Wikipedia:Pending changes/Request for Comment February 2011

Discussion

• Question Are we still expecting a new and improved version of Pending Changes to be rolled out in the future? I was under the impression that the in-limbo state we're currently in was because we were waiting for a new version with improvements in the areas people complained about. —Soap— 23:53, 16 February 2011 (UTC)

• This appears to be the last thing resembling an "official" statement on the matter, which suggested resuming discussion about a month from now. However, there have been several other conversations lately where users have expressed frustration at this delay, and at the perception that PC is being "shoehorned" into being an accepted reality on WP because the trial period has now run for seven months as opposed to two. However, if the higher-ups could give us a firm release date for said improvements as opposed to vagaries this could of course be reconsidered. Beeblebrox (talk) 00:05, 17 February 2011 (UTC)

• I always was under the impression that someone had promised a new update in November, after which time another poll would be held, but I never heard about it again. Am I imagining things? /ƒETCHCOMMS/ 03:00, 17 February 2011 (UTC)

• It appears that the trail period will run until the opposition runs out of steam. So I'm really unsure as to the point of any more RfCs. Protonk (talk) 02:52, 17 February 2011 (UTC)

• I suspect many of the supporters of Pending Changes are also "running out of steam", so to speak, as we were expecting a new and improved version to have launched by now. Currently there are no pages with PC on my watchlist and we're not allowed to add any new ones so I would have to say that for all practical purposes Pending Changes is off and to officially discontinue it would make no difference to me. I'd like to see it revived, but the last poll made it clear that the majority wanted an improved version, not just a revival of the one we have. Of course I would not be in favor of reviving a version that most people dont want.—Soap— 12:16, 17 February 2011 (UTC)

• Self correction: It seems that the wording of the page protection notice has changed from specifically disallowing new Pending Changes protection to "please don't do anything drastic". So technically we can still do it. But it seems that it is used very rarely based on what I see at WP:RFPP. —Soap— 12:57, 17 February 2011 (UTC)

I'm afraid, Protonk, that tone doesn't communicate well over the internet, so I can't tell if you're being shrewdly cynical of blatantly sleazy. Essentially, I agree (or disagree) that waiting for everyone else to "run out of steam" is not the way to solve these disputes.☻☻☻Sithman VIII !!☻☻☻ 12:30, 17 February 2011 (UTC)

I'm not making a normative statement, so I don't think either characterization applies. Protonk (talk) 16:25, 17 February 2011 (UTC)

• In the past, I've supported PC, and I would still like very much to see it work. But the delay is actually having the effect, for me, of diminishing my support. I thought that it was very clearly understood that the November "deadline" was for the purpose of allowing developers to incorporate fixes for the things that the community had identified as needing improvements. In the interim, I've seen a few things that I consider to be relatively modest improvements. But a big concern has been that PC was needlessly slow in terms of page loading etc., and that hasn't been fixed at all. Atheism is on my watchlist (ironically, now semi-protected in addition to PC, because there was still so much IP vandalism), and it remains molasses-slow to do anything there. If there's a limit to how well the developers can make PC work, I'm afraid that the community will need to insist that we reconsider whether it should be permitted to continue. --Tryptofish (talk) 20:34, 17 February 2011 (UTC)

• Please note We've gotten some word from the front office now, it seems they don't know what they are supposed to be doing and there is no giant update in the works because they didn't know if we wanted to keep using this or not. They would like us to give them some clear input on that. So, that is our first task, if we decide to keep it the next job is to define the standards to be used when applying it. Beeblebrox (talk) 18:39, 19 February 2011 (UTC)

View of Fetchcomms

We should hold a new straw poll—that seems to be the norm for gathering consensus on PC—as soon as possible (e.g., next week) to decide three things:

• Is PC working?
• Should we keep PC?
• Is someone going to apologize for breaking so many promises along the way? As stated above, the two-month agreement has stretched past half a year now. The policy on applying PC still is extremely vague (is it for BLPs only? is it being used at a decreased rate? should it still be used "normally"?), as are the details on what it's doing here still (seriously, seven months when the promise was two; is there a new version coming or not?; when will this trial end?)

Either the community agrees to break the original deal by extending the PC "trial" again or it decides to make the policy clear that it should not be used at all until a new version comes out, which would also come with a new straw poll regarding whether people like the changes or not.

/ƒETCHCOMMS/ 03:00, 17 February 2011 (UTC)

I agree with this in principle, but I don't think a straw poll would work, and I'd rather be more aggressive about Jimbo not closing it down prematurely. (See below for my full position).☻☻☻Sithman VIII !!☻☻☻ 09:05, 17 February 2011 (UTC)

Ugh. Sorry Fetch, but it was "policy by polling" that got us into this mess in the first place. That is why I deliberately opened a discussion and participants are asked not to vote. Beeblebrox (talk) 20:00, 17 February 2011 (UTC)

That's true, but my only concern is how long this will drag on. If we have a poll, we can set a predetermined threshold, a clear start/end date, etc. Discussions are more productive, but who knows how long it'll take to get consensus? /ƒETCHCOMMS/ 22:26, 17 February 2011 (UTC)

I agree with Fetchcomms. The clock is always ticking. Guoguo12--Talk--  17:47, 19 February 2011 (UTC)

RFCs usually run for 30 days unless a consensus becomes overwhelmingly clear before then. While consensus is harder to interpret than raw numbers, it is Wikipedia's primary decision making model. Beeblebrox (talk) 21:54, 19 February 2011 (UTC)

It is Wikipedia's primary decision making model, but people tend to interpret consensus differently. Guoguo12--Talk--  00:05, 20 February 2011 (UTC)

I agree with Fetch that this should be over, 2 months were promised and we're now closer to a year, than we are to 2 months. However I do not think a poll will help the mess. One unbinding poll after another caused the current disaster. What we should do is look at the origional poll, count the !votes, and if the keeps are over 75% we keep it to some extent, if not we have the whole thing deleted, removed, and blocked Sumsum2010·T·C·Review me! 23:30, 19 February 2011 (UTC)

Wait, the original poll? This poll? From five months ago? Furthermore, if 75% support means keeping it "to some extent", what extent? See? It's unclear, which means that there will be another RfC or something of the sort, and who knows if there will be any consensus there. Guoguo12--Talk--  00:05, 20 February 2011 (UTC)

Yes, that seemed to be inteded to be the all deciding, main, and only poll. To the extent of continuing testing. This has become very unclear and what is actually going on becoming more confuing. Sumsum2010·T·C·Review me! 00:22, 20 February 2011 (UTC)

Well, it is quicker than taking another poll. :/ Guoguo12--Talk--  03:09, 20 February 2011 (UTC)

I agree with the new poll, IF there is any significant push for implementation. If there is not, we can retire the PC without it. --_{Piotr Konieczny aka Prokonsul Piotrus| talk} 22:48, 21 February 2011 (UTC)

Research before more discussion?

Instead of launching into this debate again, I wonder whether it would be more worthwhile to gather some data on how the feature has worked so far. Possible questions for examination would be:

• How many articles have been subject to PC control and for how long? How has that number changed?
• What types of articles (based on category or WikiProject perhaps)?
• Percentage of proposed edits accepted vs rejected
• Time changes stay in the pending state before review: mean, range, SD, etc...
• Percentage of live edits that are reverted before/after activation of PC

It seems to me that we can't have an intelligent discussion on the future of this feature without some data on how it has been used and factors that hint at its effectiveness. Zachlipton (talk) 05:26, 17 February 2011 (UTC)

That idea has merit, but I don't see it as being mutually exclusive of having a discussion at the same time. Beeblebrox (talk) 07:28, 17 February 2011 (UTC)

Of the "Metrics" linked from the box at the top of this page, it looks like one, Special:ValidationStatistics is automatically updated, and so is up-to-date now. It can answer part of your next-to-last question: "The average delay for pages with unreviewed edits pending is 2h 58m 36s." -- Ken_g6 (factors | composites) 18:34, 20 February 2011 (UTC)

View of s8333631

A straw poll is clearly out, as the last one on this topic (which was, IMO, established rather prematurely, considering the diversity and intensity of opinions), had the effect only of polarizing users and confusing the issue. Some other, informal and discussion-oriented system should be used.

Unfortunately, the main problem with previous attempts at consensus on this was that they were ended by the Word Of God before an actual community decision could be made. This essentially shut down discussion for a long time, and this RFC is an attempt to revive it. So this time, we should let the community decide.

☻☻☻Sithman VIII !!☻☻☻ 08:48, 17 February 2011 (UTC)

Very much agreed. Especially since, usually, what Jimmy wants Jimmy gets. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 20:54, 19 February 2011 (UTC)

Since, unfortunately, the discussion appears to be turning less meta, I will also state that, for the record, I am against the implementation of pending changes. It is extremely confusing for newcomers, produces a vast, intractable backlog, and, regardless of its actual mechanics, will get us some very bad press. Also, I'm uncomfortable with the grey areas around BLPvio, blatant vandalism, subtle vandalism, hoax information, good-faith inaccurate information, good-faith and accurate but poorly formatted, etc. Where do we draw the line? Wherever we choose to, people will regularly cross it. Pretty soon we've got a systemic censorship-machine. Not cool.☻☻☻Sithman VIII !!☻☻☻ 11:14, 21 February 2011 (UTC)

View of Redrose64

Maybe my main areas of editing don't overlap with the areas subject to WP:PC, but have rarely encountered it. I was given reviewer rights on 14 June 2010 ready for PC, and in over eight months I've never used the reviewer facility (although I have used rollback several times). In all that time, I've made over 14,000 edits (an average of 55 per day), but only performed one edit to a page which was subject to PC - and that was less than a week ago, to undo an incorrect edit (to a non-BLP article) by an IP which had already been accepted as valid by another user. Maybe a broader trial was needed. --Redrose64 (talk) 13:26, 20 February 2011 (UTC)

View of TheIguana

My main concern with Pending Changes is it essentially requires a new class of users (Reviewers). I really do not think this fits well with Wikipedia as this introduces more user hierarchy and may act to discourage users from becoming active in the community by putting up barriers on the kinds of additions/help users can make. –TheIguana (talk) 20:50, 21 February 2011 (UTC)

But why does that have to be an inherent part of the system? I think of PC as an alternative to semi-protection. From this perspective, a new user can go to a PC-controlled page and edit immediately, with the review typically taking a matter of minutes and rarely more than an hour. If the page is semi-protected, new users have a hard barrier against editing and can only contribute with an edit request on the talk page. PC seems much more open in this regard to me. Furthermore, there's no reason Reviewers has to be such an explicit class: we could, if PC is formalized by policy and continued indefinitely, arrange to grant reviewer status to all users meeting certain thresholds, just as we do for autoconfirmed but more stringent. But realistically, pending changes seem to be getting reviewed quite quickly and the barrier to become a reviewer seems fairly low. Is there a real problem there? Zachlipton (talk) 21:02, 21 February 2011 (UTC)

See above, that the "average delay for pages with unreviewed edits pending is 2h 58m" - and that was on a small trial, with a huge number of people keen to try it all out. What happens when the delays creep up to days?  Chzz  ►  04:40, 22 February 2011 (UTC)

Well, it takes so long as most reviewers don't regularly check Special:OldReviewedPages. I've never seen more than five edits over there, and there's usually only just two. Being a reviewer, I don't check the place much anymore, as there's nothing to do there. ManishEarth^{Talk • Stalk} 03:16, 24 February 2011 (UTC)

Wikipedia is voluntary work, vandalism is a time sink of voluntary work. German Wikipedia is right to protect quality work from vandals, making the quality voluntary work useless, requiring voluntary work just to keep the status quo. Some vandalism has system, when it attacks quality Wikipedia, when it attacks a neutral, unbiased, quality information source. In a way, some vandalism is a form of censorship. Any tool that slows down vandalism is valid. The objective is to diminish the required voluntary work, PC and reviewing or no PC and rollback are the choices. --Chris.urs-o (talk) 19:27, 1 March 2011 (UTC)

View of Ost316

I'm stealing and adding to my opinion from my post on the poll, since nothing seems to have changed since then (unless it happened on yet another subpage): Clear rules are needed on what should be accepted and when it is best used, perhaps for vandalism and harmful libel. This may require a broader test for a longer time to weed out the early rush when reviewers are more actively engaged earlier in the trial. I can see this being useful for pages with legal concerns like BLPs—especially little watched ones where harmful information may remain undetected—but I'm not yet convinced that it should be a default for BLPs. The process also seems to increase the workload for monitoring normal vandalism. Additionally, I agree with previous comments that "Approved" seems to imply that the article has achieved some validity, while approved edits mean something different for each editor that approves. If concerns cannot be addressed, I'd lean toward stopping PC because it's a simpler scheme, easier for readers to understand, and more likely to attract to new editors. I was fine with keeping the feature enabled while changes were made or its fate decided, but I think that's its current continued use with stale discussions (before this subpage was created) is verging on making it a de facto accepted, especially since the system can remain turned on without being used. —Ost (talk) 23:27, 21 February 2011 (UTC)

View of David in DC

I'm baffled by opposition to implementation of pending changes. We want a collaborative, anyone-can-edit, open source encyclopedia. But trolls abound. And we don't want an anyone-can-vandalize or anyone-can-broadcast-nonsense-lies-or-disinformation "encyclopedia." Pending changes, in my view, moves us in the right direction.

Anyone-can-edit doesn't mean anyone-can-edit-in-real-time. My biggest fear about pending changes was that the difference between "real time" and "approved edit" would be so vast that contributors would be discouraged. Especially newbies. So I've visited the "pages with pending edits page" frequently, and generally put articles I've approved (or rejected) edits on onto my watchlist. I've been astonished at the number of times there were no pages with pending edits. I've been equally astonished that, when there ARE pages listed, none has been awaiting review for more than an hour or two.

There's no doubt that the inherent thrill of seeing your edits on the Internet is one of the things that attracts humans to collaborate here. For me, every time I see one, I feel just like Steve Martin, in The Jerk, when he discovers he's famous because he's in the new telephone directory. But we're not asking anyone to wait as long for that thrill as Steve. Telephone directories come out only once/year. Here, at worst, it takes a couple of hours between edit and fame.

In the meantime, a whole lot of carp gets caught in the pending changes nets before it goes live and discredits us. A couple of hours is a small price to pay for a much-improved, much-more-credible encyclopedia.

An unexpected (at least by me) side benefit is the weeding out of editors whose idea of delayed gratification cannot extend beyond nano-seconds. In my experience, such children are, quite disproportionately, trolls.

Already, I've seen one improvement, over the time of this extended trial. The notice that pops up on my watchlist when a page I've "watched" after approving/disapproving a pending edit keeps me connected to the page. I grow eager to keep the "pending" time short, and to keep the page in at least as good shape as I first found it. Over time, if others follow this practice, every page in the pending changes trial will be ratcheted up in quality.

I trust more improvements are to come. There ain't nobody involved in this endeavor genuinely trying to make things worse.

I cannot imagine a good reason to lift the "pending changes" status of any page currently covered by the trial, and I've seen none advanced on this kludge of an RfC page, so far. I can see why we'd want to move slowly in adding pages. I still fear backlogs. But if past is prologue, my fear is still overblown.

Now, like the Red Queen, I sometimes think six impossible things before breakfast. And I hold no corner on the wisdom market. Often I wonder if I've even got a stall there.

But the arguments against pending changes seem mostly to be of the IDONTLIKEIT and THATSNOTTHEWAYWEVEALWAYSDONEITBEFORE variety. In fairness, there are also:

The empiricists, who want more study, more statistical analysis, more data, more XXXXXX and YYYYYY. But there's such a thing as exalting process over praxis. It's a recipe for stagnation.

and

Those who feel hoodwinked by the way this started out as a short-term trial and seems to have morphed beyond that. To them I commend the wisdom of the seismologist's bumper sticker: "Shift Happens". David in DC (talk) 21:54, 22 February 2011 (UTC)

---

David in DC, thanks for offering opinions. Some seem concerned that this RfC page is getting far too long, but personally I welcome all the input. This is consensus and discussion at work.

Firstly, to address some specific concerns, I'd like to play devils' advocate, and try to counter; thus the following are not necessarily my own views - they're merely an alternative viewpoint;

Wikipedia became what it is, because it was open for anyone to edit - and I mean, *really* open - no need to create an account, and your changes are visible immediately. It is quite counter-intuitive to think that could generate anything other than garbage - yet, here we are. It works.

Semi-protection is fairly tightly controlled - it may only be applied when there is evidence of specific, ongoing vandalism. There is considerable resistance to its over-use, because it is seen as 'locking things down', and going against the original ethos.

Pending changes can offer something in-between.

One concern is, if not very clearly defined in scope, it is open to creep onto more and more articles, until Wikipedia becomes closed to new users. You might consider that they can still edit - but how many will bother, if their changes are not visible, and are subject to review? Unless we are very careful, there are many potential problems;

• If items go unreviewed for a long time, people may become disillusioned
• If well-intentioned edits are rejected, new users feel they cannot contribute - that the entrenched community is in control, and their input is not welcome
• If excessive numbers of articles are protected in this way, then we have moved away from the original "open to all" (instant edits) model, which was key to the success of Wikipedia
• It adds to the complexity, which again can put off new users (terminology, trying to understand why their edits do not appear, who gets to review them, etc)
• We don't know that it'd be "a couple of hours at worst". The trial was on a tiny, tiny number of articles, and there was huge interest generated with 'reviewer' rights; high-profile articles were chosen, which are watched by many editors. One of the intentions is/was to use it on lesser-watched BLPs. We have no idea how long it might take to review edits, if it were implemented more widely.
• Does it really weed out trolls/children? Or, do they simply go vandalise elsewhere - on something we are not watching? And if they do that, do we protect that one as well?
• Similarly, you said you monitored those articles more closely, and thought their quality improved. But, what would you have been doing if you hadn't watched them? Presumably, improving other articles. So where is the net benefit?

As I said, those are not necessarily my own views; I am trying to keep as open a mind as possible.

I don't think this is just epiricists v. the hoodwinked. My own stance, currently, is to suggest we remove it to clear the decks, and then think about possible future implementation. I hope you can tell from my responses that I am quite open to ideas; it's just that, at this time, we simply do not have enough information to delare yay or nay; nor is the remit clear. I, for one, am certainly looking to the future - but feel the way to make progress is...well, it's a bit like an edit-war. We need to stop, and discuss; come to consensus; and then we can move on.

Thanks for your input; let's hope it isn't just jam to-morrow. Cheers,  Chzz  ►  11:34, 23 February 2011 (UTC)

Chzz, this was very well argued. You should give yourself a heading above, and I'd say I support this position, even though I think PC might be usable in a limited context. Wnt (talk) 19:03, 23 February 2011 (UTC)

I'd also agree like to note that I agree with most the questions above. Particularly, I was planning on adding to my response about scope and that the trial was on a very small number of articles. —Ost (talk) 21:43, 24 February 2011 (UTC)

I think I agree with Chzz's position. The current version has three defects: too slow, too confusing, incapable of keeping up with rapid editing. Given the current emphasis--an entirely correct emphasis--on improving the transparency and ease of starting to use Wikipedia for new readers and editors, this is the worst time to adopt it. As I try to recruit prospective new editors, I find one of the key attractions is that all edits are immediately live, for people find it exciting. But I wouldn't want to close it off--there's a small segment of articles where it would be better than semi-protection. One way of keeping the use of it from spreading is to strictly limit the number of articles or the time for which it could be applied.--say 1000 articles, or 1 year at a time.--or possibly both. The only thing I might disagree with Chzz about, is that we cannot decide to adopt it without another trial. If the developers think it therefore is not worth their efforts, then either we need some volunteer developer(s) who who think it worth their trouble, or we cannot do it. Experience has made me wary of adopting their major changes until we see how they work out. DGG ( talk ) 04:26, 26 February 2011 (UTC)

I certainly agree with Chzz. One compromise I proposed was to have PC protected edits be immediately live for the editor that made them and all logged in users. It's not perfect, but without this the impact on editors will be similar to semi-protection. The problem is the technical difficulty of achieving this with IP address pooling. We could potentially make use of the temporary session cookie, but the user would still see their edits disappear if they close and reopen their browser, unless they explicitly bring up the diff listing for the latest version, which isn't very pretty. —UncleDouggie (talk) 05:31, 26 February 2011 (UTC)

How about this: a pending changes level just for IPs, not autoconfirmed users. Vandals want their edits live on articles with a lot of traffic. This would reduce the amount of work required just to keep the quality/ credibility of an article, as if the edit does not get live, it is not interesting anymore. --Chris.urs-o (talk) 19:41, 1 March 2011 (UTC)

That assumes a desire for attention, not a desire to disrupt, as several LTAs prefer to do. Against those, PC is counterproductive because a coordinated attack can lock an article down worse than any semi-protection we do ever could. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 03:18, 2 March 2011 (UTC)

View of DustFormsWords

As a reviewer with pending-changes pages on my watchlist, I have seen no significant benefit to the project or to individual pages from this system, and still hold to my original opinion that it creates an unnecessary level of bureaucracy and complexity which has the effect of widening the existing and problematic gap between long-term experienced editors and casual users. The pending changes system should be scrapped entirely; existing systems are sufficient to protect the project from excessive vandalism. - DustFormsWords (talk) 01:18, 23 February 2011 (UTC)

View of Protonk

This is a bit of a pastiche of other comments I have made here and in other RfCs. I see a few central problems:

• Scope is undefined. PCs have been sold on two premises which compete against each other. The first is that PCs represent a way to expand potential edits to indefinitely sem'd articles. Rather than regularly semi-protecting George W. Bush we could apply pending changes and disallow "bad" edits only. The second is as a mechanism to expand potential protection to articles which are not currently protected or watched as a means to prevent malign edits which would otherwise go unnoticed. The former, while somewhat dubious, represents a potential improvement to our response due to the increased granularity of options. The latter is a more sweeping change and invites other problems, but more importantly covers a broad swath of articles which we imagine as "open" now and would become closed. I don't actually think that the goals of the encyclopedia are represented by such an expansion but if the expansion was accepted by the community I could be ok with this. However I don't think the language and expectations of the original RfCs even approached this idea.
• PC is a lamppost solution. Wikipedia has serious long term growth problems. We have a decreasing influx of long term editors. We have a decreasing rate of conversion from readers to editors. We have a flattening/increasing time between edits. We have no mobile solution for editors (the mobile site has to disallow edits for tracking reasons). Top line problems which demand serious solutions for the future of the site, and we are futzing around with further reducing editor conversions because of a concern over marginal articles. Why? Because we see these problems as editors and OTRS volunteers. I don't mean to diminish the import of long term vandalism (especially on BLPs), but for the encyclopedia as a whole they are minor. When we look back 5 years from now we won't be asking about how effective we were in scrubbing "bad" edits from articles but how we succeeded or failed in reversing trends in the editor base.
• After 2 months of a real trial and 5 months of fait accompli we still have no real data on the influence of PC on edits, editor conversion, and vandalism. In this RfC and others I have seen a lot of statistically illiterate arguments. People are assuming that we need an absurd sample size (100,000+ articles), we can't determine the impact of PC on editing, we can't design a true experiment, etc. In fact, designing a real experiment would require less than 300 articles in the sample as a whole. A database report request could be made for articles which are effectively under long term semi-protection and among those articles (of which there are probably thousands) we could derive a propensity score for edits, editors, edits by editors, page views, vandalism, etc. and a sample could be chosen which is relatively homogeneous across these characteristics. Then half of the sample could have PC applied while the other half retain the normal protection. After a month or so, statistics on each set could be gathered and compared. We wouldn't be able to eliminate the Hawthorne effect and we would not be able to measure spillovers, but we could get a quick and defensible snapshot of how PC actually influences editing. By how much are those articles under PC seeing increased editing by newcomers (even if it is by proxy)? How different are the rates of vandalism (or even just reversions of unproductive edits) between semi'd articles and articles under PC? And so forth. A lot of the positives of the tool could be illustrated and the negatives could be limited or quantified.
• Related to the second point above, we are faced with a situation where editors (and the foundation) want to accept as given the growth and success of the past and move to a model which would have been repudiated 10 years ago. This is not at all unique to Wikipedia. Many organizations at the end of their "growth" phase become conservative and hope to trade growth for quality. Anecdotally we see this in conversations all the time. "Wikipedia has enough editors" "enough articles". Exceptionalism is trotted out. We are told that this time (or 2001, alternately) things are different. We needed growth in the past and now we need stability. That way lies the graveyard of institutions. Wikipedia is not so large and pervasive that it does not need to grow and react to change. If anything we need to be more open to editing and more open to change than we were when there was functionally no real competition for online information aggregation and no online encyclopedias.
• Lastly, and deeply related to growth, even if we accept that something like PC is valuable for keeping bad edits out and even if we accept that we cannot change the growth path of wikipedia, we should be fully cognizant that PC is not a force multiplying solution. What PC does is take currently unprotected or protected articles and moves them into a state where edits can be entered into a queue. If the goal (for some editors) is to use PC as a blanket solution to the BLP problem then we are talking about adding tens of thousands of articles which no one watches or cares about into a queue hoping that people will watch or care about the queue. Unlike the addition or improvement of vandal fighting bots or of the edit filter, PC adds to the workload of a shrinking group of editors (in absolute terms or relative to the task at hand). Can we imagine a wikipedia twice the size of what we face now dealing with changes via PC? Would that be more or less sustainable than dealing with changes under the curent model? My suspicion is that it would be less sustainable.
• One more point. Simplicity is a virtue. Look at the pre-Pc protection policy. there were three states (excluding move protection) that an article could exist in. It could be unprotected, semi'd or fully protected. Imagine for a minute that you undertook to explain the protection policy to a wikipedia neophyte. Pick a smart relative or friend and try explaining the pre-PC policy. It is startlingly simple. Unprotected articles mean that anyone can make any change without registering an account. Semi-protected articles mean that only registered accounts can make changes. Full protection means that no one can make changes. Obviously there are some nuances I missed. Full protection means admins can change the articles but as an empirical regularity they tend not to. Semi-protection extends to non-autoconfirmed accounts, but this is a minor quibble. Dropping those two caveats the protection policy can be explained to somone who has never before edited wikipedia inside of three sentences. You could put it on the back of a napkin and explain it to a congressman. How many more sentences need to be added before you can get a functioning explanation of PC and how it interacts with the current protection policy?

Thanks. Protonk (talk) 22:56, 23 February 2011 (UTC)

• Protonk, George W. Bush was indeed put on PC protection during the legitimate two-months portion of the trial. It got overloaded with bad edits and had to be re-semi'd. Same goes for Barack Obama. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 03:23, 2 March 2011 (UTC)

View of Rivertorch

Pending changes adds an unnecessary and undesirable layer of complexity and confusion to what should be a straightforward, enjoyable enterprise—building and maintaining an encyclopedia—while doing little to combat the subtler, more insidious kind of unconstructive edits that cannot be readily identified by editors unfamiliar with a given article or its subject. Its only potential benefit is to hide recent, unapproved changes from casual view. It does nothing to help established editors identify or revert vandalism. If the sort of blatant vandalism with the potential to harm WP's integrity—and on occasion living persons' reputations—is really unmanageable with the existing tools besides PC (and I'm not persuaded it is), then serious consideration should be given to other options. Some of those options would be intensely controversial, and I list a couple of them here only for specificity's sake, to illustrate that they exist; I am not advocating their implementation:

• Increased employment of indefinite semi-protection (including, if necessary, on all BLP articles);
• Raising the bar for autoconfirmed status (four days and 10 edits is a terribly low threshold);
• Replacing autoconfirmed status with a new status requiring confirmation by one or more human editors;
• Reevaluate what "anyone can edit" means, in particular whether the spirit of the slogan requires the instant gratification of anyone with Internet access regardless of their willingness to register for an account.

And so on. Each of those is a discussion and a half in itself (which shouldn't happen here), but I hope they provide some substance to my claim that PC isn't the only possible solution to what has been portrayed by many^[weasel] as a crisis. (P.S. I agree with s8333631's last paragraph and with DustFormsWorms and, in large part, with Protonk. That was originally the main thing I was going to say, but I couldn't find a good place to say it on this vast page. Sorry for making it even more vast.) Rivertorch (talk) 07:20, 25 February 2011 (UTC)

Is it broken?

I think the system currently works as-is and the de facto state of affairs (PC, like protection, applied at admins' discretion to articles when they need it) is fine. Stifle (talk) 10:43, 17 February 2011 (UTC)

I agree with that, the protection is being applied where required without any problem. Another tool in the box to help protect our articles. Off2riorob (talk) 16:23, 17 February 2011 (UTC)

Personally I like it and think we should keep it, but the current state of affairs has it "in limbo." The advice for admins on when and how to use it is wishy washy waffling and provides little to no real guidance. For example, when an admin goes to protect a page, a pink box pops up that says: "The pending changes trial has ended. The result of a poll was in favor of the temporary continuation of PC on most of the currently PC-protected articles until a new version is released. Please don't do anything drastic. Please don't fight. No page in the Wikipedia namespace should be protected under pending changes except those for testing." I wasn't planning to do anything drastic, just maybe add PC to this one page. Is that ok? Is that "fighting"? When is this "new version" coming out? Over at the protection policy page itself we have this "For a trial period that began on June 15, 2010 articles could be protected by pending-changes protection. The trial is now over and the Wikipedia community is currently seeking to gain consensus on where to go with the pending changes idea." Where is this "seeking" going on? Apparently nowhere for the last several months, so here we are. In my opinion these are the questions that need to be put to bed. Beeblebrox (talk) 19:56, 17 February 2011 (UTC)

Who wrote the edit note? Have you got a link to it so I can investigate it please. I agree this does need putting to bed, some admins are experianced in the tool and have a good idea where it is starting to clearly be beneficial and are adding it, as I understodd there is to be no major increase in usage and we are presently at about half the number of the trial usage on articles. Off2riorob (talk) 20:39, 17 February 2011 (UTC)

I can't even figure out how to open the history or edit the notice, maybe someone with more technical knowledge can answer that. Beeblebrox (talk) 20:44, 17 February 2011 (UTC)

Do you mean MediaWiki:Protect-text? /ƒETCHCOMMS/ 22:32, 17 February 2011 (UTC)

Yes, thanks, it is a bit outdated and not expressing current common practice so we should look at updating it. Off2riorob (talk) 22:41, 17 February 2011 (UTC) I think it should be edited to the note below , which is a better reflection of the current practice. Off2riorob (talk) 16:19, 18 February 2011 (UTC)

note - The community supported the tool in a straw poll although consensus as regards expansion is undecided. Please use sparingly if and when required so as not to greatly increase the current pool of articles under this protection.

Oppose this wording, because, that statement is simply incorrect. The community did not support it, and we are not debating expansion. The community supported a time-limited trial, which is over. And we are discussing possible future implementation. Chzz  ►  04:45, 22 February 2011 (UTC)

Before anyone does anything else on Pending Changes: Lay it all out in the open!

As Fetchcomms pointed out, this is a huge mess of contradictory, confusing, and inconsistent information. I don't believe any disucssion would be valid, let alone wise, before a clear document was composed, answering the following questions.

1. What articles are eligible or ineligible for pending changes.
2. Who is able to place and who is able to remove pending changes from articles.
3. When placed on a page, how long is pending changes to remain active on that page.
4. What determines which level of pending changes is placed on a page.
5. What recourse can the community take if it believes that pending changes needs to be expanded or has gone overboard and needs to be scaled back.
6. What recourse can editors take when they believe that pending changes has been placed on a page in error.
7. What recourse can editors take when they believe that pending changes has been placed on a page with malicious intent in order to influence the content of the page.
8. What role the WMF and WMF staff have in pending changes.

All of these need to be answered on one page, marked as policy, in a simple and coherent way.

Additionally, any data generated from the pending changes trial needs to be released in one place, linked to the above policy page, so everyone can see it.

Any further discussion or voting is completely and utterly useless until everyone is working from the same information.

Sven Manguard Wha? 21:38, 17 February 2011 (UTC)

I hope you don't mind if I answer these issues as I understand them. Off2riorob (talk) 21:52, 17 February 2011 (UTC)

• 1 - Any article where it is considered it might help to protect an article, it can just as easy be removed if it is assessed to be better protected by semi protection, or removed completely if protection is deemed to be no longer required.
• 2 - As per all protection any administrator can use all means at his disposal including pending protection to protect articles, bearing in mind from a position as to allow as open an environment as possible is the idea.
• 3 - Pending can be applied for any time short or long in the same way as other levels of protection as judged required by the administrator with a mind towards the basic principle of free to edit.
• 4 - As I understand it we only have one level of pending currently in use.
• 5 - If you edit an article and pending is in your opinion causing problems the place to go as with all protections is the admin that added it and discuss your issues with him or her. From the previous poll results large expansion is not supported, so natural slow and steady additions and removals is imo the natural way to progress with this tool.
• 6 - As with everything, go to the person that added it and explain your issue.
• 7 - There has not been a single case of this pending protection being placed in bad faith to control a page, as above, the first place to go would be to discuss with the admin that applied it, if your fears were not resolved then of to ANI it would be.
• 8 - The WMF will have to speak for themselves. Off2riorob (talk) 21:52, 17 February 2011 (UTC)
  • I was under the impression that until a new version is released, PC should only be used sparingly (limited to BLPs?) I don't think it can be used on any article, but I'm not sure. I think it will be a good idea to start a new section below on forming consensus for where PC should be used, so I'm doing that. /ƒETCHCOMMS/ 22:34, 17 February 2011 (UTC)

Sven, I think you may have missed the point that clarifying this stuff is exactly why this RFC was opened, per the header at the top of the page. I'll put it in its own section to make it more obvious I guess. Beeblebrox (talk) 22:35, 17 February 2011 (UTC)

This is a meta-comment: Sven, your list of questions makes it sound like you think this is some completely new, radically different system. This is still the English Wikipedia. We handle just about everything in the same basic ways. Pending changes is handled almost exactly like its nearest similar process, which is semi-protection: The community writes the rules, the admins push any complicated buttons, problems are reported and resolved in the usual places and in the usual ways. Pending Changes is business as usual, not some wildly different beast. WhatamIdoing (talk) 22:58, 17 February 2011 (UTC)

I would echo everything WhatamIdoing said, but I thought I would answer Sven's eighth question: the role of the WMF is the same as any other feature that's been built for the community. You all are the ones who will being using any potential or current tool on the project, so it's not up to us to decide whether it's useful and appropriate or not. (I think most Wikipedians know that, but I thought I'd go and just say it, since you asked.) Steven Walling at work 23:43, 20 February 2011 (UTC)

Where should PC be applied?

To keep discussion organized, perhaps we should start by asking where PC should be applied. My understanding is, that if/until a new version is released, PC should be used sparingly (only for BLPs) and not just in any situation. /ƒETCHCOMMS/ 22:41, 17 February 2011 (UTC)

That was not my understanding. I thought it was to be used sparingly, but slightly less sparingly in the case of a BLP article. We have been allowing and sometimes granting requests for non-blp pages at WP:RFPP, so that seems to be the current practice. This si exactly what is wrong here, everyone seems to have their own understanding cobbled together from clues because there is nothing clear and specific. Beeblebrox (talk) 22:45, 17 February 2011 (UTC)

Yes, it is basically being used sparingly and as such, it doesn't really matter where we use it, but administrators are not restricted from applying the protection if they feel it would be beneficial. You could say, we are still in a trial and gaining experience of the tools benefits but just in a smaller, more flexible way. There was no support to dramatically expand the tools usage. Its on Rubygate which although not a blp has got living people in it, and doing alright, its on George Osborne and George Michael and working fine, other places it is not so helpful, football articles because of the statistical edits. Off2riorob (talk) 23:00, 17 February 2011 (UTC)

Could Fetchcomms and Beeblebrox (or anyone) please provide diffs for the historical discussion that gives rise to their different interpretations of the agreed appropriate use of PC?--Anthonyhcole (talk) 13:45, 19 February 2011 (UTC)

Like i said, there is no clear guidance anywhere and we are grasping at clues here and there at places like the what an admin sees when applying protection along with WP:PP and WP:PC. That's the whole point of this discussion, to determine if it is to be a permanent feature, and if so to craft those clear, or at least clearer, standards. Beeblebrox (talk) 18:34, 19 February 2011 (UTC)

Based on my experience, PC must be kept and discretionary used in all articles that became frequent targets of IPs and obvious SPA accounts who are engaged in vandalism, edit-warring, sockpuppetry, etc. These are not necessarily BLP articles, although BLPs are indeed most frequent targets.Biophys (talk) 17:08, 25 February 2011 (UTC)

Is there or isn't there a new version coming out?

Someone might want to ping the devs; I'm not sure who'd be best to contact. Also, is the WMF studying PC usage/effectiveness data at all? /ƒETCHCOMMS/ 22:51, 17 February 2011 (UTC)

Already  Done [1] Beeblebrox (talk) 22:54, 17 February 2011 (UTC)

We got upgraded in some small way and its faster and some small points addressed, I don't see a new version on the horizon but as with all the other tools and interfaces, a constant re assessment and upgrades and improvements as and when possible. Off2riorob (talk) 23:06, 17 February 2011 (UTC)

The latest update was in November. This was laid out in the short term roadmap. Since that November update, it's been basically in maintenance mode, as you can see if you browse the monthly engineering updates. If the community decides it wants Pending Changes in the long run, there is some significant work that needs to be done, including forking the codebase entirely away from Flagged Revisions as it is implemented on Polish and German Wikipedia. There are also significant new improvements that could be made, as described under future releases in the roadmap. But the short answer is: there isn't another upgrade currently in the works. Promising developer time on Pending Changes means taking time away from other current or potential projects, so we have to have some signal from the community about whether it wants Pending Changes permanently or not. If that's the case, then we're committed to working on it just like any other major feature in use. If the community either says, "No, turn it off." or is unable to come to a consensus, then we can't commit more resources of course. Steven Walling at work 05:42, 20 February 2011 (UTC)

Steven, please understand that this is not at all a criticism, but just something that I want to point out. We have a kind of Catch-22 here: there are some of us in the community who would like to keep PC if it is improved, but who would rather get rid of it if it stays as is. I remember Jimbo giving the clear impression that the November update was going to include all of the more ambitious improvements, but obviously it did not turn out that way. I think it's going to be a very hard sell to the community, to agree to continuing PC on the expectation that further fixes will come later. --Tryptofish (talk) 21:00, 20 February 2011 (UTC)

The november update did improve the interface and imo speed up the tool but this isn't a sales program its a tool to help us protect articles, its working now, this is it, there will be I imagine some improvements as is normal through the normal practices and usual upgrades. Either you support its continued usage and development or you don't. Off2riorob (talk) 21:19, 20 February 2011 (UTC)

In your opinion. --Tryptofish (talk) 21:26, 20 February 2011 (UTC)

Personally I support expansion of the tool, if you reject it then I support more indefinite semi protection, I also am more and more supporting living peoples ability to opt out of the project, more and more its clear to me that not only are we unable to keep malicious and defamatory content and additions from being mirrored through the project to all corners of the world wide web especially in regard to semi notable living people but that there is also some support amongst unverified users to allow this. Off2riorob (talk) 21:19, 20 February 2011 (UTC)

I'm a WMF developer, and although I'm not directly involved with Pending Changes, I don't think there are any further updates scheduled for the near future and I don't know of anyone actively working on the project at the moment. There may, however, be a new project prototyped in the near future that could be a replacement for pending changes. Personally, I would suggest leaving Pending Changes at its current usage level (or ending the trial if that's what most people want). Kaldari (talk) 23:36, 20 February 2011 (UTC)

Tryptofish: I totally understand the chicken/egg problem you're describing here. But Pending Changes has been in use on English Wikipedia for longer than six months. It's also running permanently on millions of articles on other Wikimedia sites, albeit under a different implementation. As far as we're concerned the ball is in the community's court right now. We can't devote more manpower to a feature that still doesn't have clear support after more than than half a year of live testing and improvement. If that support can be demonstrated, we're willing to work with you all on improving it in the future, but not until that happens. Steven Walling at work 02:09, 21 February 2011 (UTC)

And I think that what you say is entirely reasonable. I'm saying that this probably means that we won't continue to use it at this project. --Tryptofish (talk) 22:46, 21 February 2011 (UTC)

going forward

• - Support , as is, slow and steady development and assessment, the wheels aren't dropping off. A clear net benefit to the protective toolbox. Off2riorob (talk) 02:03, 18 February 2011 (UTC)
• This RfC is to clarify PC first and to determine whether it should be continued second, so perhaps we should work on getting consensus for extending it/ending it after we figure out a clear policy on it? /ƒETCHCOMMS/ 21:07, 18 February 2011 (UTC)

OK, would you please look at the edit notice update in the section above to reflect the current actual practice and edit it for me please. Off2riorob (talk) 21:13, 18 February 2011 (UTC)

Like this? I'm not certain what you mean. /ƒETCHCOMMS/ 03:31, 19 February 2011 (UTC)

What people have decided should mean something

The experiment was originally supposed to end in two months.

Wikipedia:Pending changes/Closure said "The scheduled two-month trial has ended. The community should now decide if the implementation is to be continued, and it should discuss possible adaptations, in terms of policy. Developers have indicated it would be too complex to turn off the feature, then turn it back on in case the decision is in favor of continuing the implementation, so they will wait for the community decision — unless it takes more than a month, in which case they will turn off the feature."

Wikipedia:Pending changes/Straw poll on interim usage said "a hard stop date of December 31, 2010 will be set for a new poll on interim use of Pending Changes in the event that the release of the new version is delayed"

Now it's not necessarily wrong for people to come back and keep extending a trial period. Some of the reasons why we might want to could be:

• The experiment is visibly producing useful ongoing results
• The experiment is allowing changes by developers or MediaWiki software to be evaluated as they are made
• The Pending Changes trial is unambiguously useful to Wikipedia.

The problem is, I don't see evidence of any of these, and I don't see any sign that community discussion has called for anything but a hard stop date which has simply been ignored. This project is displaying some weird kind of continuismo that just can't be justified by saying "They might fix it any time, who knows". The problem is, when the community can't stop an experiment despite clear resolution to do so, it creates general resistance to any new experiments, e.g. Wikipedia:Templates_for_discussion/Log/2011_February_7#Template:Invitation_to_edit. I think that unless there's some sudden strong consensus right now to extend the trial, the original poll's decision to end it as of December has to be honored. Wnt (talk) 15:25, 18 February 2011 (UTC)

• Seconded, Wnt. There was discussion for a trial, nothing else. We never had a vote for it to be fully enabled, or to continue as a fixed thing. A trial is a trial, the consensus was to continue only the trial, not for it to be taken as read that it was a done deal. The hard stop date MUST be applied. Now. BarkingFish 01:15, 19 February 2011 (UTC)

Well, the thing is, it isn't fully enabled either. Less than 1000 articles are affected, and admins apparently get some sort of a message discouraging them from adding any more. So the people who think it should be rolled out now aren't getting what they want either. It's just a bureaucratic anomaly. Wnt (talk) 05:09, 19 February 2011 (UTC)

It looks like a failure to adhere to consensus. Perhaps this is an instance where consensus should be ignored, but no one has made the case for that in this discussion yet. --Anthonyhcole (talk) 13:58, 19 February 2011 (UTC)

• "Pending change protection" should be removed from all articles, and then we can have a discussion about possible future use. There is currently no consensus for it. The community agreed to a time-limited trial, and were explicitly told by WMF, At the end of that, unless the community clearly requests otherwise, we'll turn it back off.ⁿ¹

When that did not happen, we had a straw-pollⁿ², which was "inconclusive" to say the least. Jimbo instigated another pollⁿ³ which said it would set no precedent for future use with a hard stop date of December 31, 2010; that poll received 60% support, and that was deemed enough to continue usage. The current continued use of PC makes a mockery of due process and consensus. The only sensible course is a) remove it, and then b) discuss if we want to conduct a further trial implementation or not, and what the remit of such a trial might be. I believe this is the only way we can ever truly resolve this messy situation.  Chzz  ►  14:50, 19 February 2011 (UTC)

Jimbo wants pending changes to continue, so of course he's not going to stop it on December 31 to have another poll. That was up to the community that made the decision. But it seems the community forgot or maybe didn't know how go about it. There was no new poll, and if there was discussion at any of the village pumps I didn't see it. I didn't see any requests on Bugzilla to have it disabled, and checking now I see that the most recent reference to pending changes is bug 26335, a tracking bug filed prior to the "hard stop date". If we never actually requested for pending changes to be turned off, we shouldn't be surprised that we still have it. Reach Out to the Truth 16:54, 19 February 2011 (UTC)

The last poll, run on a headcount, not consensus poll, was 289 for keeping the tool switched on and 199 against. All I see here, is people fixated in process, no good reasons to actually switch it off are being presented here at all. Its become a useful addition to the toolbox of article protection and ignoring that for the sake of historic process is pointy imo. Off2riorob (talk) 17:04, 19 February 2011 (UTC)

That's very black and white. The last poll was not simply about "keeping the tool" and "switching it off", it was much more complicated than that. Please clarify how POINT applies here? - Kingpin¹³ (talk) 01:02, 20 February 2011 (UTC)

I didn't link to any guideline or essay, thats my opinion that objecting to a useful tool because of issues about it was supposed of be switched of after two months is pointy. and that is also my opinion about what the last poll was basically about, users can read up on all the complications if they are interested. Here it is - Wikipedia:Pending changes/Straw poll on interim usage interim usage the title says, so it was probably about that. Off2riorob (talk) 01:46, 20 February 2011 (UTC)

Disagreeing with you on what is and is not useful is not pointy, it would help if you assumed good faith: I'm sure all of us here want the best for Wikipedia, even if we have different opinions on what the best way to achieve that is. - Kingpin¹³ (talk) 14:11, 21 February 2011 (UTC)

Yea, it looks to me like it was never turned off because nobody ever actually asked for it to be turned off. The people who didn't want it turned off probably assumed the people who did would make such a request. In any event, I agree it's not worth getting caught up in that now. Perhaps process failed us there and everybody thought somebody else was minding the details or making the final determination. It doesn't really matter, we need to focus first on if we are keeping it or not, not whether it should have already been turned off. It would be silly to turn it off only to turn it back on again later, and there's no rush that I can see. The crux of the issue should be this: is pending changes a good thing or not? Is it causing harm? Is it preventing vandalism? If it isn't causing harm and is preventing vandalism then we should keep it, regardless of any past errors in process. Beeblebrox (talk) 20:17, 19 February 2011 (UTC)

How could we have made such a request without the supporters claiming that we were asking another parent, given they had a majority in the poll? —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 21:04, 19 February 2011 (UTC)

I don't know, but the point is the focus of this discussion should not be "why wasn't it turned off" but rather "are we going to keep it." Lets not dwell on how it might have been done differently and instead focus on moving forward. Beeblebrox (talk) 21:46, 19 February 2011 (UTC)

Isn't the argument of "stupid to turn it off only to turn it back on again" exactly the reason it wasn't turned of immediately last time, and why it's still on (without consensus) now? I do not think it would be pointless to turn it off, as it would finally bring a halt to this headlong charge and allow users to talk about this logically and fairly. - Kingpin¹³ (talk) 01:05, 20 February 2011 (UTC)

What we all need is a clear, current consensus to come out of the community so that we can end the perpetual trials and move to either turn it off permanently, or improve and use it in the long run. Not a running debate about the past. (Just to be clear: I'm here to act as a community liason for tech staff, not to be spokesman for every Foundation action or decision of the past months or years.)

As far as turning off Pending Changes while we have a discussion, I fielded the idea to tech staff today. If it were as easy as flipping a switch, then I think we'd be comfortable potentially doing it as a show of good faith. But we've always maintained that if we turn it off, we should leave it off, not flip-flop. Though it's not extremely hard to turn Pending Changes itself off, there are some issues around it that would require time and energy, such as maintaining the test suite. In general, tech feels that if they're going to expend resources shutting it down, it should be your permanent choice, not an interim one. If you don't want Pending Changes used while you have a discussion about its long term future, then as always the community can simply remove it from all articles. Steven Walling at work 02:02, 20 February 2011 (UTC)

I think that temporarily removing it from all articles would be entirely sufficient to fulfill the promises that have been made, even if it is retained on some test pages outside the main namespace. It may not be very important to remove it from the small list of articles where it is currently active, and it may not be beneficial, and the discussion below does need to happen; but it will be much easier to have that discussion and get people to work through the myriad policy issues necessary for Wiki-wide adoption if they feel that the outcomes of previous discussions have been respected. Wnt (talk) 08:11, 20 February 2011 (UTC)

If removing it from every article it is currently on and banning its further use for the duration of this process is what it will take to convince everyone to stop arguing about the past and focus on the relevant question of what to do in the future then I'm all for it. Beeblebrox (talk) 18:20, 20 February 2011 (UTC)

We just keep going in circles. Doc James (talk · contribs · email) 23:34, 20 February 2011 (UTC)

Until this fundamental issue is solved we will never get anywhere. So why doesn't some admin step up and remove pending changes from all remaining articles so that this becomes a moot point and we can discuss where to go from here? ~~ GB fan ~~ 15:40, 21 February 2011 (UTC)

That seems like an unnecessary disruption of Wikipedia to prove a WP:POINT. We all know what Wikipedia looks like with and without pending changes and can discuss the relative differences. In the meantime, pending changes is protecting articles now, and removing it immediately puts those vandalism-prone and/or sensitive articles at risk for misinformation and BLP concerns, especially if it is not replaced with semi-protection or another mechanism. The goal here is not to blindly "fulfill promises," but rather to discuss the future as you say, so why put the encyclopedia at risk while we have the discussion? Zachlipton (talk) 15:46, 21 February 2011 (UTC)

You're talking about several hundred pages out of three million! If it is so terribly disruptive to give up PC, how can we stand having all those other pages out there? The pages were subjected to this as part of an experiment - they can do without it. Wnt (talk) 05:56, 22 February 2011 (UTC)

I don't like the accusation of point; perhaps the message I wrote on 25 August 2010 might clarify my feelings; I don't want to unnecessarily repeat things here, but please read Fait Accompli. Thanks. (the subsequent thread is now archived in Wikipedia talk:Pending changes/Straw poll/Archive 1).  Chzz  ►  18:10, 21 February 2011 (UTC)

Fair enough, and I'm sorry if my tone was unnecessarily accusatory. I can certainly understand the perspective you're coming from. My point really is that the discussion we need to have is what to do with pending changes in the future, and whether its turned on or off while we have that discussion is a secondary issue that reframes the main debate. Zachlipton (talk) 18:55, 21 February 2011 (UTC)

I can tell you this admin hasn't "stepped up" and removed PC from every article it is on. Because, as has been mentioned, admins have been repeatedly advised not to fight or do anything drastic in this regard. So far a few users are arguing that they don't believe it should be discussed until it is off, but that argument is based on emotion, hurt feelings that nobody turned it off when the trial was over. What would it actually accomplish to turn it off, other than to soothe a few disgruntled users? It is entirely possible to discuss permanently retaining or rejecting PC whether it is turned on currently or not. This is just a political point, it is not really relevant to the larger question of whether we want PC or not. If you want someone to take the blame for it not being turned off, it would be nice if you could identify who that someone is. Nobody seems to know the answer to that, but again it has no bearing on the wider issue and is only a distraction from what we should be discussing. Beeblebrox (talk) 19:21, 21 February 2011 (UTC)

I resent the accusation that my argument is "based on emotion, hurt feelings". It is based upon the core principles of consensus. I am, in fact, trying to help move things forwards; I feel the way to do so is, to accept that - currently - there is no consensus for using it, and thus, accept the community has not approved its use beyond a time-limited trial. I am not trying to pin any blame on the current situation; I am merely trying to move things along: we previously tried to discuss possible implementation, and for whatever reason we have not come to an agreement (yet). To me, the best way forwards is to adhere to the consensus. The community has not yet agreed to its use. If the community is given a reasonable proposal for its use, in whatever form, and across whatever range of articles - whether that be a further time-limited trial...then we might be able to discuss this. Zachlipton, thanks for your understanding. Beeblebrox, please reconsider your undue emphasis of your views, in this RfC - in terms of the bolding of comments, and addition of somewhat slanted L2 headings - Wikipedia:Shouting things loudly does not make them true. Quietly, all I am asking is: let's start afresh; let's forget the previous poorly-managed trial - with no need to apportion any blame, but let us learn from it; set our remits clearly, agreed by consensus, and make some progress. Best,  Chzz  ►  22:43, 21 February 2011 (UTC)

Ditto Chzz, in addition, at the moment one of the main arguemnts for keeping PC at the moment seems to be "well it's on at the moment, what would be the point of turning it off?" Which isn't really just when you consider one of the main reasons it got turned "on" in the first place is because there was the understanding that it wouldn't stay on indefinitely - Kingpin¹³ (talk) 06:21, 22 February 2011 (UTC)

Poll wording

So, the word from the front office is that we need to work out if we want this tool or not - so it looks like we need to work out the format of the poll and the percentages we are looking for to keep on or switch off. I think the wording should be simple and along the lines of Do you support pending protection - yes or no. Without any conditions as to time limits or usage or any more trial - we have had a long enough trial now for people to have made up their opinions as to its benefits or not as the case may be and then if the answer is no, we can just forget about it and move on or accept the tool and use it and develop it as required. Off2riorob (talk) 20:14, 19 February 2011 (UTC)

But do we have to do it by polling to reach some specific percentage threshold? I realize that is easier to measure than consensus, but I'd rather we didn't just vote on it. Beeblebrox (talk) 20:19, 19 February 2011 (UTC)

Having got in over my head in creating the first poll I am not going to get over involved in this set up, but we don't have to poll, we can discuss and convince and look for consensus. The main objections don't seem to actually be about the tool but about the process. Off2riorob (talk) 20:25, 19 February 2011 (UTC)

Aren't we getting a bit ahead of things? There's been a more-than-two-month trial - what were the results? Has anyone figured out how often good additions were reverted by accident? And what about the technical upgrades that everyone is supposed to be waiting for - are they done, in progress, abandoned?

There's also the larger question: why should people have to decide to abandon the PC project just in order to end the trial? There's a short list of <1000 articles subject to PC - if this mechanism is even conceivably usable on a fraction of Wikipedia as a whole, hundreds of thousands of articles, it shouldn't be a big deal to strike a few hundred articles off the list for a few months until the software is working optimally. No one is saying you have to forensically erase all the software off the Wikipedia servers. Wnt (talk) 21:08, 19 February 2011 (UTC)

I don't believe we are getting ahead of ourselves. Further development of PC has been put on the back burner exactly because the trial has petered out and a new discussion, like this one, had not yet taken place to decide whether to even keep using it or not. There's no sense working to improve something that might not be used, so they have been holding off. Therefore, the first and foremost decision that needs to be reached here is if we are going to even keep PC. Beeblebrox (talk) 21:44, 19 February 2011 (UTC)

Perhaps it's not an either/or proposition. I, too, see a lot of value in going ahead with a community decision about whether to keep or to discontinue, but I also think that it would be a lot more productive to conduct that discussion based upon the kinds of information that Wnt would like to see collected. --Tryptofish (talk) 21:58, 19 February 2011 (UTC)

If somebody wanted to go ahead and collect that information and post it here I agree it would be very helpful. Beeblebrox (talk) 22:02, 19 February 2011 (UTC)

I don't think there are any figures, especially not as regards, the number of good edits rejected or the number of bad edits accepted. The only assessment is the users that have been active with the tool, users experience. I would like to see figures of the users most active in accepting or rejecting pending edits if there are any as those users would be good to get comments as to the operation and interface questions. There are some users that just object to pending protection as a matter of principal in regards to free to edit and article control issues. In reply to Wnt's comment about the upgrade there is a link to a front office comment User_talk:Risker#Long_post_is_long. that until the community decides if they actually want the tool or not any upgrades are not being worked on, why waste their time if it is not even clear what the community wants to do with the tool. Off2riorob (talk) 22:09, 19 February 2011 (UTC)

I personally would like to establish a clear policy on applying PC before deciding whether to continue it and then making policy changes. In addition, I would like the finalized details of the poll agreed upon by the community before any poll starts. I think the "jury" idea is not bad and that the community/voters should be voting based solely on whether PC is helpful or not (e.g., technical (dis)advantages, effectiveness, trueness to Wikipedia's mission, etc.) and how it should be used (hopefully, after this RfC, the only issue about this is what scale it should be used on—all articles, BLPs only, also non-mainspace pages, etc.). I don't like the idea of "yes/no" because it doesn't mean anything—we should have separate categories for "does it work [technically/effectively/efficiently/etc.]" and for this we would need some sort of data analysis. I'm sort of disappointed the WMF didn't hire a data analyst to examine PC. /ƒETCHCOMMS/ 03:59, 20 February 2011 (UTC)

I totally agree that it's more complicated than just yes or no. It's more like, "How?", "Where?", and "Should we at all?" This kind of complexity is exactly why we agreed with Risker's idea about a community jury of sorts, since their first task could be to make "the finalized details of the poll" and get them "agreed upon by the community before any poll starts", like you said. If we could do it without pre-selecting community members responsible for that, then that's fine too. However we do it, the ideal situation is that the next poll answers both the core turn it off/leave it on question and the more complex questions about how and where to use it. Hopefully without confusing all those factors. Steven Walling at work 06:06, 20 February 2011 (UTC)

Thats a difficult ( impossible ) thing to assess with a single poll and also imo completely unnecessary. What I see as the simple version, is a simple yes or no to the tool and then just allow normal editing usage to take over, there is imo no need to set any guidelines as to usage, if the answer is yes, keep the tool then just allow normal editing and ongoing assessment and experience of where it is useful and where it is not useful to settle where ever it wants to naturally. If the answer is no then just switch it off, attempting to use community polling to set such guidelines for usage levels is simply a wrong idea and likely never going to find any consensus either. Off2riorob (talk) 14:36, 20 February 2011 (UTC)

I'm still not convinced a poll is a good idea at all. I don't think it is needed, because so far no objections to the actual use of PC have become apparent, all the objections have been of a procedural nature, hurt feelings at the perception that consensus was ignored and so forth. I am trying to convince everyone to let that go and address the real question. If real, substantive objections are raised then maybe we should move ahead with polling. If none are manifest by the end of the week I say we move ahead with clarifying how it is to be used and forget the poll on if it to be used. Voting would allow people to oppose solely on procedural issues. Discussion requires them to provide a real, substantive reason not to continue using it. So far no such reasons have been presented. Beeblebrox (talk) 18:16, 20 February 2011 (UTC)

I would rather get 5 root canals than endure yet another round of torture by PC polling. —UncleDouggie (talk) 11:32, 21 February 2011 (UTC)

Yea, me too. People come to enjoy and all this naval gazing about such as this just turns them off, and I suspect and it seems apparent that there will be less interest in this discussion than the previous ones. Off2riorob (talk) 17:15, 21 February 2011 (UTC)

Do we really need to do this?

Honestly. I know this is all done is good faith and with the aim of having consensus behind implementation... But really. Pending changes, in my eyes, has proven to be nothing but beneficial to every page I have seen it activated on. The current system works well. Why not instead put out a call for anyone to voice their concerns with the current system, now that we've had a many-month trial? Does it need to be improved; has it impeded editing; has anybody complained about it? This would show the general disdain. I honestly don't think there would be any anecdotal evidence of pending changes causing issues with good-faith positive and constructive contributions to the encyclopedia. As such we should be moving forward, not asking if we should move forward. - ʄɭoʏɗiaɲ ^τ _¢ 03:33, 20 February 2011 (UTC)

Isn't that part of this RfC? To see what the community thinks about PC? My concern is that a. there is no major update coming that further improves loading time/more functionality (PC articles have to be loaded by ajax or javascript or something, it seems; old revisions still take forever to load on long pages), and b. that the community's trust has been exhausted given that this two-month trial has turned into a seven-month one. If the WMF wants to turn on PC, they have the authority to do so. If they want to leave it on indefinitely without any attempts to further divulge information on continuing/ending PC until after this RfC was started, then I see a serious problem. /ƒETCHCOMMS/ 03:51, 20 February 2011 (UTC)

I work on PC and I haven't noticed any recent loading problems with any pages. Also you claim the communities patience has been exhausted, this is not true at all, it might be your feelings but not overall. I don't see a massive outpouring of objections to this tool here or anywhere else. All I see are a few users with their own personal objections to the tool being left on without a consensus. Off2riorob (talk) 12:37, 20 February 2011 (UTC)

Again, major revisions to PC are on hold pending the resolution of the question of whether or not we are keeping it. If we decide to leave it turned on it then they will turn their attention back to it. We're talking about people who are in the small circle who are paid to be here. Paid with the money from the fundraiser we just had. Money they are being careful not to waste developing a tool that may not be used. I have to agree with Rob, the main objection I am seeing to keeping PC is that the trial didn't end when it was supposed to. That is an established fact. Who was responsible for making sure something happened when the trial was supposed to be over is a much more murky area. Everyone seems to feel it was somebody else's job to take the next step. So none got taken. Until this RFC was opened. The WMF are asking us to tell them what to do. That should make most of you happy I should think. So I'll say it again, our first order of business has to be to decide whether to keep it or not, and "it wasn't turned off when the trial was over" is not a response that helps us come to that decision. Beeblebrox (talk) 18:07, 20 February 2011 (UTC)

I'd like to point out that I made a comment earlier in this discussion, that was a substantive complaint about the performance of PC (and, Rob, it was about page loading), and it was not at all about any of the consensus stuff. Ironically, we are now in a loop, where some of the community want to see improvements to PC, without which we do not want it to continue, and the developers are not going to improve it, without a consensus to keep it. --Tryptofish (talk) 20:54, 20 February 2011 (UTC)

Thanks, I am getting a lot of current issues with my wikipedia interface at the moment as I heard some alterations are happening/have happened and I can't get rid of this beta thing on some articles. Anyways, I haven't really noticed this loading large pages as an issue in relation to pending particularly, I get problems with loading large articles in general, which articles with pending protection on them have you been experiencing this issue on? Off2riorob (talk) 17:21, 21 February 2011 (UTC)

During the trial (the actual trial period) I had a problem with World War I, giving a SQL error, and discussion with techs told me that it 17 seconds to render that page, that PC "often tries to parse things twice", and the squid caching servers timeout at 30 seconds. Despite the error it seems that sometimes the change is actually accepted. I am not sure if that bug has been resolved, or if there are still problems with PC on large pages. See here for the original reported issue. There's also this bug which may or may not be related; sorry that is all the info I have. Maybe someone here knows?  Chzz  ►  18:46, 26 February 2011 (UTC)

Thanks for these links, I will have a look in the next couple of days. Off2riorob (talk) 19:38, 26 February 2011 (UTC)

One problem with PC: the default term (indefinite)

Hello. I do not have a great deal of experience in PC, and so my opinion is only partially formed. In a nutshell:

1. It was disrespectful of the original collegial decision to allow the trial to effectively carry on beyond the agreed term of two months. Not only PC protections are still there, but new ones are being liberally (my opinion) issued on a daily basis.
2. In some cases it is preferable to semi-protection, but this can be deluding us into making WP more closed and protective (in the bad sense of the word 'protective') instead of making it more open.
   OK, so far so vague and probably trite. However, what I have not read about in my skimming of the various straw polls is:
3. Term of protection. I have reasonable experience with indefinite semi-protection, mostly from a point of view of challenging unreasonable or old ones and have them reversed/unprotected. One experienced admin in this field recently argued that indefinite semiprotections are very often forgotten about, and they practically turn to mean "infinite" protections. I agree with this strongly, and I think that PC suffers from the same fundamental problem. Therefore, I think the proposal to a complete ban on indefinite protections (or, more weakly, to have all of them obligatorily peer-reviewed), in order to prevent abuses, applies to PC as well. 220.100.87.222 (talk) 12:40, 20 February 2011 (UTC)

Good point. Start another proposal. Dokter Zombieman ^brains.../_{the infected} 12:43, 20 February 2011 (UTC)

That proposal was made about a year ago and failed to gain consensus. If PC is kept it will be subject to the protection policy the same as the other forms of protection, and any protection can be appealed at any time by contacting the protecting admin or filing a request at RFPP. The most common request related to PC I have seen has been users who don't feel it is doing enough and they want semi-protection. 17:55, 20 February 2011 (UTC)

Right, you can appeal at any time, but to an admin protection is just a click, whereas to undo it it takes a lot of energy and hardly anyone has the skill and stamina to do it. And even when you do manage to present a case, you get this kind of crap, which evolves in this other kind of crap. 220.210.177.127 (talk) 23:25, 20 February 2011 (UTC)

If you want to edit in areas subject to vandalism, the requirement that you obtain a pseudonymous account is not too high a bar, compared to the downsides of lacking effective prior restraint on vandalism of known targets. Jclemens (talk) 15:44, 22 February 2011 (UTC)

The point here is not protection of pages that genuinely require protection. The point is admin abuse and dubious protection of pages that have not suffered enough disruption to justify protection. 219.111.127.34 (talk) 12:13, 23 February 2011 (UTC)

Dedicated discussions

• It seems we are all over the map here, in the interest of focusing each thread on a specific issue I am adding dedicated sections for the main topics. Obviously, it is all dependent on what the result of the first discussion is, but we can discuss the other issues anyway so as to be better prepared to move forward if it is decided it is to be kept. Beeblebrox (talk) 23:32, 20 February 2011 (UTC)

Current settings

Interaction of Wikipedia user groups and page protection levels

	Unregistered or newly registered	Confirmed or autoconfirmed	Extended confirmed	Template editor	Admin	Interface admin	Appropriate for (See also: Wikipedia:Protection policy)
No protection	Normal editing						The vast majority of pages. This is the default protection level.
Pending changes	All users can edit Edits by unregistered or new editors (and any subsequent edits by anyone) are hidden from readers who are not logged in until reviewed by a pending changes reviewer or administrator. Logged-in editors will see all edits, whether accepted or not.						Infrequently edited pages with high levels of vandalism, BLP violations, edit-warring, or other disruption from unregistered and new users.
Semi	Cannot edit	Normal editing					Pages that have been persistently vandalized by anonymous and registered users. Some highly visible templates and modules.
Extended confirmed	Cannot edit		Normal editing	Normal editing*	Normal editing		Specific topic areas authorized by ArbCom, pages where semi-protection has failed, or high-risk templates where template protection would be too restrictive.
Template	Cannot edit			Normal editing			High-risk or very-frequently used templates and modules. Some high-risk pages outside of template space.
Full	Cannot edit				Normal editing		Pages with persistent disruption from extended confirmed accounts. Critical templates and modules.
Interface	Cannot edit					Normal editing	Scripts, stylesheets, and similar objects central to operation of the site or that are in other editors' user spaces.
* In order for a template editor to be able to edit extended confirmed protected pages, they must also be extended confirmed, but in practice this is almost always the case.
Other modes of protection: Create protection Move protection Upload protection Office protection Cascade protection

• Metrics
• Log
• currently pending
• All PC protected pages

Should we approve pending changes as a permanent part of our protection tools or have it turned off?

Is pending changes preventing vandalism? I don't know. We have no way of knowing how many vandals decide not to bother saving their changes once they realize they won't be visible right away. I imagine this is frustrating to vandals, but have no way of knowing for sure. Is it driving away ip users? Again, hard to know as it is impossible to tell if an ip has left or just rotated to another number. My personal experience with articles on my watchlist has been that PC generally slows down the vandals but does not stop them altogether. And ip users who understand how it works will add their constructive edits on the assumption that they will be seen as such and accepted. Beeblebrox (talk) 23:46, 20 February 2011 (UTC)

Yes we should approve pending changes as part of our tool kit. It is much easier to pick up and revert vandalism before it is followed by half a dozen other edits. PC will not be a cure all but it has a limited place just as semi protection and protection. Doc James (talk · contribs · email) 23:48, 20 February 2011 (UTC)

• We should first remove PC from all articles. Then, we could discuss possible future implementation. It does not have to be an all-or-nothing decision, as the wording of this question indicates. If the community are appropriately consulted, then we might be able to agree to some form of trial - whether that is on 100 articles or 1 million articles, I do not know. We could propose and agree to some series of phased implementation, perhaps - with key measures, milestones, along the way. We could try it out on some selected articles of various kinds - and we could gather stats on the success or otherwise. We could work with WMF, to examine the benefits and problems.

What does "part of our protection tools" mean? PC can be implemented in so many very different ways - and that is a key point in these discussions. I advise anyone who has not already to read this blog about it.

PC can mean anything, from dewiki system of protecting all articles, through to perhaps using it only on BLPs on the main-page. It is not possible to give a "yes/no" answer - for a "permanent commitment" to a partially-developed system, without defining the scope of usage.  Chzz  ►  03:19, 21 February 2011 (UTC)

I meant deploying it in a manner consistent with out current protection policy, I don't think anyone is still advocating for de-style "flagged revisions," that went away a long time ago. I think it is possible to give a yes or no answer to the question of whether we keep PC in the form used in the last trial, it's exact scope and use, if not subject to the normal conventions of the protection policy, can of course be clarified, and in fact that is the purpose of the sections below this one. Beeblebrox (talk) 19:05, 21 February 2011 (UTC)

One of the original trial's condition was that PC will be taken off as soon as the trial ended. Sadly, many are still in place and then WMF promised to have a solution by November. That's two broken promises already. Take it off first before any meaningful discussion can continue. OhanaUnited^{Talk page} 19:07, 21 February 2011 (UTC)

We have had two major polls since then anf both supported continued use of the tool, its time ot let go of such emotional thoughts and look at the value of the tool now. As for the upgrades, the tool is better and thanks to the people that worked on that but until the community either accepts it or rejects it their will be no more improvements, so its make your mind up time. Off2riorob (talk) 19:31, 21 February 2011 (UTC)

We have no reason to look at the value of the tool because the Foundation and supporters like you lied to us twice already. Time to end this attempted fait accompli. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 22:06, 21 February 2011 (UTC)

Aw, diddums, welcome to the adult world. Off2riorob (talk) 22:09, 21 February 2011 (UTC)

Mind not being condescending? —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 23:13, 21 February 2011 (UTC)

Only if you don't mind not referring to me as a liar. Off2riorob (talk) 23:17, 21 February 2011 (UTC)

But you have. You promised the community two months and reneged hard on that; the devs promised a fixed version but from what I'm hearing they've more-or-less reneged on that as well. How can you not be a liar? —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 04:06, 22 February 2011 (UTC)

I am trying my hardest not to respond in kind, please stop calling me a liar it is a personal attack and you are repeating it again. I am a supporter and a user of the tool only, I am just a volunteer editor, not a paid employee or even an administrator, I have lied about nothing. I don't care about the two months trial and that there is or was no consensus to continue. usage, ignore all rules if it benefits the project would be my comment. I would just roll out the tool and wouldn't even listen to your comments but that does not make me a liar, please stop your personal attacks. Off2riorob (talk) 10:34, 22 February 2011 (UTC)

• I'd say turn it off. At this point, it's much less useful than semi-protection. And I'm saying this as someone who used to strongly endorse it. --Tryptofish (talk) 22:49, 21 February 2011 (UTC)

I have to mention - you voted to switch it off five months and four days ago. Off2riorob (talk) 22:58, 21 February 2011 (UTC)

Please look at what I actually said in that link. I guess no attempt at good faith goes unpunished! --Tryptofish (talk) 23:04, 21 February 2011 (UTC)

No punishment intended, you refer to no consensus to continue and to close it on good faith and you wanted a major improvement which I will agree hasn't occurred although this version is better than the original version. Personally imo if you switch it off discussion and the tool will vanish completely. Shall we switch it off and see if I am right? What conditions are required to switch it off at this stage, there has been more people support the tool in all the polls than reject it? Off2riorob (talk) 23:14, 21 February 2011 (UTC)

• I would still like to see it turned off. I've tried to work with it, and have even added it to articles a couple of times, but it's rarely helpful, it's slow, and I still find it awkward to use. SlimVirgin ^{TALK|CONTRIBS} 23:22, 21 February 2011 (UTC)
• Yes, I'd like to see it turned off. I admit I'm one of those who would like to see all articles with some kind of protection, but like SlimVirgin says above, I find it very awkward to use, and I don't see it helping much. I did not notice a decrease in vandal activity in the articles I watched that had it enabled. — Huntster (t @ c) 00:50, 22 February 2011 (UTC)
• It should go off. Consensus was never reached for it to remain on, and is tecically still in the way overextended trial phase. Sumsum2010·T·C·Review me! 04:16, 22 February 2011 (UTC)
• Turn it off. Half a year into a two month trial, and I still haven't seen a single actual benefit from it, and lots and lots of time going into it. Contributor time is the single most limited resource this project has, and this feature burns through that priceless resource for no particular benefit. Courcelles 07:49, 22 February 2011 (UTC)
• Kill it. Hell, if you do, you may win back some of the community you screwed over by extending it well past it's two month period! —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 10:22, 22 February 2011 (UTC)
• Turn it on and turn off the people opposing it. ;) It's not perfect but it does add a layer of protection that is most definitely needed on some articles. Like many others, I found it confusing to use but I'm sure it can be fixed. If anyone doubts that it's needed, try editing the articles related to any Disney TV program for a while and you'll see it's a far better option than semi-protection. What we do need to do, first, is tighten up and formalise WP:PC. --AussieLegend (talk) 11:35, 22 February 2011 (UTC)
• Keep it and improve it. It's a useful tool to deal with vandalism and while it's substantially more complex than semi-protection, it also allows editing that semi does not. Jclemens (talk) 15:42, 22 February 2011 (UTC)
• Keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it keep it! It allows IPs to edit, gf's! --Perseus8235 16:34, 22 February 2011 (UTC)

  I can be just as vehement about PC/FR: DieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDieDie... —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 23:08, 26 February 2011 (UTC)

• I'd like to remind everyone that we are not voting, and to ask again that discussion in this section be limited to reasons to keep or reject pending changes. Reasons based on the tool itself and its usefulness, not reasons to be angry at whoever it is believed is to blame for it not being turned off already. Beeblebrox (talk) 16:37, 22 February 2011 (UTC)

I would say:

1. Keep the software for pending changes throughout - the only question should be where it is enabled.
2. Remove pending changes from all existing articles. It was turned on before as an experiment for a limited trial which is supposed to be over. Some of those groups might, arguably, "need" some sort of protection, but their status needs to be gone over one by one, starting from the default unprotected state. It's only several hundred articles - there are probably more that were started last week and haven't been gone over.
3. Conditionally, allow implementation of Level 1 only on specific pages for limited intervals as an alternative to semi-protection. Talk pages should not ever be considered for PC. Pending changes should be considered only by admins capable of imposing semi-protection, who feel that the normal criteria for imposing semi-protection have been met, who can set a fixed end date for each affected article, and who are willing to report at the time of shut-off to a central discussion area about whether PC was an effective alternative and any problems or suggestions. It is expected that the results from the current experiment, if any, will be reported to the same forum, and that conclusions from this forum will be reported by some direct liason to the MediaWiki developers, and their progress likewise will be reported back to that forum.
4. As a condition for the above, the WMF must strongly assert and defend the position that the pending changes are "published" when the IP submits the edit, and that the editor who accepts the changes is only helping to quickly screen published material for errors - and is not legally liable for any libel or other misrepresentation contained in an edit, even if he should have been able to see it. This must be made clear by policy, and assented to by the WMF, before the PC mechanism is reapplied to article pages. Wnt (talk) 18:23, 22 February 2011 (UTC)

• I don't think this needs to become another giant poll, but we really ought to consider removing PC. Both sides have good arguments. There is a claim to be made that placing PC on currently semi-protected articles allows wider editing of those articles than we now allow. There is also a claim that PC represents a function creep--much like Tasers allow for the use of non-deadly force but also result in higher levels of "force" being used because the options available are more varied. We don't have any strong evidence that PC prevents the sort of vandalism we care about (subtle vandalism to articles with few active watchers) but we also don't have strong evidence that PC represents a net loss in new editors or readers. Part of the reason we don't have any evidence like this is because we didn't have a real trial. We just had a "provisional" adoption of PC (which was never meant to be provisional at all) with ad hoc application all over the wiki. A true trial where currently long term semi-protected articles are switched at random to PC would give at least some evidence of efficacy or impact. Instead we just had a 2 month toe in the water period which now appears to be indefinite; the principal finding being that PC didn't solve all of our problems and didn't break the wiki. A tool like this which both complicates readers' conversion to editors and represents a potential shift in how we approach open editing needs more of an endorsement than "doesn't break the wiki". Because once it is in place (more) permanently we won't be able to remove it and we will not be able to reliably determine the tool's impact on overall editing trends.
• Another reason for a true trial is that it will avoid this mess of constant polls, whining about data, and proxy fights over ideology. The foundation can simply design an experiment, get the consent of the community to run it (much easier than the process we are in right now), and run it. Once the experiment has run its course then we can make clear claims about impacts on editing behavior due to the tool alone.
• Absent a real experiment or any show of good faith all I see is a tool which the right people want to have implemented being kept around un the assumption that attrition will finally wear away at the opponents. We don't allow that kind of behavior from regular editors or admins so I'm at a loss to see why we should expect it from the foundation. Protonk (talk) 21:51, 22 February 2011 (UTC)

• Pending changes is much better than the persistent proposal "Stop all IP's from editing". One thing I've noticed is that PC kinda got abandoned, because there just weren't enough things to be reviewed. Maybe if there was an IRC channel where new PCs could be reported, PC would be much more efficient. ManishEarth^{Talk • Stalk} 03:22, 24 February 2011 (UTC)

• Keep it. I am a strong supporter of PC, and feel that it should be a permanent part of our protection tools. More so, I believe that it should gradually take over and possibly even replace indefinite semi-protection. I believe that full protection and semi protection go against what makes Wikipedia special- being the encyclopedia that ANYONE can edit. However, I understand that, for many different things, protection is necessary to preserve the quality of the encyclopedia. I support keeping full protection as it is now, but phasing out indefinite semi-protection and replacing it with PC. Whenever you indefinitely block a page from being edited by IP editors, who can offer a wealth of knowledge that any encyclopedia needs in order to grow, you block the possibility of expanding and improving Wikipedia. Temporary semi-protection should remain as-is as a response to vandalism and other things, which is why I believe that semi-protection has its use as well. What I want to point out very strongly is that I do NOT believe that PC should be applied to every page, or even to every BLP. It should only be applied to those pages that would otherwise be placed under indefinite semi-protection; otherwise, we risk destroying Wikipedia's reputation and the freedom to edit that many have come to identify with it. Many people have also stated concerns about a delay in PC acceptances. However, I believe that if PC was expanded, it could become something like counter-vandalism efforts are now, and could actually be done very quickly. If it is only applied to a couple of pages, then it isn't worth having, because nobody will want to monitor it on the off-chance that someone will edit those pages. However, I can't say that I even agree with the statements that it takes a long time for pending changes to be accepted, since every time I manage to see them, they're already "under review". --Slon02 (talk) 23:20, 24 February 2011 (UTC)

• Keep If I can add a completely outside view of this, I was on a pretty long wikibreak when a lot of the trial happened (so much so that I didn't get reviewer!) and there is only one page in my watchlist that was in the trial. From looking at the discussion here there is one thing that sticks out to me the most - the chart in the Current settings section about what level of protection accomplishes what end. It seems to me that semi-protection applied without pending changes is a useless level of protection to maintain. The gradual ramping of protection in that chart would be logical without that level. Keep pending changes, and modify the protection policy so that semi-protection is a level above, and only to be applied, after PC level 2 has failed to do the job. The only other way to achieve equal results would be to drop PC altogether and greatly increase the autoconfirmed requirements. I do not think that is a desirable change. Autoconfirmed is such a low barrier to editing, that it really isn't any barrier at all, and I do not believe that serious contributors have the same instant gratification needs that vandals, trolls, and spammers have. Those with a true desire to add useful content will not be as easily turned away as those with nefarious intent. Jim Miller ^{See me | Touch me} 16:35, 25 February 2011 (UTC)

No, not in the form used in the last trial. Not yet ready: too slow, couldn't handle multiple edits, not transparent to users. Possibly in some other form, but , as Chzz says, the first step in getting a form with good functionality is to get rid of the present one. DGG ( talk ) 05:05, 26 February 2011 (UTC)

• Keep it but improve it and expand to double/triple the amount of articles and with some way of collecting data to prove its usefulness. IP users who want to vandalise and those who want to edit both go "Huh? where's my edit?" but at least the ones who are editing correctly can have theirs put through.

It must be better to have both stopped and a small wait on good edits rather than having vandalism going live straight away, relying on someone to come and fix it at some time in the next year. Any outstanding revisions can surely have a system set up so they can be collected like we do with categories? (Or even a category for "outstanding revision"?)

Really we need closure on "has it worked", "how much has it worked" as that is the only way to judge how bad it would be if we took it off. Chaosdruid (talk) 18:48, 28 February 2011 (UTC)

See [2] for current pending edits. As of this writing there are but two edits awaiting review, both less than an hour old. Beeblebrox (talk) 19:41, 28 February 2011 (UTC)

Sorry, but that stat is utterly meaningless.

PC is on 953 pages right now. We have 515,258 pages in Category:Living_people - although, OK, let's say "only" 100,000 articles. 60 minutes, 953 articles, = 3.78 seconds per article. Times by 100,000 means 104 hours to process each change. Yes, this is meaningless; 42.9% of statistics are made up on the spot. Conclusion: we've absolutely no idea what affect it will have, how long the queue will be.  Chzz  ►  04:03, 1 March 2011 (UTC)

No. Get rid of it. It increased overhead, it didn't decrease workload. I saw unhelpful changes approved which still had to be backed out. Articles with persistent vandalism problems still have recourse to semi-protection. Tarl.Neustaedter (talk) 00:42, 2 March 2011 (UTC)

In what types of situations should pending changes be used?

This is a sticky one. Generally it has been used significantly less than semi-protection, but that is likely because of the "trial" and the advice not to do anything drastic. This is supposed to be a "soft option" to semi protection so it could see greatly expanded use if it is kept. Obviously the general advice of the protection policy applies, are there other points that should be added specifically for pending changes? And are those standards to be interpreted more liberally if the article is a WP:BLP? Beeblebrox (talk) 23:39, 20 February 2011 (UTC)

Some of the article that are currently semi protected could be switched to PC and some of the borderline cases could have it applied. It allows a more nuanced approach to vandalism.Doc James (talk · contribs · email) 23:51, 20 February 2011 (UTC)

• This is where we desperately need some analysis of how it has performed, in the trial. In which circumstances was it beneficial, and which were not so good? Surely it isn't too hard to gather such information; I really did think that was the point of a trial. We should be able to compare articles with a similar prior history of vandalism which a) had PC b) had semi, and compare and contrast? And then we might be able to make informed comments regarding possible future implementations. We're shooting in the dark here. We had a trial, but we haven't analysed the results. If the trial didn't produce any meaningful results, then perhaps we can plan a new trial which will do so. If we remove it from all current, we'd have additional data to look at - seeing if the article problems are increased or reduced when they don't have PC. Chzz  ►  03:25, 21 February 2011 (UTC)

  Agreed. Someone neutral needs to crunch the numbers and come up with hard data. Until then, this section is sorely premature. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 22:08, 21 February 2011 (UTC)

• I think this needs to be left up to admins' discretion. Inevitably, if we start drawing up criteria for articles to be eligible for PC, not all that are eligible would benefit from it and there will be some (possibly many) that could benefit from it but are excluded by the criteria. HJ Mitchell | Penny for your thoughts? 14:55, 21 February 2011 (UTC)
• Only when it meets the present requirements for semi-protection and is a high profile BLP article or one about a currently popular individual, and possibly in similar articles that are not BLPs but have BLP problems. There never was consensus for using it in non-BLPs--every admin who has done so has applied it without justification in policy & provided an argument why leaving to admin discretion is dangerous. DGG ( talk ) 05:08, 26 February 2011 (UTC)
• PC1 on BLPs of high vandalism and on high edit warring pages, both where IPs and confirmed users are causing problems. This allows reviewers to take some pressure off admins, especially at low attendance troughs, by constructively stopping live edits while notifications/requests are made to PC2/Prot/ANI/3RR boards etc.

PC2 on same where IPs, confirmed and auto-confirmed warring, high vandalism etc.

This would not stop any additions, more edits would be made, just not seen live until checked over.

Obviously after that it's over to the prot and other measures. Chaosdruid (talk) 18:48, 28 February 2011 (UTC)

What kind of time frame is appropriate for individual uses of pending changes?

I think this should be the same as with any other type of protection. The response should fit the problem. If protection is being applied for the first time, it certainly should not be added on an indefinite basis, which is what has often happened up until now. Short periods should be tried first, and only extended if they fail to curb the problem. Beeblebrox (talk) 23:35, 20 February 2011 (UTC)

Indefinite PC makes the most sense. Remember, in BLPs, indefinite PC for problematic BLPs is an alternative to indefinite semi for problematic BLPs. I don't use PC on non-BLP issues, but throw about 1-2 indef PCs on BLPs per week--but then, as ArbCom and OTRS, I see more complaints from BLP subjects than anyone. Jclemens (talk) 15:39, 22 February 2011 (UTC)

No indefinite PC's were ever approved, because they all were limited in duration to two months... I think that this mechanism needs to be kept on a tight leash until more is known about it, even if that means continuing to rely on semi-protection on articles marked for indefinite protection. Wnt (talk) 18:29, 22 February 2011 (UTC)

• No protection of an article should be indefinite. The need will generally diminish rather rapidly. For the few that need longer, the protection should automatically expire in no more than a year, so it can be reviewed and reapplied in still actually needed. DGG ( talk ) 05:15, 26 February 2011 (UTC)

What kinds of improvements or new features would we like to have added to pending changes?

What can PC do for us that it doesn't already do? Are there ways to make the interface more user-friendly and to avoid long load times on PC protected pages? Beeblebrox (talk) 23:47, 20 February 2011 (UTC)

I thought these had already been solved? We now have a reject button and it's a lot faster than when it started. HJ Mitchell | Penny for your thoughts? 14:32, 21 February 2011 (UTC)

It's a question, is there anything it doesn't do right now that you would like for it to do? The devs have stopped work on any improvements pending a consensus on its continued use, but are willing to make improvements should it be kept. Beeblebrox (talk) 18:58, 21 February 2011 (UTC)

Could someone in the know post a cute little table of the problems people reported and which ones have been fixed? Wnt (talk) 05:49, 22 February 2011 (UTC)

• I would like a Twinkle-esque system for adding pending changes to articles without going through the protection tab. NW (Talk) 20:37, 25 February 2011 (UTC)
• We need some way of actually explaining it to new users--it has proven hard enough to explain to the most experienced. DGG ( talk ) 05:16, 26 February 2011 (UTC)

Should we define our requirements first?

The biggest problem with PC is that the community didn't have any control over what was provided. The proper process is to lay out our requirements, get them implemented, test that implementation, and get fixes made prior to expansion. The comment above that the developers are working on something new and secret shows that the same thing is happening again right now. It's time for us to say enough. Let's end this fight over the current system and instead guide development of something useful. I've proposed many ideas in the past, as have others, but discussion went nowhere because we were forced into the PC tunnel. I truly think that working together we can define something that will get consensus. The first step is to ask if we have a problem worth solving. We presently invest a lot of effort in developing and operating automatic, semi-automatic and manual tools to reduce the frequency and lifetime of vandalism. Do users feel that we are wining or loosing the war? If we are losing, clearly we need to talk about what to do differently. If we are winning, we still need to ask if we are willing to continue the current level of effort or do we want more tools? Also, to what extent do the current vandal fighting efforts cause collateral damage, including participation in the project? —UncleDouggie (talk) 11:51, 21 February 2011 (UTC)

I am inclinded to agree with UncleDouggie. I reckon we could come up with an idea that is better than PC. Possibly one that uses ideas from existing vandal fighting tools such as STiki. Yaris678 (talk) 13:14, 21 February 2011 (UTC)

I seem to be having trouble getting this point across no matter ho many times I repeat it 'There is no update in the works because it would be a waste of resources to work on something that may be shut off at any time. Iff we decide to keep PC switched on there can be future updates, but the Foundation isn't going to pay anyone to develop software that may never be used. Beeblebrox (talk) 19:01, 21 February 2011 (UTC)

We have been told on this very page that there is no work being done on PC and there won't be no matter the outcome of this discussion because the Foundation is already at work on their next grand scheme with no input from us. Furthermore, I must assume that the recommendation against expansion means that the new magic bullet is planned to put PC out of its misery. We either get together with a coordinated voice and demand that the Foundation involve us or we file this and all future PC discussions in the DOA file right next to every attempt that's ever been made for improving RfA. —UncleDouggie (talk) 05:01, 22 February 2011 (UTC)

Does the token few hundred pages on which PC is currently active really persuade someone over at MediaWiki to work on this harder? Wnt (talk) 05:51, 22 February 2011 (UTC)

I think the idea is not that people should work harder on PC. The idea is to come up with something that people actually want, and hence something that someone might want to work hard on. Yaris678 (talk) 08:18, 22 February 2011 (UTC)

The question here isn't PC or no PC, it's really what, if anything, do we want to do with our core operating system (MediaWiki) to improve the process around key areas like vandalism control/prevention, BLPs, page protection, barriers to entry for new users, accuracy, etc...? PC is just a tool built into the software. If we don't like the tool, we get to change it, because software is awesome like that, but I'm not hearing even bluesky ideas here as to what people want besides "everything is fine as it is now," "turn off PC and never speak of it again," and vague requests to "make PC work better." I suppose this is really a way to structure this discussion going forward:

1. Do we want any sort of MediaWiki feature to help with the tasks PC has set out to aid? If not, we turn PC off and drop the whole thing.
2. What is our absolute ideal pie in the sky version of such a feature?
3. Can we continue to use PC as a reasonable starting place to move toward #2 as pieces of that feature are slowly developed, understanding that perfection is the enemy of the good?

Zachlipton (talk) 09:13, 22 February 2011 (UTC)

Sounds reasonable to me. Although I might do something like:

1. Do we want any sort of MediaWiki feature to help with the tasks PC has set out to aid? If not, we turn PC off and drop the whole thing.
2. What sort of things would we like the software be able to help with? I would try to keep this at a fairly high level like "we want to discourage vandalism" and "we want to encourage good-faith editing, including by IPs"
3. What sort of solutions can we envisage? This would include:
   1. Adaptation of PC
   2. Other changes to MediaWiki (e.g. an improved diff functionality)
   3. Tools that interface with MediaWiki (e.g. an adaptation of Twinkle)
   4. Some combination of the above
   5. Blue-sky ideas
   6. Pie-in-the sky ideas
4. How do we make sure any of it happens? Ask the foundation nicely? Do it ourselves?

I would also ask if we want to use PC in the meantime. We would probably be able to get a sensible answer to that question after question 2, but I can see reasons for asking it earlier and later.

Yaris678 (talk) 12:11, 22 February 2011 (UTC)

Just get rid of it

or put it on every page. Whichever you want. Mono (talk) 01:58, 22 February 2011 (UTC)

Keep it simple

This just makes Wikipedia more complicated. Ditch it. Apuldram (talk) 09:52, 22 February 2011 (UTC)

I second this sentiment. Pending changes complicates the matter without reducing the workload. On articles with pending changes, vandals can still vandalize, and we still have to revert it, and at the same time, we have to manually approve legitimate edits by the designated underclass. Thus the workload is not lessened (compared to semi-protection, which locks out the designated underclass), and the matter is made more complicated. Thus, pending changes must die. SchuminWeb (Talk) 04:17, 23 February 2011 (UTC)

Agreed, too. wvbailey's rebuttal of TeleporterMan below is a laundry list of reasons why I've staunchly opposed PC from the start, not the least of which is the "fuck IPs" mentality. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 19:10, 23 February 2011 (UTC)

Proposal for this RFC

I know it gets into the meta-meta discussion but I see giant walls of text with many people making the same point over and over. I'd like to propose that 2 "floor managers" be nominated to present the pro/con arguments (preferably on linked subpages) that are organized, and at a high glance explain each side's arguments, refutations, counter-arguments. I'm all for a healthy debate, but the wide ranging discussion about all parts of PC that if we sat down and split out the issues we'd have a much cleaner concensus about what to do with this. Hasteur (talk) 16:32, 22 February 2011 (UTC)

I've been trying to contain things and keep it on track, but it's not working out so well. More like herding cats. And there doesn't seem to be any amount of pleading that will convince some to stop complaining about the extension of the trial and focus on what we do next. I'll be mostly out of touch the next few days but I agree something needs to be done to further structure this thing, my attempted refereeing doesn't seem to be working. Beeblebrox (talk) 16:43, 22 February 2011 (UTC)

I wonder if the real problem with this might not be that the discussion is too high–level. That is, that it presumes that the part of the community who are interested enough in this to bother giving an opinion about it have uniform opinions about, or have at least accepted consensus about, underlying issues such as:

• whether IP or newcomer editors ought to be allowed to edit at all,
• whether vandalism is or is not a real problem, or at least enough of a real problem that it needs additional measures to address it,
• whether or not the decline in the number of active editors is a real problem and, if so, whether or not it is capable of being fixed,
• whether or not WP is overly bureaucratic,
• whether or not WP is too hard to edit,
• whether or not good faith contributions by newcomers shouldn't be so quickly subject to deletion,

and I could probably add half a dozen more issues if I thought about it long enough. Everyone who enters this discussion comes in with an opinion on those issues and nuances his or her position here to fit it. While WP may, in theory, have current consensus on all those issues, as reflected in policy or guidelines, for each one there are many who feel that one or more of those consensuses ought to be changed and see this proposal as a way to implement, influence, or express their opinion without, in many if not most cases, clearly stating the theoretical or philosophical presumptions on which they base their position in this debate. I'm afraid that this may be one of those (rare?) instances in which the consensus model for WP simply breaks down and Jimmy Wales or the WMF has to either step in and break the impasse or let the current status quo become permanent by default. There are simply too few who are willing to definitively say keep or delete and far too many who want to instead say "keep if" or "delete unless," a situation not helped by the tech staff (per Steven Walling at work's 02:02, 20 February 2011 (UTC), statement above) taking the position that they're only willing to make changes if there's a definite consensus to keep it but also aren't willing to delete it without a definite consensus to do so. I would suggest (and !vote) that:

• removing PC from all the articles to which it has been applied and
• prohibiting its use

until and if consensus has been reached about it going forward, and if so where and how, would satisfy the tech staff's desire not to spend time on it any more, while at the same time preserving the programming, and allow the discussion to go forward without having to decide whether or not to "extend" the trial any further. If I were to guess, however, I strongly suspect that there's an extremely high chance that the current status quo is going to become permanent by default. Best regards, TRANSPORTERMAN (TALK) 22:27, 22 February 2011 (UTC)

re TransporterMan's list of questions: wvbailey's opinion

• whether IP or newcomer editors ought to be allowed to edit at all: NO. There should be a "published version" which is (somehow) determined by "the community" to be solid enough to publish, and behind it a shadow version that is editable by all. How the shadow copy would turn into the published version would be an interesting process, indeed;
• whether vandalism is or is not a real problem, or at least enough of a real problem that it needs additional measures to address it: YES, and YES, and YES, many of my recent "edits" are merely reversing vandalism of the most trivial, bathroom/potty-language/kissy-kissy type. But there is another kind of vandalism that is more serious -- the religiously- or culturally- or politically- or commercially- or self-advancement motivated sort of vandalism. In fact, unless something changes, as I get older (I'm a geezer now and it's all downhill from here, I give myself about 10 years more in the saddle) I can see myself just spending my time as "he who reverts vandalism", a kind of Yoda job, just a guardian at the gate kind of guy. Warning: I'll be harsh. Be very afraid.
• whether or not the decline in the number of active editors is a real problem: NO. How many articles do we need? Carry it to the limit? Billions? Trillions? Eventually mankind's knowledge -- at least the part that is worth keeping (no, it's not a picture of your kitty-cat, nor a picture of yer mom (dad, sis, bro, uncle, flying monkey) with a beer in hand at the barbeque last spring, nor your ancestors' wrinkled, faded sepia photos etc) is codified.

• and, if so, whether or not it is capable of being fixed: It will fix itself, one way or another . . .;

• whether or not WP is overly bureaucratic: NO,
• whether or not WP is too hard to edit: I don't know what this means. After a certain time and/or a certain number of edits a person becomes slightly more capable, and certifiably crazy, than a newcomer . . .,
• whether or not good faith contributions by newcomers shouldn't be so quickly subject to deletion: Back 5 years ago, any newbie contribution (I too was a newbie back then) might have had a certain contstructive weight. But now, in established articles with 100's or 1000's of edits, the role of the newbie when editing established articles becomes less certain. New articles: go for it. Hence my proposal for a published version versus a shadow-copy.

wvbailey's opinion: WP is a good thing, good but just a start toward a more mature wikipedia. Bill Wvbailey (talk) 23:06, 22 February 2011 (UTC)

• Mind if I rebut your points?
• There should be a published version which is... determined by the community... and behind it a shadow version which is editable by all. The Foundation's three steps ahead of you w/o the shadow version and not likely to actually do that wiki-side.
• But there is another kind of vandalism that is more serious -- the religiously- or culturally- or politically- or commercially- or self-advancement motivated sort of vandalism. PC won't help against this if some of the party *is* religiously- or culturally- or politically-motivated. Take note of those three areas I linked - They are areas where PC would fail, even with the Arbitration Committee sanctions in place. I do agree that vandalism is serious, but if we ramp up our response, the vandals will match within a month.
• [The decline in active editors is not a real problem]. How many articles do we need? You're conflating active editors with article writers. Most of us don't write anew; we mainly edit what's already there. A decline in active editors is an issue. And how will it fix itself?
• [Wikipedia is not overly bureaucratic]. This is disingenuous to say the least, as you have to perform the Vogon Dance (send edits in triplicate, etc. etc.) to edit unless you're tenured. We need less bureaucracy, and Pending Chains/Fragged Revisions adds more bureaucracy.
• Back 5 years ago, any newbie contribution (I too was a newbie back then) might have had a certain contstructive weight.[sic] Baloney. Out of all the IPs I've seen, I've seen quite a few dedicated good-faith editors who have refused to register. Are you suggesting that they get thrown under the bus? —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 19:27, 23 February 2011 (UTC)

PC was a trial, so where are the results?

It is completely pointless discussing where we should go next without some data. I haven't read all of this massive page but it seems to be repeated over and over that there is no analysis being produced. The trial was pointless if it cannot produce any data. What I am looking for is some metrics on performance, especially comparing articles protected by PC against articles protected by some other means or unprotected. SpinningSpark 00:37, 23 February 2011 (UTC)

Results is a bit like fishing for ghosts. You might be wanting to know - how many defamatory comments especially in relation to living people were stopped from being published via wikipedia and mirrored around the www on articles that would not have been protected at all, and considering the more open edit situation that pending offered in comparison to semi protection, how many unconfirmed IP users were able to make an accepted addition when they may have not bothered when the had to request it on the talkpage, questions regarding such figures are unanswerable. We can say, the more low watched BLP articles we add pending protection to the more defamation we will stop publishing all over the web for sometimes lengthy periods of time, and we know that unconfirmed IP users have benefited from the ability to post direct to the article when semi protection would have stopped them. Off2riorob (talk) 00:49, 23 February 2011 (UTC)

It is only like fishing for ghosts because we never got a proper trial. Without a proper trial all we have are people like you who in good faith can make claims about how the unknowables point in your favor and people like me who can (again, in good faith) claim the opposite. If we had an actual trial of the tool we could make real empirical claims. Until that point we ae really just arguing from priors, which will never result in one side convincing the other of anything. Protonk (talk) 01:19, 23 February 2011 (UTC)

I don't think you could ever get the statistics you seem to want, what specific figures would you have liked to have recorded? What opposites are being claimed - the only negatives being presented is that some users didn't like using it and that it loads slowly, and comments like, the tools we have a fine, and I didn't see no benefit etc, imo such negatives and personal opinions are far outweighed by the benefits above. The benefits as I have commented above are indisputable and perhaps its time we rolled the tool out to trial one of the things it was created to do, to protect little watched BLP articles from unnoticed defamatory additions, lets trial on 100,000 such articles. Off2riorob (talk) 01:22, 23 February 2011 (UTC)

There are a few points to be made here. First, if the downsides of PC represent intangibles or unknowables then we should be more cautious about implementing it rather than less. We should be respectful of the unknown, not derisive. Second, statistical inference is not alchemy. It is entirely possible to create a sample of articles subject to long term semiprotection which are similar along a number of dimensions (e.g. page views, reversions, edits per editor, edits as a whole, etc.) and split that sample into a control and treatment. The treatment group is assigned PC for the duration of the trial. The control remain semi-protected. You can then compare outcomes for the two batches of articles. Obviously there will be a Hawthorne effect, but this can be minimized by not publicizing the articles in the treatment (even though people will find out) and never revealing the articles in the control. compare this to the mess we had with the actual PC "trial" where the tool was turned on and demoed but no inference about its efficacy or impact is possible. Protonk (talk) 03:07, 23 February 2011 (UTC)

I should also point out that the total sample doesn't have to be very large. Something like 200-300 articles total would be more than sufficient, even if the effects we are measuring are small. Protonk (talk) 03:22, 23 February 2011 (UTC)

Off2rioro complains that we are not specifying what kind of statistics we are looking for. Let me be clear, my question here is are there any statistics whatsoever giving any kind of metric. This discussion is woefully short of facts. Apparently, the only fact that can be stated with any certainty is that there is no evidence that PC has any benefit. SpinningSpark 08:08, 23 February 2011 (UTC)

I can give you some statistics. I've used PC on some 40 pages. Each page (100%) contained (search-indexed) vandalism before it was PCprotected. None of the pages (0%) showed any vandalism while PC was in effect - in Google or to casual browsers. 100% of editors were able to edit each article while the protection was in place. Those benefits seems fairly convincing to me. -- zzuuzz ^(talk) 08:45, 23 February 2011 (UTC)

This is a fallacy. Spammers take the easiest path. In this case, they just go spam something else instead. Do you want to take on a trial of putting PC on all of Wikipedia? We could never keep up with the reviewing load and the distraction would be equivalent to having everyone fight at the front door while the vandals go blow the entire back wall off the place. How many subsequent good IP edits were lost because editors didn't see that their efforts had any effect? These problems make a statistical analysis of PC impossible. Please don't waste your or our time. —UncleDouggie (talk) 09:12, 23 February 2011 (UTC)

So we don't semi-protect anything because vandals will pick an unprotected article instead? Nonsense! Targeted protection has its benefits. -- zzuuzz ^(talk) 09:25, 23 February 2011 (UTC)

I disagree. Flexible vandals will go somewhere else whether a page has PC or semi-protection. The only answer is to block them. Vandals determined to attack a specific page mandate that some level of protection be applied. The trial showed that many of the high vandalism PC pages had to be reverted to semi-protection during the trial because the reviewing workload was out of control. That leaves infrequently accessed BLPs, for which semi-protect with {{edit-protected}} works just fine. Most of those edits will need work before being accepted anyway, so the convenience of the "accept" button in PC doesn't do much for us. The trial actually showed it did some harm because all it takes is for one inexperienced reviewer to accept an edit not realizing the BLP implications. Editors tend to be more careful carrying out {{edit-protected}} requests. —UncleDouggie (talk) 13:28, 23 February 2011 (UTC)

Data is not the plural of anecdote. Protonk (talk) 18:04, 23 February 2011 (UTC)

Second that. SpinningSpark 20:48, 23 February 2011 (UTC)

• - As I said, and as Zzuuzz appears to have understood very well, the benefits to the project are loud and proud. The issue is do you care? Are you bothered that defamatory content sits in our articles about semi notable people unnoticed for months? Do you think publicity about such things are bad or good for the project? If you gave the living subjects a vote for this protection to be kept or switched off what do you think they would vote? Do you feel any responsibility to the subjects of our articles? How many of our living subjects do you think would like to opt out of the project if we refuse to protect their article from defamatory additions? Ask yourselves questions like this. Its a difficult position to be in imo when we have an almost totally unidentified not notable user base rejecting additional protection for the identified notable subjects of our articles. It is absolutely undeniable that pending protection keeps defamatory content from being inserted and remaining in the BLP articles of semi notable low watched articles and through insertion and publication in our article also mirrored all over the world wide web. If you reject this then what do you support to stop such additions? Liberal indefinite semi protection? Allowing living subjects to opt out? Account only editing? Have a look at this, the latest of a constant stream from the BLPN, defamed as a criminal on the 24 November, removed three months later.Off2riorob (talk) 11:41, 23 February 2011 (UTC)

  So let's define our requirements. PC is a rather inefficient, error-prone tool. If you're so worried about BLPs, and with good reason, let's find a real way to protect them. How about starting with the low hanging fruit like having the edit filter reject any BLP edit without an edit summary? The next step is a big red warning/captcha/flashing neon thing if an unconfirmed user changes anything more than punctuation in a BLP without including a reference. —UncleDouggie (talk) 13:28, 23 February 2011 (UTC)

  I'm sure a very astute programmer could come up with a 'bot that crunched through all the 100's of millions of edits pulling apart those instances when the edit was accepted, and those when it was rejected. I can do this manually for Al-Khwarizmi (see its revision history, it's almost unbelievable, and a classic example of why PC or an equivalent is necessary). But I don't have the programming skills to do this on a wide basis. Maybe a call for volunteers is required. Bill Wvbailey (talk) 14:53, 23 February 2011 (UTC)

Not even that, Wvbailey; if I'm reading the database correctly, there are 'only' about 162,000 PC revisions. There's less than a thousand pages with PC. So yes, I agree - some analysis would be very helpful. I thought that maybe WMF could help out with that, but possibly we need to do it ourselves.  Chzz  ►  15:01, 23 February 2011 (UTC)

• I don't agree with Douggies comment that "PC is a rather inefficient, error-prone tool". - IMO Pending protection works already and keeps defamatory content out of articles and is a more open editing environment to unconfirmed users than semi protection. Clearly there are issues with the interface that can and imo if we commit to it, will be improved. Off2riorob (talk) 15:08, 23 February 2011 (UTC)

Any such dry accepted/rejected ratio statistics is useless and dangerous. Loads of edits are being rejected without necessarily being vandalism, and our goal is to understand how good this tool is at protecting the articles against vandalism, as opposed to e.g. good faith edits and content disputes with article "regulars", who regularly incorrectly reject these as "vandalism". 219.111.127.34 (talk) 15:43, 23 February 2011 (UTC)

This is simply not true, no one is calling vandalism when its not, that is an editing offence and taken quite seriously. Content is assessed if pending is on the article or not and rubbish uncited detrimental additions although not vandalism are rejected as a matter of course. The occasions when a good edit is rejected through user error is minimal and occurs at exactly the same rate irrelevent of the type of protection on an article and also if there is no protection on an article. Off2riorob (talk) 18:04, 23 February 2011 (UTC)

I am not disputing that there is margin for human error, that is fine. But it is a proven fact that some regulars (some of which admins) have strong ownership of "their" articles, and it's not unusual for content disputes to be resolved with semi-protections to leave unregistered users out of the story, branding them as vandals or socks. I have been quite active in this area and I can cover you with evidence supporting my claims. You can start with this list. I can also think of at least one case where the admin semi-protected (and indefinitely, at that) an article with zero history of vandalism pre-emptively, which is possibly even worse because you are effectively calling nothing vandalism.

Don't get me wrong, I do agree that PC has its place, I just think that there should be more control and accountability around the use of both tools (PC and semi) and that looking at the raw numbers of rejections and acceptances can make for a biased and misleading statistics. 113.197.194.241 (talk) 22:39, 23 February 2011 (UTC)

• More information would perhaps be helpful. There are certainly limits to what we can learn though. 1.)Page-load and performance statistics are liable to change if we commit to PC and development subsequently continues. 2.)As a tool, PC is only as good as the editor using it (like most tools), and so figures regarding failure by reviewers to stop bad edits are liable to change as familiarity with the tool increases and as the interface is improved. 3.)Statistics regarding attempted vandalism of PC protected pages may change as vandals become more familiar with PC and change their tactics in response.

One of the few statistics unlikely to change over time is the effect PC has on the number of contributions made by IP users (who are presumably new, and therefore unfamiliar with PC, and WP policies in general) on PC protected pages. I would be especially interested to see this.

Regarding workload: Am I misinformed about the use of PC on German Wikipedia? I was under the impression that PC (or something like it) was applied to all articles there. How do they cope with the workload? Please clarify the differences between the German implementation and the proposed English version. Revcasy (talk) 15:49, 23 February 2011 (UTC)

• One approach to this problem of "scale" is what public auditors do all the time. They take random samples and then audit the samples manually. A statistician could tell us how many articles to audit, and then the volunteers would go through them revision by revision trying to tease out what's going on. Problem here is: such an analysis would involve judgment and experience and good intentions: the auditors need to know the content well enough to know when an edit is (i) frivolous (e.g. see Division by zero. How many stupid kid-jokes about Chuck Norris dividing by zero do we need? And Al Gore Rhythm as a pun for Algorithm? Thankfully we don't see that too much any more), (ii) in good faith and an improvement, but rejected by a bully, (iii) in good faith and not an improvement and rejected for reasonable cause, (iv) vandalism pure and simple (erasing articles or sections, inserting porn-links or flammatory statements, inserting names of girlfriends). I don't know statistics well enough to design a sampling plan (we could find someone to help, though).

But there's another more philosophic way (using inductive reasoning): in general because there are so many edits even without sampling we can be sure that in the aggregate population there will be some false negatives, and some false positives. But, unless someone can demonstrate to me by argument (not statistics) that the percentages of false negatives and false positives are onerous, then I'm going to accept the success of PC. For instance: (here's the induction) why would you believe that experienced, competent editors (those with a track record hence allowed to approve edits) would, when given "the power", suddenly, and thereafter routinely, disobey their hither-to honest natures/consciences and the principles of wikipedia which they've hitherto followed? Previously they've shown themselves to be competent and suddenly they go wonky and turn evil? Inductive reasoning assumes that what has been happening usually kind of continues on (as if it has inertia). My experience is that we should assume good faith of PC-enabled editors until their actions show otherwise. And we can assume that the actions of good editors will continue to be good editing. Inductive reasoning demands counter-examples to call into question the premises/assumptions. Does anyone have any counter-examples? Bill Wvbailey (talk)

I don't think those words mean what you think they mean. You haven't offered a chain of inductive reasoning in support of a proposition. You have simply stated your priors (PC is a good thing) and declared that unless someone makes a counter argument you are going to continue to believe those priors. That's fine, but it isn't an argument from first principles at all. Protonk (talk) 18:22, 23 February 2011 (UTC)

• - If you apply pending to the 100,000 least watched BLP articles the workload will not be greatly increased because those articles receive few edits, but all those unwatched or little watched articles about living people will never get deaths added and never get defamed by an unconfirmed account, correct me if I am wrong but it is unconfirmed users that add the vast majority of the defamation. Its a simple win win situation, you don't need statistics to work it out. Either that or just semi protect them and be done with it and don't allow the unconfirmed accounts to edit on those articles without an edit request template on the talkpage. Off2riorob (talk) 18:32, 23 February 2011 (UTC)

This is the sort of mission creep that gets people worried. It's one thing to use PC as a substitute for semi-protection, but another to go festooning it over large portions of the encyclopedia. The whole magic of the idea is still that any kid can come on and edit an article and see his changes come up right away. Wnt (talk) 18:56, 23 February 2011 (UTC)

That magic is kinda spoiled when you are a minor notable person that is demeaned across the www for months by content published through wikipeda as someone that fucks his mother. Not very magic is it? Are you happy with that magic? I tell you, you are not notable and if you were notable to be soiled like that you wouldn't think it was very magic at all. Off2riorob (talk) 21:32, 23 February 2011 (UTC)

That blade cuts both ways - say you are a minor notable person who wants to hide the fact he diddles little kids... —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 21:41, 23 February 2011 (UTC)

Yea, right, we have a lot of that, please provide diffs. Firstly you need to accept that these less notable BLP articles are not watched by anyone sometimes and accept that defamatory content is inserted to some of these articles by unconfirmed accounts. Then you need to ask yourself - do I give a fucking damn about that, many people don't, they are just users here and have no responsibility to anything, we have some users that in all good faith are against the project, unidentified users and many are children and they simply don't give a fuck about anything, never mind some subject being labeled a pedophile for months, so what. It will be removed sooner or later, so what why should I care about jonny smith being a minor celebrity and wikipedia publishing he fucks his mother, I don't like pending and there is no consensus to trial anymore...yada yada yada. Off2riorob (talk) 21:47, 23 February 2011 (UTC)

I'm glad we are having a rational discussion. Protonk (talk) 21:48, 23 February 2011 (UTC)

Well, I don't like it, we have enough vandal fighting tools is just not cutting the ice. I think I have expressed my position pretty clearly, so, removed from my watchlist, let me know what you decide. Off2riorob (talk) 21:50, 23 February 2011 (UTC)

How often does this actually happen? How much does it matter when it does? How often is an IP responsible? How many IP vandals wouldn't set up accounts to do it anyway? Seriously, if you want to protect frequently vandalized articles, surgical use of PC as an alternative to semi protection might make sense, but when you're talking about finding miscellaneous slanders out of the whole database, maybe searches for "pedophile" and various vulgar terms might work better. Speaking of which, we could use some kind of special terms for the Search box that limit a search to BLPs, rarely edited articles, newly created articles, uncategorized articles, etc. so that we can search "Special:BLPs Special:unedited-7-days fucked" and get back an answer. Wnt (talk) 05:07, 24 February 2011 (UTC)

Caution needs to be exercised providing tools that find weak articles. For instance identifying unwatched articles has always been considered a bad idea as it is more useful to trolls than those trying to protect articles. SpinningSpark 08:02, 24 February 2011 (UTC)

I think Wikipedia is too deep into the Knowledge Is Good business to be having second thoughts now. It's not just that the whole database is public and someone else could offer the tool, but more that we otherwise assume that good editors can outwit vandals, with positive results. In any case you should at least have a tool that looks for unwatched BLPs with juicy terms in them and sends those out to a group of editors to go over. Wnt (talk) 16:32, 24 February 2011 (UTC)

We have tools that do this already. WP:STiki flags high risk edits for review using it's own algorithms. It also now has a mode that routes the "overflow" from User:ClueBot NG for review that consists of edits below the automatic revert threshold and reverts that would otherwise violate the 1RR rule for bots. —UncleDouggie (talk) 21:30, 24 February 2011 (UTC)

Without getting into BEANS territory, the unwatched pages special page is so terrible that it doesn't have any to admins. It was the first thing I looked at when I got the bit and I haven't been there since. But wikipedia would see a serious return to supporting or building tools for editors to find and correct articles with problems. The current system of tags and fairy dust doesn't work for very obvious reasons. And there are plenty of degrees of usefulness between tags which nobody cares about and a tool which could be used for largely malicious purposes. Protonk (talk) 16:52, 24 February 2011 (UTC)

I haven't seen it, but I seem to remember that for the longest time "Watch this page" was a default for every edit. Later I shut it off and blanked it all out (generally I go back via "my contributions"), but looking just now I still see my watchlist ended up with five entries on it somehow. My guess is that articles are "watched" almost entirely by random editors with unwatched watchlists. This is why I suggested above going by when the articles are last edited. Wnt (talk) 17:32, 24 February 2011 (UTC)

Alternatives to PC

Are any alternatives to PC being considered? As Wnt stated above, the major downside of PC is that it prevents a good faith IP editor from immediately seeing the result. Wiki is meant to be a form that does just that - immediate. Any protection scheme that builds walls around articles is going to be bad from that point of view. Far better that the guilty are directly attacked rather than the innocent editors harmed through colleteral damage. It seems to me that the biggest problem is shared IP addresses and IP hopping connections. Problem editor on a fixed IP or with an account are rapidly stamped on. Technical means of distinguishing different editors on the same IP or the same editor on different IPs would largely obviate the need to protect articles at all. I am thinking here, for instance, of AOLs X-Forwarded-For. I really don't have much clue over what else may be available but ir may be worth opening this discussion up a bit beyond the narrow polarised for/against PC this has became. SpinningSpark 22:57, 23 February 2011 (UTC)

Makes sense to me. See also #Should we define our requirements first? Yaris678 (talk) 12:40, 24 February 2011 (UTC)

PC is less restrictive than semi-protection, which is the main alternative. It enables unregistered and new users to make edits directly to the article, whereas they would otherwise be locked out completely. We have already WP:XFF, active checkusers, some very skilled and dedicated admins, swift indefblocks, IRC feeds, various automated screening tools, and massive rangeblocks in place. None of these can prevent vandalism, sockpuppets, dynamic IPs, or open proxies slipping through. Blocks and patrollers are not all they're cracked up to be. Protection is an essential tool, and PC is just another form of protection. -- zzuuzz ^(talk) 13:52, 24 February 2011 (UTC)

How many more of these [3] have to occur before we can protect all BLPs with some form of editorial oversight? Notice what the user said in his report: "[This is] inaccurate and potentially defamatory", "Presumably, the contributors to these programmes have used Wikipedia as a tool for research. This is unacceptable as it may affect David Cox's chances of enhancing his career due to malicious rumours that are wholly untrue. [emphasis mine]", and "I look forward to hearing from you at your earliest convenience." That last sentence, taken together with the first, indicates that he may have been considering legal action against the wikimedia foundation. Reaper Eternal (talk) 21:32, 24 February 2011 (UTC)

Legal questions can get kicked back to the foundation. We shouldn't be engaged in amateur legal theorizing in order to determine a course of action, because we are liable to give undue weight to certain options which without any real cause. For instance, how would you suggest we deal with all potential future BLP violations? We could fully protect all BLPs, but this would enshrine errors or problematic content as it exists now? We could delete all BLPs and refuse to host future BLPs, but this would result in removing articles on presidents and prime ministers as well as articles on music producers etc. I'm not trying to construct a straw man but when we start talking about legal threats and responses to them it is nearly impossible to avoid running into absolutes. Instead we should really only be talking about what makes sense as editorial policy in the face of the problems with BLPs and our mission to remain an open encyclopedia. Will PC fix problems like the one you mentioned? All of them? Perhaps in 6 months time will people be clamoring for another stronger solution to the BLP problem when PC fails to magically fix them? Those are conversations we need to have and we can't have them under the threat of legal action, especially as interpreted by amateurs. Protonk (talk) 21:59, 24 February 2011 (UTC)

We are not looking for absolute solutions, just improvements, step by step. Off2riorob (talk) 23:05, 24 February 2011 (UTC)

What ended up happening with the Pending Changes Compromise idea? --Yair rand (talk) 20:51, 25 February 2011 (UTC)

I think I said this at the time, but I think that compromise suggestion is a good one. Reading towards the bottom of the page, it looks like the idea is to merge PC with WP:STiki. I have been having a go with STiki this evening and it's awesome. Yaris678 (talk) 00:31, 26 February 2011 (UTC)

I recently restarted my research on the compromise after a bit of a wikibreak. I have been working with STiki for a long time now, including reporting the majority of the bugs, all of which have been fixed by Andrew West in the latest release. Note that STiki currently supports three data sources for risky edits. You can select the active queue from the menu. If there a lot of users draining the Cluebot-NG queue, the others will sometimes offer better hunting. The Cluebot-NG queue is still good because among other things, it contains all blatant vandalism that the bot was prevented from reverting due to 1RR. This combined bot/human approach has tremendous benefits and it's probably just the start of what we can do. At the moment, I'm working on a big Twinkle update before it dies again, so further progress will be somewhat delayed. —UncleDouggie (talk) 04:55, 26 February 2011 (UTC)

more magic

I can show you many many of these - Ardashir Vakil BLP, low notability almost no watchers, this addition was added .. - "His daughter, Tara Vakil, has also achieved notoriety as a result of starring in several adult films." - diff - added november 15, 2010 - removed over three months later on feb 24 2011 by someone claiming to be the subject of the BLP with the edit summary - "I am Ardashir Vakil ...I am removing slanderous comments...and out of date info" - magic isn't it. Pending protection would have stopped this defamation attack content added by an unconfirmed user from being published by wikipedia and mirrored around the multiple sites and the whole of the world wide web for over three months. Thats a statistic for you. We should trial pending on a hundred thousand of the least watched BLP articles and most of you won't even notice it, there won't be much extra work as those articles don't get a lot of edits and all those articles will be protected and yet still editable, and...the pending will attract experienced editors and those articles will without doubt benefit from any attention they get..Add it to the hundred thousand low watched BLP articles and let editors add or remove the protection if and when required and see where the level naturally settles down. Off2riorob (talk) 22:37, 24 February 2011 (UTC)

Semi-prot would have stopped that too, as would flagged revisions. What we do not know is, how the application of PC to a hundred thousand articles will affect Wikipedia. And it isn't just BLPs, there is plenty of potential for BLP-related issues on many many more articles. I mean - if we reached a consensus to simply lock down Wikipedia to all IP edits, I'd be fine with that. What I am not fine with is, sneaking it in without due process.

Note, I do agree that a further trial might be the way forward (and we could discuss how many articles, and all the other details). But you also know my stance about removing it and then forming consensus - so I won't repeat that.  Chzz  ►  00:21, 25 February 2011 (UTC)

I'm worried about this system scaling anywhere near 100,000 articles, much less 1,000,000... where are we going to find the reviewer time to do all that extra work? I don't think we will find it, because it doesn't exist, not in the quantity that would be required. And what time this new enormous backlog did get would be drawn off from some other activity... Courcelles 00:40, 25 February 2011 (UTC)

The wheels won't drop off and the level will naturally stabilize where it is controllable, lets give it a try, I don't see any backload worries, there is not time limit for additions to the BLP articles of semi notable people bots will do most of the start up work and we can take it from there. Off2riorob (talk) 20:11, 25 February 2011 (UTC)

Firstly, I would remind Chzz that removing PC would necessitate Semi-prot on a few thousand articles, making them unaccessable to IP's until the entire community is able to agree on something with what would probably need to be 67% consensus. The last poll got around 65%, so it might not be that hard, unless the bar is raised (which would make consensus impossible via the law of very large populations.) Further, Courcelles, if you check the PC page, the Devs are working on a Huggle-like review system, and bots could easily patrol edits for blatant vandalism (in fact, CluebotNG has been hugely successful in reverting actual vandalism with a false positive rate of .1%.) My primary concern with PC is deployment on articles for which it is not suited, like high traffic BLP's and other high risk articles that simply have to be Semi-Protected. I'm all for another trial, but use it on low traffic articles, and set the consensus bar at a reasonable level. Ronk01 talk 04:11, 25 February 2011 (UTC)

• Not "thousands" - less than 1,000 articles are protected by PC. (951 articles, 7 test pages, 2 other ^1,2, 10 redirects, as of now)
• We don't necessarily need to semi-protect them all; in some cases, no protection may be fine.
• There are about 4,700 articles using semi-protection, which can be indirectly edited with an accepted system. SQL query for articles only, not redirects, semi'd
• The last poll did not get 65% support; it got just under 60%.289 votes for temporary continuation of PC, and 199 for temporary removal - noting the word "temporary"  Chzz  ►  19:35, 25 February 2011 (UTC)

1. So we open these articles to IP editing, knowing that they will be vandalized?
2. Yes, and that system is rarely used, slow to respond, and almost unknown.
3. I apologize, I miscalculated. There was and is no supermajority (the real world word for "consensus") favoring any action regarding PC. Thus the second poll was conducted, which ran under simple majority by fiat (something that we don't like, but is nonetheless policy) And indicated that 60% of editors support PC in is current form (if you count the editors who voted close because they wanted to fix a technical error, or only read WP:Consensus and WP:Poll, it goes up to over 2/3 support of PC in some form) I agree that there are a few things that need word, better metrics are needed, and policy needs to be rectified; but that all justifies (or necessitates) continuing use, not removal and exposure of hundreds of articles to libelous content. PC does work, it may not be efficient yet, but it does work. Ronk01 talk 19:56, 25 February 2011 (UTC)

You're concerned about some 900-odd articles being exposed - and not even that, 'coz they can be semi'd. We have, at the very least, over 100,000 with BLP content that currently have no protection at all. Chzz  ►  20:00, 25 February 2011 (UTC)

• So Chzz, you say.."We don't necessarily need to semi-protect them all; in some cases, no protection may be fine." - are you happy then with the example that I have shown you here? You know there are more, many more. Nothing you are offering would have helped that at all. In your world, semi notable people are defamed and that defamation is left sitting there for months until the subject or a friend sees it and calls him or her up and they have to come here and I will tell them - you are the unidentified internet user that stopped their article being protected because you felt slighted about consensus, .. the living people being defamed will spit in the face of your consensus worries. Off2riorob (talk) 20:01, 25 February 2011 (UTC)

No, Off2rioriob. I am extremely concerned about the (hundred/s of) thousands of vulnerable articles. It is not in "my world" that semi notable people are defamed (for months); it is in the current Wikipedia world. I am looking for a way for us to form consensus on how we can improve the situation. Not all of those 951 will need semi-protection. Even if most do - and if some would have been better-off with PC (debatable, but almost certainly true in some cases) then that is a small price to pay, to get over the current impasse and move along to better, consensus-based solutions to the critical problem of BLP vandalism. Chzz  ►  20:09, 25 February 2011 (UTC)

These 900 are just the tip of the trial iceberg, there is no impasse at all, just an ongoing desire to improve and protect the project. So if you are concerned then lets trial on 100, 000 BLP articles that have almost no watchers - do you have another idea to protect them from being defamed? Off2riorob (talk) 20:15, 25 February 2011 (UTC)

I'm not discounting that as an option, Off2riorob. I suspect I'd push for a more phased test before planting it on such a huge number of articles - and certainly I'd push for much clearer policies, some tech improvements, and working out measurable results. But I think - despite all appearances - we're singing from the same hymnbook. I also appreciate that the problem is urgent - but, it's been urgent for years. Let's keep talking. My current thoughts, from this RfC, are that we might be able to firstly come to a consensus agreement to remove it (end the first trial; draw a line under that), and then to start building consensus for what we'd need, in order to evaluate it more thoroughly (a new trial, of, IDK, top-of-my-head, 3333 articles per month for 3 months, running for 6 months in total, gathering certain pre-agreed stats, comparing with a broadly-equivalent set of another 10,000 articles that we leave? These are embryonic thoughts in my mind only, not to be taken too seriously; I'd want to see each specific debated and consensus formed. I'd hope it possible that even some of the most staunch nay-sayers might agree to something like that if they can help us set up the measurables. If nothing else, it'd be a chance for them to show that PC doesn't work. Am I dreaming, in thinking we could form consensus about something like that?  Chzz  ►  20:38, 25 February 2011 (UTC)

I can assure you that based on the last poll, consensus to remove PC will be impossible to gain, unless there has been a significant change in attitudes toward PC in the last couple of months. I do like the idea of a fresh trial though, as I've noted as an option below. Ronk01 talk 20:51, 25 February 2011 (UTC)

I've tried to avoid complaining about the past polls, but the issue that we now need to meet some special threshold to remove PC after it was snuck in under a 2 month trial really rankles. When the trial was proposed this exact issue was brought up by opponents and dismissed as fantasy by proponents. We could NEVER start a 2 month trial and then just continue it on the basis of status quo bias. But that is exactly what happened. Protonk (talk) 04:04, 26 February 2011 (UTC)

I agree. However, at this point I don't care anymore how we got here. The proponents have invested a lot of energy in wearing us down, so we might us well let them claim a victory. Why they put all their energy into this particular battle is beyond me. Some of them have even recently said that a new trial should be on different articles, which was one of my main points for shutting down the old trial. The end result is of course that no one will believe any time limits on a new trial, so I don't see how it's even possible to propose such a thing. They have forced us into a situation of debating full implementation, whatever that may be, without another time-limited trial. —UncleDouggie (talk) 04:31, 26 February 2011 (UTC)

Agreed only partially with UncleDouggie. The only reason we're in this situation is via fiat of Jimbo Wales. And as a counterpoint, how long until someone approves a stealthy BLP violation, Off2riorob? And I do mean "When", not "If". As above, so below. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 22:53, 26 February 2011 (UTC)

True, but how long until a stealthy BLP violation is added to a page without protection? Oh, wait, that's happening all the time now. Ronk01 talk 23:39, 26 February 2011 (UTC)

Ronk, don't act stupid. You and I and practically everyone else knows that PC is an alternative means of protection, so comparing it to unprotected articles is comparing apples to oranges. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 23:50, 26 February 2011 (UTC)

Actually, some of the more recent proposals have indicated widespread use on currently unprotected BLPs. Ronk01 talk 23:56, 26 February 2011 (UTC)

PC is still a variant of protection, Ronz. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 00:03, 27 February 2011 (UTC)

Yes, I never debated that. Thus I fail to see your point. Ronk01 talk 02:31, 27 February 2011 (UTC)

What exactly is the PC reviewer's responsibility to prevent such additions? If a vandal writes that the daughter is a porn star, followed by a reference to an offline book or magazine article, is the PC reviewer supposed to look up and find out if it's a real book? Read it and see if the article says what it says, even if it's offline or behind a paywall? Because a vandal can learn quickly enough to make up such detail, if it really is required to keep a change in. Meanwhile, if a reviewer isn't required to make such a detailed research, is he still liable for "publishing" the acceptance? Especially if the IP editor is from, say, an internet cafe in Indonesia, while the reviewer lives in the same litigious Western country as the subject of the article? Probably it will be safer for the reviewer to simply reject any unflattering sounding change, regardless of validity, without examination. And I think that's what will happen. Wnt (talk) 15:50, 27 February 2011 (UTC)

Im my personal experience, I've never hit a paywall when researching a claim, but I suppose it could happen. Those kinds of claims generally require more than once source anyway, so I would reject based on policy, not CYA. Ronk01 talk 18:11, 27 February 2011 (UTC)

As a reviewer you should operate basically as you would on any other article. In the hypothetical case mentioned, you have the added benefit that while you review the content is not on display to the www so I would google search the claim and if there was absolutely nothing in the returns then I would perhaps attempt to access the book online if I couln't and I had been unable to find an accessable online verification of contentious desired addition then I would remove it with an edit summary of, controversial claim with unaccessible supporting citation please move to discussion on the talkpage and post a comment on the users talkpage and the article talkpage asking for some discussion of the addition. Vandals can learn to be tricky but a bit of talkpage discussion and the realization that their desired vandal edits are no longer appearing in the search result or the article usually gets them moving on. I have found that there is a degree of feedback amongst editors and occasionally I give and get a note that a review I did was incorrect, but if you take the usual vandal fighting and editorial guidelines into consideration you will never accept anything that would be unacceptable content.Off2riorob (talk) 01:25, 28 February 2011 (UTC)

(reindent) "[...]a bit of talkpage discussion and the realization that their desired vandal edits are no longer appearing in the search result or the article usually gets them moving on." You're assuming attention-seekers, Off2riorob, not hit-n-run, LTA, or stealth vandals, which are just as common if not more so. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 03:52, 28 February 2011 (UTC)

Yes, there are different types of vandal type additions and levels of detrimental additions, I myself deal with all edits on all articles I watch in the same way whatever the level of protection. Hit and run - they just get reverted and I think they would be less attracted to bother on articles where their vandal addition never gets published but that is currently a hypothetical statistic. LTA - is a new one on me.. Stealth vandals - imo these guys are dealt with in the same way as an attention seeker but they never show up for the discussion. The issue for me is that we know that articles with few watchers are having vandal additions remain in them for lengthy periods so what are we going to do about it? Pending protects them from such additions. Off2riorob (talk) 09:58, 28 February 2011 (UTC)

Off2riorob, that's an interesting interpretation, but it is not in-keeping with the trial guideline, which says it is just to catch and filter out obvious vandalism and obviously inappropriate edits. Perhaps you could make these suggestions on WP:VPP.  Chzz  ►  04:13, 1 March 2011 (UTC)

How do we determine "effectiveness"?

I'm starting to not dislike PC so much anymore. What annoys me, though, are three things:

• No one knows for sure what's going on. This is (supposed to be) solved after this RfC.
• No one has apologized for making a two-month trial into an eight-month "trial". This is solved by someone apologizing. Except, I don't really know whose fault this is.
• There is no documented evidence on determining PC's effectiveness (what is the point of it—to reduce vandalism? What about preventing live edits without reliable sources; is that an existing side effect and something that is good or bad?) and technical issues. How do we prove it is working, not working, too buggy, etc.? How can PC be improved; what's currently wrong (specific issues) with it, etc.?

I'd like to propose the creation of some group of editors or even a WMF-hired analyst to exactly determine if PC is working, and see if a pre-compiled list of issues/bugs outweighs the benefits. In this case, I would strongly support a new trial that would lead to definitive evidence about PC's effectiveness.

/ƒETCHCOMMS/ 23:38, 24 February 2011 (UTC)

The current FUD can be resolved, if we can agree to remove it pending discussion. We got into this whole mess of crap because of polling and unclear remits.

I encourage everyone to read m:Polls are evil, which is so appropriate here.

I'd support discussion of a possible new trial, and I think that - given the chance - we could come to a consensus about how such a trial should work...I really do. If we had a discussion about "OK - how many articles do we need to try it on?" - some would say 100,000; others would say 300. I think, through discussion, we could come to some agreement. We could decide "OK, how do we pick the articles", "How long a trial do we need", "What information are we looking to get from this trial", and "what exactly will we do at the end of the time period".

I'm not interested in blame. We're here, so let's make the best of it.

The first rule in an edit war is to stop, and discuss, get consensus, and then boldly go ahead. That's exactly what we need to do with PC.

The PROT policy right now makes very little sense; it admits the trial is over, but then talks about cautious use during the "current Interim period". I've no idea what period that refers to.

For all of these reasons, I've been insisting since the start of this - that we remove it, now, from all articles. And then we try to learn from our mistakes, and move onwards. FWIW, I would also support some form of new trial, on condition that we had a very clear remit for it, arrived at through consensus. As the meta essay says, Yes, establishing consensus is a lot harder than taking a poll. So are most things worth doing.  Chzz  ►  00:36, 25 February 2011 (UTC)

<unorganized rant> Chzz, would you care to explain how you wish to conduct a discussion among 600+ editors (basing this on the last poll) There are many, many editors here with strong opinions on PC, and it seem that they are the only ones who really comment on PC (in fact, nearly all of the editors involved in the first poll voted in the second) RfC's tend to raw similar crowds. Another measure needs to be taken. Perhaps some form of Signpost promoted ongoing Strawpoll (no real meaning, just so we can get an idea of what the community at large thinks.) </unorganized rant> Ronk01 talk 04:20, 25 February 2011 (UTC)

I have great faith in consensus, Ronk01. We've dealt with other contentious issues, e.g. images of Muhammad or Gdańsk/Danzig. The latter did go through some voting but, with quite clear, specific points - and after considerable discussion.

The 600+ people don't have 600+ totally distinct viewpoints. There are certainly some quite different views, but I'm still confident that consensus can be accomplished. I think that the comments in this RfC already have brought out the issues, and lead to a better understanding - and I thank everyone who has contributed. We have no deadline; we'll get there.  Chzz  ►  19:10, 25 February 2011 (UTC)

Yes, and in the meantime, libelous comments are being added to BLPs and instantaneously mirrored across the web before anyone gets to them. Trying to establish consensus here is like herding cats. I would remind you that Wikipedia is first and foremost an encyclopedia, not a social experiment, not a club, and most certainly not an anarchy. Regarding viewpoints, yes, there are 600 different viewpoints, every edeitor has some image of what they want PC to be, or indeed not be. Reconciling those viewpoints is difficult, because they run from Jeremy's "never talk about PC again!!!" to the views of others who refuse to accept anything other than flagged revs on every page. The unfortunate fact is that the extremes of opinion are represented here, not the means. Ronk01 talk 19:52, 25 February 2011 (UTC)

With or without PC, defamation is posted on Wikipedia and mirrored all over the net - for example, every few minutes a new attack page is created. We zap 'em pretty quickly, but still, Google Cache has them for a while - and other sites do too.

Reconciling the views is indeed difficult, but not impossible.

I have not seen anyone flat-out refuse to accept anything other than flagged on all pages.

I do not understand the last bit - "extremes of opinion are represented here, not the means" - could you clarify, please?  Chzz  ►  20:17, 25 February 2011 (UTC)

Yes, but PC can stop some of it.

It depends on what level of consensus you want, if you want 2/3, that's entirely possible, but very difficult. IF you want RfA style consensus, that's never going to happen sanas-miracle.

Unless Jeremy has changed his mind, he was still flat refuse.

I suppose I should be more clear, the editors of this RfC have a vested interest, or some string motivation for expressing their opinions.

Ronk01 talk 20:23, 25 February 2011 (UTC)

• Chzz, do you understand that until a desired addition from a unconfirmed ip address is reviewed and accepted by a reviewer it is not visible? not published to the web? Off2riorob (talk) 20:24, 25 February 2011 (UTC)

I understand that, with current implementation, revisions that have not been accepted are not visible to logged-out users, only to logged-in, except when they click 'edit' (in which case they are editing the latest revision). That isn't quite the same as 'not published to the web'. Websites wishing to monitor the most controversial aspects of Wikipedia could easily see non-approved revisions if they wished, and–at the risk of sounding argumentative–one distinct possibility is, that it would make it easier for them to see which edits were more controversial, ie those flagged as needing review. Chzz  ►  20:53, 25 February 2011 (UTC)

:::User:Chzz - that is simple nonsense - please provide a singe citation to support this has ever happened - don't bother it hasn't. Are you as a living person prepared to take responsibility legally for your attempts to remove protection form articles that are being defamed, if so please provide your contact details. Off2riorob (talk) 21:25, 25 February 2011 (UTC) Off2riorob (talk) 21:23, 25 February 2011 (UTC)

May we please avoid Legal discussion. Let's leave that to General Counsel. Ronk01 talk 21:27, 25 February 2011 (UTC)

note - my comment has been pointed to as a legal threat, its not at all its a reminder of the facts. You are all legally responsible for your contributions, you are not faceless users, and no one is protecting you, on request the foundation will give your details to anyone that has a decent reason, if that threatening then then that is just the truth. Users seem to think they are faceless and protected by some imaginary legal team, well your not. This is not a threat its something you should bear in mind when you contribute. Off2riorob (talk) 16:40, 26 February 2011 (UTC)

I was not suggesting that your comment was a legal threat, merely that such legal concerns are not something we need to be concerned about. If the foundation thought that there was a legitimate legal concern, they would take action. With regards to the release of personaly identifiable information, the privacy policy clearly states that this cannot occur without a valid subpoena, which would not be granted unless the user was directly responsible for a clearly illegal action, never because they wanted to discuss a form pf protection for a little longer. Ronk01 talk 17:10, 26 February 2011 (UTC)

No not you Ronk it was user Kingpin on my talkpage - editors details will be released on request of a valid request that is the simple fact of play, ands all users should follow as responsible a position as possible to protect themselves against that. The bottom line is - We all have a duty of care to protect the living subjects of our articles to the best of our ability and with all means at our disposal. I doubt if the legal advisors to the foundation would come out and say it, but as I see it the foundation is further protected from defamation and libel suits by implementation of such an additional protection tool. The idea is not to legally threaten anyone but to point users towards the bigger picture than their individual positions and personal objections to completely minor issues, such as the historic main objection here that the tool should have been switched of five months ago and its a bit slow.User should ask themselves - ok, I don't support this tool to help protect living people from being defamed through wikipedia, then, what do I support to protect those living people.Off2riorob (talk) 17:14, 26 February 2011 (UTC)

• It was not a legal threat. It's fine. Just discussion.
• I'm well aware of data protection legality and surrouding legal concerns
• WMF already has my contact details
• WMF will not release them to "anyone that has a decent reason"
• I make no claim that websites have already used PC to monitor articles; I said it was a possibility, it could be done
• If "attempts to remove protection" might result in law-suits, then we're all in a heap of trouble. Fortunately, it won't, no more than if I give you a pen and paper, and thus 'allow' you to defame someone.  Chzz  ►  17:55, 26 February 2011 (UTC)

  (edit conflict) Off2riorob: You are saying that we will all be sued if we refuse to vote in favor of permitting the use of PC and you seem to double-down on this by saying that we would have already been sued had the trial actually been stopped when it ended. You may very well just be warning us about something that you're personally concerned about. In this case, I would prefer for you just to state that you're freaked out by the risk and plan to quit the project if PC isn't implemented. That would unquestionably be a non-threatening statement. Saying we will be sued, but you're not really threatening to do so, isn't a good way to go. I wouldn't count on wrapping yourself in WP:AGF with regards to lawsuits. —UncleDouggie (talk) 18:01, 26 February 2011 (UTC)

I am simply using my finger to point at the moon and asking you to look at the moon and not to focus on my finger. IMO there are legal issues and dangers to the foundation if additional protection of such BLP articles is rejected. The bigger picture, because we do as individuals have responsibility, if you don't support this added protection then don't you support more protection? Be a bit responsible towards the bigger picture. If you reject this simple addition that I can show you many many times that living people were protected from defaming content then I would urge you to support allowing the living people the decency of opting out of the project or much more liberal semi protection. If you don't support any more protection and you are happy with the defaming additions then that is your responsibility and state and accept that. The educational benefits of the 100, 000 low notable living people is quite close to zero, actually its just project bloat. Very poor articles , weakly cited and with so few watchers as to be open to unwatched defamation, if you don't support some kind of preemptive protection for these people then let them protect themselves and opt out. Off2riorob (talk) 18:25, 26 February 2011 (UTC)

Perhaps you should read the rest of the page before accusing me of not caring. I was only responding to the use of a lawsuit to shutdown this debate. Also, we're not responsible for second guessing the foundation's legal problems. I assure you that if their lawyers determine that PC is essential for them to avoid being sued, it will be implemented on every article tomorrow. —UncleDouggie (talk) 18:44, 26 February 2011 (UTC)

Can we return to my original question? It was "How do we determine "effectiveness"?" /ƒETCHCOMMS/ 21:46, 27 February 2011 (UTC)

If it has been designed to allow unconfirmed accounts more possibility to post than semi protection would only allow an edit request on the talkpage, and to keep untrue and malicious content from the article then it is effective. First you need to determine what you want to compare it to? Off2riorob (talk) 22:23, 27 February 2011 (UTC)

Hoi, Off2riorob, remember PC's only as good as CRASH's membership is. It's not infallible because humans aren't infallible. How long until a reviewer approves a stealth-vandalism edit? —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 03:54, 28 February 2011 (UTC)

I imagine that has likely happened already, as you say reviewers are not perfect by any means. I know at least one reviewer that doesn't review edits in anything close to the way I described in the section above and I have seen false additions and uncited false content accepted but I have never seen content accepted that could be described as harmful to the subject or demeaning. Off2riorob (talk) 09:46, 28 February 2011 (UTC)

But how long until vandalism is inserted into a previously unprotected page? Oh, wait, it's happening right now. Ronk01 talk 14:15, 28 February 2011 (UTC)

Does anyone here think that there is any possible validity to Off2rio's sort-of-struck-out legal claim that Pending Changes is addictive? That if it's applied to an article, you risk legal liability for yourself or Wikipedia if you remove it? Because if that were true, obviously we would need a hard, all-out, eternal ban on the application of Pending Changes to any new article, and we'd have to delete all the articles to which it had previously been "experimentally" applied and start over, or maybe salt the topics permanently. A permanent policy ban on experimentation in general might also be appropriate. I don't think it's true, but just for purposes of discussion, we have to consider what would be required to save Wikipedia. Wnt (talk) 17:58, 28 February 2011 (UTC)

A Proposal Compiled From Ideas in This RfC

If PC must be removed:

1. A new Straw Poll (no meaning, a real Straw Poll) with the question "Do you support PC on its technical merits? (no ideological arguments, or complaints about "broken promises")" This would have to be widely advertised to the Community, so as to acquire as many !votes as possible.
2. If PC is removed from all pages, Semi-prot them. This prevents the pages from being vandalized while we discuss.
3. Do not Turn Off PC totally. Keep it on test pages
4. Request further development of the tool pending a third trial.
5. Develop a metric to determine the absolute effectiveness of PC in usable terms.
6. Set a definitive bar for consensus for a new trial and, eventually retention at 2/3.

If we keep PC on all pages:

1. A new Straw Poll (no meaning, a real Straw Poll) with the question "Do you support PC on its technical merits? (no ideological arguments, or complaints about "broken promises")" This would have to be widely advertised to the Community, so as to acquire as many !votes as possible.
2. Request further development of the tool for immediate implementation upon completion
3. Develop a metric to determine the absolute effectiveness of PC in usable terms.
4. Set a definitive bar for consensus for retention at 2/3.

In either case, it would be appreciated if we could discuss the merits, not the politics of PC here, it seems to me that we might actually get something done if we do. Ronk01 talk 20:22, 25 February 2011 (UTC)

It seems much more likely that we'd get something done if there weren't people putting the words "broken promises" in quotes when referring to the events surrounding PC, and downgrading the significance of these events... --Yair rand (talk) 20:55, 25 February 2011 (UTC)

Plea for discussion before !voting

Please, before people start support or oppose-ing the above, can we discuss the points you made? Ie, before we have any kind of poll, can we talk about it?

I'd like to respond to some of the points above - and I intend to, ASAP. I hope others will want to discuss the proposed poll too.

I just ask that, for the time being, people don't start !vote-ing again. This is a personal request, from me. Please let's discuss things, before calling for !votes.

I will write more here as soon as I can. Thanks.  Chzz  ►  20:47, 25 February 2011 (UTC)

I should mention that the above is not a poll, rather it is a proposal for one, along with several other things. I very much want to discuss my proposal, and work out something functional and agreeable. Ronk01 talk 20:52, 25 February 2011 (UTC)

Many thanks, Ronk01! Phew. My quick posting, above, was aimed at heading 'em off at the pass, before we got a whole string of "support", "oppose", etc! -I am very grateful to you, for drawing up those concise points-for-debate. But yep, we need to discuss the points before people start counting up numbers again! Thanks for your understanding.  Chzz  ►  20:55, 25 February 2011 (UTC)

• Yes, I too think we need to start with discussion. I'm not sure what is meant by the two "If" sections, because it seems to me that any straw poll would seek to determine which of the two scenarios would be the case, so we would not choose one beforehand. But I like the idea of evaluating it only on the technical merits and not the other issues, and I like the idea of setting, at the start, a defined threshold. --Tryptofish (talk) 20:56, 25 February 2011 (UTC)

Yes, the lack of a threshold on the first poll was a major issue. The bar kept being moved.Ronk01 talk 21:05, 25 February 2011 (UTC)

Points 1,2,3 are useless, unless WMF will accept #4. We need input from Steven (WMF) (talk · contribs), on that concern. Ie, are WMF willing to compromise on continued development of the extension if enwiki support further trials but not necessarily acceptance of implementation? We need an answer to that, before we can form this proposal.  Chzz  ►  23:46, 25 February 2011 (UTC)

I definitely don't mean to speak for the developers, of course, but I think they already have answered that: they are not willing to do more unless the community commits to implementing whatever they will come up with. As I've pointed out above, that creates a Catch-22, because it forces us to decide based on what exists at present. --Tryptofish (talk) 19:49, 26 February 2011 (UTC)

Just to respond: it's not that the staff require that the community implements "whatever they will come up with". When it comes to any proposed new features or work, there is no absolutism in accepting new alterations to Pending Changes. It's simply that right now there are no full time developers devoted to Pending Changes, so if we're going to take time away from other projects to work on it going forward then we need to have a clear picture of where and how the community wants to use the feature, if at all. To answer Chzz directly about point four: if the community comes out of this with another trial as the response, then development is probably not going to move past maintenance. The community is welcome to continue using it however you want, including in trial mode, but that's the likely response when it comes to developer time at the Foundation. Steven Walling at work 20:37, 27 February 2011 (UTC)

Yes, thanks, I should have worded that better. I guess the better point is that the community will have to commit to PC in its present form if the community hopes, realistically, to be able eventually to use PC in an improved form. --Tryptofish (talk) 21:04, 28 February 2011 (UTC)

No worries! I just wanted to be clear. Anyway, the commitment we're talking about here doesn't need to be absolute or wide-ranging by any means. I've had other people ask me, so I should probably say: there is no specific threshold of articles (by number or by type) that the Foundation is shooting for here in evaluating whether we need to restart development or not. If the community can only get consensus for limited use of some kind, then that's okay, and we can discuss what coding needs to be done and on what time frame. It's only trials that are unlikely to provoke more staff development of the feature. Steven Walling at work 04:24, 1 March 2011 (UTC)

I'd actually like a clearer answer, Mt. Walling, as "directly about point four [..] probably not going to move past maintenance" - 'probably' is not very clear. In this RfC, we are discussing our options. Are you saying that, if we will not accept full implementation at this time, WMF will give up on PC?  Chzz  ►  04:29, 1 March 2011 (UTC)

My instinct is to say "no" to that exact question, because it implies we're demanding some kind of sweeping use of Pending Changes to work on it more. That's not the case. And "full implementation" is a very vague concept at this point which could mean any number of things. Also, we shouldn't "give up" on features currently being used, trial or not. If it's used in any format, it's our job to maintain it or at least see that it's not impairing the regular functioning of the site. Anyway, let me try to put it another way: if the community wants major development to start again, then you need to demonstrate there's a real need for it. Another limited trial for a month or two (after half a year of trials) says to us that consensus has not reached the point where we should take resources away from other critical projects to devote them to Pending Changes again. Steven Walling at work 04:56, 1 March 2011 (UTC) (P.S. You don't have to call me Mr. Walling. Steven will do.)

• Comment. The cut off should be 50/50. Not 2/3rds. We all get equal say. Not one group gets two votes for every one vote of the other group. --Doc James (talk · contribs · email) 00:26, 26 February 2011 (UTC)

  This is why polls are evil. We would do better with an expert mediator. —UncleDouggie (talk) 04:19, 26 February 2011 (UTC)

• When major policy changes are in question, we have always required a supermajority. It's not giving one group more voice than another, its rather a way of achieving stability. What is 51% tofay will be 49% tomorrow, and we can't keep flipping policy back and forth. DGG ( talk ) 05:19, 26 February 2011 (UTC) .

• Just because we have required a supermajority for other stuff does not mean that it is appropriate here. 50% plus one is best. I guess we could have a poll on what number to us but I am sure all can see where that would lead... Doc James (talk · contribs · email) 17:42, 26 February 2011 (UTC)

Numberwang, I imagine. Sorry, couldn't resist trying to lighten the mood. The trouble with 50%+1, Jmh649, is that a week or a month later, a new poll might show 50%-1 and then we'd have to remove it.  Chzz  ►  18:00, 26 February 2011 (UTC)

Oh and this will be different if we use 66.6%+1 verses 66.6%-1? One person one vote is called fair. One way to determine fairness is to ask what if the positions where reversed "what if 2/3rd where required to get rid of PCs". This of course is a version of the golden rule. Doc James (talk · contribs · email) 19:30, 26 February 2011 (UTC)

(edit conflict) Yes, kinda. a) Proposal to turn on the lights, 66.6% in favour, lights turned on. <one person changes their mind, then...> b) Proposal to turn off the lights, 66.6%-1 against, only 33.3%+1 in favour, which is less than 2/3, proposal fails. Lights remain on.  Chzz  ►  19:36, 26 February 2011 (UTC)

Great so PC is currently on ( like the light above ). Therefore we need 66.6+1% to vote against it to turn it off. I guess I could live with that but still do not think it terribly fair. :-) Doc James (talk · contribs · email) 19:40, 26 February 2011 (UTC)

When I said earlier that I like the idea of a fixed threshold, I wasn't sure what that threshold should be. The most important thing is to decide the rules at the start, instead of changing them mid-process. But I now think the 2/3 number makes better sense than 50/50. It's clear that everyone, on both sides of the discussion, regards this decision as an important one. For issues that require unambiguous consensus for change, we have plenty of precedent for requiring a supermajority—just consider RfA/RfB. --Tryptofish (talk) 19:45, 26 February 2011 (UTC)

Yes, but remember that RfA/RfB are generally small discussions compared to this, so 2/3 is perfect here, anything higher, like the 70% needed for RfX would be unreasonable. Ronk01 talk 20:06, 26 February 2011 (UTC)

Sorry, I wasn't clear. I agree with you. I meant that 2/3, not something even higher, would make good sense, but I was just pointing out that we use even higher thresholds in some well-known cases. But higher than 2/3 would indeed be too high here. --Tryptofish (talk) 20:41, 26 February 2011 (UTC)

Not a problem, it's nice to know that there is a level of approval for 2/3 consensus. (Am I the only one who is amused at the idea of consensus for consensus?) Ronk01 talk 22:15, 26 February 2011 (UTC)

While everybody here seems to be happily discussing voting thresholds, I feel I should just share my view on this, which is I do not think polls are the way forward for PC. No matter how well planned, they just seem to me to be fundamentally the wrong way to go about this. I can explain my thoughts on why polls are a bad move, but it's fairly typical stuff we've all heard before, one only needs to read these two discussion pages to see all the things which can go wrong. - Kingpin¹³ (talk) 22:52, 26 February 2011 (UTC)

No polls. The first one had mercurial guidelines, and the second one was instituted to save face after the first one failed. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 22:59, 26 February 2011 (UTC)

I would appreciate the restriction of ideological complaints and arguments here. Thank you. Ronk01 talk 23:55, 26 February 2011 (UTC)

The first poll, however, *did* have mercurial guidelines. The rules frequently changed in the course of polling to address concerns, and the second poll was then used after the first poll closed in an effort to wash away the taste of the first poll by making sure the rules didn't change. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 02:22, 27 February 2011 (UTC)

That it did. The second poll was an attempt to prevent mass craziness. I would advocate more discussion before any more polling. Ronk01 talk 02:33, 27 February 2011 (UTC)

Enough polls, !voting, discussion... May I suggest what some will consider a radical approach? Just dump PC and make editing registration manditory. And here's a fearless prediction: registration is going to be required sooner or later. Let's do it now. Jusdafax 03:53, 27 February 2011 (UTC)

Not going to happen as that's above all of our paygrades. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 06:56, 27 February 2011 (UTC)

Jeske is quite correct, that is actually not a decision we can make or enforce. Unfortunately keeping this RFC on point has not so far been achievable, we keep getting off-the-wall proposals like this instead. I certainly don't see any consensus evolving here, just the typical mess that all the recent major policy discussions eventually become. Apparently it is too much to ask for people to limit their comments to specific topics so that we can actually get something done. The one thing I am sure about is that more polls are a very bad idea, but I'm unsure what to do to put this RFC ack on track as nobody will listen to the repeated pleas to limit the conversation to the actual topic of the RFC. Maybe we should start moving threads that do not directly discuss whether we keep PC or not to the talk page? Beeblebrox (talk) 19:50, 28 February 2011 (UTC)

I went back to the top of this talk and re-read #Purpose in that context. It seems to me that we have, by now, discussed to death whether or not to keep PC, and we have established that there are various opinions about that—thus the many talk threads. I fear the only way we will get from a discussion that lacks a clear outcome, to a decision about whether the community wants to keep PC or not, will be by way of a poll. --Tryptofish (talk) 21:09, 28 February 2011 (UTC)

(edit conflict) - Yes, I agree, we are not getting anywhere fast, we need to focus. (I accept my part in the lack of focus) Would you attempt to put us back on track and do the honors of moving new stuff or old stuff that is off topic to talk? imo we have a few options that have clearly had some support and some more discussion of those is now needed - for example -

• - 1 - turn off and continue discussion.

• - 2 - turn off and reject.

• - 3 - trial on a larger number of the lowest watched BLP articles (open to discussion but perhaps 50,000 or five percent of such articles).

• - 4 - continue as we are with no major expansion a simple acceptance of the tool and allow normal usage without restriction as and when required as per current alternative protection alternatives.

• note - these four options give us two keep on and two turn off options a npov balance.Off2riorob (talk) 21:12, 28 February 2011 (UTC)

Just to be clear in light of that ec, "you" means Beeblebrox, not me (I hope!). --Tryptofish (talk) 21:14, 28 February 2011 (UTC)

Yes, excuse me, I ended with an additional indent. I am up for a poll format discussion as well - Off2riorob (talk) 21:29, 28 February 2011 (UTC)

• How is number 3 a balanced option rather than a hundred-fold increase in the incidence of PC to articles not presently semi-protected (the apparent "advantage" of PC over the current system in terms of editor engagement). Protonk (talk) 21:51, 28 February 2011 (UTC)

Option 3 is a trial expansion. Its as I remember a 25 fold specifically directed expansion focused through the results and experiences gained in the original 2,000 article trial. This option has a degree of support in this discussion and at a previous poll. I am unsure what you are questioning, please clarify, do you disagree that this option has been raised as a possibility in discussion or do you have specific problems with it? As I see it, these are the four arisen from the trial and previous discussion options, lets lays them out either for more discussion or polling and a talkpage. If you have strong objections to one of them or all of them then lets clear that up first. Off2riorob (talk) 22:01, 28 February 2011 (UTC)

I have specific problems insofar as it (1) proposes a >10-fold (I don't have exact numbers on the number of articles under PC at any given time in the trial but it has to be at least an order of magnitude lower than 50,000, probably more) increase in a program whose existence is heavily contested. If we are having issues with the program being approved on the basis of status quo bias now it will be impossible to deal with later. (2) Because it conflates the BLP problem with the protection mechanism. We should be asking if this is an appropriate tool for use on the entire encyclopedia. Bringing in the BLP bogeyman (and making vague comments about legal liability) only clouds the discussion just as it did with the request to autmatically delete unreferenced BLPs. (3) The sample size you are proposing is unnecessarily large for a true trial unless that trial is meant only to stress the PC queue. A trial to determine the effect of PC on article quality, editor engagement and any other feature of the protection system need not be in the thousands of articles let alone the tens of thousands. Protonk (talk) 22:11, 28 February 2011 (UTC)

The original trial was around above 2,000 and currently there are around just under 1,000 articles Pending protected. Please correct me if I am wrong someone but there are around a million BLP articles, that would mean adding pending to the least watched five percent of them for this option. The option to just switch off also has some support, if you just want to continue a bit more discussion we can put these four options in sections for individual specific continued discussion and you can add some more if you can think of any that have had support. Off2riorob (talk) 22:13, 28 February 2011 (UTC)

I would reject outright any proposal to make PC encyclopedia wide, but expansion to our least watched BLPs might not be the worst idea. Ronk01 talk 22:21, 28 February 2011 (UTC)

You aren't wrong about the percentage of articles (at least as a gross estimate), but that isn't the variable of interest. The variable of interest is the number of articles in the PC "trial" and proposing a trial of 10-20 times the number of articles in the current trial is tantamount to proposing to indefinitely extend the program. that's only my first objection. Protonk (talk) 22:22, 28 February 2011 (UTC)

These are only the supported positions as I have seen them, if you have others please add them. Perhaps we could then focus by opening individual sections to discuss the specific options??? Off2riorob (talk) 22:26, 28 February 2011 (UTC)

• note - as previously commented on - there has been a lot of discussion - vague additions such as this from user Cenarium at such a stage will be moved to the talk page here - please attempt to stay on focus and read and assimilate previous consensus and support, thanks. Off2riorob (talk) 03:44, 1 March 2011 (UTC)
  • Unacceptable, especially when you have just yourself made a proposal here. I reverted, and do not even think of reverting. This lack of AGF is alarming. I read the discussion, and this is entirely on purpose. And considering that your unconsidered actions, Off2riorob, are largely responsible for the present mess, you should certainly not take the initiative here. I remind you that I am the one who, alone, made the proposal for this trial which was approved, so I do know what I'm talking about. Cenarium (talk) 04:23, 1 March 2011 (UTC)

Really go on then, you reverted, because you can, and threaten, but so what, your additions are worthy of the talkpage only , please consider that and attempt to jion in . and help focus. Off2riorob (talk) 04:25, 1 March 2011 (UTC)

Not trying to play the big mediator here... but calm down guys. I think Cenarium's proposal is okay personally, but I understand why Off2riorob is wanting to try and keep things cleanly on topic. Both of you are right, and acting in good faith. Steven Walling at work 04:31, 1 March 2011 (UTC)

• I endorse the suggestion to avoid any kind of polling before proper discussion on the next steps has been undertaken, and to this effect I made a proposal below. Cenarium (talk) 04:37, 1 March 2011 (UTC)

Regarding the initial proposal of the section, I think we should analyze the trial before any poll/discussion on PC's future. Cenarium (talk) 05:00, 1 March 2011 (UTC)

Proposal for way forward

A few ground rules I view as necessary: first, no rushed things, especially no rushed big poll; second, we should analyze the trial before any substantial discussion on the future of PC; third, we should develop through compromise a proposed implementation with its usage policy and see if it has consensus. It is absolutely necessary to have only one option, otherwise it will be impossible to reach a consensus, as we have seen already. We must work out a proposal then put it to a vote, that's it. We cannot afford multiple proposals, we must find a compromise proposal then see if it has consensus, that's how we've got this trial, we should go about the same way for a proposal for an implementation of indefinite duration. If people have different proposals, they'll have to wait until either the compromise proposal is approved or not, because we need to reach a decision.

Concerning the analysis of the trial, which should take 2-3 weeks, one way to go about it is to make a list of specific questions (e.g.: was lvl 1 PC efficient at preventing obvious vandalism ? was lvl 1 PC efficient at preventing clear violations of BLP ? what can we conclude of articles where vandalism is high with PC ? what can we conclude of articles where vandalism is low with PC ? what can we conclude of articles where vandalism is average with PC ? was lvl 2 PC protection useful in preventing persistent vandals ? was lvl 2 PC protection useful in preventing sockpuppets who are not obvious vandals ? overall, is the second level a useful addition ?), then have people give their answers on those, with examples to back up their claims (also, we must think of enough questions before starting answering them, then more can be added).

Then, we work out a proposal for an implementation of pending changes which should form the basis for a policy on its use and detail its implementation. This must be a compromise proposal. When this is done, we discuss with the community at-large to see if there is consensus for this proposal, culminating in a poll, where roughly 2/3 is needed in support.

1. If there is consensus for the proposal, we adopt the policy with its implementation.
2. If there is no consensus for the proposal and no considerable support (a 'strong majority') for pending changes, we ask to have the implementation turned off.
3. (Alternative 1) If there is considerable support (a 'strong majority') but no consensus, we freeze the implementation and try for a last time to find a consensual proposal. We submit it to the community, and if there is still no consensus, we ask to have the implementation turned off. Further proposals can be considered.
4. (Alternative 2) If there is considerable support (a 'strong majority') but no consensus, we freeze the implementation and try for a last time to find a consensual proposal. We submit it to the community, and if there is still no consensus, we remove pending changes from articles. Further proposals can be considered with the goal of reaching consensus on one in particular. If no consensus is reached within a period of two years, the implementation is turned off.

Cenarium (talk) 01:14, 1 March 2011 (UTC)

I would object to your third option, as we would essentially be denying the rights of a majority in favor of a minority, essentially giving them two chances to get what they want. That may be fine for AfD, but major policy changes need to be discussed until consensus is reached. Ronk01 talk 01:30, 1 March 2011 (UTC)

Isn't it well established that no consensus amounts to status quo in Wikipedia tradition ? At some point if still no consensus comes, we need to roll back to status quo. It doesn't prevent from continuing discussion though, which I've noted. Cenarium (talk) 04:28, 1 March 2011 (UTC)

Yes, in AfD's and minor discussions. However, in a major policy change like this, consensus is a must. Ronk01 talk 04:47, 1 March 2011 (UTC)

But what if we can't reach it ? There has never been a consensus for using PC indefinitely, since the trial ended we failed to reach any kind of consensus, so if we still can't reach it, at some point we have to roll back. Then we can continue discussing, but there is presently no legitimacy for continued use of PC. I can see your point though, so I've proposed a softer 'roll back', see alternative 2. Cenarium (talk) 05:21, 1 March 2011 (UTC)

I would remind you of the Devs position that when PC is turned off, it stays off. We can remove it from all pages, but not actually turn it off until we have consensus to do so. Ronk01 talk 12:31, 1 March 2011 (UTC)

• - Really - this is talk page fodder - please consider and join in - really you are almost not even editing in this discussion or contribution presently to the wikipedia.Off2riorob (talk) 04:31, 1 March 2011 (UTC)

I don't have time to edit these days, so what, this is just a proposal to consider. Cenarium (talk) 04:33, 1 March 2011 (UTC)

Cenarium, I applaud your efforts to try to bring resolution; however, there are a number of specifics to which I object. Firstly, the proposal begins, "we should develop through compromise a proposed implementation " implies it is already accepted that we will accept implementation; there is not. Secondly, that it is "absolutely necessary to have only one option, otherwise it will be impossible to reach a consensus - that is partially valid, in regards to polls, but you are confusing the term 'poll' with 'consensus'; if you have not already, please consider the views expressed in meta:polls are evil. We got into this current situation because of polls. Thirdly, what is "considerable support" exactly? You said, ""If there is no consensus for the proposal and no considerable support for pending changes, we ask to have the implementation turned off." - if 'considerable support' is numeric, that is bizarre, because if 2000 people vote 'yay' and 3000 vote 'nay', clearly the 'yay' has considerable support. The last poll scraped up less than 60%, which I personally do not consider evidence of clear approval - certainly, nobody would pass RFA with 60% - thus, there really is no consensus for it (yet), thus as far as I am concerned, it should not be implemented (as of now).

I suggest - as I did from the start of this RfC - that, firstly, we turn it off. Because its use, now, is anachronistic; no consensus has formed to use it (and definitely not beyond a brief trial).

If we can do that, we can then discuss possible future use.

I am disappointed at the seemingly intractable stance that WMF will only help develop if we accept implementation; however, I will not permit that to sway my opinion. Chzz  ►  05:13, 1 March 2011 (UTC)

This does not presuppose acceptance. What I mean is that, when people propose a new policy, they write a policy, often several users get involved and through compromise and discussion modify the proposal, and when it's considered ready, they submit it to the wider community. This doesn't presuppose that the community will find consensus for it.

I mean that we should determine whether there is consensus for one particular proposal, because it's extremely difficult to determine consensus when several proposals are available. Unfortunately, polls have become almost inevitable in policy discussions of a large magnitude (and are used for those repeatedly). I agree though that it's not optimal, and that polls should only be an indicator. So we could ask a panel of bureaucrats (or admins, or trusted users to be chosen beforehand) if in lights of discussions and the poll, consensus exists.

Considerable support should be understood as "a strong majority", I know it's not unambiguously defined, we may leave it up to the interpretation of the panel of bureaucrats.

Unfortunately, I do not think it is possible to have it turned off like that, considering the position of Jimbo and lack of momentum in the community. (For what it's worth, I would have preferred that it be turned off, at least removed from articles, at the end of the trial, it would have made things easier indeed. - see here for my position at the time) Cenarium (talk) 05:56, 1 March 2011 (UTC)

• Would you agree with analyzing the trial as I suggest above, by making specific questions then having people answer ? I soon won't be able to edit for a few days, so if it sounds ok, go ahead and feel free to add the questions I proposed. Cenarium (talk) 06:42, 1 March 2011 (UTC)

My whole issue with this is I couldn't work out what you want to do and it seems overly complicated and vague and three steps back from where we had come to, but I will try to get it. The first part is imo unobtainable here through discussion. As I think Chzz said, there are users here that strongly want to first switch off the tool and then suggest discussion will be better. Myself I object to switching off the tool without a community poll, both polls have been in support of continuation of the tool in some form and as such I think switching off should require support, not consensus support but just more than fifty percent. I also feel that once the tool is switched off discussion will evaporate. IMO a handful of vocal objectors here are not representative of the wider picture of general acceptance of the tool, please remember that four hundred and seven wikipedia editors supported the tool going forward in one way or another. Off2riorob (talk) 13:21, 1 March 2011 (UTC)

• - 1 - You want to form a single proposal for implementation here

• - 2 - you want an analysis of the tool done here using the questions stated - lasting three weeks

• - 3 - you want to poll the community for two thirds support to implement - if this is not formed you want to follow the four options above.

was PC efficient at preventing obvious vandalism ?

• This is a tough question to answer as we mostly only see our own edits and edits on our watchlist but in my experience obvious WP:vandalism was rejected by reviewers on sight as they would on an unprotected page on their watchlist. Their is no doubt that having an editor look at the desired addition of an unconfirmed account is more likely to keep vandalism out of the article than a little watched article with no protection and while pending has been on trial I have seen multiple BLP articles vandalized with attack content added and not removed, sometimes for months. Off2riorob (talk) 16:10, 1 March 2011 (UTC)
• Definitely. I have never seen a case where a reviewer accepted obvious vandalism. Reaper Eternal (talk) 17:17, 1 March 2011 (UTC)
• Agreed. Not sure how any Reviewer could miss obvious vandalism. Especially since many are also Rollbackers and thus experienced vandal fighters. Ronk01 talk 17:22, 1 March 2011 (UTC)
• Yep. --joe decker^{talk to me} 17:23, 1 March 2011 (UTC)

was lvl 1 PC efficient at preventing clear violations of BLP ?

• IMO the answer to this question is exactly the same as my answer to the question above. Off2riorob (talk) 16:12, 1 March 2011 (UTC)

what can we conclude of articles where vandalism is high with PC ?

• In my experience pending was removed when vandalism was high as semi protection became better for the situationOff2riorob (talk) 15:30, 1 March 2011 (UTC)
• Agreed, the article Barack Obama is a perfect example of one that needs semi, not PC. Ronk01 talk 15:36, 1 March 2011 (UTC)
• Agree, getting large numbers of revisions needing review where they are only vandalism wastes time. Those articles should be semiprotected. Reaper Eternal (talk) 17:18, 1 March 2011 (UTC)

what can we conclude of articles where vandalism is low with PC ?

• Generally, I have found that articles that do not get much vandalism do not need PC, with the exception of BLPs. If somebody inserts "penis" into an article on a random village, that will merely be annoying and might make readers doubt Wikipedia's accuracy. However, if defamatory content is inserted into a BLP (in other words, libel) and the subject sees it—that is something we must try to prevent at all costs. Reaper Eternal (talk) 17:22, 1 March 2011 (UTC)
• Agree, minor vandalism like random word insertion is of little concern, since it usually gets reverted by a bot or an RC patroller within a couple of minutes or less. BLP attack vandalism can be more subtle, and would likely escape a bot or perhaps even an RC patroller. PC provides a method for slower, more methodical analysis of an edit. Ronk01 talk 17:36, 1 March 2011 (UTC)

what can we conclude of articles where vandalism is average with PC ?

• See above comment for information. Reaper Eternal (talk) 17:23, 1 March 2011 (UTC)

was lvl 2 PC protection useful in preventing persistent vandals ?

• I have really only seen one case where PC2 was really necessary: the Justin Beiber article. Autoconfirmed accounts repeatedly came out of nowhere to vandalize the article. In my opinion, PC2 should only be applied with semiprotection to high-risk BLPs where autoconfirmed accounts are vandalizing. Reaper Eternal (talk) 17:25, 1 March 2011 (UTC)
• Agree. If the concern is Socks, eliminating concerns from IP edits is a good idea. PC-2 would be great on articles that get quite a bit of sock vandalism, but in combination with semi, it would probably be more efficient. Ronk01 talk 17:33, 1 March 2011 (UTC)

overall, is the second level a useful addition

• That is a good question. PC-2 of course bars all non-Reviewers from making direct edits to a protected page. Though the PC policy specifically disallows its use in disputes, it has seemed to be useful (in those rare cases where it is applied, as far as I know, fewer than 100 pages have ever been PC-2 protected, and none right now) for preventing sock vandalism. Of course, PC-2 is by no means acceptable as a long term method of protection, as it ostracizes many perfectly good editors by refusing to directly accept their edits. However, I would be in favor of retention of it as an alternative to anti-sock Full-prot in some instances. Ronk01 talk 15:42, 1 March 2011 (UTC)

An alternate way forward

The "way forward" proposed above continues to confound the question of what to do with PC with that of what to do with the trial. I don't understand why you think you need that. As confirmed by above discussions, this would only drag out and confuse discussion, make editors feel like the results of previous discussions are irrelevant (and therefore this one also), and possibly scupper the whole idea as editors just give up arguing about it, except to vote a knee-jerk No to anything. Plus, the "proposal for implementation" doesn't need to be made in a single step - it would be more logical to proceed in stages, focusing all effort to getting each one right.

I would suggest:

1. Remove all current mainspace articles from the PC list. People voted, the poll was closed, that's what it said. You need to restore the credibility of the discussions. You don't have to rip out the software.
2. The end of the trial is part of the trial. What happens to level 1 PC articles once PC is ended? Do people report an upsurge of vandalism and call for semi-protection? Does vandalism linger unnoticed? Do things quietly go back to normalcy? Or do some people get so hysterical about the possibility of vandalism that they refuse to let these random articles go back to unprotected status even despite a lack of vandalism...?
3. Propose "Pending Changes Step A". Only level 1 protection is allowed, and it is applied only to selected articles currently subject to semi-protection. A fixed stop date is set for each such article, with a maximum end date limited to the lesser of (a) some fixed number of months and (b) the original end date for the semi-protection. When the stop date comes, the article will be returned to semi-protection if (b) has not yet expired. A readable central log of all PC articles, duration of protection, and complaints/problems should be kept.
4. Propose "Pending Changes Step B". This permits admins who normally semi-protect articles to apply Pending Changes level 1 directly, and to set later (but finite) stop dates. The total number of PC + semi-protected articles and the terms for which these are applied should be watched.
5. Propose "Pending Changes Step C". Applications for indefinite Pending Changes Level 1 and short-term use of Pending Changes Level 2 can be entertained. Each article should be discussed and voted at WP:ANI, like article probation/general sanctions - which might be a preferable alternative.

Recognizing, above all, that consensus may be "no" to any of the later steps. Wnt (talk) 20:01, 1 March 2011 (UTC)

Endorse in general. (#2's not really a process, more of a comment.) If Steps A and B are supposed to be sequential in time, I'd recommend that they be simultaneous (which, in effect, would eliminate the difference between A and B). Either way, in B would recommend elimination of "who normally semi-protect articles," since all sysops can do that, or if not eliminated, then define what it means. — TRANSPORTERMAN (TALK) 22:20, 1 March 2011 (UTC)

I'll generally endorse this because I feel it is important to disambuguate the "trial" with the "introduction" of the tool. We actually need some data on effect on vandalism, queues, editing, etc (see the drum I'm banging above about this) and a real trial can provide this. Protonk (talk) 22:38, 1 March 2011 (UTC)

'Limited endorsement. I have to be concerned about 2 though, as that could lead to BLP libel. Also, as discussed above, PC-2 is best implemented alongside semi-prot. Also, the community did not form any kind of consensus in either poll, thus the claim that "People voted, the poll was closed, that's what it said" is only partially true. The credibility of the discussions themselves seems less damaged than that of the process. Ronk01 talk 23:03, 1 March 2011 (UTC)

Support #1 - removal from all articles, and removing the vague policy permitting its possible use - clarify that it is to be used only for testing purposes, not on any live articles (until a possible future date when of course the policy can again change). #2 is not really a proposal as such; of course we should be analysing results, but I don't see why 'proposing' it here will help. For 3,4,5 I'd like to see discussion and consensus on a proposed further trial, with very carefully defined and consens-agreed remits and limitations. But I don't think any of it can happen whilst we remain in this state of flux. Hence, yeah; let's remove it from existing articles to 'clear the air', and then look at it afresh.

I also quite like notion, mentioned, of setting up some kind of 'task force' or whatever; a small group of users who could work on the options for possible use of PC; people who could carefully review all the debate (here and elsewhere), look at statistical evidence, look at the direct effect it has had on articles, and help put together further (nice and clear) proposals to the community, with their clearly stated findings.

Setting up such a group might allay the concerns of some, that the whole debate will "fizzle out" one PC is removed. Chzz  ►  23:36, 1 March 2011 (UTC)

• This is a very difficult proposal to get my head around. Which steps are you suggesting a two thirds community consensus for? You appear to want to switch the tool off and then seek a two thirds majority for this - # Propose "Pending Changes Step A". - to summarize - you suggest removing semi protection from some articles and for a set period of time replacing with pending protection and collecting data and evaluating the data. If two thirds support is not attained then that will be the end of the issue? or will you then poll the next option and again require a two thirds majority? Also what number of articles are you considering this on and for what time limit? If the first option is rejected you will poll the next option, allowing admins to use pending as they want to - you say the numbers and times lengths should be watched, what figures are you considering and for how long? and then you propose that indef pending can be added but a discussion should be had for each article at WP:ANI again this option will require a two thirds majority. Off2riorob (talk) 23:36, 1 March 2011 (UTC)

• This looks like a pretty well-formed proposal to me. Points one and two give us time for a breather and to think about the relative merits of Pending Changes. The rest looks like the kind of very gradual rollout (with plenty of time for feedback and broad community oversight, including to step back if need be) that is appropriately cautious. Even a limited, step-by-step commitment to long term use will show the developers it's time to recommit time and energy to Pending Changes. If there's volunteer involvement, we could also probably the time that it's not being used/used in level one only to flesh out a new roadmap and ideas for future development. Steven Walling at work 00:56, 2 March 2011 (UTC)

This is just a switch off and poll the community to retrial - it will never get two thirds support - users have had enough of trailing a simple protection tool, we all have experience of it, its time to poll yes or no to the tool and move on. Off2riorob (talk) 01:01, 2 March 2011 (UTC)

Inferences of an unused resource

I am a bit dismayed that the IRC channel Wikipedia-en-pending has been so under used as a resource which I believe could have been beneficial in developing this program. If the entire log from the inception of this channel were reviewed, the inference would be contrary to reasonable expectations. Nevertheless it could still serve a purpose, if the people who care about Pending Changes would consider the benefits of real time discussion. Other than that observation, I support Pending Changes as a valuable protection mechanism. Thanks for considering this as well. Regards My76Strat 20:39, 1 March 2011 (UTC)

keep or turn off

Lets just go for the gravy - the editors know what the tool does now, lets just do a simple poll - Keep or Reject. Looking for a two thirds support for it. as simple as this below, we can hold it here with community notices. Off2riorob (talk) 23:53, 1 March 2011 (UTC)

I think people still aren't ready for a simplistic yes/no poll. Hasty polls caused a lot of bad feeling the last two times around, and I would hate to endure that again if new proposals for options are being put forth (like above). Steven Walling at work 00:39, 2 March 2011 (UTC)

Hi Stephen, people are ready, they have all experienced the tool, they don't need any more polling to see if they want to add the tool instead of semi protection in another trial, the proposal above is totally unworkable and without the possibility of ever getting anywhere, apart from the switching off of the tool. There is nothing really more to trial, we just need go for the end game now, the naval gazing turns people off. Off2riorob (talk) 00:51, 2 March 2011 (UTC)

The statement "the proposal above is totally unworkable and without the possibility of ever getting anywhere" is not borne out by the fact that there are several perpetual participants in this RFC who support it, in part or whole. Steven Walling at work 01:03, 2 March 2011 (UTC)

There will never ever be a 66 percent support for a new trial as is described in the proposal. Users have little energy for this issue as it is, the trial is over so lets see if the community supports the tools usage or not. Off2riorob (talk) 01:10, 2 March 2011 (UTC)

The trial only ends once PC is disabled and metrics taken. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 02:58, 2 March 2011 (UTC)

As the trial is over and editors have experience of the tool, do you support the Pending protection tool being added and accepted to the article protection toolbox.

Yes

• Off2riorob (talk) 00:00, 2 March 2011 (UTC)
• I assume this is the place to support Pending changes but this page is so chaotic it is hard to see where to vote. But please do consider this my total support for Pending changes and copy this vote to where ever else needed. Thanks. I will not be reading this page that often, but I really like Pending changes. History2007 (talk) 00:20, 2 March 2011 (UTC)
• Support using Pending Changes on an ongoing basis. Collect (talk) 01:51, 2 March 2011 (UTC)

No

1. The things I see fraggedrevs/pending chains as needing to fix are still ever-present - ignorant and elitist reviewers/supporters and ineffectiveness as an antivandal (or even a BLP) solution compared to everything else. —Jeremy ^{(v^_^v Hyper Combo K.O.!)} 02:55, 2 March 2011 (UTC)
2. Pending changes has just never sat well with me. It defeats Wikipedia's motto: The free encyclopeida anyone can edit. Before you start applying this argument agianst semi-protection, consider this: With semi-protection editors are not led to belive they can edit something, and then have their good edit not appear, simply because someone else is in a bad mood, dosen't like the topic of the article or the editor, or dosne't know enough about the subject to decide if the edit is ok or not. Even with PC installed reverting vandalism takes just as much, if not more, clicking. Vandals should not have to be told not to vandalise Wikipedia. Vandalising Wikipedia is just like taking someone else's homework, that they worked very hard on, and scribbling all over it just before it is turned in. Hope that comunicated my view clearly, and polietly. Sumsum2010·T·C·Review me! 03:03, 2 March 2011 (UTC)

Abstain

• Polling is evil. This "Yes/No" proposition is impossible. It is like asking, "Should Wikipedia has pictures on it?" - and then using the resultant "50%+1" to conclude we should permit galleries of hardcode porn. PC is not a 'yes/no' choice; nor do we have any idea yet if it is 'good' or 'bad'. Chzz  ►  01:49, 2 March 2011 (UTC)
• Voting gives a clear yes or no, but in this situation, there are too many options, too many shades of gray. Guoguo12--Talk--  02:45, 2 March 2011 (UTC)