as they become part of the article history. A vandal can access a prior version, and amaze their friends at what the article says, or the image shown. The small rise gained is perhaps enough incentive to vandalize the page. If the reviewer can simply cancel the edit as in show changes, and the "bad changes" are never apparent on the article history. I suppose some record of the "reviewer canceled" edits would have some productive purposes and could be accomplished. I believe this would also serve as a disincentive to any malformed ideas which lead one to vandalize, in addition to perhaps being more productive. In fairness I have also contemplated some of the negative potential associated with reviewing the edit from show changes. Admittedly its own problematic potential may serve as reasons for not reviewing in this manner. In any case it is my intention to stay as informed as possible during this trial, because I want to see it succeed. Accusing no-one of anything, I believe there are some who would rather see this trial fail. That in it self is problematic. My76Strat (talk) 19:34, 23 June 2010 (UTC)
In other words, you're wanting WP:RevDel to be used much more widely? That's a proposal for elsewhere. Anomie 19:44, 23 June 2010 (UTC)

Help link in reviewing interface

Would a help link in the reviewing interface next to review this revision in the form of a question mark be useful (as in the protection inteface) ? Cenarium (talk) 02:30, 24 June 2010 (UTC)

There are some articles on which this just isn't going to work

I think that applying pending changes protection to Adolf Hitler, Bible, Barack Obama and a few others was an extreme lapse of judgement. Bible was re-protected almost instantly, Hitler was re-protected yesterday and there's a current request at RfPP for Obama to re-protected. Surely it would be obvious to anyone that those three articles (along with some others) should just have been left protected? From what I've seen it appears to be working on a few articles, but Taylor Swift and Selena Gomez have also both had to be re-protected. HJ Mitchell | Penny for your thoughts? 16:50, 23 June 2010 (UTC)

I wish we could just go back to the way it was before the pending changes nonsense. It was much simpler then.--William S. Saturn (talk) 16:59, 23 June 2010 (UTC)
I actually like the reviewer thing - more people are looking out for spam being added, and then any edits that are spam are not immediately available in the article. Granted, obvious spam gets blacklisted quickly and admins. often pick it up, but a point I make so far: the New Pages list continues to have very old newly published pages that people have not marked as being patrolled, while the reviewer list - thus far, though still early in testing - tends to be cleared within a matter of minutes. Maybe with more articles it would be not as routine, but what do I know?
As for having to reprotect certain articles that this was tested on - keep in mind reviewing only applies to IP edits in most cases. If an article needs to be protected because IP edits routinely vandalise them, keep in mind that several vandals also vandalised Cat yesterday (might be very early this morning Wikipedia time) after my review. Reviewers stopped and reversed them, no admin. intervention steps were required; does that mean we need to autoprotect Cat as well? It's no different than Barack Obama in that case, and I also reviewed a vandal attempt there yesterday (I was beaten to the Undo button). Vandals will randomly try to attack somewhere and try to avoid being caught. The only way to stop vandals completely is to forbid IP edits entirely and force everyone to use a username. This might be counter-intuitive, however, and I know this would never happen NEway, but it means that we'll always disagree over the best way to catch them. Reviewers reviewing IP edits is a good step; it allows for us to not have to autoprotect (all) articles, and anyone trying to vandalise them will not be able to do so because the crap never makes it to the article, even if it's only found an hour later. Before this, it would appear in the article for a moment, and occasionally some vandalism DOES slip through and stay there for a while before someone researches it normally and says, "That ain't right". I still think for instance that Taylor Swift is something that can always be improved as she is just beginning a very solid career that might someday rival Shania Twain and others...and forbidding IP edits entirely makes it seem like they are unwelcome even if they are editing in good faith. CycloneGU (talk) 19:04, 23 June 2010 (UTC)
Well, frankly, I would love to make account creation mandatory for editing. In my experience, 85% of IP edits are vandalism. You could go to the recent changes, revert 100 IP edits without looking and only the vast, vast majority would be perfectly good reverts. Take a look at Taylor Swift, there were one or 2 decent IP edits, but they're not worth the extra work created by having to revert so much shit and drive by libel. I for one would rather be worrying about the sourcing or the prose or something than what some horny teenage boy want to deface the article with. HJ Mitchell | Penny for your thoughts? 19:26, 23 June 2010 (UTC)
What is considered "working". I thought the idea that while vandalism edits could still be made, their reversion before the public could see was "working as intended"?--Cube lurker (talk) 19:40, 23 June 2010 (UTC)
Noting the post you are replying to here is the topical one..."working" means that spam is not appearing in the article, and someone like you or I is catching it.
Going back to my other example, here is the "Cat" revision history. As I read it, the last four edits at the time of my typing are spam edits. All four were caught, not accepted, and quickly reversed in 1, 2, 5, and 3 minutes respectively. The one before that was the one I reviewed, and that one was NOT spam. That is the result of the "bug" I posted a little above this chain, but aside from that, I accepted it (though it shows I accepted my own, not the other one, because of the "bug"). The review option gives us a chance to stop spam before it appears in the article. While there is the possibility that a rogue reviewer could be assigned and that person starts approving spam, clearly that spam will eventually be caught by another reader/editor and edited in any case, but the rogue reviewer will not be allowed to continue reviewing and probably placed on alert NEway after all the spam is cleaned up, so I don't think that is even a valid concern. I think the review feature is working as designed, and perhaps the complaint about protected articles being placed as review articles (which is only a test, keep in mind) might be misplaced. Let's keep some (not all, clearly) of them review articles for two months and test them, mkay? Give it a chance before immediately shooting it down.
As a side note, featured articles should quality automatically for review articles (again, some may need to stay protected), so these are good ones to run the test on if they can be located. I know we have more than 2,000 of them, though, so being selective in choosing them is a key principle. CycloneGU (talk) 20:07, 23 June 2010 (UTC)
Please don't even suggest it. I'll fight that proposal tooth and nail; I understand you need to run your experiment somewhere, but mass-disrupting some of the few decent quality articles on Wikipedia is certainly not the place to do it. – iridescent 20:14, 23 June 2010 (UTC)
Clearly I'm inexperienced, then. Featured articles are high targets for spam, I'm sure; wouldn't testing this on SOME featured articles be feasible in conducting a full test of different types of pages? Cat in fact used to be a featured article (checked the talk page), so it clearly has been a spam target and the reviewer efforts are clearly working. Where do you think the experiment should be conducted? CycloneGU (talk) 21:46, 23 June 2010 (UTC)
Quick point here: Ernest Hemingway, today's featured article, is in reviewer's rights to review IP edits. Let's see how that one does, and let's see how other articles that get front page rights do as well. I think it's being changed in a little over a half hour from now and I'll see what goes up then, and monitor that as well! CycloneGU (talk) 22:24, 23 June 2010 (UTC)
I don't agree with all FAs getting pending changes automatically for several reasons. Not all of them warrant it. Some are just not edited that much. I have concerns with what happened in Hemingway's article, per the section I just started below. I have other issues with this idea as well. --Moni3 (talk) 23:48, 23 June 2010 (UTC)
Adding pending changes en masse to any group of articles indiscriminately is a terrible idea. Some FAs suffer next to no vandalism, making it totally unnecessary while others need to be kept on permanent semi-protection. By adding this to high profile articles, all we're doing is creating more work for people to revert the shit and making it harder to improve the articles. Just because the shit isn't visible to logged-out readers doesn't make it any less work. HJ Mitchell | Penny for your thoughts? 01:04, 24 June 2010 (UTC)
All right, I'll clarify here. I'm not saying put ALL featured articles under review ability; I agree that would be a TERRIBLE idea, and some benefit better from semi-protection. Even those that get next to no vandalism, however...I think some of these would be good to use for testing purposes. After all, aren't all reviewers also here to help fight spam? We're here to relieve some of the admin. work in pages we otherwise don't have access to or don't properly understand to use properly, correct? I like to think when reviewing articles, even under this test, that I'm helping with some of the work that needs to be done fighting spam; that is why I think it's a good idea, even on featured articles. But again, some articles should stay protected, I will not argue this. CycloneGU (talk) 01:33, 24 June 2010 (UTC)

There's a provision in the suggested policy and original proposal that when vandalism levels are too high, SP should be used instead, or it would indeed be counter-productive, by driving away community resources for no substantial benefit. There are several competing needs to balance, this is a matter that should be handled on a case by case basis. I expected that it would happen to some of the articles that have been put on PC. Problems are likely to occur on articles on popular subjects since they attract more editing, and more vandalism. But it's okay to try (to a certain point), also it's a common practice for articles, even long-term protected, to be unprotected to test if protection is still needed, we should just be ready to put back under SP. Cenarium (talk) 02:11, 24 June 2010 (UTC)

Has a threshold of failure been determined for articles like Barack Obama and Adolf Hitler? In 24 hours only one IP edit to Gay has been marginally constructive. Everything else is ridiculous. At what point can an admin just change an article back to semi-protection? --Moni3 (talk) 16:15, 24 June 2010 (UTC)

I removed pending changes from Gay. The pending changes template thing is still on the article. I'm not sure what I'm supposed to do there. A little help and pointer about how to fix it would be appreciated. --Moni3 (talk) 20:03, 24 June 2010 (UTC)


Is there going to be one, as for admin?

  1. Non-reviewers could ask for reviewers to look at something
  2. Reviewers could ask for advice in unclear cases

Peter jackson (talk) 09:57, 18 June 2010 (UTC)

I support this proposal; it could be useful. Salvio ( Let's talk 'bout it!) 13:12, 18 June 2010 (UTC)
I support it also. Would be very helpful. Orphan Wiki 13:03, 23 June 2010 (UTC)
Yes, it would be helpful. I support this idea also. SwarmTalk 18:44, 25 June 2010 (UTC)
Created at Wikipedia:Pending changes/Noticeboard, I had suggested it here as well. Please feel free to format and reword. Cenarium (talk) 22:24, 25 June 2010 (UTC)

Time on pending status.

I don't know if it's been yet considered, but I think it is reasonable to set a goal as to the maximum amount of time an edit should remain in a pending status. With coordination, the project can be continuously staffed, perhaps incorporating the Wikipedia-en-review channel at IRC. Consensus is emerging that the review itself should be a very quick verification of no vandalism or contentious attacks. Checking the difference for this purpose should be quick and efficient. Therefor I propose that a consensus goal be reached as to what is acceptable in this regard. I propose that 30 minutes or less be the stated goal. My76Strat (talk) 02:08, 21 June 2010 (UTC)

Here's a radical idea. Edits not reviewed within a time limit (30 mins - 8 hours) get autoreviewed by the software. The tool allows a window to catch vandalism and abuse. The odds are good that if a significant time has gone by and an edit hasn't been reverted, it's valid, just hasn't had manual review yet. Track record shows that on the whole, vandalism catching is very effective.
So perhaps if nobody reverts an edit after X time, and nobody's reviewed it, the system itself marks it as accepted. That way there's a maximum guaranteed backlog/delay. Nothing stops people patrolling Special:System accepted current pages and re-tagging them as "accepted by <username>" if they want, or adding a checkbox for "Auto-accept after a maximum of (time)?" when applying Pending Changes to a page, to make this optional on the few pages we really do need human scrutiny. FT2 (Talk | email) 03:31, 21 June 2010 (UTC)
I don't think the idea is at all radical and in fact was contemplating these same ideas. Naturally for purposes of this trial and project goals, human intervention would be preferred. But it has to be managed properly within the stated objective that this action be accomplished within a time-frame. If human intervention does not accomplish this goal, then an automated process to accomplish it would be favorable to me opposed to an article stagnating in some pending status. My76Strat (talk) 04:11, 21 June 2010 (UTC)

A different idea, proposed by gmaxwell, is that unreviewed edits are hidden from the "oldreviewededits" page for some period of time. My guess, and his, is that there will be too many people eager to patrol that page, not too few. The main thing we want to avoid is people patrolling articles at some high rate (say several articles per minute). It's better to let most edits be reviewed by people who have the page on their watchlist already. — Carl (CBM · talk)
This negates the whole purpose of the test (test, not FR or PR). First you wanted to test human review process, arguing that auto tools aren't enough. Now you fall back to just another automated tool (whether it keeps or kills the unsighted edits). East of Borschov (talk) 03:39, 21 June 2010 (UTC)

PC is still evolving. I don't see a problem with logic that goes "Most vandalism is caught quickly. PC on a page deprives vandalism of public viewing or usefulness. One concern is the possible review lag if used on a large scale. But the few edits on PC pages not patrolled within a couple of hours will be a small proportion and exclude the vast majority of vandalism. We can fall back on usual process for that, PC will have done its workin almost all cases. So if we have an auto-review function after a while, it limits the risk of queue lag and keeps the benefits of PC. FT2 (Talk | email) 03:45, 21 June 2010 (UTC)

This may be the case for 100% deployment. The test, among other things, must evaluate review lag and any problems with review queue. Either the proposal is premature (made before the test has been deployed to its 2,000-article sample) or the test has already failed. East of Borschov (talk) 03:59, 21 June 2010 (UTC)
I fully agree that there are many lessons yet to learn during this trial. I also believe its chance of success depends largely upon the foundation. Stating goals, by metaphoric comparison, is only an effort to ensure the foundation is well built, and not a precursor to failure. IMO. My76Strat (talk) 04:27, 21 June 2010 (UTC)
I've changed my opinion on that during the last week. What I've seen is that pending edits are reverted much more aggressively than anon edits on normal pages. I'd like to get away from that, and rather keep an edit pending until someone has the time or expertise to competently review an edit than forcing a decision one way or the other.
After a quick look at the pending edits we get, and how it works out on deWP, I would expect that we could review the bulk of all trivial, pending edits with Huggle or a more specialized tool even if we significantly widened the number of PC-protected articles (not that there are plans to do that). Non-trivial edits on our highly-watched core articles would also be reviewed within minutes. We'll always have a long tail of pending edits of low-key articles, but I don't have an issue with that. How problematic is it really if edits to an obscure BLP pages remain unreviewed for even a few days? If one goal is to keep BLP violations down to a minimum then forcing the issue by time pressure or automatic reviewing is counter-productive.
Of course, I tend to focus on the maintainance side here, not the editorial side. It is hard for me to judge how off-putting it is if an edit remains queued for hours or days. Personally I think I would have accepted it, as a new user, and am actually accepting it when I edit deWP today. Amalthea 10:22, 22 June 2010 (UTC)
I support the position that it's better to leave an edit pending rather than taking a questionable action simply to get it off the list. --Gmaxwell (talk) 18:52, 22 June 2010 (UTC)
Comment. In the chat room, the bot blurts out, "There are three pages waiting to be reviewed. Please do so now!" (Or something like that.) Something like that almost belies urgency, doesn't it? CycloneGU (talk) 21:40, 25 June 2010 (UTC)

Who Will Review the Reviewers?

What is the process for reviewing the use of Reviewer status? Is there one? I ask because I was brought to this Reviewer page by following a link on the page of an editor who had requested Reviewer status. "Never give power to somebody who wants it," is an old and wise proverb. Several editors have found this particular editor to be an extreme ideologue. Reviewer status is not much power, granted, but it would give this editor a veto over new contributions he would not otherwise have. If he abuses it, what is the process for reviewing his edits, and removing "review status? ElijahBosley (talk) 14:18, 23 June 2010 (UTC)

One can use Special:Log to review the actions of reviewers, maybe a link could be added to the user/talk/contributions of each reviewer. Here you can review my reviews. Something like this: 84user (talk/contribs/reviews) Can anyone tell me how to place that suggestion in my signature (note, html not allowed)? -84user (talk) 18:47, 23 June 2010 (UTC)
Thanks. I should clarify for anybody reading this that 84User is not the ideological editor I had in mind. In fact I recently awarded 84user a wikipedia:barnstar under another account alias in gratitude for a fascinating image 84user found and posted on an article I've been working on. ElijahBosley (talk) 21:01, 23 June 2010 (UTC)
"Reviewer" status is a privilege. Ideally a reviewer has previous experience in an area that has shown an administrator they can be trusted with the task of reviewing. Basically, if an admin trusts someone enough to give them the bit, no one should need to review their work, they should be capable of doing it well. Furthermore, if someone's anti-vandalism work is enough to give them rollback, it should be enough to give them reviewer. Granted, there will be bad reviewers, just like there's bad rollbackers, but for the most part, I don't think we'll have a big problem with them. SwarmTalk 18:54, 25 June 2010 (UTC)
What nonsense. It's not a privilege at all, as until a week or so ago any autoconfirmed user could do what now only this new caste of reviewers can do. It's a joke. Malleus Fatuorum 19:24, 25 June 2010 (UTC)
Yes. Gwen Gale (talk) 19:31, 25 June 2010 (UTC)
What the hell are you talking about?? All a reviewer can do is accept or revert "pending changes". "Any autoconfirmed user" could never previously do this. It's a new feature entirely, so I find it confusing what you mean by "any autoconfirmed user could do what now only this new caste of reviewers can do". As for "privilege", any user group, be it autoconfirmation or adminship, is, by definition, a privilege. SwarmTalk 19:54, 25 June 2010 (UTC)
What the Hell are you talking about? Mad as a fucking hatter. Malleus Fatuorum 19:57, 25 June 2010 (UTC)
Prior to pending changes, autoconfirmed users could approve (tacitly - by not undoing) edits by non-autoconfirmed users. –xenotalk 20:03, 25 June 2010 (UTC)
Right, but I think Malleus' concerns simply lie more along the lines of how 'the ever-expanding elitist hierarchy of Wikipedia is destroying the project'. SwarmTalk 20:16, 25 June 2010 (UTC)
No. Gwen Gale (talk) 20:18, 25 June 2010 (UTC)
In other words, you have nothing to contribute to this conversation. CycloneGU (talk) 21:38, 25 June 2010 (UTC)
We both know what you mean is "I don't agree with you, so I'll try to attack the shortness of your post as a weakness." Gwen Gale (talk) 22:11, 25 June 2010 (UTC)
You may think what you like, but I wish you'd think about it before you did. I have never made any comment about the "ever-expanding elitist hierarchy of Wikipedia", and I dont even believe that there is one. Unlike you, apparently. I have explained my objection to this nonsense elsewhere. Malleus Fatuorum 20:28, 25 June 2010 (UTC)
I for one do not understand why everyone thinks there is an elite group. I have the privilege of being a reviewer based on my past history. If I were making disruptive edits, I would not be a reviewer; it wouldn't be my God-given right. CycloneGU (talk) 21:38, 25 June 2010 (UTC)
If you think it's a privilege to be allowed to do what you were allowed to do before, then I just don't understand you at all. Malleus Fatuorum 23:10, 25 June 2010 (UTC)
...HUH? We're being given the privilege of trying out a new way to possibly do something better. New people have access to something to help watch for spamlicious edits, and I previously did not have the ability to automatically know where IPs were posting to help revert spam. CycloneGU (talk) 00:58, 26 June 2010 (UTC)
Believe whatever makes you feel happy. Malleus Fatuorum 01:49, 26 June 2010 (UTC)
Since I don't understand what you're trying to convey, and since I don't care to start an argument here, I'm just going to completely ignore this exchange. CycloneGU (talk) 01:56, 26 June 2010 (UTC)
'Twas a joke, Malleus. I've never been a supporter of the "pending changes" proposal either. It's just hard to cooperate with someone when they call you 'mad as a fucking hatter' while you're trying to explain something to someone.

Main page featured articles and pending changes

Per the conversation at Talk:Ernest_Hemingway#Pending_changes_and_main_page_day, Hemingway's article was on the main page today. It was not vandalized as much as I had anticipated, perhaps because U.S. schools are out for the summer, but when edits were coming quickly, the pending changes software made it difficult to maintain the article's integrity. I have the following concerns:

  1. The load time to see the diffs is much slower than what normal load times seem.
  2. When I accepted edits, I refreshed my watchlist to see the same edit still waiting to be accepted.
  3. Main page FAs often have quite a few editors watching them to remove vandalism. I edit conflicted several times. I believe the lags caused confusion and edit conflicts. Vandal edits were accepted. Earnest edits to remove vandalism were overturned. Vandalism had to be manually reverted several times.

This was my first experience working with these pending changes on a main page article. If these issues aren't improved, it does not seem as if allowing anonymous editors access to edit should be more important than maintaining the article's content integrity. I'm interested to see where the discussion goes. --Moni3 (talk) 23:46, 23 June 2010 (UTC)

Contrary to my comments in the discussion above this one, I also have my own concerns about the pending changes feature. Some of them mirror exactly what you state in your points. Let me give what information I came, either supplementary or refutative, regarding each point.
  1. The load time to see differences between ANY two edits I usually find to be slow. If I head to Cat (heck, I've used it several times on this page, let's stick with it), go to the edit history, and click any two random edits, it will take a while to load the differences. Thus, this issue isn't a pending changes issue; it's a Wikipedia-wide issue in my experience.
  2. I haven't had this issue yet. Are you giving it the full time to finish submitting to the server? I found last nite, during my first real evening accepting and refuting edits, that it might take up to a half minute to submit my acceptance of the edit. Then there's the "bug" I reported a few threads above this when you edit an edit that you're reviewing and then YOUR edit becomes the one to be accepted, causing otherwise unnecessary steps.
  3. I trust that admins. and other special task editors always monitor the main page; this goes without saying. I'll ignore that part, thus. I will discuss the problem with edit conflicts and such, however. Namely, I went to undo an edit last nite on Domestic sheep. I submitted, and got an edit conflict. I thought, "All right, so it's going to just bring up the text I've entered and compare it to the recently submitted one, so I'll just go ahead and submit it again just to be sure." This...resulted in this somewhat embarrassing edit, merely restoring the spam that another editor had just removed moments ahead of me. I undid it one minute later, but multiple reviewers can cause this problem. My only apparent solution: if OldReviewedPages shows an edit as "Under Review" (in yellow), do not touch it; it's already being handled, and this will reduce edit conflicts. I'm guilty even still of going into one already being reviewed, and it's possible two will see the list at the same time without the header and click it...but other than that, problems should be averted. (Meanwhile, if I now get an edit conflict on an undo, I merely cancel my edit.)
So that sums up some of my experiences, which in general mirror some of yours. I like the system and the idea behind it, but it needs refining. CycloneGU (talk) 00:48, 24 June 2010 (UTC)
The load time issues certainly need addressing. I myself have struggled with the first two problems, but I thought the load time was my laptop playing me up, evidently not. Any page with PC protection on it seem to take much longer than others to load and the diffs take forever to load, then accepting them takes bloody ages and on one or two occasions, after spending about 5 minutes just trying to accept one edit, I go back to my watchlist and it tells me there are still pending changes. HJ Mitchell | Penny for your thoughts? 00:57, 24 June 2010 (UTC)
On the third point, I found a confusion at 01:11-01:14 after two consecutive vandalism edits [5] [6], the second one was reverted but not the first one, and it was accepted, though immediately self-reverted (unacepted) so they haven't been visible long. Conflicts seem inevitable, maybe we should give some advice on this.
We should probably have a place where technical issues can be specifically discussed. Cenarium (talk) 01:08, 24 June 2010 (UTC)
I don't know whether that was an edit conflict or not. Here is the vandalism, which is then reverted, but then another editor, reverted again thereby replacing the vandalism, and finally a third editor seems to have fixed it manually. This happened a number of times. If the pending changes message isn't immediately turned off, or if some sort of lag exists, then these types of problems will persist. Truthkeeper88 (talk) 01:16, 24 June 2010 (UTC)
Yup, the only difference between that one and my own such example was that I caught my own example and fixed it myself. This will be a constant thorn until something is changed in the reviewing system to flag reviews already reverted. CycloneGU (talk) 01:58, 24 June 2010 (UTC)
This kind of things can happen on any article, I don't have specific examples but I see it occasionally, conflicts between RC patrollers. People should be careful when reviewing if there are intermediate edits, bug 24007 may help, and to check the changes in case of conflict. I don't think there's a technical issue, the software won't accept the new revision in case of new edits, but the one the user selected. Cenarium (talk) 02:08, 24 June 2010 (UTC)
Point taken. So the problem, thus, is that more people are encountering the problem for the first time. =) CycloneGU (talk) 02:19, 24 June 2010 (UTC)
My experience at Ernest Hemingway (and other PC-protected articles) is largely what Moni3 and CycloneGU mention:
  1. Load times longer than I experienced at Gaza flotilla raid when this former current event was current.
  2. Accepting a review took "a long time". I typically found it to be an apparent waste of time: by the time I'd entered a review message and hit accept and waited for the review to be accepted - another reviewer had already accepted.
    I'm too stupid when it comes to Ernest Hemingway to do this, but I found on other articles (e.g. South Africa) I was accepting really daft edits on the basis that there were no good reasons for unaccepting or reverting the edit. I'd then spend time trying to copy edit a new section about Baseball in South Africa. Maybe I should have just left it for a more knowledgeable editor, but it offends my inner editor to leave really ugly sections in a major article.
  3. I don't think I edit conflicted, but I certainly saw it - I looked at an example Truthkeeper88 showed us at Talk:Ernest Hemingway#Pending changes and main page day and I saw other examples or possible examples after that.
    ...and I take CycloneGU's point that this may simply be part of the learning curve ;-)
(This all on Firefox 3.6, on Ubuntu 10.04, and a fast, fast PC (quad core, etc)).
There's an obvious advantage with PC-prot, as PC1 lets us "protect" TFA without actually protecting it: "everyone can still edit Today's Featured Article". That's an advantage for casual, unregistered readers - but it's a pain in the proverbial for those editors monitoring TFA. Much more so than semi-protection. On the plus-side, I'll be watchlisting TFAs more carefully now ;-) Today's doesn't seem to have PC-protection, however...
TFOWR 09:26, 24 June 2010 (UTC)
I'm not sure if it's what you mean by "edit conflicted", but there is a bug (bug 24043 where if two editors undo an edit at the same time and the second uses the "accept pending changes" checkbox, then the edit both were trying to undo gets "approved" by that second editor even though that was the intention of neither. Anomie 11:00, 24 June 2010 (UTC)
bug 24043 is titled "Review edit conflicts are not gracefully handled" and in fact shows what happened during the June 23 TFA. A solution would be to advertise this more widely so that reviewers are aware of it. Until the bug is fixed, reviewers should also have a look at the article history to be certain they "reverted", "undid" or "accepted" what in fact they intended. Truthkeeper88 (talk) 12:20, 24 June 2010 (UTC)
This then leads back to the slow load times and the fact that even after I accepted edits they still appeared to need accepting. I couldn't be sure when I refreshed the article history that I was seeing the updated version. I may have been seeing a version several minutes old with several other edits already made. I hope these bugs can get fixed. If not, it does not seem to be an efficient system to use on TFA. --Moni3 (talk) 12:26, 24 June 2010 (UTC)
I had a problem after an attempted edit revert on Julia Gillard last nite (EDT) - there was one edit that I went to work with, and while I tried to revert it something else saved. Eventually it said four edits. I here tried to revert the article back to the previously accepted version. Nope...went up to 7 unaccepted edits. Somehow, at 06:12 Wiki time, Mkativerata managed to fix it, but something strange happened and I know I submitted my edit three times to not have it go through. Note this wasn't a featured article! CycloneGU (talk) 15:23, 24 June 2010 (UTC)

This vandalism just occurred on Ernest Hemingway. I had a pending review message, and the pending review edit was highlighted on my watchlist. I looked at the diff, saw vandalism and almost rollbacked but decided instead to undo. That showed that Cenarium had already reverted the edit, so I did nothing. On refreshing the watchlist the pending change message was still highlighted, another refresh and it disappeared. Had I rollbacked, I would have undone Cenarium's edit and reinstated the vandalism. I believe Michael Jackson is scheduled to be TFA for tomorrow - my suggestion to get around the bug is to avoid rollbacking and instead to use undo, which displays the diffs and allows the use of the edit summary. Truthkeeper88 (talk) 14:32, 24 June 2010 (UTC)

And moments later this from an IP gets through? Not making sense at all... Truthkeeper88 (talk) 14:35, 24 June 2010 (UTC)
Regarding tomorrow's TFA, it HAS to be Michael Jackson. It's the anniversary of his death. It'd actually be funny if Staples Center shows up early in July...on the anniversary of the memorial of his death. *LOL* CycloneGU (talk) 15:16, 24 June 2010 (UTC)
Jackson's article is still on partial protection, which I agree with. I don't think pending changes should be imposed on Jackson's article, and especially not tomorrow. --Moni3 (talk) 15:21, 24 June 2010 (UTC)
Full agreement. If it weren't protected, I'd propose it today. But fortunately, semi-protection was just added early this morning before I went to bed; I was there right afterwards looking for protection on another page. CycloneGU (talk) 15:24, 24 June 2010 (UTC)
I've tried rollback on several which hasn't worked because someone else has got there first. That's fine, but it's alarming if you could find yourself reverting the other reviewer.Fainites barleyscribs 20:31, 24 June 2010 (UTC)
Also - several times I have accepted with the intention of then undoing with an edit summary but the load times have been so slow someone else has used rollback before I've finished! Not impressive.Fainites barleyscribs 21:22, 24 June 2010 (UTC)
If you try to rollback an edit and someone else edited the page in the mean time, it fails, the message displayed is MediaWiki:Alreadyrolled. There seems to be a problem with undo though, or how edit conflicts are handled, but I'm not sure that it's specific to pending changes. Cenarium (talk) 15:49, 25 June 2010 (UTC)

New page patrols

With the number of reviewers approaching 5,000 I thought it might be appropriate to allow page patrols to be only allowed for reviewers. At present, any autoconfirmed user can patrol a page and page patrol sprees often go unnoticed. -- Marcus Qwertyus (talk) 06:07, 26 June 2010 (UTC)

Broken Link

For the duration of the trial, the permission may be removed only in the circumstances specified at Wikipedia:Pending changes#Reviewers.

This link in the Becoming a reviewer section of the project page is broken (as the #Reviewers section does not exist at that place). Only wanted to report the small issue.

mfg, OldDeath - 00:41, 27 June 2010 (UTC)
Apparently merged into the section "Reviewing" in this edit. I've retargeted the link accordingly. Anomie 03:36, 27 June 2010 (UTC)
Thanks ;) - mfg, OldDeath - 20:18, 27 June 2010 (UTC)

File:Redaktor Wikipedia 600px.png, which I guess is being used on de.WP for FR and now here for PC, looks way too much like the American CBS logo, among other things. Why not a green checkmark on the WP globe and be done with it? Gwen Gale (talk) 00:15, 16 June 2010 (UTC)

I agree. It actually looks creepy, even cult-ish. Like the eye of Sauron, done in the naked-baby style of Tweety. I remember first seeing it on Wikibooks and scratching my head. --an odd name 00:41, 16 June 2010 (UTC)
I have opened a thread here. --William S. Saturn (talk) 00:56, 16 June 2010 (UTC)
Not a trademark or copyright worry, that's not why I brought it up. Gwen Gale (talk) 08:24, 18 June 2010 (UTC)

AN discussion moved here

right|100px I received Reviewers access recently and while adding the userbox to my userpage, I noticed something that some US editors will notice....the Reviewers logo looks just like the CBS logo. Take a look. The "eye" in the Reviewers logo is slightly smaller, but it is just like the CBS logo. The Reviewers logo was created by Polish user Derski, so he probably wasn't aware of the CBS logo. Also, the Reviewers logo has a "This media file may meet the criteria for speedy deletion" message and a "This file is a copyright violation..." message. With those, I feel that we should get a couple good logo makers and create a new Reviewers logo that isn't close to the CBS logo for obvious copyright reasons. - NeutralHomerTalk • 01:29, 18 June 2010 (UTC)

Agreed....and preferably one that is a little less creepy (that eye is....) -- AnmaFinotera (talk · contribs) 01:42, 18 June 2010 (UTC)
Second Great Seal of the US BAH-p257.png
Yes, that is way too close. In addition, the Foundation has sometimes been known to object to modifications of the globe logo as well. Mind you, my views of the whole "pending changes" matter are well enough known, so I'll just offer my suggested logo without further comment. Gavia immer (talk) 01:48, 18 June 2010 (UTC)
I second Anma's "creepy" comment; it's like the Information Awareness Office logo. Why not just a little globe with a checkmark superimposed? No need to add the eye of Mordor. --Floquenbeam (talk) 01:57, 18 June 2010 (UTC)
Except that's the symbol for the Autoreviewed flag, a different matter. Courcelles (talk) 02:00, 18 June 2010 (UTC)
Another brilliant idea, shot to hell. --Floquenbeam (talk) 02:03, 18 June 2010 (UTC)
How about a globe superimposed with a float? --Floquenbeam (talk) 02:07, 18 June 2010 (UTC)
I like it. Might find a thumbs up that can "blend in" to the "globe", but I like it. - NeutralHomerTalk • 02:10, 18 June 2010 (UTC)
Just a clarification to what Gavia immer brought up, I am only meaning change/remove the "golden eye" part, not the "globe" part. - NeutralHomerTalk • 02:01, 18 June 2010 (UTC)
Maybe this image? File:EgyptianEyeGold3D.png Obviously it has to be cleaned up a bit, the rusty look didn't quite turn out. Maybe we should continue this discussion elsewhere, as well? —Preceding unsigned comment added by Quinxorin (talkcontribs) 03:55, 18 June 2010 (UTC)
Even better! Who likes what? Should we have a !vote? - NeutralHomerTalk • 04:58, 18 June 2010 (UTC)
I've commented on this where perhaps this thread should carry on (and on the img talk page). I don't think this image is at all fit for a reviewer icon on an opening editing project. Not only does it look too much like the American CBS logo (I don't say that from any trademark or CR worry), but I think any eye graphic would be far too big brotherish and otherwise unsettling. Besides, I think the thing is ugly (and eye graphics are very hard to render in an appealing way). Why not a green (or come to think of it, more neutral blue) checkmark arrow or something like that over the globe logo? Gwen Gale (talk) 08:22, 18 June 2010 (UTC)
Sorry Gwen, was unaware that there was a previous thread on this. Either way, I think blue or green arrow would work just fine. I know I have said that for the other ideas, but they are good ones. We could use any of these arrows and they are allowed under "GNU Lesser General Public License"....whatever that is. So, if we could get a good image marker, I think we could use Gwen's arrow idea and have this thing up and running within the hour. :) - NeutralHomerTalk • 08:32, 18 June 2010 (UTC)
I'm thinking blue, since green could stir up misleading notions of "editorial approval" rather than "acceptance," blue also matches the colour seen in the article edit history. Gwen Gale (talk) 08:36, 18 June 2010 (UTC)
Excellent, all the ones I listed above all blue, so you can choose from them. - NeutralHomerTalk • 08:39, 18 June 2010 (UTC)
IMO the eye graphic in itself may well have captured the doubtful ambience created by the whole PC experiment but I agree that it would be best to remove it given its substantial creepiness factor. Dr.K. λogosπraxis 08:50, 18 June 2010 (UTC)
Truth be told, Dr K, my first thought about the "CBS eye" logo was spot on the same, it canny does "fit" from that kind of outlook. Gwen Gale (talk) 10:04, 18 June 2010 (UTC)
Pending changes logo arrows.png
You know Gwen, how the saying goes about ...minds thinking alike and all the rest. :) Dr.K. λogosπraxis 14:51, 18 June 2010 (UTC)

I threw this together... Gwen Gale (talk) 09:02, 18 June 2010 (UTC)

  • Comment—get over it. Is it really that important/prominent/dreadful? ╟─TreasuryTagconstabulary─╢ 09:11, 18 June 2010 (UTC)
    Thanks for your kind and thoughtful input, TT :) Gwen Gale (talk) 09:13, 18 June 2010 (UTC)
    :P I am serious, though. This is such a minor issue to be having a long discussion about! ╟─TreasuryTagRegent─╢ 09:46, 18 June 2010 (UTC)
    It is a big deal with it verges on copyvio. - NeutralHomerTalk • 09:49, 18 June 2010 (UTC)
    It does not verge on copyvio simply because in the history of humankind, more than one organisation has had the idea of using an eye as their symbol. ╟─TreasuryTagestoppel─╢ 09:52, 18 June 2010 (UTC)
    Another reason we shouldn't have the "eye" in the logo. If other companies have "eye" logos, more copyvios. Bad idea. We have a new idea below, let's go with that. - NeutralHomerTalk • 09:54, 18 June 2010 (UTC)
    No, that's nonsense. Are Crown Finance breaching Royal Mail's copyright because their logos feature the same device? Of course not. I understand that Gwen's concern is about the icon's philosophical implications, and I have to say that they do not strike me as majorly problematic. ╟─TreasuryTagRegent─╢ 10:02, 18 June 2010 (UTC)
    I am not talking about any other company but Wikipedia, because I am only associated (as an editor) with Wikipedia...don't care what the others do. At present, with have a logo that is very close to the CBS logo plus the image has a "This media file may meet the criteria for speedy deletion" message and a "This file is a copyright violation..." message. So these are good reasons to do away with the image and get a new one. We have three ideas (counting Gwen's) given on this thread and they are better they what we have. I am sorry, but this will be an agree to disagree situation. - NeutralHomerTalk • 10:07, 18 June 2010 (UTC)
    It is not very close to the CBS logo; they both feature an eye ZOMG COPYVIO ALERT. Er—no. The easy solution to the copyvio tags on the page is to remove those tags. Oh, they've been removed already. ╟─TreasuryTagquaestor─╢ 10:13, 18 June 2010 (UTC)
As I've said elsewhere, CR and trademark vios are not at all my worries, never were. Icons do have much sway over how users think of software, moreover UI graphics indeed set many and sundry moods so I think this is very much worth talking about. Gwen Gale (talk) 09:53, 18 June 2010 (UTC)
  • I like the new logo. No copy-vios there, made with exsisting icons, not creepy (as others have said), and looks good. I think we should put it in use. - NeutralHomerTalk • 09:19, 18 June 2010 (UTC)
  • I like too. Not only do I like the translucency thing going on (I'm a sucker for alpha and mixing effects), but I can kinda see the two arrows portraying the PC process: one version (the left triangle) waiting for approval to replace the one in front (the right triangle). I doubt that's why Gwen Gale used that design, but I'll think of it that way anyway. --an odd name 09:40, 18 June 2010 (UTC)

Looks like another logo is already being used here and there. I think that one's ok from a marketing outlook ("it's only a time delay thing"), although the clock is a bit misleading as to what an editor can do with this tool. Gwen Gale (talk) 12:06, 18 June 2010 (UTC)

  • I like the proposed logo better than the one we have now; and I concur with Gwen that the one with the clock might be a little misleading. Salvio ( Let's talk 'bout it!) 12:59, 18 June 2010 (UTC)
I don't think we should use the old globe; it's outdated. Also, rollback and other images like these have been updated to fit with the new globe. Quinxorin (talk) 17:41, 18 June 2010 (UTC)
That's unfortunate, especially in light of how poorly received the new logo was. –xenotalk 17:50, 18 June 2010 (UTC)
  • Per Salvio and Xeno, anything but the eye. Just kindly get rid of it. Dr.K. λogosπraxis 15:11, 18 June 2010 (UTC)
  • I like the blue arrow logo. I would also be fine with a clock logo, as long as it looked 3D. Quinxorin (talk) 17:41, 18 June 2010 (UTC)
  • I like the eye, the blue arrows make me think that we should be fast-forwarding something. VernoWhitney (talk) 17:45, 18 June 2010 (UTC)
  • Done. Since there seems to be consensus that the image needs to be changed and that the blue-arrow one is better (but not the best) I have changed the image. If you disagree, feel free to revert. —Preceding unsigned comment added by Quinxorin (talkcontribs) 18:47, 18 June 2010 (UTC)

I have to agree with verno the fast forward button is a bit stranger then the eye. If your going to change it from the eye at least pick something other then the fast forward button. -Tracer9999 (talk) 20:24, 18 June 2010 (UTC)

The yellow eye is a MediaWiki icon that is part of the flaggedrevs extension. [7]. You will "see" it when you view the checked version of a page. That is kinda why it is an eye, the page has been reviewed and accuracy affirmed.
As to the blue fast forward arrows, that is the last thing one should be doing when they review changes. delirious & lost~talk to her~ 22:56, 18 June 2010 (UTC)
They managed to embed the eye into the PC code? What an eyesore. Next step? Embed the eye into the vector skin. Now that would be a sight for sore eyes. Dr.K. λogosπraxis 23:30, 18 June 2010 (UTC)
You have it backwards. The yellow eye has been a part of the flaggedrevs extension for almost 3 years as best as i can tell. The yellow eye is the logo for the extension not some sneaky trick just to offend all those in favour of the blue fast forward arrows. I for one am right fed up with having to change logos in my user box to keep removing the fast forward arrows. Those who knew it was me who brought the logo out of Wikimedia Commons have been asking me what happened to the logo.
Considering i am the one who put the logo into use on English WP and it is me who is to blame for the yellow eye being everywhere it would have been nice if someone told me there was mass objection at copyright, ANI, here, and the Speedy deletion tags on Wikimedia Commons. I did contest the deletion on Commons. If you look at the reviewer user box talk page you would see where the discussion actually started on WP; noöne has replied there. It began on IRC between myself and the one who created the user box and made use of a new page patroller logo. Another hint would be that i created the user category.
Overwriting the file on Commons does not have the consensus of Polish, which is where the logo is for and is accordingly named in Polish. Its also used on German and Arabic. Do you have consensus on those languages too? If not then you might want to revert, upload anew to its own name for English usage, and get a bot to go rename the file everywhere it appears on this project. delirious & lost~talk to her~ 12:33, 19 June 2010 (UTC)

In any case, the logo should be made with vector graphics rather than bitmap. Bitmaps look bad when resizing. --MrStalker (talk) 23:41, 18 June 2010 (UTC)

I agree. Anomie 00:49, 19 June 2010 (UTC)
Nice svg conversion Anome. Looks great. I recently saw it in action. Thank you. Dr.K. λogosπraxis 10:07, 19 June 2010 (UTC)

FYI, I've reverted the changes at commons. You shouldn't upload a new file that's anything but a clear clean-up over an existing file. Among other things, it also changed the image on lots of other sites based on this single short conversation here. VernoWhitney (talk) 12:32, 19 June 2010 (UTC)

Thank you. Among a couple of other things i just pointed that out too Face-wink.svg delirious & lost~talk to her~ 12:36, 19 June 2010 (UTC)
FYI, according to Special:GlobalUsage/Wikipedia Reviewer.svg, the file File:Wikipedia Reviewer.svg is not currently used on any other Wikimedia wikis. Although that file had problems of its own before your revert, in that it lacked the {{LGPL}} license from the Crystal Clear arrows used and didn't reference the authors of the source images. The revert of File:Redaktor Wikipedia 600px.png was entirely justified, of course. Even though I made File:Pending changes logo arrows.svg, I personally don't care which image is used; I just enjoy occasionally converting an image to SVG. Anomie 13:47, 19 June 2010 (UTC)
Yeah, the svg file isn't used on other wikis, but it still shouldn't be overwritten with a different image, at least per unwritten commons tradition, although I'd be surprised if it wasn't written down somewhere. VernoWhitney (talk) 14:01, 19 June 2010 (UTC)
I wasn't speaking of the .svg file with the English name. I was most particularly speaking of discovering that File:Redaktor Wikipedia 600px.png suddenly had blue arrows despite it being the Polish WP/WT logo for their "editor" user group of pretty much the same rights. I honestly am not sure how many people opposing the yellow eye were aware when they opposed that it is actually from the extension itself and has a very direct relation to the reviewer/editor user group. Complaints that it looks like CBS and that eye on the pyramid are true. If you realise that then either CBS or the US government is violating the (perceived) copyright or/and trademark of the other, or neither are and it is likewise not a concern to use here. Considering it has spent about 3 years as the icon for the checked version of page in the FlaggedRevs extension i am pretty sure that both MediaWiki and the WikiMedia Foundation have looked into this. I have known of the logo for over two years now myself, hence i brought it into use when the rights went live on this project. I realise some people like the various icon groups out there but in this instance we have an actual image from the related extension to use. The "reviewer" right and the "eye" for checking/reviewing just make sense to pair together.
For anyone who wants to see where the yellow eye comes from here is the url not hidden inside [ ].
delirious & lost~talk to her~ 14:49, 19 June 2010 (UTC)
  • I agree with NeutralHomer. Whether it's a copyright violation is not the point. It is too similar to the CBS logo. This flagged revisions thing has gotten some publicity, and I don't see the point of using a logo that might get negative attention. ScottyBerg (talk) 15:00, 19 June 2010 (UTC)

+ Globe w Thumbs-up, ala User:Floquenbeam above: "a globe superimposed with a float". This signifies a more positive approach than the big-brother eyeball. Jojalozzo 00:22, 20 June 2010 (UTC)


Looks like {{Template:User reviewer}} has been changed back to the "eye" logo. Copyvio anyone? Doesn't matter if it is SVG or PNG, it looks like like the CBS logo above. We have a new idea, let's stop edit warring over it and just put it into practice. If someone has a better idea (besides "Let's Keep the Old One") add it now, otherwise I think the "arrow" logo should be put into practice, a bot to correct any mistakes on other Wikis and the "eye" logo deleted as a copyvio. - NeutralHomerTalk • 17:11, 19 June 2010 (UTC)

Go tell it to commons if you think it's a copyvio. If not, then just leave it alone. VernoWhitney (talk) 17:17, 19 June 2010 (UTC)
First things first. - NeutralHomerTalk • 17:18, 19 June 2010 (UTC)
I also posted to the only admin I know on Commons, Siebrand, about the problem. - NeutralHomerTalk • 17:28, 19 June 2010 (UTC)
I'm an admin at commons (Siebrands talk was on my watchlist..). I think that the "eye-logo" is clearly PD-inelligible; One circle with an oval (correctly: Vesica piscis) inside, besides that simplified shapes of an eye have been used for centuries. Apart from these issues, i feel that the current logo is more appropriate for reviewing than the fast forward symbol. --DieBuche (talk) 18:12, 19 June 2010 (UTC)
Just want to make sure I understand....does that mean the "eye" logo (the current Reviewers logo) should stay or go? If "go", which one do you think we should add if not the "fast forward" logo? - NeutralHomerTalk • 18:16, 19 June 2010 (UTC)
"Stay"; sry if I was unclear :). btw, I'm not sure how many users from outside the US would immediately associate this as the "CBS logo".--DieBuche (talk) 18:20, 19 June 2010 (UTC)
I am from the US (Virginia to be exact) and immediately thought "whoa, CBS logo". But then again, I work with TV station articles most of the time, so I pick up on things like that. :) - NeutralHomerTalk • 18:28, 19 June 2010 (UTC)

A MediaWiki icon is not copyvio!!! Seriously, do you really think MediaWiki and the WikiMedia Foundation would disseminate copyvio all over the world and use it themselves too, and have done so for years now? I have seriously become tired of it but once more, with feeling i offer the link to the source file that proves its origin and thus implicit recognition as not being copyvio:

As to reverting the image in png format, it was inappropriately overwritten in the first place. That is a global image with origin from and extensive use on Polish projects, as well as others. Noöne had any business overwriting it with a different image per any discussion here. The proper course of action would be to upload the alternative image to its own file name. The png file is even titled in Polish and the description says it is for Polish WP.

Even though you are not so bold now, I can't believe you had the arrogance to advocate getting a bot "to correct any mistakes on other Wikis" and then deleting a MediaWiki icon as copyvio wherever it may be. We have no business telling other projects they are wrong and we're going to fix them automatically. Deleting the image would hit into the software too because this wiki and others that run FlaggedRevs have this icon. It is smaller but it is out there and all over this project too by now i would imagine. Read the checked version of a page; you will see the yellow eye in the notice that it is a checked page. So far as i can tell those crying "copyvio" are the only ones who are making mistakes.

I am not from USA but i too see the similarity to the CBS logo, and the scores of other icons and items in the shape of an eye. It is really more common than you first might think if TV is your focus of WP editing (i too do mostly tv articles). Both reviewer logos can legitimately coëxist. I brought out the original one for the user box because it seemed wrong to use a Patroller logo and odd to use a different logo from what the extension itself and a few other projects are already using, File:Redaktor Wikipedia 600px.png. delirious & lost~talk to her~ 18:56, 19 June 2010 (UTC)

FYI, the new image is up at commons at File:Wikipedia Reviewer 2.svg, so there shouldn't be any more issues with overwriting the existing images. VernoWhitney (talk) 19:00, 19 June 2010 (UTC)

I've never thought of it as a copyright or trademark worry at all but I'll hazard this, the CBS logo gets away with its dodgy symbolism because it was very keenly designed by William Golden. As for that "redaktor" graphic, Frank Stanton would never have allowed it to become the new logo for CBS's TV network in 1951. It's clumsy, unsettling artwork. I don't think PC should have an eye logo at all but if editors want an eye logo, I think it should be redone with a bit of flair into something that looks cool. Gwen Gale (talk) 19:00, 19 June 2010 (UTC)

As it happens, the guy who drew it, William Golden, did have misgivings about the eye he'd drawn as a logo for the whole network. When Golden began dropping hints he wanted to design a new logo, Stanton said, "Just when you're beginning to be bored with what you've done is when it's beginning to be noticed by your audience," and stuck with the eye.[8] Gwen Gale (talk) 19:21, 19 June 2010 (UTC)

I look at the CBS logo and pretty much think of it what you think of the flaggedrevs icon. I don't mind the symbolism (i rather prefer it) but the CBS logo is far from ideal and optimal in its design - take black circle, centre overlay with white oval, repeat with small black circle. Flat eye. The animated one conceived by Mr Golden sounds like it had much promise. I wonder what Mr Golden would think of the "Golden Eye". The more you rework it the more it becomes a derivative and possibly considered original work if it goes far enough from the icon; in equal proportion it begins to loose relevance and relation to the flaggedrevs extension. One could very well end up modifying the yellow eye so much that it becomes a purple heart. :P I made a few based on the rollback logo.[9] If you hate the eye as already seen on the projects you will probably not be fond of these, "This user keeps an eye on Wikipedia as a pending changes reviewer". I like the right view on the first line. I could play with sizing and such but it is an idea.
However the fight over trying to claim copyvio continues. The reverting on Commons is sadly still going on. This is starting to sound like a blog post. Cheers delirious & lost~talk to her~ 23:21, 19 June 2010 (UTC)
Good, I am glad the fight continues. Others on this thread have seen this as a copy of the CBS logo (and in that a copyvio) so it shouldn't stand. I am open to revamping the eye if people like it, to get it away from the CBS look, but I don't feel we should get into WP:ILIKEIT and WP:IDONTLIKEIT territory. I feel we should work to a common goal and get a logo that has zero chance of even being closely considered a copyvio and something everyone will like. Obviously the "fast forward" logo wasn't that, but I think we could get something all will like. - NeutralHomerTalk • 23:34, 19 June 2010 (UTC)
Oooo. A party: commons:Commons:Deletion requests/File:Redaktor Wikipedia 600px.png, commons:Commons:Deletion requests/File:Wikipedia Reviewer.svg, andcommons:Commons:Deletion requests/File:FlaggedRevs-2-1.svgVernoWhitney (talk) 23:41, 19 June 2010 (UTC)
Well done William S. Saturn. Now with that out of the way, we work on a new logo. - NeutralHomerTalk • 23:47, 19 June 2010 (UTC)
  • This is in NO way a copyvio or a trademark violation in regards to the law. There is a strong case to say that the cbs logo itself isn't eligible for copyright (many trademarks aren't still trademarked obviously) but even ignoring that fact the current ones are different enough to not be a problem. They also do not infringe in ANYWAY on trademark law and to say so only shows a misunderstanding of the law involved. Trademark does not in any way = copyright. Also there seems to be alot of comments about "a lot of complains on enWiki" about the logo, as a quick aside reading this thread seems to be a VERY few people (like on one hand) making lots of comments against. That does not equate to a lot of unique comments. I'm not saying you can't decide to replace it, it ain't preferct, but copyvio it ain't either. James (T C) 01:38, 20 June 2010 (UTC)

Break 2

  • OK, since everyone but VernoWhitney and Deliriousandlost, even the people saying "not a copyvio" say the image isn't perfect...what would you like to see? - NeutralHomerTalk • 03:03, 20 June 2010 (UTC)
    Not everyone. VernoWhitney (talk) 03:26, 20 June 2010 (UTC)
    There, happy? - NeutralHomerTalk • 03:37, 20 June 2010 (UTC)
    Deliriousandlost appears to be fine with it too. Just accept that it's not everyone. VernoWhitney (talk) 03:45, 20 June 2010 (UTC)
    Um, no. 2 out of alot, still needs to be replaced. I don't step down from battles easily. - NeutralHomerTalk • 03:46, 20 June 2010 (UTC)
    I was unaware this was a battle. Silly me. VernoWhitney (talk) 03:48, 20 June 2010 (UTC)
    Battle to get rid of something that is too damned close to the CBS logo. Think we should get Godwin involved? Makes things more interesting when we get the legal guy involved, I think. - NeutralHomerTalk • 03:51, 20 June 2010 (UTC)
    For the pure amusement I would get from Mr Godwin giving you instruction in trademark and copyright law I welcome your suggestion to request he comment on this. delirious & lost~talk to her~ 04:16, 20 June 2010 (UTC)
    Will do that on Monday (Wikimedia isn't open on the weekends). - NeutralHomerTalk • 04:20, 20 June 2010 (UTC)
    This is not the first discussion on this logo. I read one long ago which is why I think I knew of this logo. I will be honest in saying I am not sure if it is here, on meta, on mediawiki, or somewhere else. If I can find it i'll link it. delirious & lost~talk to her~ 04:38, 20 June 2010 (UTC)
    If others have raised concerns in the past, then this is even more reason to find a logo that all will like and raises no concerns of any kind. - NeutralHomerTalk • 04:42, 20 June 2010 (UTC)
  • just a note: I'f you are going to send Mike a note I would send it to him as soon as you can so that he can answer it on his own time. Believe me he checks and responds to email at all times of the week/day :) James (T C) 05:12, 20 June 2010 (UTC)
    • Email sent. Wasn't aware the legal staff worked on weekends. - NeutralHomerTalk • 05:42, 20 June 2010 (UTC)
      • The below is the response I got from Mike Godwin, posted with permission. - NeutralHomerTalk • 06:27, 20 June 2010 (UTC)

With regard to similarity between the CBS logo and this one, the issue isn't really one of copyright but of trademark. CBS would almost certainly argue that there's some kind of trademark infringement going on in the reviewer logo. If the reviewer logo were a better one, it might be worth arguing over, but there's a lot wrong with the logo overall from a design standpoint (combination of color and grayscale, failure to abide by puzzle-globe design guidelines, overall lack of aesthetic simplicity, etc.) that I would probably just remove it if CBS ever complained, which they are likely to do.

Better to start over -- come up with a reviewer logo that isn't a knockoff of either the puzzle globe or CBS's eye.

        • I think with that, I do believe we should try and find a happy medium. Let's get some logos together that are already on Wikipedia (like the blue arrows, those were licensed under "GNU Lesser General Public License" hence we could use them) and put them on the Puzzle Globe and put something new together that everyone likes and has no possible way being recognized as something else. Why not something with this logo since we are reviewing, viewing, scrutinizing an article. - NeutralHomerTalk • 06:30, 20 June 2010 (UTC)
          • Would you please ask some followup questions: What exactly are the puzzle-globe design guidelines? And should various other "something superimposed on the puzzle-globe" images, such as File:Wikipedia Rollback.svg or File:Admin mop.svg, also be redesigned to not be a "knockoff" of the puzzle globe? And if not, why not? Anomie 12:41, 20 June 2010 (UTC)
            • Given the comment attributed to Mr Godwin above i have emailed him asking for a direct comment that also addresses the fact that the icon is in broad use by the WikiMedia Foundation and has been for some time now. delirious & lost~talk to her~ 14:35, 20 June 2010 (UTC)
              • Mr Godwin replied to my email & confirmed the comment above quoted. Mr Godwin also suggested taking this to Jay Walsh and the design team since it would be best to replace the icon in the FlaggedRevs extension too. I have accordingly emailed Mr Walsh. delirious & lost~talk to her~ 16:03, 20 June 2010 (UTC)

As I've been saying all along, it's not a copyright or trademark vio, but it looks too much like the CBS logo. It's also ugly and unsettling. Gwen Gale (talk) 08:16, 20 June 2010 (UTC)

I just wanted to be certain on the vio part, hence my going to Mike. I agree with his "better to start over" suggestion. - NeutralHomerTalk • 08:17, 20 June 2010 (UTC)
As Mike said they would likely claim it would be (which they will always claim doesn't mean much) I have no problem with starting over but we aren't going to delete all of them from commons because we have to remember this is part of the software as well (the eye is). My guess is Mike didn't realize that part just saw the logo you pointed out, and yea he is good at responding quickly :) James (T C) 08:30, 20 June 2010 (UTC)
Software graphics aren't too hard to update. I would think that, one way or another, it needs to be dealt with meta-wide. As for the copyrighted puzzle globe, it's also used in the rollback logo. Gwen Gale (talk) 08:32, 20 June 2010 (UTC)
I have no control over Commons (just pop up to upload an image there and that is it) so if Commons wants it, they can have it. I think something with this logo would be a good start. - NeutralHomerTalk • 08:33, 20 June 2010 (UTC)
Oh, let me make a clarification, I have ZERO problem with the Puzzle Globe and think it should be used for the new logo. - NeutralHomerTalk • 08:34, 20 June 2010 (UTC)
The worry about the puzzle globe is that it's copyrighted by the WmF, it's not free. I ran into that when I uploaded the quickie "FF-PC" logo I'd done and was trying to put together the licence info. Gwen Gale (talk) 08:41, 20 June 2010 (UTC)
I don't think we should worry about the globe as we use it on every single page of Wikipedia, plus many userboxes, barnstars and such. So using it now isn't a worry I don't think. - NeutralHomerTalk • 08:50, 20 June 2010 (UTC)
I agree: logos for admins, rollbackers, autoreviewers and so on all have the puzzle globe... I'd support a logo with a magnifying glass on the globe. Salvio ( Let's talk 'bout it!) 11:31, 20 June 2010 (UTC)

Pending changes logo mag.png I've done another, how 'bout something like this? Gwen Gale (talk) 09:19, 20 June 2010 (UTC)

Wasn't it you that said these should be on the Globe or is it not necessary for these to be on the puzzle globe? - NeutralHomerTalk • 09:23, 20 June 2010 (UTC)
Mike said above, "come up with a reviewer logo that isn't a knockoff of either the puzzle globe or CBS's eye." That means no puzzle globe. The worry would be that the puzzle globe is copyrighted and shouldn't be mixed with other stuff. Gwen Gale (talk) 09:24, 20 June 2010 (UTC)
OK, then I am cool with your version. Sorry about making it smaller, was just trying to keep all the pics on the right so they didn't bunch up the text. Didn't mean to goof things up. - NeutralHomerTalk • 09:30, 20 June 2010 (UTC)
Only so you know, you made it bigger, which is why it jagged when displayed. :) Gwen Gale (talk) 09:32, 20 June 2010 (UTC)
Oooops! I was intending to make it smaller (like 150px). Won't do that again. :) Either way, I think it is a good logo, just need others input. Probably won't be up for awhile. - NeutralHomerTalk • 09:41, 20 June 2010 (UTC)
It's 128x128. Gwen Gale (talk) 09:46, 20 June 2010 (UTC)
Oh, I don't care if this one ever goes up as the PC logo, I'm only trying to help out. Gwen Gale (talk) 09:42, 20 June 2010 (UTC)
Pending changes logo viewmag globe.png

If the globe's ok to use, here's a take I've done on the globe with the magnifying glass. Gwen Gale (talk) 12:06, 20 June 2010 (UTC)

IMO, that implies searching for something rather than reviewing. Anomie 12:42, 20 June 2010 (UTC)
One is searching/taking a closer look, for vandalism, BLP vios and so on. Me, I like the one above with the blue and goldenrod lines. I only put up the globe (which Godwin already said shouldn't be knocked off for this) to show how it might be done with the globe. Gwen Gale (talk) 12:46, 20 June 2010 (UTC)
I like the one with the magnifying glass and the globe (just as much as the one with the "fast forward" sign); I don't really like, on the other hand, the one with the blue and golden lines, because I think that it would be inconsistent with the other logos, all of which display the globe. Salvio ( Let's talk 'bout it!) 12:54, 20 June 2010 (UTC)
Mr Godwin confirmed to myself that he would not litigate the matter should CBS object to the eye icon/logo. Accordingly and on Mr Godwin's direction i have emailed Jay Walsh to seek the input of Mr Walsh and the Wikimedia Foundation design team for design advice on selecting a new icon for the extension that can also be used as the logo for the user boxes on all of the projects that have the FlaggedRevs extension installed and where there are "editor" or "reviewer" user groups. delirious & lost~talk to her~ 16:03, 20 June 2010 (UTC)
That one I like, Gwen, but if the eye logo is to remain, I want two fangs under the eye, and an atomic bomb cloud rising out of the hole as an alternative to be displayed for reviewers that are self described as extra rabid reviewers! Rursus dixit. (mbork3!) 17:34, 20 June 2010 (UTC)
Note about the Magnifying Glass version above: I believe this was discussed before (at the very least as part of the software) the problem is that it would regularly appear right below the search box which also uses a magnifying glass. James (T C) 19:31, 20 June 2010 (UTC)
Understood, although how likely would someone be to muddle one of those teeny tiny badges (and an even more itty bitty magnifying glass within) with the big search function icon? Gwen Gale (talk) 20:23, 20 June 2010 (UTC)
I definitely like the magnifying glass on the globe. Nicely done! :) - NeutralHomerTalk • 23:32, 20 June 2010 (UTC)
Pending changes logo stripes globe.png

Here's yet another. Gwen Gale (talk) 20:02, 22 June 2010 (UTC)

Wow...I'm not sure where to jump in to this conversation. I'll start by saying that the eye logo isn't my favorite, and isn't one that we've been intentionally using on English Wikipedia. The icon we have been using also isn't terribly popular, so we're rooting around for an alternative. If you have ideas, please let us know on flaggedrevs.labs - RobLa (talk) 22:59, 22 June 2010 (UTC)

I still like the one with the magnifying glass. Not liking the one with the globe with the stripes....I think we should call that the Zebra Globe :) - NeutralHomerTalk • 02:06, 23 June 2010 (UTC)
Wikipedia Checkuser.png
Wikipedia Checkuser.svg

I, personally, prefer the icon to have the globe as all the other userrights do. The magnifying glass above is a nifty icon, although since the Checkuser icons have the same symbology I am not 100% on it. It is obvious different, but some people may still have concerns. We also want to make sure we don't get sued by any magnifying glass manufactures. :) Just my 2 cents. Avicennasis @ 16:32, 24 June 2010 (UTC)

Well, I saw this discussion earlier, but, since I didn't have anything to contribute, I didn't feel the need to join in. However, I was sad to find out that this discussion never produced any results. What happened? Could we get this discussion going again to possibly change this thing? SwarmTalk 09:16, 5 July 2010 (UTC)

Don't change it, stop freaking out. I read that WHOLE thing above and it's a few editors (being generous with a few) whining about it. Love, Beam 00:06, 7 July 2010 (UTC)

Really? Looked to me like a very large portion of editors were in favor of changing it. SwarmTalk 02:33, 12 July 2010 (UTC)
You know what, whatever. I went ahead and created my own damn topicon. Problem solved. SwarmTalk 03:08, 12 July 2010 (UTC)

Maybe this is too late

I am new to the discussion because I simply did not know it was going on (and was preoccupied). Have people reached a consensus about how reviewers will be selected etc.? I have concerns about creating more hierarchy at WP (although I think it is a step forward, to have finer gradations in the level of protection of articles, and a desire to create a mechanism by which protected articles can be edited - if I get the gist of what is going on accurately) Slrubenstein | Talk 00:35, 12 July 2010 (UTC)

In case you're not aware, WP:pending changes is in a trial run right now, so the reviewer usergroup is active. Initially, the reviewer right was apparently granted randomly to rollbackers and autopatrollers and other trustworthy users (many got it without asking for it). It can be requested at WP:PERM. All admins are automatically reviewers. Basically, any trustworthy user who knows core policies can get it. The basic criteria are listed here. The "how to" for reviewing is here.SwarmTalk 02:49, 12 July 2010 (UTC)
Let me rephrase: Yes. SwarmTalk

Sooooo.... the answer to my question is yes? Slrubenstein | Talk 10:33, 12 July 2010 (UTC)

The short answer is yes, but:
  1. This is just a trial.
  2. Any consensus on Wikipedia can change over time.
Yaris678 (talk) 18:57, 14 July 2010 (UTC)

Do intermediate versions matter if you are going to accept a change?

With regard to the step-by-step guide... Do intermediate versions matter if you are going to accept a change? I am trying to think of why. It may be that someone has removed valid info that someone else has just added… but that may just be an ordinary difference of opinions between editors… can we assume that it is vandalism, or any of the other problems this feature is supposed to keep out? Yaris678 (talk) 19:26, 14 July 2010 (UTC)

If the question is "Do I have to worry about accepting any of the intermediate revisions?", the answer is no. If you're going to make an edit to the latest revision, there's not even a need to accept that before making your edit, although you certainly may accept it first if it meets the reviewing criteria and probably should if there is any chance your edit would be legitimately reverted.
If the question is "Should I check the intermediate revisions to see if Editor A added some useful information and then Vandal B removed it?", the answer is yes. But that's really no different from what you should theoretically be doing anyway. Anomie 19:46, 14 July 2010 (UTC)
But how would we know that Vandal B is actually a vandal, and not just someone who disagrees with Editor A? Obviously, we could get involved in a discussion between the two of them and get them to explain there different positions, but this seems far from the purpose of reviewing.
Yaris678 (talk) 21:10, 14 July 2010 (UTC)
How do you know any editor is a vandal? You look at their edits. Anomie 22:58, 14 July 2010 (UTC)
And the edits are the removal of something that looks valid. How do you know this is vandalism and not just a disagreement between editors? Yaris678 (talk) 07:08, 15 July 2010 (UTC)
If there is a list of waiting pending changes, I just start at the oldest and work through to the latest and revert or accept as I would reviewing a normal edit to an article I am watching. There are many reasons not to accept an edit, not only vandalism. Off2riorob (talk) 11:39, 15 July 2010 (UTC)

I had a conversation with Yaris678 about what he was trying to actually get across here. You can read the conversation we had, but here's the basic idea. Suppose you have two pending edits from two different editors, where the earlier edit added some content and the second edit removed some or all of it for a personal reason, not because it violated policy. These two editors are involved in some kind of edit war, so as a reviewer, should you approve the latest version (with the content removed), or do you approve the intermediate version, or do you just do nothing?

My personal opinion is that, if you approve any of the versions, that's implicitly picking one side of the argument over the other, so I would say that it's better to err on the side of caution and just not approve either one. Yaris, did I represent the issue correctly above? — HelloAnnyong (say whaaat?!) 12:04, 16 July 2010 (UTC)

Yes. That is right. The only twist is that I see reviewing as mainly about stating whether the latest version is fit to be shown to non-logged in editors and hence the reviewer should just approve the latest version. That is where I differ from Annyong. Neither the approach advocated by myself nor Annyong is described in the step-by-step guide as it stands. Yaris678 (talk) 13:23, 16 July 2010 (UTC)
(edit conflict) Or you could avoid taking sides by approving both the versions, as long as they're both free of vandalism, copyvios, BLP issues, etc. Anomie 13:24, 16 July 2010 (UTC)

So I think we see the issue. Do we keep reviewing purely for the stated purpose?

Or do we also see the acceptence of a version as having a political dimension? Do we have to mark other acceptable versions as such, even if they are not the latest version, so as to avoid being seen as taking sides in an edit war?

Yaris678 (talk) 17:21, 18 July 2010 (UTC)

Confused, sorry

If I leave a comment in the comment box while reviewing, where does that go? Right now it appears to disappear into the ethers... CastAStone//(talk) 04:34, 19 July 2010 (UTC)

It goes into the logs its not a summary that goes with the page.--Steam Iron 05:36, 19 July 2010 (UTC)

Granting reviewer to users with a history of copyright issues?

While I understand that we're still in a test stage, I begin to be quite concerned about reviewer rights being handed out to contributors who turn out to have quite a history of issues with copyvio - something that is apparent by looking at the contributor's talk page history for instance.

As the reviewing process makes it clear that revisions with obvious copyvios should not be accepted, I get sceptical when the rights are granted, in particular, those users who reacted badly when informed of the issues they had created. Multiple CorenSearchBot warnings should probably also at the very least raise an eyebrow. Trialling it out as widely as possible is one thing, but I believe it might be worth having a thought or two on criteria which should disqualify (at least until further discussion) users from getting the reviewer status. MLauba (Talk) 22:48, 22 July 2010 (UTC)

I have also come across a worrying number of users who have been granted reviewer rights - not only for copyvio but with multiple warnings on their talk pages for various issues - including senseless edit warring to the point of vandalism. I know the lists were bot generated, but I suggest that the allocation of rights has been carried out too loosely.--Kudpung (talk) 05:06, 27 July 2010 (UTC)
I know its too late to remove the privileges from a lot of the users who were granted it initially, but as I just came across a user who was granted it only a day after CorenSearchBot tagged a newly created article which was almost entirely a copyright violation, and copyvios are one of the 4 things that reviewing is supposed to stop, is it too much to ask that not having created copyvios is one of the prereqs for being a reviewer (just like vandalism, legal threats, or violating BLP should all be)? VernoWhitney (talk) 13:54, 3 September 2010 (UTC)

Twinkle Undo's

I've seen a situation where it looks like another reviewer's twinkle undo didn't work (not only did it not undo the revision but the "undo" of the revision was now also a pending change), and while I realize that the page says "native" undo, I think it would be helpful to point out that twinkle undos don't work. jheiv talk contribs 22:07, 25 July 2010 (UTC)

Last time I looked they do work. Got a link? Amalthea 22:10, 25 July 2010 (UTC)
Well I approved them all (the article was John Isner) so it no longer looks like that -- I did take a screenshot though jheiv talk contribs 22:15, 25 July 2010 (UTC)
That's normal behavior: There were two pending IP edits, Andyjsmith reverted the latter, i.e. reverting to a pending edit, so MediaWiki did not auto-review the edit, and it remained pending. The same would have happened with manual undo. You can see the same behavior with Huggle if you look at the revert from 16:05, 24 June 2010, while if you look at the Twinkle revert from 20:25, 6 July 2010 instead you'll see that this one was auto-reviewed since it reverted to a previously accepted version.
Amalthea 22:52, 25 July 2010 (UTC)

Stop reviewing edit if unsure?

How would folks feel about an addition to the guideline that suggests that if an edit isn't obvious vandalism etc, but you're not sure whether it's a good edit (e.g. it may be subtle vandalism that you can't detect because you're not familiar with the article's subject matter), then you should leave the edit to be reviewed by someone more familiar with the subject?

This has been a topic of conversation on the Feedback page. The reasoning behind this suggestion is that many of the articles now in pending review used to be in semi-protection. Subtle IP vandalism is now being allowed to appear in these articles where they couldn't before, until knowledgable followers of the articles spot them and revert them. So why not recommending leaving any uncertain edits for the article followers to review? Ryan Paddy (talk) 20:07, 27 July 2010 (UTC)

Seems to be the slipery slope toward turning it into an article ownership tool as opposed to an anti-vandalism tool.--Cube lurker (talk) 20:13, 27 July 2010 (UTC)
Guideline should remain as is. Changing it now will confound interpretation of the trial results. Gerardw (talk) 20:20, 27 July 2010 (UTC)
I can see both sides of the issue, but I think it should probably stay as it is. For me, the big reason to change the guideline is that, as it currently operates, PCC could provide an incentive to sneaky vandals. They will feel some kind of false satisfaction in having got their vandalism past someone. It may well be reverted by someone else later but the satisfaction is enough for them to do it again. I think the best approach is to say that if a page suffers several bouts of sneaky vandalism then it should be put on semi-protection. Yaris678 (talk) 21:09, 27 July 2010 (UTC)
Leaving it as is feels dangerous; I see no issues with not accepting pending changes if you are unsure - because that way someone watching the article (or another reviewer) may be more confident to accept it. The whole point of a page being under pending protection is, surely, the high level of obscure vandalism. For example a lot of the football articles I watch are under pending changes because people keep adding team numbers, transfers and other information that is wrong or unsourced. So many times we see reviewers accepting edits which are wrong - and anyone closer to the page would have seen that immediately. It's not such an issue on a common problem like that but as you go out to the edges it risks giving incorrect additions more weight. I think simply expanding the advice to say if you are unsure about the addition simply leave it for someone else to review --Errant Tmorton166(Talk) 09:28, 28 July 2010 (UTC)
Agree ...with what Tmorton166 said. I add pages to my watchlist if I protect them: there are plenty of football articles on my watchlist as a result. I see edits that I simply can't review, so I leave them unreviewed. Advising reviewers to also do this seems sensible. TFOWR 09:34, 28 July 2010 (UTC)
There is a huge diference from someone making a personal judgement to leave an edit for someone who feels more confident in that area, and making it policy that if you don't have knowledge in an area, you should not review the edit.--Cube lurker (talk) 11:18, 28 July 2010 (UTC)
I don't think it needs to be such explicit policy. Surely it is sensible just to add (in the guidance) if in doubt, rather than simply approve the edit consider leaving it for others to review. --Errant Tmorton166(Talk) 11:55, 28 July 2010 (UTC)

I agree with Tmorton166's formulation. It should be left up to the reviewer what they do, but the option of leaving it for someone else should be made clear. Right now, the reviewing guideline gives me the impression that if I start reviewing an article, I must complete the review one way or another. In user design terms, there is currently no affordance for the reviewer to stop reviewing a pending change and leave it for someone else, it simply doesn't appear to be an option either in the guideline or in the interface. I think the guideline should be updated to make it clear that it's an option to stop reviewing if you aren't sure whether a pending change is a good edit, especially if the content of the article is very unfamiliar. (I also think the interface needs a "stop reviewing" button that marks the pending change as no longer "under review", and I've raised that on the Feedback page.) Ryan Paddy (talk) 22:15, 28 July 2010 (UTC)

This issue was raised by Off2riorob (talk · contribs) at Wikipedia:Pending changes/Feedback#Users accepting false additions who argued that "false additions" should also be a criteria for not accepting the edit. I don't agree with changing the guidline. The intent of pending changes is to avoid libelous or patently false information or vandalism to be publically available. Vandal patrol's objective is to quickly do the same after it becomes publically available. The page editors are responsible for confirming they are accurate or, more importantly, properly attributed. If the pending change reviewer is expected to be knowledgeable about the subject, then they would be unable to approve the majority of the edits.
For instance, I see an unusual number of football/soccer articles in pending changes. I see a lot of IPs modifying the stats that are not obvious signs of vandalism. I would approve them -- the responsibly is on the page editors to check for accuracy. If they believe that there are an inordinate number of IP changes, then they should elect for semi-protection. Pending changes are given to a group of reviewers who can't be expected to be page experts. And pending changes should be reviewed and accepted/reverted quickly so as not to frustrate the good faith of many IP/new users -- those who may become productive editors. If a pending change had to wait until an "expert" reviews it, it could sit in the queue for who knows how long. If pending changes are expected to be reviewed by experts, then they should only show up on users' watchlists and not the general "Pages with pending changes". But that would assure that many pending changes wouldn't be approved/denied for a potentially lengthy period of time.
Finally, if pending change reviewers are expected to be experts, then inevitably we will be chastened if we allowed a "false addition" in good faith. I'd stop being a reviewer if my good faith volunteer efforts would create friction because I'm expected to be knowledgeable about a particular player's goals per season statistic.
The logic is: Pending change review → Page editor review → Frustrated? Then semi-protect. There is little difference between vandal patrol and pending change review except we improve the project by not having potentially negative edits become public. Accuracy is the responsibility of page editors. I frequently go a step beyond -- will read the reference, study the edit, etc. But then I've stopped being a pending change reviewer and taken on the responsibilities of a page editor. But that has nothing to do with my job as reviewer but simply my interest in editing a page -- same as when on vandal patrol. ∴ Therefore cogito·sum 18:48, 2 August 2010 (UTC)
So pending changes will be most useful for articles with occasional vandalism that's usually obvious. For articles that frequently have subtle vandalism, semi-protection is likely to be preferable? Ryan Paddy (talk) 20:13, 3 August 2010 (UTC)
Yes, frequent vandalism would be handled by semi-protection. ∴ Therefore cogito·sum 01:51, 4 August 2010 (UTC)

RV to a version from an unconfirmed user

When I (confirmed user) reverted an edit (using WP:TW) to a version from an unconfirmed user my own edit was marked as "pending review". Is this intended? --Old Moonraker (talk) 05:35, 30 July 2010 (UTC)

That has happened to me. ScottyBerg (talk) 16:31, 3 August 2010 (UTC)
And a request for review has just turned up on my watch list for an edit by a contributor—a reviewer in his/her own right—who was adjusting an addition from an unconfirmed user.--Old Moonraker (talk) 17:22, 3 August 2010 (UTC)
The page will always have pending changes until the latest version has been accepted by a reviewer. Whether this should be the case is arguable if the page is identical to a previously accepted version. Yaris678 (talk) 04:48, 4 August 2010 (UTC)

I don't know if it's just me, but the reviewer "eye" logo (File:Wikipedia Reviewer.svg) looks a lot like the CBS "eye" logo (File:CBS.svg). My first reaction in seeing it was that it looks like the CBS logo - enough to concern me about copyright/trademark issues with CBS. What do others think? Seems to potentially open a can of beans that we might not want to open... SchuminWeb (Talk) 23:43, 3 August 2010 (UTC)

There was a whole massive discussion of this (now at Wikipedia talk:Reviewing/Archive 2#the logo) that didn't really get anywhere. That said, I would personally be in favour of changing the logo; so if you have any more thoughts on the subject... Yaris678 (talk) 05:01, 4 August 2010 (UTC)
I read through that whole discussion (thanks for linking). After reading through it, I thought binoculars would be appropriate, and no one ever brought that up. Better than the Wikipedia-on-CBS thing we have going on now, that's for sure. Funny how some people thought the eye logo seemed cult-like. I saw CBS in it. But yeah, I thought binoculars and the Wikipedia globe logo would go nicely together for reviewer. SchuminWeb (Talk) 13:34, 4 August 2010 (UTC)
Yes... Binoculars seem to make sense. They imply looking at the change from some distance. The reviewer is not supposed to get too involved with everything that is happening on the page so the analogy is better than the magnifying glass one used in some of the suggested logos. A magnifying glass also has the problem that it often means search. Yaris678 (talk) 11:24, 5 August 2010 (UTC)
Indeed. Let me cobble something together and see if I can come up with some concept art. Of course, a real graphics editor should refine any illustrations that I do, since I am not a graphic designer, nor do I play one on television. But at least I can hammer out some concepts. SchuminWeb (Talk) 15:05, 5 August 2010 (UTC)
This is a good idea. I like the binoculars proposal and hope that it will soon be implemented.--William S. Saturn (talk) 17:05, 5 August 2010 (UTC)
Pending changes logo binoculars.png
Here's the concept that I came up with. What do you think? SchuminWeb (Talk) 23:40, 5 August 2010 (UTC)
I think it's hideous... :/ it looks dumb. Try something else. Use something like a gavel or something, like a "final ruling" type thing. --A3RO (mailbox) 04:13, 6 August 2010 (UTC)


I think a gavel is not the right metaphor. Reviewing is not making a final ruling - Wikipedia is a work in progress. In terms of the binoculars logo, it doesn't look great as it is but, as SchuminWeb implied, we are just throwing around concepts at the moment. My own thought is this: Don't have the globe in the background. Instead, have it reflected in the lenses. Obviously, the binoculars would take up more of the logo... I would also probably put then at a different angle so that a greater proportion of the logo was lens. Yaris678 (talk) 12:26, 6 August 2010 (UTC)

I still prefer the eye. It's simple and easily recognized at any reasonable size. VernoWhitney (talk) 12:41, 6 August 2010 (UTC)
At 24px (a common "top of the user page" size) it looks like a blob (e.g: Wikipedia Reviewer.svg), the binoculars actually looks a lot cleaner at that size Pending changes logo binoculars.png. I dislike the eye quite a lot. It does look a lot like the CBS logo, it's not very "descriptive" and the colour is too washy for my tastes. --Errant Tmorton166(Talk) 12:52, 6 August 2010 (UTC)
Indeed, that binoculars concept took all of five minutes to put together. Just a concept at this point. I do like the idea of putting a reflection of the globe in the binocular's lenses. I think that would be a great look. SchuminWeb (Talk) 20:33, 6 August 2010 (UTC)
Metaphor for what? Yaris, you're looking at this bigger than what it is... the pending revision can't go through with it being accepted first, right? So it makes perfect sense to use it. The binoculars make absolutely no sense, you could use a telescope and it would still be the same meaning witch a stupid picture. Keep it simple, yesh. --A3RO (mailbox) 03:29, 7 August 2010 (UTC)
A gavel implies more power than the reviewer has. But maybe you are right to stress acceptance, rather than looking at things. Here are some ideas that use that:
  • A tick (already suggested in previous conversation).
  • A gate, such as
  • like one between two fields in a farm
  • like those barriers that go up to let cars through (might be hard to make work)
  • A gate house, like one at the entrance to an old country estate.
  • A stamper, possibly with the word "accept" on it.
Any other ideas?
Yaris678 (talk) 05:42, 7 August 2010 (UTC)
I don't like that word... power... no one on here is in a position of power or authority so it shouldn't really matter what picture is used but lets not make it look dumb, yanno? --A3RO (mailbox) 20:28, 7 August 2010 (UTC)
Yaris has some good ideas. I'll just point out this image, which garnered a reasonable amount of support during the last discussion. I've, personally, been using it to represent my reviewer status because I refuse to use the eye.
Pending changes logo arrows.png
Also, for the simple solution of a tick, this image with a black tick can be used.It represents a patroller and isn't used for anything as of now. It's distinctively different than {{autopatrolled}}.
Wikipedia Patroller2.png
Other than these, I like the idea of a stamper; it accurately represents what we do and doesn't have any other implications. Note: We need to remember to use the new globe logo when designing new images. SwarmTalk 07:14, 14 August 2010 (UTC)
The eye is big-brother-ish. Binoculars are a bit ornithological. The arrows look like fast-foward, which is not the idea. Checkmarks 'might' overstate the accuracy of the review. How about a magnifying glass?Ocaasi (talk) 07:25, 15 August 2010 (UTC)
The checkuser icon already uses a magnifying glass, so we probably should stay away from it. I think checkmarks are a relatively low-impact symbol that don't really imply anything other than "reviewed" or "done" or "checked" etc. I would prefer a symbol that better represents reviewing if we can think of one, but if not it's a safe alternative. SwarmTalk 00:51, 16 August 2010 (UTC)
Checkuser icon
I wouldn't want to step on the checkusers, but, there are like 10 of them? We've already got 5,000 reviewers, so there might be some basis for stealing it. Maybe we could combine something with a magnifying glass, like a magnifying glass with a check-mark 'inside' it? Ocaasi (talk) 20:19, 19 August 2010 (UTC)
That would require a significant increase in size for the magnifying glass then since the topicon is a size of 20x20 px. The current logo while it looks similar to that of CBS is accurate since revisions need to be sighted or checked. A magnifying glass does not have that same symbolistic effect. Ғяіᴆaз'§Đøøм | Champagne? 10:31, 15 September 2010 (UTC)
The double-arrow logo and the black "check" logo are both accurate as well, and they look nicer.
Alright, per the above, I'm going to change the image to the black check mark. It's distinctive, unique, fitting and doesn't have symbolism that might send the wrong message. We'll see if anyone feels strongly about it. Swarm X 21:56, 11 December 2010 (UTC)
I reverted the userbox since the above conversation has been stale for three months and there didn't seem to be a strong consensus for any of the proposed icons. A check is also used for the autopatrolled/autoreviewer logo if you're looking for something unique. VernoWhitney (talk) 00:11, 12 December 2010 (UTC)
  • Here's a bunch of userbox icons used for reviewers on the Russian Wikipedia, if anyone's interested. I personally like the Eye of Horus one. --illythr (talk) 00:43, 12 December 2010 (UTC)
So much for boldness. I would point out that there's no consensus in favor of the current image. Thanks for the list. Swarm X 05:18, 12 December 2010 (UTC)

Context guidance for reviewers?

Is there (or should there be) any way to provide reviewers with some brief context for oblique but common vandalism on an article, so that an innocuous-seeming edit isn't waved through by a reviewer unfamiliar with the subject or the article history? The Milkybar article gets occasional but endless vandalism from jokers adding their friends' names to a list of Milkybar actors (it's been a common playground taunt in the UK, for decades) - checking the history, one such edit was accepted by a reviewer, presumably because they were unfamiliar with the context and had no reason not to assume good faith. --McGeddon (talk) 09:38, 11 October 2010 (UTC)

Scope, deployment, and removal

verdy_p (talk) 05:57, 14 October 2010 (UTC)


I've opened a conversation about unilateral rights removal here. --Moonriddengirl (talk) 16:30, 23 October 2010 (UTC)

Technical problem

Pardon me if there's a better place to post this, but I'm encountering a problem that others must face as well. Yesterday, at least twice I encountered multiple edits by users in which I wanted to accept the first one or two but not subsequent edits.

I can manually accept the early edits, but when I attempt to rollback subsequent edits, it rolls back all of them. You can see an example in the history of State-sponsored terrorism, where I manually accepted the first edit of IP However, when UplinkAnsh (talk) correctly attempted to rollback the two subsequent edits, the system rolled back all three, which is unfair and discouraging to the editor.

In summary, the situation is we can't seem to revert 1 or 2 spurious edits without deleting all of a user's consecutive edits. Can this be retooled to allow more detailed managing in the interest of helping users see why parts were rejected when other edits should pass?

--UnicornTapestry (talk) 13:15, 5 November 2010 (UTC)

It is possible to revert without rollback. There is no problem with the review feature, the issue is that rollback reverts all a user's edits - which is what it's supposed to do. Hut 8.5 13:34, 5 November 2010 (UTC)

Rejecting earlier edits, possibly allowing latter edits

World Wrestling Entertainment had six pending edits, with one being the earliest made and six the latest (the version of the article you'd normally see). I started stepping through the revisions and it was soon obvious that the first three comprised vandalism. The rest, however, were [Edit:] good according to this website and this other website. I saw no way to mark it as such. The [pending review] tag was still there in the history for all of the edits, but it apparently had stopped doing anything -- there were no more confirm/reject boxes until I went to confirm my own edit. (Edited for grammar.) Banaticus (talk) 11:21, 20 February 2011 (UTC)

All the edits are uncited, you could revert them all and then add the citation and any updates. Reviewing sometimes takes a little extra work and sometimes an ip adds something without a cite and it is worth googleing it to see if it is true and if a cite is found then add the cite and the addition. I have had the same issue and got a bit lost in the complications, I have sometimes reverted them all and then replaced the one that had value but this is not really an optimum solution. Off2riorob (talk) 13:13, 20 February 2011 (UTC)
I think the main problem is that reviewing is exactly like the rest of Wikipedia -- basically built on Subversion with a strict order of changes. If you don't like one change in the stack, you have to toss the entire stack. I think reviewing needs a more modern versioning system like Mercurial (or something, I'm not saying that such is the answer, but the move from something like Subversion to something like Mercurial is what I'm talking about) which could merge the changes together then allow me to knock out individual changes from within that stack before I committed the remaining stack (presuming that latter changes hadn't impinged on previous changes). I realize that such would be a fundamental change to how Wikipedia works and would require much programming, but I think this wholesale "either reject the entire stack or commit the entire stack" system that we have now just isn't really tenable for every page in the long run with a reviewing system. Banaticus (talk) 21:29, 20 February 2011 (UTC)
But you can undo an old revision as long as later changes haven't impinged on the change from the old revision. As for reviewing, you're not reviewing an edit or a set of edits, you're reviewing the entire revision of the article. So if edit A introduces vandalism and edit B is a good edit, you don't approve the revision after A+B because the revision as a whole isn't good. Instead you revert A, and then approve the resulting revision that has B's changes without A's vandalism. Anomie 22:16, 20 February 2011 (UTC)
So instead of rejecting A's revision, I should rollback/undo A's edits, then accept B's revision? Was my problem with accepting/rejecting revisions to the WWE article that I rejected the first three revisions in a block instead of rollbacking them? Banaticus (talk) 22:00, 21 February 2011 (UTC)
It doesn't quite work that way. Let's try this with a more concrete example.
  1. Let's pretend there is a stub with the following text: "Foo Corp manufactures widgets since 1963."
  2. Then editor A comes along, and changes the text to the following: "Foo Corp manufactures widgets and poops since 1963."
  3. Then editor B comes along, doesn't notice the vandalism, and changes the text to the following: "Foo Corp manufactures widgets and poops since 1963. Foo Corp reported widget sales in 2009 totaled $1 million."
Then you come along to review the changes. You cannot approve the state of the article in revision #3, because it contains vandalism, even though B's edit is good. What you need to do is
  1. You revert editor A's edit, which changes the text to the following: "Foo Corp manufactures widgets since 1963. Foo Corp reported widget sales in 2009 totaled $1 million."
Revision #4 is free of vandalism and other major problems, so you can approve revision #4. At no point do you approve revision #3, and at no point do you approve any individual edit. It looks like that's what you did in this edit. Hope that helps. Anomie 23:37, 21 February 2011 (UTC)

Why a "Captcha" on every edit review?

Each time I review an edit, I get a Captcha. "No, you don't." Yes, I do -- let me explain. See, I mark an edit as good/bad and put in a description and then a new page loads where it again asks me if I really want to edit. This new page, though, fills the description bar with even more text and once again I have to click in the summary box and jump to the end to make sure that what I was saying wasn't cut off. Can't we skip one of these steps? Banaticus (talk) 11:25, 20 February 2011 (UTC)

Special:PendingChanges doesn't list how many revisions are on a given page

Special:PendingChanges doesn't list how many pending revisions are on a given page that is awaiting pending review. So, to really get an understanding of what happened, you can't just click on the "review" link, because you might be missing previous edits made since the last time the article was reviewed. You have to click on the history link and then review whatever edit(s) might be there. This is clunky -- there should be a way to display on Special:PendingChanges next to each article that's up for review how many unreviewed edits are sitting on each article. Banaticus (talk) 21:44, 20 February 2011 (UTC)

Test edits: accept both or reject both?

Editors often make a bad edit and then undo it. Should we accept both edits or reject both? -Colfer2 (talk) 23:35, 11 March 2011 (UTC)

Just accept the latest (undo) edit. That, I think, should resolve the issue. Orphan Wiki 23:40, 11 March 2011 (UTC)
As stated here. Orphan Wiki 23:42, 11 March 2011 (UTC)

Do not accept the revision if it is racist or sexist?

Under Reviewing process, a recent edit added "You should not accept the new revision if ... 5. it is racist or sexist". This is obviously in good faith, but which policy does it violate? All other items on the list link to specific policies or guidelines which might be violated, eg WP:BLP etc. Do we have a specific policy about racism or sexism? Should we? Should item 5 link to something? Or is item 5 not necessary because it is covered by something else? Mitch Ames (talk) 12:39, 4 April 2011 (UTC)

Since this was added a few days ago without any discussion beforehand I've removed it. Hut 8.5 13:40, 4 April 2011 (UTC)

Let the reviewed author communicate with the reviewer?

I may be too inexperienced with the process, but is there a direct way to let the author being revised communicate with the reviewer? That. of course, for a registered author who in good faith changes a page and is being rejected by a reviewer? JanEnEm (talk) 08:19, 18 July 2011 (UTC)

Help:Using talk pages. Anomie 10:38, 18 July 2011 (UTC)


Since pending changes just passed will this be started up again soon? Dan653 (talk) 18:56, 23 June 2012 (UTC)