Wikipedia talk:Pending changes/Straw poll on interim usage

How close will we get to 50/50?

This theory suggests that you can expect outcomes very close to 50/50 when large numbers of people vote, because the consensus forming process can proceed by moving to a metastable state of perfectly balancing opinions first and staying there for long times. Count Iblis (talk) 18:14, 20 September 2010 (UTC)

I Think its going to be close 50/50 sound like the norm here LOL...but am still wondering y this is going on when it was stated to be a trial period in the first place with a defined end time...think we are going about this wrong but what choices do we have.Moxy (talk) 18:21, 20 September 2010 (UTC)

It does indeed look like it could be close to a statistical tie. I doubt it will reach a true 50/50 though. I'm expecting a +5/-5 spread, give or take. ⒺⓋⒾⓁⒼⓄⒽⒶⓃ② ^talk 14:08, 21 September 2010 (UTC)

See also Duverger's law and Black's theorem. CRGreathouse (t | c) 03:12, 22 September 2010 (UTC)

Last straw poll

It looks like the infobox has been modified. For those interested in the discussion, the last straw poll was at Wikipedia talk:Pending changes/Straw poll. User A1 (talk) 18:37, 20 September 2010 (UTC)

Consensus interruptus?

It seems that if the opposes win here, then the situation will end up much as it would have if the process of consensus building on the old talk page had been allowed to continue without intervention. On the other hand, if the supports win here with something close to a 51/49 majority (which seems likely), then the controversy continues unchecked.

It seems obvious to me that we should have gone on with the original consensus. It would have been more efficient, quicker, and would have resulted in less drama. Revcasy (talk) 20:21, 20 September 2010 (UTC)

Oh and in case the poll page does not make it clear. Our course of action on this issue will be decided by simple majority. Revcasy (talk) 20:30, 20 September 2010 (UTC)

I think it should have been turned off after its trial (thats were the drama is...looks like they are tiring to sneak one by us)...if the vote is 50/50 thus no consensus reached ..it should be turned off as originally stated ....That said .. i see that a third vote option would have been a good idea as many Keeps are saying to keep it only after its fixed..so this arent real keeps vote for the current configuration, but rather for an updated and bug free new version.Moxy (talk) 20:34, 20 September 2010 (UTC)

Let us hope this will be closed by an admin who will notice this - and not by Jimbo. Septentrionalis PMAnderson 21:32, 20 September 2010 (UTC)

I'm going to quote Lumos3 here.

I am unmoved by the "broken promise" argument. This poll seeks consensus for an extension to the original trial period and there's nothing wrong with that. We have to manage Wikipedia in its best interest day by day. We agreed the PC is beneficial, and since we are going to implement an improved version in November what's the point in switching it off for 2 months?

Granted, replace "implement" with "trial" in that sentence and it's closer to fact. CycloneGU (talk) 13:37, 21 September 2010 (UTC)

What is the point of this poll?

If voters for the original straw poll are asked to vote for the continuation of the PC feature once again, why should we be expecting a change in votes? The people who voted to keep will vote to keep here. Therefore it should turn out roughly 65% for and 35% against. Again.

This is NOT a question of majority / minority. The straw poll was to decide whether or not the feature was to be kept. It was clearly established (even by Wales) that there was no consensus to do so. So why are we issuing another straw poll to decide whether it should be kept again? This is idiotic! Turn it off as agreed upon, fix it, and then issue another trial. _EricLeb⁰¹ (Page | Talk) 23:57, 20 September 2010 (UTC)

Side note: This poll should run on consensus, NOT simple majority based on headcount. Otherwise it's as if we are scraping the entire first straw poll, and saying "okay, so we'll should try again, but this time make it easier to get a Keep vote". _EricLeb⁰¹ (Page | Talk) 00:09, 21 September 2010 (UTC)

A poll on holding this poll was held on Jimbo's talk page. The decision there was to go ahead with this poll in the current format (i.e. with two options: keep or close). Count Iblis (talk) 00:02, 21 September 2010 (UTC)

And how was the poll for this poll an accurate representation of what the community thought we should do? _EricLeb⁰¹ (Page | Talk) 00:09, 21 September 2010 (UTC)

You'll have to check the lengthy discussions on Jimbo's talk page for that. I've seen that a RFC was posted there, but I haven't checked what happened since then. Count Iblis (talk) 00:13, 21 September 2010 (UTC)

There is no poll. There was a brief discussion and a decision. Unbelievable. _EricLeb⁰¹ (Page | Talk) 04:15, 21 September 2010 (UTC)

Why vote?

Maybe I'm letting my cynicism get the better of me in this matter, but am I the only one who doesn't see the point of voting in this "straw poll"? The point of a straw poll is to get a rough sense of who believes what, & to either begin the discussion based on that or to use it to bring the discussion to a conclusion. But more importantly, I can't help thinking that regardless of how anyone votes in this poll, we're going to have Pending changes up & running on Wikipedia very soon -- whether it is actually a good idea or not. (And to repeat my original statement on the matter, I'm all for it, if someone can show that it helps in some measurable way.) -- llywrch (talk) 00:02, 21 September 2010 (UTC)

As a newly registered editor who’s never encountered PC ‘in the wild’—or indeed heard of it outside of the two poll pages—I don’t consider myself well enough informed to vote. But it seems peculiar to me that (AFAICT) this poll is being held after the expiry of the trial period, and that the experiment didn’t end as scheduled. Odysseus1479 (talk) 00:21, 21 September 2010 (UTC)

Truth be told, I found this peculiar on August 16 as well and determined work was being done to remove articles from the queue slowly. I did not realize the project was being extended. Nonetheless, I was happy for its extension because it does work in some places, but I support removing it from places where it does not work. CycloneGU (talk) 05:31, 21 September 2010 (UTC)

Option 1 has won

Option 1 has just crossed the 2 sigma mark. I suggest that an univolved Admin prematurely close the vote per WP:Waste of Time and certify option 1 as the final result. Count Iblis (talk) 00:37, 21 September 2010 (UTC)

I suggest you wait for a bit, the 2 camps are barely 20 odd !votes apart at this time. It says 7 days, so wait them out instead of trying to push shit through. BarkingFish 00:46, 21 September 2010 (UTC)

So many votes. But let it run for a week and give everyone a say who wishes.Doc James (talk · contribs · email) 01:00, 21 September 2010 (UTC)

Dabomb87's nomination for adminship was decided within six hours - heck, it set a record - but it still ran for seven days. This will happen here as well, and it won't be fully supported like the nomination! CycloneGU (talk) 05:23, 21 September 2010 (UTC)

Calm down, it's been five hours. But I agree, this is a waste of time. _EricLeb⁰¹ (Page | Talk) 02:15, 21 September 2010 (UTC)

61% in favor now, and a 4 sigma deviation from perfect 50/50. That's almost as strong a satistical result as the LHC needs to claim the discovery of the Higgs boson. Count Iblis (talk) 15:43, 21 September 2010 (UTC)

Count, why do you want to invalidate the poll by closing it against its own terms? DuncanHill (talk) 15:46, 21 September 2010 (UTC)

Count, I actually agree with Duncan here. There is no reason to rush closure of the poll. That will just lead to even more drama, and this is already becoming too much. If there is another poll after this of near-identical type, while I will still support PC, I will happen to oppose that poll. Of course, if the topic of the poll does differ enough, I'll still support it, but we've had enough polls. CycloneGU (talk) 17:20, 21 September 2010 (UTC)

Yes, and User:Harris has told me that she won't certify the poll results prematurely, like she did last time. Count Iblis (talk) 17:40, 21 September 2010 (UTC)

If your goal is to create more drama, please continue pushing for an early close. Nothing is hurt by letting this run to it's obvious conclusion. --Onorem♠Dil 17:43, 21 September 2010 (UTC)

I'm a supporter of keep, but I agree a premature close is a silly idea and sure way to generate more drama, ill feeling and wasted time then keeping it open for the specified timeframe ever will Nil Einne (talk) 21:30, 21 September 2010 (UTC)

Further, I believe this poll was under Jimbo's discretion. Early close might aggravate your founder. =) CycloneGU (talk) 22:26, 21 September 2010 (UTC)

Question

Where/when exactly was there consensus for starting a new trial on November 9? --Yair rand (talk) 02:20, 21 September 2010 (UTC)

The great, grand owner of this place, Jimbo Wales' head of course. Mr. R00t ^Talk 03:00, 21 September 2010 (UTC)

They also held a poll on whether or not to hold a poll to replace the previous poll on Jimbo's talk page. Mr. R00t ^Talk 03:02, 21 September 2010 (UTC)

Seriously? And it passed? Could you provide a link? (And Jimbo is not the owner.) --Yair rand (talk) 03:18, 21 September 2010 (UTC)

There is no poll. I see absolutely no poll anywhere on his talk page. I just see a brief discussion and a decision by Wales to start a poll. This is pure lunacy. I'm getting sick of Wikipedia. _EricLeb⁰¹ (Page | Talk) 04:13, 21 September 2010 (UTC)

It was a request for comment, actually, not a poll. CycloneGU (talk) 05:24, 21 September 2010 (UTC)

Ericleb01: There's an easy solution to getting sick of Wikipedia (besides quitting), just do a personal dramaout. It's not for everyone, but then some people actually thrive on drama. Interestingly, WP:DRAMA defines perpetrators as "banned editors, stymied POV warriors, malcontents, and those with personal grudges." Much of the PC drama doesn't really don't fall into these categories. —UncleDouggie (talk) 06:55, 21 September 2010 (UTC)

I don't need a dramaout because I've already decided to leave once I get my article through FAC. This place is way too binded with too many vague guidelines and isn't ideal for new editors at all (although I've been here for four years). This poll is just pushing everything over the edge for me. So I'm going to leave, concentrate on my studies, and visit as a WikiGnome. _EricLeb⁰¹ (Page | Talk) 20:15, 21 September 2010 (UTC)

Hey Ericleb01, If you are getting sick of Wikipedia, what is the point of pushing an article through FAC? I have an alternate suggestion. Admit that Wikipedia does not matter much and is deeply flawed, then do not let the flaws bother you because it does not matter much. If you find any jewels in the dunghill, you can enjoy their beauty. Ignore the rest. --Fartherred (talk) 14:51, 23 September 2010 (UTC)

Liability issues?

I'd like to hear an answer for the record: If some anonymous brand-new account posts a libel against a BLP as a prank, and you confirm the changes, do you take on formal "joint and several" i.e. equal liability for everything they said? Wnt (talk) 04:54, 21 September 2010 (UTC)

Oh, and if so, is there any policy against someone just refusing to confirm any changes made as a matter of routine, to avoid taking on unexpected liability? Wnt (talk) 04:55, 21 September 2010 (UTC)

If you're okaying unsourced controversial BLP edits, what the hell are you doing deciding on it anyway, be it approving a {{editsemiprotected}} request or anything else? I don't know what the exact legal bit on it is, but you should be disapproving, reverting, whatever the individual case requires, those types of things. Seraphimblade ^{Talk to me} 04:58, 21 September 2010 (UTC)

Well, with PC it's a user, not an admin, who might be less skillful; and it's a simple button to accept a change, not an edit that you copy and paste and put in under your own name. I don't think it's obvious with the interface that you're assuming actual legal responsibility for the IP's edits — in fact, I'm not at all sure that you are. Some (many?) might say or think that confirming the edits is just a "quick check for vandalism", rather than a guarantee of perfect truth. Wnt (talk) 05:04, 21 September 2010 (UTC)

Initially, Pending Changes was considered a vandalism check and acceptance of an edit does not guarantee its accuracy. CycloneGU (talk) 05:34, 21 September 2010 (UTC)

It might be good to see that formally disclaimered someplace, for the benefit of all involved? Wnt (talk) 05:36, 21 September 2010 (UTC)

One would assume that refusing to confirm any changes would be acceptable; one doesn't have to be flagged with the reviewer status. As to the larger legal question, I imagine that any prosecution of a reviewer would end up being a totally novel court case. "Acceptance of an edit does not guarantee its accuracy" might not be enough cover. I would hope that if a reviewer was caught frequently confirming BLP violations, they'd lose access to the confirmation process. Archaeo (talk) 13:00, 21 September 2010 (UTC)

If no one has reviewer status, then every single non-admin edit would have to be checked. That would take us backwards. But yes, I do believe a reviewer accepting bad reviews continually can lose reviewer privileges. CycloneGU (talk) 13:27, 21 September 2010 (UTC)

The problem I'm seeing here is that it seems like Wikipedia could end up addicting itself to this process, to its great detriment. At first people use it simply as a "quick check for vandalism", the effort required is minimal, and the loss of useful content is small. But then someone starts a legal case, and suddenly a reviewer is getting sued for something that an IP out of an unrecognized open proxy said. Suddenly reviewers are being told, when you make this text publicly viewable, you're the publisher; you're fully responsible for any libel, any copyright violation, any inclusion of photos that dunceling prosecutors find to be obscene, etectera. They don't feel safe to OK any edit where they would have to walk to the library or go behind a paywall to read the source of it. And at that point, obviously, you have people finally saying no, this is a really terrible idea, let's end PC - but then the lawyers come out and say, wait, because you had this tool available and you didn't use it, that'd make Wikipedia liable for all the resulting liability claims. And then you're sunk - you start by losing anons, then new editors, then all editors. Wnt (talk) 15:10, 21 September 2010 (UTC)

Noting the reply already below here, which merits separate discussion, I am replying to Wnt. I do agree with you, and my stance as a reviewer is that if an edit cannot be properly sourced, then there is no confirmation that it's valid information and should be published in the first place. It goes back to verifiability. CycloneGU (talk) 16:23, 21 September 2010 (UTC)

This is a very good point, and needs answering. What advice has the Foundation received about the effects on editors' liability of Pending Changes? DuncanHill (talk) 15:48, 21 September 2010 (UTC)

This is in fact a serious problem. If someone edits a BLP and cites the information to a book (which is not online), how should a reviewer know whether or not to accept the edit? Sjakkalle (Check!) 16:27, 21 September 2010 (UTC)

None of the implications of Pending changes have been considered in sufficient detail. Not in regards to Wikipedia philosophy ("The encyclopedia anyone can edit"), nor to practicality (e.g., how many edits will appear & need to be approved in a given time slice, & whether there are enough reviewers to do the work), nor even to implied legal responsibility. We have a few people who fervently believe it is a panacea for all what ails Wikipedia, & a larger number of people who have put faith in their opinion -- as well as a number who oppose it on some equally irrational premise. But no one for this change seems at all interested in performing due diligence or even setting forth an argument for this. -- llywrch (talk) 16:44, 21 September 2010 (UTC)

I think they may have been, by at least some people. The fact that it seems to have been approved, and somewhat functional, elsewhere is one thing that has contributed to my own support. Factually, and maybe luckily, in my case, few if any of the articles I have on my watch list are apparently "tagged" for pending changes, so maybe I am luckily one of the least involved people and can thus pontificate to others about things that they should be doing which I myself don't have to do. I've always loved doing that sort of thing whenever the opportunity arises. ;) I am more than willing to adding to my watch list relevant articles which are included, however. Basically, I see it as a maybe final step toward ensuring that articles which are frequent targets of vandalism or other disruptive editing become articles which new or IP vandals can disrupt to their heart's content, because none of the vandalism will appear until someone else "approves" it. Personally, I would use it myself only on articles in which a living person has recently been involved in a substantial, potentially controversial, matter, has recently died, articles about organizations or groups in similar incidents, or maybe articles which, like maybe Satanism (and maybe periodically Santa Claus?) tend to attract disruptive editing on a predictable basis. I would love to see some sort of standard procedure involving hopefully multiple editors to approve articles for being included in the PC list, and would "lobby" for such where possible. Anyway, anyone is free to take this as arrogant, ill-informed pontificating by someone who himself has been lucky enough to have never actually dealt with the matter in practice. John Carter (talk) 16:56, 21 September 2010 (UTC)

I agree that any practical use has to be extremely limited, with rigorous standards for use. Llywrch, I can only assume you know better than I do to what degree the Foundation has considered the issue; has Mike Godwin weighed in on this liability issue? It seems to me that if the Foundation hasn't fully considered this, then it shouldn't be put to use. It almost looks like we're trying to find problems to use this technology on. Archaeo (talk) 22:08, 21 September 2010 (UTC)

I doubt I know any more about how much the Foundation -- or anyone -- has discussed the matter than you, Archaeo. All I know about any discussions on Pending Changes is what has appeared on Wikipedia -- although I have done my own pontificating on this proposed change elsewhere. After making repeated claims that no one has adequately discussed this change, the only counterproof I have been given was a single page -- Wikipedia:Pending changes/Metrics/Preliminary Analysis, which is little more than a bare beginning in considering the data. I only guess that either I'm being ignored because I'm a considered a troll, or that the discussions have never taken place. I don't know which I want to be true. -- llywrch (talk) 02:24, 22 September 2010 (UTC)

So is there an answer to this question? I find this to be very scary. There is no way that I will ever review again if there is a chance I am made liable for what someone else posted. In fact, now that I've seen this unanswered question, I will not review anything until there is a clear and proper answer stating that no liability will fall to the reviewer. HumphreyW (talk) 07:41, 24 September 2010 (UTC)

Is there no answer to this? Doesn't anyone know? HumphreyW (talk) 03:44, 3 October 2010 (UTC)

IANAL, and nor is most of the WP population. You'd have better luck contacting Godwin on his talk page about this, but I doubt we'll know the absolute answer until Miles Edgeworth, Prick Esq. decides to try and make a quick buck for Mr. Whiny Celebrity. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 05:00, 3 October 2010 (UTC)

There has a lot of high level discussion about this topic, along with in-depth discussion. The conclusion is that no one knows because there has never been a case. —UncleDouggie (talk) 05:21, 3 October 2010 (UTC)

Didn't I just say that? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 17:59, 15 October 2010 (UTC)

Poll opposition section

Please do not attempt a "closure" of this section. Everyone posting in it is well aware that it is not a counted section. It is still, however, a valid response, even if it will not affect the outcome. Seraphimblade ^{Talk to me} 05:14, 21 September 2010 (UTC)

I am aware of that, but I think most editors would appreciate it if the discussion took place on this talk page. Ronk01 talk 05:25, 21 September 2010 (UTC)

The fact remains that altering instructions in a poll, when a huge number of people have already voted and commented is incredibly bad practice. Doing it because you personally believe that "most editors would appreciate it" hardly seems to negate that. Nor, indeed, does altering it because of some possible discussion/agreement during the prior discussion on setting up the poll. I understand that you helped in the setting up of the poll, and so have some involvement, and I'm not trying to start a disagreement, but please just consider that altering the instructions during a poll, for whatever reason, even to correct what you believe it should have originally said could be seen as invalidating the poll for many people. Even if you don't believe your change would have altered the votes/comments preceding it, you have no way to be certain, since you have not asked them. If you think I'm creating an issue over nothing, then consider whether offering more ammunition to people who already think this process is a disaster is wise. Sorry for the long post, but I wanted to explain my take on it, rather than just saying "I feel this is wrong"  Begoon•talk 08:40, 21 September 2010 (UTC)

Support;)--intelati^(Call) 05:42, 21 September 2010 (UTC)

I support having the interim poll, but I also strongly support the right of others to express their opposition to having the poll at all. The answer to free speech is more free speech. --DThomsen8 (talk) 12:27, 21 September 2010 (UTC)

Not a problem. When I did all of that I had just finished a ten hour surgery, and I was just a little tired. This whole thing is just frustrating to me. I don't like the idea of a majority vote (check Jimbo's talk page, there was a thread where a I was arguing with him about that a few days ago) But if it keeps a valuable too alive for a few more weeks, with what seems like wide community support (As I write, the vote is 61%-39%, edging close to consensus). I especially objected to those who said that the initial poll, which only missed consensus by 0.03 percent "represented broad consensus to permanently shut down PC" which it clearly did not. Perhaps I am just frustrated to resistance to what is nothing more than a new, less strict form of page protection. It isn't going to mae the wheels fall off, and I certainly can't see how it could "make Wikipedia into [more of]] an Oligarchy" as once editor said. Again I understand those editors who want to close PC until the new trial, and respect those who simply want it gone, but I respect the rights of, but cannot comprehend the motivation (I will, however continue to assume good faith) those who oppose the poll. Ronk01 talk 12:52, 21 September 2010 (UTC)

At this point the issue, to many editors, is larger than just PC. I was never of the opinion that PC would make the wheels fall off. Also, we did not need broad consensus to turn off the two month trial. We needed broad consensus to keep it beyond 2 months, which we have still not achieved. And, once again, majority is not consensus.

I know that all of this has been said before. It makes me think that some editors simply will not ever see the points being made, or will continue to discount them as unimportant. However, it is galling for them to continue to profess ignorance of the opposition's viewpoint and motivation when it has been stated clearly many times. Revcasy (talk) 13:36, 21 September 2010 (UTC)

A reminder for those who cite WP:Consensus

An excerpt from WP:Consensus:

"Declarations from Jimbo Wales, the Wikimedia Foundation Board, or the Developers, particularly for copyright, legal issues, or server load, have policy status."

He can do it without consensus, I don't like it, you don't like it (I don't even think he likes it). But he can. Therefore, this poll is a manifestation of Jimbo's refusal to become fiat. Just a note. Ronk01 talk 05:22, 21 September 2010 (UTC)

Even though we support the same side, and I might just be starting a new thing to bicker about...how is Pending Changes related to copyright, legal issues, or server load? CycloneGU (talk) 05:26, 21 September 2010 (UTC)

Particularly does not mean only, he can make a declaration that every Wikipedia page has a duck on it and (provided the foundation didn't object) it would be policy. Though that never happens thankfully. Ronk01 talk 05:34, 21 September 2010 (UTC)

Duck

Touché, touché. And just for fun (and hopefully a little much-needed humour), let's add a duck. CycloneGU (talk) 05:38, 21 September 2010 (UTC)

But, not a quack :)--intelati^(Call) 05:40, 21 September 2010 (UTC)

(edit conflict) Theoretically, yes. In practice, Jimbo's been overruled by the community before, and it would probably happen in such a case as your hypothetical ducks. Generally speaking, in matters of things like copyright or legal issues, Jimbo or WMF receives extremely wide deference. And well that they should—in such a case, they probably know something we don't and that they can't freely discuss, and they have to take action to avoid placing the project in jeopardy. In contrast, when they try to make regular old policy, they can and have faced opposition. Wasn't too long ago that there was a massive incident over Jimbo unilaterally deleting a bunch of free images off of Commons, and that certainly did not end with the images remaining gone. Seraphimblade ^{Talk to me} 05:40, 21 September 2010 (UTC)

I think the problem with citing that exception is that Jimbo hasn't actually said he's overruling the community, and has said that this requires community consensus (and is obviously trying to make it look like that exists). - Kingpin¹³ (talk) 08:21, 21 September 2010 (UTC)

Wasn't that all repealed after the WP:COMMONS censorship situation? I vaguely recall Jimbo being asked to give up status. User A1 (talk) 08:47, 21 September 2010 (UTC)

Firstly: I know that this is more like a decision to override one policy, not creating another. Secondly, the WMF could overrule all of us. Thirdly I would point out that the poll is edging dangerously close to 66%, which was the consensus limit of the first poll. Ronk01 talk 12:41, 21 September 2010 (UTC)

Jimbo has said that consensus is important to him, on this subject, and that he thinks we need consensus to implement PC, so that he can technically override consensus is irrelevant, the foundation has also said that they will put as much effort into PC as they think is justified by the support it receives from the community, aka they won't be overriding the community in this either. So it's irrelevant that they can technically. I also don't see your'e point about this poll being almost 66%... so what? Sadly, I think this poll has been ruined too, by canvassing (see below), as well as that I'm still opposed to the idea of using polls at all, in particular using them repeatedly because one doesn't get the result wanted. - Kingpin¹³ (talk) 12:54, 21 September 2010 (UTC)

If you read my comment on my vote, I said that the WMF will not be putting much effort into PC if they see that it is not being actively used on-Wiki. That being said, I would object to automatic talk page notifications of all previously involved editors being characterized as canvassing. Canvassing is defined as an attempt to gather support for one side, not just telling people to vote (regardless of who is in the majority). Even if the last poll had resulted in a minority for PC, I would still want to see all of the other editors vote in this poll. (Which I don't like, but I also like PC, so it's the lesser of two evils). Ronk01 talk 13:09, 21 September 2010 (UTC)

I don't really think that it would make much difference if we are using PC up to November or not, as the foundation have set out a clear summary of the changes they are going to make, and when it will be done by. It's called votestacking, when you inform a certain group of people who's positions are already known. - Kingpin¹³ (talk) 13:22, 21 September 2010 (UTC)

200 users who were going to oppose the poll were also informed. I am however afraid of real offwiki canvassing, on both sides. In relation to the WMF, if they don't see that the community supports PC enough to keep it running, they might slow down on the development. Ronk01 talk 13:26, 21 September 2010 (UTC)

I realise a minority of opposition was also notified, now if only 200 supporters were notified too that's less suspect (although WP:CANVAS does warn against messaging large numbers of users). But to your most recent point, yes exactly. If people don't want PC enough, then WMF won't want PC enough, that's how this process should work. You are working on the assumption that PC must be implemented regardless, and if lacking support hinders the process it's a problem? Jebus989^✰ 13:36, 21 September 2010 (UTC)

If you look, most of the Close votes are in favor of PC, they just want to close until the trial. If you took all of those an put them in with the support column, that would be significant consensus in favor of PC. Most objections to PC in fact, are technical, not philosophical. Ronk01 talk 13:42, 21 September 2010 (UTC)

Ok - so now I really am lost. This is a poll on a technical consideration concerning the trial, yet we should move all the opposes that are "technical" to the support column (including mine - I would really hope not)? I thought you were in favour of just counting them - which would remove the possibility of even judging their merit in this rather odd way. Would we be moving the support votes that are "technical" to the oppose column, too, under this brand new approach to result determination?  Begoon•talk 17:21, 21 September 2010 (UTC)

No, I was not saying that they should be moved (in fact, I would be pretty peeved if someone did). I was just making the point that most users who oppose do so because of technical reasons. I was almost a technical oppose myself. Ronk01 talk 22:58, 21 September 2010 (UTC)

Ah, ok, thanks for explaining. It still doesn't make any sort of useful "point" to me to apply a parameter for switching the meaning of a vote from oppose to support when it matches a criteria, but not applying the same filter or consideration to support votes. Selectively filtering one of the two groups doesn't seem to achieve much, other than to bias the outcome one way. Still, it's your "point", so you can construct it however you like - it doesn't require me to understand it.  Begoon•talk 05:09, 22 September 2010 (UTC)

Perhaps the best way to explain it would be to say that only a small group of editors oppose PC on principle, most oppose it on grounds of policy (like this poll), or for the aforementioned technical reasons. Cheers! Ronk01 talk 18:38, 22 September 2010 (UTC)

Thank you again. Whilst there may be some truth in that, I still don't see how it's particularly valid to examine it this way. You seem to be neglecting the many votes which have said they support keeping the trial going without consensus (in some cases even acknowledging that it should really be discontinued, but for these technical reasons they support leaving it on) purely because it's "easier", or "difficult to turn off and back on again" (which, if true, is a design fault that needs to be added to the list for the next trial). They seem like pretty "technical reasons" to me, yet are unconsidered in this analysis. Sorry to keep replying here, but it just seems terribly wrong to me to construct a hypothesis that doesn't apply the same criteria to all the votes here, rather just those in one camp. It also seems a little odd that in most of the points you have made on this page, you advocate pure numerical vote counting without weighing the strength of votes to gain consensus, yet in this thought experiment you analyse the meaning of the votes if they can be moved in one direction, but not the other. I'm genuinely not trying to "get on your back" about this - I just find it a little confusing.  Begoon•talk 23:23, 22 September 2010 (UTC)

Yeah, I'm not so sure if we can rely on simple numerics anymore, since the replies are so diverse. But I do agree with you that many votes are based on technical reasons on both sides (and the whole thing about it being hard to turn off actually hasn't been fully investigated, but something tells me that we won't ever find out) so it is clear that a second trial would likely have significant community support if the proper fixes were made. I do think that there needs to be traditional consensus to open the second trial (I don't really think getting it would be a problem, as I mentioned above) though the disgust of some editors with the system is apparent. To continue the thought experiment, I wonder how the current support and oppose votes would look if all those who supported PC (not the poll, or keeping the trial running, but PC itself) were moved to the keep column, and those who oppose PC were moved to the close column? Thoughts? Ronk01 talk 00:38, 23 September 2010 (UTC)

Thanks. I'm glad you can see a little of where I'm coming from. The problem with continuing this particular thought experiment, from my point of view, is that it would just add to the confusion, and since I voted here partially because of that confusion, and its implication on long term support for PC, I'm loathe to add to it. There already seem to be a good number of people confused about the purpose of this poll, so I'll pass, if that's ok. I wouldn't have replied to the initial post, if it'd hadn't contained some assumptions which I thought it was important to clarify.  Begoon•talk 00:51, 23 September 2010 (UTC)

Unintended consequences?

I'm kind of wondering what the unintended fallout may be from this type of ramrodding. I'm seeing a lot of very upset people here, and I'm sure not real happy about the way it's been handled myself. Do people really think this is acceptable? What I can see happening, is that next time there's a request to trial something (even something worthwhile), a lot of people will be very gun shy about allowing it to even get off the ground. And they should be. A trial means a trial, and if there isn't a clear consensus to keep going with it, you cut it off as soon as the trial period's over. A majority is not the same thing as a consensus, and Jimbo, of all people, ought to know that. It really does feel to me like this is an attempt to railroad this in when it's clear it doesn't enjoy consensus. If Jimbo plans to put it in regardless of what anyone says, he should just say so, and spare us a lot of futility. If he doesn't, he needs to accept that "no consensus to continue the trial" means "the trial is discontinued at once", not "keep rephrasing the question until it returns the answer you agree with". I'm just wondering if anyone besides me will be very hesitant about trying any new features on a "trial" basis in the future. There's always the test wiki, and I think going forward, we might be better off to test things there and then see if there's consensus to implement here or not. But it's kind of a shame. A live test can give better data. But we must be able to trust that it is a test, and that absent exceptionally strong consensus, the test will be ended on schedule while we figure out where we go next. Seraphimblade ^{Talk to me} 08:06, 21 September 2010 (UTC)

You are right in that this will leave a significant number of editors behind who will oppose future trials out of fear of a repeat of this. Another consequence will be to serve as a(nother?) precedent for making decisions by majority vote instead of consensus. The question is whether that is unintended. --Xeeron (talk) 10:40, 21 September 2010 (UTC)

It has been made explicitly clear that this will not set precedent for future discussion. Remember, this is only a short extension of the trial until PC 2.0 is released. I don't like majority voting either, but I also don't like discontinuing a trial that has supprot that is only 0.03$ from consensus. Could you imagine what that many angry editors would be like? Ronk01 talk 13:01, 21 September 2010 (UTC)

Since when has consensus been measured in numbers? Consensus isn't something which is measurable - Kingpin¹³ (talk) 13:03, 21 September 2010 (UTC)

Consensus is defined by numbers. There are three levels of consensus that are used on Wikipedia: 66%, the lowest level, often used for very large proposals where higher levels would not be possible. 70%-80% is good enough on RfA (It depends on the biases of the closing Crat). 80-85% is consensus for a new policy, and consensus on the talk pages of most of the more controversial articles. Ronk01 talk 13:19, 21 September 2010 (UTC)

Well, isn't that really having it both ways? You either accept that consensus needs to be assessed by a neutral party or parties, giving appropriate weight to participants, and not just "weighing the numbers", or you don't. If you instead insist on a particular numerical result (on what basis, I'm not clear) then you cannot subsequently say "it nearly passed". It either reached your arbitrary number or it failed. Personally, I don't agree with your "counting" option at all, except that the closer can use it as a guide to the level of support required after assessing the strength or otherwise of the contributions, but supposing one were to accept that strict numerical method, then, by its very definition, the result is binary, yes or no. Since consensus was required to continue, in this case, and not the other way around, it was not reached.  Begoon•talk 13:39, 21 September 2010 (UTC)

Excellent point Begoon. Also, consensus is not the result of a vote in a straw poll. Consensus is an agreement about what to do, arrived at by discussion and compromise. I strongly disagree with the idea of redefining consensus in terms of percent majority in a vote. As stated above, consensus is not a numerical concept. It is not defined mathematically. Revcasy (talk) 13:45, 21 September 2010 (UTC)

Any form of support-oppose system can be defined mathematically, otherwise, it is not a support oppose system. I will repeat myself again- The first poll was near consensus, I would have been in favor of temporarily turning PC off then, because consensus was not reached (very close though). It was Jimbo's decision to keep PC active, and to repoll. I simply participated in the discussion to make the poll fair. I voted keep because I like PC. On the note of discussion: How do you plan on coordinating discussion between 600 editors! That is utterly impossible. That kind of consensus building approach is only appropriate for XfD and articles. If you oppose polling, then what do you call RfA? The kind of slow, methodical consensus building that was used back in 2005 is no longer possible on these massive discussions, Wikipedia is simply too large now. It's like a U.S. presidential election- there are so many opinionated people on both sides that massive agreement is impossible. (Look at Obama, he had unprecedented support, but his "landlside" was well under 60%) It is nothing more than a simple Anthropological fact: As a community grows larger, it is more and more likely to fractionate itself into opposing groups, making mass consensus impossible in whole-group decisions. We need to accept that, as a side effect of our massive growth, we've outgrown mass consensus in this type of discussion. Ronk01 talk 14:13, 21 September 2010 (UTC)

If WmF wants to keep it running, it's their website, they can do that. Editors can each make up their own minds how to deal with it for themselves. Is this being shoved through in hopes someone can claim consensus, maybe from a marketing/PR outlook? There are hints that's happening. Gwen Gale (talk) 14:27, 21 September 2010 (UTC)

I actually support this position. I was looking at the names and numbers of the users actually working with the tool yesterday and am in the process of creating a list of those users. People that don't support or don't like the tool are not even using it or reviewing anything so why there is so much complaining I have no idea. Personally I don't like huggle or rollback, so I don't use them. Off2riorob (talk) 14:46, 21 September 2010 (UTC)

err anyone who has had an edit tagged as pending is effectively using the tool. Will you create a list of those users or do they not matter?©Geni 23:15, 21 September 2010 (UTC)

Unconfirmed users are the contributors of the desired additions. As I see it as regards this discussion they do not require listing at all. If you see any use as regards listing their IP addresses or whatever please explain your reasons for that as I see nothing of value at all in doing that.Off2riorob (talk) 23:22, 21 September 2010 (UTC)

I removed pending changes protection from several articles that I watch over because it was interfering with my ability to block and detect sockpuppets. Does that make me a "non-user"? Does disabling the feature on articles that it was causing problems on make my opposition to it invalid or of less merit than someone elses support?—Kww(talk) 23:28, 21 September 2010 (UTC)

I don't know your involvement, was it just to remove the tool or did you actually do any reviewing? Your edit history appears that you didn't like it and that you removed it from multiple articles. Did you do any reviewing or was it just the removals? Off2riorob (talk) 23:34, 21 September 2010 (UTC)

I did both, but how does that matter? If the only thing I had done was figure out that people were approving edits from socks while permitting them to edit anonymously, why would that make my participation less valid?—Kww(talk) 23:46, 21 September 2010 (UTC)

You are the only one that opines your comments are not valid. Did you review many additions? Off2riorob (talk) 23:49, 21 September 2010 (UTC)

"People that don't support or don't like the tool are not even using it or reviewing anything so why there is so much complaining I have no idea." would imply that people that don't use it frequently shouldn't have a strong opinion. I reviewed every addition that showed up on my 11,176-article watchlist. Didn't seek it out, but didn't avoid it either.—Kww(talk) 23:54, 21 September 2010 (UTC)

Perhaps you would have had a better experience if you had tried the tool as an ordinary editor, forget the socks, poor edits are mostly totally reverted under pending. As I experienced it reviewers also improved as they began to understand and see the tool in use.Off2riorob (talk) 23:49, 21 September 2010 (UTC)

Preventing sockpuppets from editing is one of my major focuses, and I think that anything that makes it easier for them to edit is inherently bad.—Kww(talk) 00:00, 22 September 2010 (UTC)

As I said, perhaps you need a break from that, sock puppets are everywhere here, personally if they add a decent edit cited I am not much bothered about that, my focus as regards this tool is the founding goals and the future, pending allows contributions from unconfirmed users to a page that would otherwise be semi protected long term, this supports a benefit to free to use, not perfect but better. The tool has many benefits in multiple locations, in others it is a nuisance. Also as I have hundreds of reviews on multiple articles and can speak truly from experience. Off2riorob (talk) 00:07, 22 September 2010 (UTC)

And that's exactly the problem: people reviewed edits and evaluated whether they were good edits, but made no effort whatsoever to determine whether the edit was from a sock or not. Just because you don't care whether socks edit or not doesn't make socks editing acceptable: I block them whenever I detect them, and spend considerable time reverting and undoing each and every edit they make that I can detect. PC makes my tasks harder, and thus I oppose its deployment. The whole feature is based on the erroneous assumption that there is something desirable about it being difficult to associate edits with the individual editor that made them. There's nothing inherently bad about long-term semi-protection. In fact, I would argue that it should be the default state for most articles. Perhaps you should spend a year or two working at WP:SPI and your perspective would change.—Kww(talk) 00:16, 22 September 2010 (UTC)

If a sock isn't interfering with discussion or consensus building, let them edit! (unless they are evading a block). Simply put, going on sock hunts is often counter-productive. I've dealt with real socks, you know the ones that actually cause problems by interfering with the community, and I would rather that admins spend their time going after them, and not reverting the positive edits of non-problematic socks. If a sock is causing problems, block it, if it isn't leave it alone and do something worthwhile, like preventing AIV backlogs . As for your assertion that semi-protection should be the default state of most articles, what happened to "the Encyclopedia that Anyone can edit?"Ronk01 talk 03:23, 22 September 2010 (UTC)

I'm sorry that you think that the socks I seek out are somehow harmless. Editors that have to put up with Brexx, ItHysteria, Wiki-11233, ItHysteria, Xtinadbest, and others would think otherwise. Socking is the single most destructive thing an editor can do, and is never harmless.—Kww(talk) 03:51, 22 September 2010 (UTC)

I know you are in the midst of a War against socks, but consider: There are far worse things than socking. Vandalism, perhaps. Vandalism with socking? I know how problematic socking can be, but is there really a need to hunt down every edit they make? (even if they are positive?) I'm not saying that socking is harmless (I've had a few problems with them myself,) but don't oppose a useful tool that would open up pages to good faith IP's just because it makes your work of reverting positive edits harder. Ronk01 talk 04:13, 22 September 2010 (UTC)

Ronk, I've done sockpuppet work myself, and I'll tell you right now, socking is extremely disruptive. There's a very good reason it's looked upon with such extreme disfavor here. Simple vandalism, or even creating socks to vandalize, actually isn't so bad—blatant vandalism is caught and reverted quite literally every minute around here. It's the socks that continually make the same poor but not quite vandal edits, and suck up everyone's time by repeatedly "not getting it", that are a real drain. Or like one I chased, Runcorn, who abused socks (including an admin account) over a period of years to subtly introduce bias to a good number of articles. These are not "positive" contributors, even if they make themselves out to be, and rooting out abusive socks is in no way something to be vilified, as you seem to be asserting here. Seraphimblade ^{Talk to me} 04:38, 22 September 2010 (UTC)

I am most certainly not glorifying socks, or vilifying SPI, I've dealt with Jossi socks, I know how bad it can be. But basing opposition to PC because it might make it a bit harder to revert sock edits? Come on, give me a better reason. (by the way, vandal socks are a huge problem, think Grawp) Ronk01 talk 04:48, 22 September 2010 (UTC)

If you're asserting that it's probably not the strongest argument there is against it, I'd probably tend to agree. I was just a bit taken aback by what seemed to be an equating of "rooting out socks" and "reverting positive edits". And yes, I know vandal socks can be interesting ones, especially when they IP hop—more brings to mind Ararat arev and Ockenbock for me, but they're just a pain. The real nasty ones are the subtle types. Seraphimblade ^{Talk to me} 05:01, 22 September 2010 (UTC)

I disagree. There's nothing worse than socking.—Kww(talk) 05:10, 22 September 2010 (UTC)

Politely, you're all missing the point. Kww said: "The whole feature is based on the erroneous assumption that there is something desirable about it being difficult to associate edits with the individual editor that made them." This is an interface problem, a design issue that can be fixed in a new UI. The question is: how can we design pending changes so that it doesn't interfere with attempts to catch sock-puppets? Let's ask the developers and come up with some ideas. Ocaasi 13:47, 22 September 2010 (UTC)

Bluntly, there is no way. Murphy's Law in regards to all forms of Wikipedia vandalism-preventing/-masking has never been disproven - and thanks to User:NawlinWiki and Park51, the same is true of Pending Changes. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:02, 23 September 2010 (UTC)

Canvassing?

Very rough graph showing votes over time, please feel free to make a better looking one! ;).

Hi, I've just drawn up a really rough graph trying to show the votes over time, along with the EdwardsBot delivery (initiated by Off2riorob, see the discussion at Jimbo's talk page). The graph shows that after the delivery there was a large increase in votes, with a larger increase in the votes for option 1, to be expected since about double the people notified previously supported keeping PC in some form. However, just before the delivery it also looked like option 1 might just be starting to gain a bit more support. Obviously this graph is very rough, and again, I don't guarantee it's right, but it does show it might be worth somebody's time having a proper look at this. Thoughts? - Kingpin¹³ (talk) 11:30, 21 September 2010 (UTC)

Actually, I think your graph is rather well done, and helpful for the discussion. I had voted Keep before the bot posted on my talk page, but then again, I am a PC reviewer interested in the process.--DThomsen8 (talk) 12:23, 21 September 2010 (UTC)

Thanks :). Also, taking a look at the more recent voters, nearly all of them (certainly in the support column, which is where most of the new ones are going) have received the message from EdwardsBot. - Kingpin¹³ (talk) 12:39, 21 September 2010 (UTC)

Thanks for doing this. The fact that the message originated from the designer of the first poll and a supporter of PC should also be taken into consideration Jebus989^✰ 12:45, 21 September 2010 (UTC)

Consider also that the majority of voters who voted in the first poll supported PC, so I'm not surprised that the majority of the editors brought here by the bot support PC. Ronk01 talk 12:56, 21 September 2010 (UTC)

Yes, but that's exactly the point, the bot notified ~400 editors supporting PC, and only ~200 opposing it, doesn't really seem fair to me - Kingpin¹³ (talk) 12:57, 21 September 2010 (UTC)

Indeed. Given the atmosphere surrounding these polls, the notification probably should have been done by an uninvolved admin at the very least. I also happen to think a talk page notification was unnecessary. Revcasy (talk) 12:58, 21 September 2010 (UTC)

Well put, Jimbo's talk space and 3 involved editors was not enough of a forum. I left the discussion there when Off2riorob agreed and said

Yes, perhaps we don't need to bother notifying after all, seems to be well linked up I appreciate and accept your points about the previous poll, thanks Off2riorob (talk) 19:32, 20 September 2010 (UTC)

Jebus989^✰ 13:09, 21 September 2010 (UTC)

Kingpin13If the previous poll had 400 users supporting and 200 against, then what else would you expect? The message could have been delivered only to the 200 people who did not support and not a single one to support. Then would that haver been fair? The message was sent to ALL VOTERS in the previous poll. No exceptions. And even here we still have 17 abstain votes, which is a section that should not even exist; this is Jimbo's poll, after all. Also, I gave the message in a previous discussion for use in such a delivery, and it was 100% neutral; I was surprised when I received it myself as I was not aware the delivery had been permitted or agreed to by Jimbo yet.

Now, if you find off-Wiki canvassing for either side, that is a problem. My worry is that this might exist. CycloneGU (talk) 13:18, 21 September 2010 (UTC)

Yes I expected it to cause an increase in the supporters, how exactly does that make it okay? What would have been fair is to not deliver any messages. While it's sometimes okay to leave messages to people who have previously voted, in this case it's clearly inappropriate, as the numbers are so large, and as shown by the graph the delivery caused a change to what was previously an unbiased sample. - Kingpin¹³ (talk) 13:32, 21 September 2010 (UTC)

Just to clarify (I'm doing a lot of that lately...), we did not canvass any users to the original poll. Likewise, we have not canvassed any users to this poll either; simply notified those who voted at the old poll of the new one. We simply sent a message to people who came, on their own, to the previous poll. This was done to ensure that they got the opportunity to have their vote count and not accidentally think they have already participated in the current poll. Arguing that it's better not to notify them is akin to saying you do not want then to speak up a second time. CycloneGU (talk) 13:40, 21 September 2010 (UTC)

I'm not sure who you mean by we, but Off2riorob clearly says on Jimbo's page that he notified people in the same manner for the previous poll. That said, I really don't think that canvassing is a problem. I'd expect that most people would have found their way here through the watchlist notice. --Onorem♠Dil 13:47, 21 September 2010 (UTC)

CycloneGU, you realise that this is almost the exact same discussion we had yesterday? Now that two completely independent parties are raising all the same points maybe you should accept they have some validity. It's interesting you state they came on their own to the first poll (except for a couple of TP messages), but that the same global watchlist notification was not enough to make them partake again? No-one is saying people shouldn't be notified, just in a fair and balanced way like the same watchlist notification to everyone, which was already done in this case. Looking at the keep reasons, many people who have been notified don't seem to realise that this is a different poll, their keep reasons are near identical without realising this is about closing the trial period rather than the merits of implementing PC Jebus989^✰ 13:52, 21 September 2010 (UTC)

Announcements on Watchlist, WP:CENT, AN and the VPP, and the Pending Changes talk page, should have been enough, but if anyone felt they were insufficient, further notifications should also have been universalist (i.e in public fora, rather than to the talk pages of selected individuals). DuncanHill (talk) 13:56, 21 September 2010 (UTC)

The bot sent messages to a skewed sample, those who had voted in the last poll. It wasn't the most blatant canvassing I've seen here, but I think it was canvassing. Gwen Gale (talk) 14:21, 21 September 2010 (UTC)

Jebus989, while I assume good faith, I am appalled at your comments above.

• "Now that two completely independent parties are raising all the same points maybe you should accept they have some validity." - I never denied your points having validity.
• "...the same global watchlist notification was not enough to make them partake again" - technically, it is but it also isn't, why not notify all interested parties? There was no exclusion of either side of votes. If the message was only sent to supporters or only to opposers, then there indeed would be a canvassing claim as mentioned earlier.
• "...their keep reasons are near identical without realising this is about closing the trial period rather than the merits of implementing PC." - I've also looked at oppose reasons. Many still comment that PC should be shut down permanently, which is not the point of the poll at all - it's only a discussion of the next six or seven weeks.

I can't help but feel you are viciously attacking my views here, but in particular, trying to suggest my points have no merit because two people think otherwise. That is a ghastly thing to do. CycloneGU (talk) 16:16, 21 September 2010 (UTC)

Ii don't see anything wrong with notifying the people that were involved previously, as the two polls are closely related and so close together it would seem a shame to me that users who expressed a fair bit of interest should not be notified that a connected poll is underway, in fact it seems almost necessary to notify them as this poll is so close to the last and so connected to the original. The comment posted was totally neutral and was sent to all users that commented on the previous straw poll page. Off2riorob (talk) 14:33, 21 September 2010 (UTC)

Do you think it's fair that one user can have such an impact (as shown in the graph) on a decision like this? Also, the polls aren't really that similar, as mentioned above a lot of the supporters seem to think this is about PC in the long run (as the previous poll seemed to be), rather than over the next few months. - Kingpin¹³ (talk) 14:36, 21 September 2010 (UTC)

(e/c)Of course those who voted in the previous poll, and who have given a fair amount of thought to this process already, should have been notified. It's not canvassing, it's simply informing people who have previously invested some time and energy into it. This isn't uncommon, unfair, or devious, or canvassing. Like some others, I had already voted by the time I received the bot notice. But I certainly appreciated knowing that editors are notified when something they are directly involved with comes up for a poll, and would hope this sort of thing will continue. First Light (talk) 14:37, 21 September 2010 (UTC)

Personally although the graph is interesting I think to assert too much into the small increase in polling numbers is applying undue assumptions. Off2riorob (talk) 14:39, 21 September 2010 (UTC)

With respect, you performed the actions we're debating, so the fact 'you don't see anything wrong' is no wholly a useful addition. And the graph is a straightforward case of cause and effect, with no other notable variables having been changed in this time period. Since you're involving yourself, following the discussion yesterday, why did you say you say "perhaps we don't need to bother notifying after all, seems to be well linked up" and then some hours later silently invoke a bot? Why was your username appended the the messages rather than EdwardsBot? The signature of an outspoken and heavily involved member with a strong opinion in favour of PC does not add to the neutrality of the message Jebus989^✰ 15:02, 21 September 2010 (UTC)

Well the graph is basically a straight line imo. I had users commenting in support of notification and one user, saying against, none of which imo were close to a strength of objection equaling do not notify, as it was I had as you say come to a point of no objection to not running the bot but as I said there was actually very little objection to the notification. As for my signature on the message, I am not an expert I am a volunteer, the thought that my involvement will have any net acceptance gain is highly improbable indeed, in fact I could easily make a case for the opposite being more likely. Perhaps for next time I will ask to have the bot sign its own message, all of this is a sideshow, this is not about me but about the tool. Off2riorob (talk) 15:12, 21 September 2010 (UTC)

Aside from your unbeleivable graph comment, there was 1 user in favour of yoursuggestion of notifying by talkpage (CycloneGU) and 1 opposed (me). There was an equal amount of both support and objection, and you went ahead and did it anyway, after saying you wouldn't. It is baffling Jebus989^✰ 15:22, 21 September 2010 (UTC)

Jimmy commented in support of broad notification here also NuclearWarfare commented here in support of notification and another uninvolved user ran the bot after my request. Off2riorob (talk) 15:53, 21 September 2010 (UTC)

All Reviewers should also be notified about this on their talk pages. The more informed opinions we can get on this the better, which is why the first bot-notice was such a good thing. First Light (talk) 15:29, 21 September 2010 (UTC)

Throw in all autopatrollers and rollbackers too. As it is, the assumption that people are incapable of seeing a watchlist notice, a post at AN, a post at the Village Pump/Policy, the WP:CENT notice and a post on the Pending Changes talk page is a bit rich. Why not everyone who commented on the original proposal which came up with the two month trial? DuncanHill (talk) 15:33, 21 September 2010 (UTC)

I agree, why not? Let's throw Users in while we're at it. The reason I mentioned Reviewers is that they all had direct experience of the pending changes process, and would give more of an informed opinion. First Light (talk) 20:01, 21 September 2010 (UTC)

A point that's being missed so far is that the people who want PC shut down are in general much more vocal and angry about it than those who generally support it! In both this straw poll and the last, much of the support votes were of a "this seems like a good idea, why not?" ilk, while the oppose votes have been far more passionate. They are therefore more likely to contribute disproportionately to the debate in a way that belies their actual numbers - for instance, looking at the breakdown of editors on the talk page of the first straw poll, one might easily be mistaken into thinking that the vast bulk of people were fiercely opposed to PC and would take on anyone to make sure their message got heard. (I think feeling in a minority, or not being listened to, tends to get people galvanized into action; a few months before this trial, there was a pro-PC burst of frustration from editors who felt like the promise of PC being delivered to enwiki would not be delivered by those on high and the technical staff had been ignoring their pleas, and they kicked up a very vociferous protest themselves. Now a form of PC has been delivered, they've calmed down, and it's those who dislike the dawn of PC, and feel unlistened to, that have got their tempers up.) It's unsurprising that people who felt betrayed by the last poll were exactly the kind of people likely to comment first when given a fresh opportunity to air their grievances here. They were in less need of a prod than the "I quite like it" support voters, who therefore benefited disproportionately from the mass mailing.

I'm wary of calling this "canvassing" (for the sake of appearances it would be better for such messages to be mailed by someone seen as unaffiliated); the message was genuinely neutral in wording and targeting, and it's no surprise that a disproportionate amount of "oppose" voters got their say in first (bear in mind that many of them had hung around on the talk pages in disgust whereas many "supporters" at the last poll voted only in passing; information about the new poll would therefore have been spotted first by those in opposition, and being more fired up, they needed less of a prod to comment anyway). An apposite question is: in what way could the final results of this poll have been different, had the mailing not taken place? I think the objective answer to this, is that the only way this could have ended up in an "oppose" majority had the mailing not been sent out, is if the previous "support" voters had not noticed this poll and therefore not commented here. Even if you're opposed to PC, you have to admit that that's a pretty shoddy way to win a poll. The mass mailing gave everybody who had previously expressed an interest in the matter, an equal opportunity to be heard here too. There's no point crying foul that there are more supporters than opposers - that's just the reality of the balance of numbers, and had the poll ended more evenly simply because a lot of people who would have liked to express an opinion, weren't informed that they had the opportunity, that wouldn't have been any fairer. The thing that is worth crying foul about is the way that the numerical balance is translated into a result: is consensus broken? Particularly on controversial binary issues where a middle-ground compromise is (physically/technically) difficult to obtain? Are we approaching a "majoritarian dictatorship"? How can the views of large minorities be best respected? Those are good, tough questions which deserve an answer. "Wouldn't it have been lovely if a whole bunch of people interested in this topic but who disagree with me, had not known about this poll, thereby allowing my side to win?" (I'm paraphrasing, don't actually think that anyone here is really asking this, but it's worth being wary of unintentionally sounding like it!) is not such a great question, and can't form the basis of a fairer settlement between the opposing views. TheGrappler (talk) 15:32, 21 September 2010 (UTC)

That is a very even-handed analysis. I too am wary of calling it canvasing, though I oppose PC (slightly), and the poll (strongly). "...and can't form the basis of a fairer settlement between the opposing views," unfortunately it does not seem that a fair settlement between the opposing views is on the table, only a simple majority wins here, because this is not a discussion, an effort at consensus, or any such thing, but a vote. Revcasy (talk) 15:47, 21 September 2010 (UTC)

I agree with much of what you say, but your statement "and it's those who dislike the dawn of PC, and feel unlistened to, that have got their tempers up." ignores a good number of people who have said things similar to my comment in my vote. I support PC, I am an active reviewer, and I am in favour of another trial when the agreed issues are fixed. The reason I opposed this is because running one vote after another with ever diminishing hurdles, which is how this is perceived by some, is to a large degree amplifying the anger and distrust you describe. A long term view would imo be to concede that there is significant opposition here and regroup prior to a new trial. The potential damage to long term perception of PC is far more important than whether a trial is kept running for a few weeks without consensus. That's my opinion, and it comes from no sense of anger, or resentment, purely a fear that in order to achieve a short term goal of little consequence, we risk sacrificing an amount of long term trust and support for the entire concept.  Begoon•talk 15:59, 21 September 2010 (UTC)

Yes, it was oversimplification not to mention people like you, but I think it's fair to say that most of the anger is coming from people opposed to PC, unless I'm reading this all wrong? A few months back there was a lot of anger and frustration from pro-PC people who felt that the developers had been ignoring them, and now it mostly seems to be coming from those who are opposed to it. There are also folk like you who like PC but are worried about the community-split aspect, and on the other hand, some previously opposed to PC but who feel that now it's better to accept it rather than make a running battle out of it. But I think those are minorities; a fair assessment seems to be that the real passion or "debate heat" (not that this is either a good or bad thing) is largely being generated at the moment by those opposers who feel let down by the last poll. TheGrappler (talk) 22:54, 22 September 2010 (UTC)

Well, again, I almost agree :-) I think that many of the votes here are indeed being cast on purely pro/con PC terms. That's not the question being asked. I can also understand why it is being done in that way. Human nature, for one, and confusion over why the poll exists for another. In all honesty it was the utter confusion and anger that seems present that caused me even to vote at all. When something is causing such a rift then it's time to stop, and assess your next steps carefully, not add to the confusion and anger by being widely considered (rightly or wrongly) to be continually altering the "rules" to obtain the ability to push on regardless. I'm certainly not saying there is any sort of "grand conspiracy" or bad intention - just that once there is so much demonstrated confusion and anger over a, (probably unimportant) short term goal, it's better to take a longer view. I fear that PC, whilst quite rightly eventually implemented, will lose significant long term support from all this nonsense, and, in the end, that's just damn silly and disappointing.  Begoon•talk 23:58, 22 September 2010 (UTC)

In your first paragraph you come dangerously close to describing true consensus as opposed to vote counting. The uninterested "this seems like a good idea, why not?" votes should rightly be weighted less than the lengthy, explanatory arguments from involved parties who have a greater knowledge and experience of the debate (this vote is an exception having been stated to be won by a numerical majority rather than consensus). Your later 'paraphrasing' falls into what wikipedians like to constantly refer as as a straw man argument. No-one is trying to hide this poll from anyone, there is no comment on this page implying anything of the sort. I, for one, would have wholeheartedly backed watchlist notifications (though this was rightfully done anyway) and even talk page messages to all active account holders (which would never happen for obvious reasons). That's honest, unbiased notification. If you want yet another rhetorical question to balance this argument: "Would pro-PC user x have been as keen to message a 400:200 majority who would knowingly oppose his own opinion rather than reinforce it?" Jebus989^✰ 18:42, 21 September 2010 (UTC)

That raises some interesting questions, though I actually think "consensus" is a little more complicated and I very strongly disagree that the argument is a straw man. One issues is how to go about vote-weighting - you seem to conflate passion/strength of feeling/commitment to an issue with the opinion being better informed/better reasoned. But that ain't necessarily so. It's certainly true that a lot of "keep" voters in the last poll were more casual: they didn't engage with the talk page discussions (dominated by "opposers"; a similar phenomenon seems to be happening on this talk page), they didn't feel particularly fired up about it, they seemed to think it was a pretty good idea that had worked well overall and didn't feel the need to write a screed in defence of it (perhaps partly because they're in a clear majority in the first place; if most people seem to agree with you to start with, then why do you need to defend your opinion?). Also worth pointing out that in a straw poll people aren't expected to write a defence of their position; simple statement of preference is all that a straw poll is calling for, and such votes shouldn't be ignored entirely even though no argument has been put forward. (In my opinion that's actually a good reason not to have vote-like "straw polls" at all; perhaps "tightly-structured discussions" would be better, more like RFA, or a sectioned-by-!vote-type AFD would be, where quality of argument is more strongly taken into account.) On the other hand, people who feel more fired up and passionate, and also feel in a minority, are more likely to write longer explanations - if you recognize you're in a losing/minority/ignored position, it pays to make the extra effort to be more persuasive, and also because you have more of an emotional engagement with the issue, you feel like you have more of a point you want to communicate anyway. You can still be well-informed about an issue and fairly casual about it, if it's just not a very big deal to you. And "looked okay to me, on reflection" is a perfectly valid argument, just a compressed one: if pressed for more details, I'm sure such a voter could provide a more detailed rationale, but being in a majority they may not have felt they really needed to defend the position. Contrast this to the time when PC looked like it would never come: people who felt like the devs and WMF had abandoned them, voiced detailed and vocal concerns about the damage that the lack of PC was wreaking on Wikipedia, in their opinion - so the pro-PC crowd are certainly capable of making both detailed and passionate arguments, but in the last poll most people don't seem to have felt the need to do so. Weighing up strength of arguments is hard: even if a minority have objectively stronger arguments than the majority, one wouldn't say that the minority represent consensus. But separating strength of opinion from strength of argument is also tricky. I do think there's an argument that a particularly passionately held opinion might need to be weighted more in the interests of community cohesion (especially if it represents a large minority chunk rather than a fringe one) but that's community-minded pragmatism speaking, rather than a philosophical quest for the true meaning of the word "consensus".

As for the mass mailing: it's a great shame the question of who would have been more likely to do it and in what circumstances even arises. It should have been done by somebody neutral and previously uninvolved. In fact in my opinion it should have been done automatically by a bot (and I wish it happened for repeat AFD and RFAs too). As far as I'm concerned, everybody who has taken the time and effort to participate in a poll or similar discussion, should be personally invited to participate in any follow-up or repeat. You can claim all you like that this page wasn't being hidden, but the graph tells you everything you need to know: watchlist and WP:CENT notifications, though valuable and important, are simply nowhere near as effective as actually telling people that a follow-up discussion is taking place. Failure to mass-mail is a failure to put the page on full view to precisely the group of people the poll is targeted at (i.e. folk with an interest in PC implementation); remember that not all editors use their watchlist or keep an eye on WP:CENT. If it attracts previous supporters and opposers in a ratio that is not 50:50 then that's fine: it will attract them in whatever ratio the subsection of the population that expressed prior interest in the issue held (in this case a clear but not overwhelming majority of supporters). That's not disproportionate which is why I think it's unfair to call it "canvassing". Had the majority here voted "oppose" simply because several hundred previously interested supporters were not informed about it, it wouldn't have been any fairer, it would just have reflected the fact that many of the opposers were more engaged with and upset about the issue and had stayed in touch with it, whereas the "support" voters were largely more casual. How "consensus" works when you have one group of people who feel more passionate about an issue than the other, I don't know: it reminds me a lot of the Ireland naming discussion where the majority of the interested "hardcore" were strongly opposed to "Republic of Ireland" as an article name, yet the majority (of mostly casual, pass-by but not necessarily uninformed or stupid) voters preferred that form. That also produced a lot of bad blood.

I don't know what the answer is on this, but I do know that a lot of our discussions about what "consensus" means are very naive, imagining that we simply weigh up both the relative numerical and argumentative strength of opinions, when people's depth or strength of feeling about an issue has such a big impact on how the community functions, and also on both whether people vote/participate in a discussion and the effort they go to when phrasing their arguments (a longer, more passionate argument may not actually reveal a rationally stronger case, but may be an indication of negative effects on the community if that view is not incorporated in a final solution). And aside from "how we weight up consensus", the "how do we incorporate minority views" is something we're struggling with here. Sometimes on a controversial new initiative like speedy deletion (which put a lot of power into the hands of admins, and scared a lot of inclusionists), compromise could be built around making the terms of its use very limited, so even non-admin inclusionists could "consent" to its introduction, even if they didn't particularly "like" it. (In the most difficult scenario, is this ultimately what "consensus" means? A compromise that people may not like but can consent to living under, rather than quitting wiki?) But in some cases like "Should we have PC?" and "Should we have an article at Republic of Ireland?", how do people compromise? I'm asking this as a serious question, not a rhetorical one. For physical/technical reasons, those are binary, yes/no questions, so what is the model for compromise-building if there's no stool in between? Do these questions require a completely different form of consensus-building process? Those are really serious questions (certainly not rhetorical ones!) and I haven't seen anybody yet come up with an answer to them. TheGrappler (talk) 22:54, 22 September 2010 (UTC)

Not appropriate for PC?

I am a reviewer for PC, but I am rarely able to find an article to review. I did see one particular case, where the historical subject is highly controversial among certain editors, resulting in many changes and sometimes edit warring. My opinion would be that this article is not a good candidate for PC, but I am not sure how to express that to the powers that be. Where can reviewers, or other editors, suggest that certain articles should not be in the pending changes list? --DThomsen8 (talk) 12:20, 21 September 2010 (UTC)

You can request that protection levels or types be changed at WP:RFPP. Hut 8.5 12:27, 21 September 2010 (UTC)

I have decided to do nothing about the particular situation I saw early today. --DThomsen8 (talk) 21:19, 21 September 2010 (UTC)

An excerpt from WP:POLL

Wikipedia works by building consensus, generally formed on talk pages or central discussion forums. Polling forms an integral part of several processes, e.g. WP:AFD; in other processes, e.g. article editing, polls are generally not used. In both cases, consensus is an inherent part of a wiki process. When conflicts arise, they are resolved through discussion, debate and collaboration. Polling, while not forbidden, should be used with care, if at all, and alternatives should be considered. In addition, even in cases that appear to be "votes", few decisions on Wikipedia are made on a "majority rule" basis, because Wikipedia is not a democracy. Apart from that, on occasion, "higher" bodies (e.g. the Arbitration Committee, Board of Trustees, or Jimbo Wales) can impose decisions regardless of consensus. Ronk01 talk 14:31, 21 September 2010 (UTC)

Yes, it would be fine it Jimbo just came straight up and said that he was overriding consensus. But he hasn't, he keeps trying to hide behind consensus. Calling what is happening here consensus is disgusting. Please admit that this isn't consensus decision making and stop trying to class it as such. - Kingpin¹³ (talk) 15:47, 21 September 2010 (UTC)

Its not disgusting at all in any way. A blind person could see that there is basically a level of support for the ongoing trial and improvement of this simple not wheels dropping off tool. Off2riorob (talk) 16:29, 21 September 2010 (UTC)

"Basically a level of support" is a long way from consensus. Supporting continuation of this tool at this level is supporting convenience over ethics. The trial ran two months, there is no consensus to continue, and running poll after poll after poll in an effort to drag a failed trial on indefinitely is disgusting.—Kww(talk) 17:02, 21 September 2010 (UTC)

An important reason voting doesn't work for online communities is that the losing side may not consider it binding. Or they may react by voting with their feet -- that is, leaving. Or they may remain, & make the working environment less pleasant, despite their best intentions, because they feel their opinions no longer matter & act accordingly. Building consensuses works, at least in theory, because everyone feels they have been heard & thus have some share in the result. Jimmy Wales is so convinced that this change is needed that he hasn't seriously listened or considered any objection to it; Wikipedia's going to get it, it's not going to be the panacea he has been quoted it would be,[1] & not only will volunteers leave but it will be harder to attract new ones. -- llywrch (talk) 17:11, 21 September 2010 (UTC)

Indeed, or an equal possibility is that active reviewers will just stop looking at the PC outstanding page since they have become disillusioned with the feature. I admit that's an option I'm considering for the first time now after seeing the nonsensical, ever shifting ways to attempt to justify ignoring the lack of consensus here. And I'm upset about that, because it's a good concept, and I'm certain will work well in the long run - but I can't sanction the refusal to consider the long term implications of this ugly squabble on ongoing perception of the feature. Follow consensus, or admit you're not prepared to, and implement it regardless - either of those would be (at different levels) far preferable to this fiasco.  Begoon•talk 17:39, 21 September 2010 (UTC)

• Please note that, in the previous poll, there was a two-thirds majority (Approximately, since it was ~ 65% to 35%) that supported the use of PC. This sort of majority works the same way as consensus in most ruling bodies, America's own Senate works the same way. Thus, there was, most definitely, support for the continued use of PC. Silverseren^C 17:45, 21 September 2010 (UTC)
  • Are you referring to the poll that had 3 options for continuing with some sort of PC and 1 option for disabling it entirely? --Onorem♠Dil 18:02, 21 September 2010 (UTC)

• Please note that the original statement required 2/3 support, or 66.67%, not 65%. Note that even that threshold was jiggered low in advance: RFAs normally require 70%, RFBs 80%. Virtually nowhere is 65% considered a consensus of any kind. Supporters set the threshold low, failed to meet that, and, rather than doing the honorable thing, continue to set new and lower standards. This poll, set in advance to succeed at bare majority, is a classic example of setting new standards to justify something they were determined to do anyway.—Kww(talk) 18:05, 21 September 2010 (UTC)

Please also note that of the four options in the previous poll, the one which received the most votes was close Jebus989^✰

And whilst we're noting things, please also note that it didn't achieve the level you suggest. I'm fairly certain that the US Senate doesn't very often look at the results of a vote that fails to meet the threshold and say "Oh, ok, that was pretty close, so we'll just pass it anyway". I also utterly disagree that selecting one legislative body's voting threshold amongst many is at all appropriate for the kind of community consensus traditionally sought here, even if, all of a sudden, consensus has started being a straight number count. If it has, I missed the memo. If it helps you though, you could check the UK out - I think that's still simple majority, so might give you even more leeway to pursue this line of thought. Or would that just be silly?  Begoon•talk 19:39, 21 September 2010 (UTC)

To be fair, the total of the Keep votes combined was 65%. So Keep still technically got more votes overall. I think the idea was still to count Keep against Close and take the Keep option with the highest vote, but it was a poor design. My attempt to salvage something useful out of it came too late. CycloneGU (talk) 19:32, 21 September 2010 (UTC)

Jimbo's eventual conclusion (keep Pending Changes enabled, and to increase the hard-coded limit of pages as the performance characteristics of the system allow it) bears a strong resemblance to option 3. Now of the 408 people who supported keeping pending changes in some form 110 of them did not support option 3. If you exclude these people from the support then the support level goes down to 57.9%. If you count them as opposers then the support level goes down to 47.8%. Hut 8.5 19:53, 21 September 2010 (UTC)

That is a hypothetical assumption that people who support Option 2 may not necessarily support Option 3 or even outright refuse it. Some said so, many others didn't and we can't assume otherwise. But this kind of poll setup was what doomed the poll immediately in the first place; that poll had no way of establishing a clear consensus and was designed very quickly and abrasively. Hopefully the new one will lead to a more clear outcome for Jimbo and the WMF. CycloneGU (talk) 20:06, 21 September 2010 (UTC)

Stop !!! PPL in power never want NPOV. So Wikipedia must be destroyed. So PC must be made useless by anonymous IPs. This is not about democracy. This is about keeping Wikipedia alive or let it be destroyed. 2/3 majority is ridiculous anyway, the golden ration is 0.6180. This is a lure, do not help switch off a feature to defend Wikipedia. --Chris.urs-o (talk) 18:16, 21 September 2010 (UTC)

There is an alternate option. to the "leave" option mentioned above. Simply those who vote close should simply boycott any subsequent WP:PC actions. Any of Template:Retired, industrial action, boycotting and just living with it should be suitable for those voting close. User A1 (talk) 18:50, 21 September 2010 (UTC)

Semi protection ain't so bad. Gwen Gale (talk) 19:47, 21 September 2010 (UTC)

Unless you're an IP. =) CycloneGU (talk) 20:06, 21 September 2010 (UTC)

Yeah, it's true. But maybe for some BLPs, opening a user account is worth the woe. Gwen Gale (talk) 20:09, 21 September 2010 (UTC)

For the most part in terms of actual value they probably fall bellow the likes of Asbestosis and Mesothelioma. Astroturfing has the potential to be a valuable target for differing reasons.©Geni 23:35, 21 September 2010 (UTC)

Change the watchlist notice

I don't know where the watchlist notice originates from, but I disagree with the current notice. A straw poll is currently being held to decide how pending changes should be utilized until the release of a new version of the software - To me, this suggests that PC is going forward regardless, and the poll is here to determine what its use is. It's not at all clear from the watchlist notice that this is a straight vote about whether or not PC should be implemented. --Onorem♠Dil 18:46, 21 September 2010 (UTC)

Apparently, Jimbo intends to force the new version on the English Wikipedia regardless of consensus. The purpose of this vote is not to decide something, the purpose is to allow Jimbo to say that it was a community decision. --Yair rand (talk) 19:31, 21 September 2010 (UTC)

Yair, I intend no such thing and you do great damage to your overall position by going on in this way. Let me be clear: I do not intend to force the new version on the English Wikipedia regardless of consensus. I have stated, repeatedly and with significant detail, that I believe very strongly that the feature should gain consensus support in order to be kept.--Jimbo Wales (talk) 19:57, 21 September 2010 (UTC)

8) He's alive!!!--intelati^(Call) 01:21, 22 September 2010 (UTC)

But you're doing whatever you can to sway that consensus, however you can. Why not be straightforward about this? Say the WmF wants PC and will get PC, flaws and all and be done with it? Gwen Gale (talk) 21:12, 21 September 2010 (UTC)

I won't say that, because it is false. WmF would be more than happy, as far as I can tell, to stop development on PC and do something else, if that's what the community prefers. In my view, if PC is not the right thing, it is not the right thing, and we move on to the next idea. It is completely false to say that we will "get PC, flaws and all" - if we don't want it, we can reject it. But the way our community works is not to have a single vote on something and stop there, but rather to proceed iteratively to find a solution that has broad support. I seek your help in doing that.--Jimbo Wales (talk) 01:18, 22 September 2010 (UTC)

But the way our community works is not to have a single vote on something and stop there, but rather to proceed iteratively to find a solution that has broad support. This has always been our policy. How is it consistent with the present poll, which proposes to act on the basis of a majority vote - that is 51% support?

The way to reach broad support is to change the proposal so that those who oppose it can be persuaded to join a broad support. This means producing a better version than the present one, presenting a test which will be genuinely closed and genuinely effective, and not claiming that a majority which would not approve an admin is "broad support". Septentrionalis PMAnderson 20:41, 22 September 2010 (UTC)

The new version is scheduled to be ready on November 9. If there is no consensus for the new version to be turned on, will Pending Changes be active on the English Wikipedia at that point? --Yair rand (talk) 20:01, 21 September 2010 (UTC)

The poll currently is not for that purpose. It's to determine what happens before November 9. I think there was enough overall generic support (for the WMF that is) on the what-the-heck-happened-here poll to satisfy a new trial on November 9 for the new version. CycloneGU (talk) 21:08, 21 September 2010 (UTC)

That wasn't an answer to what I asked. --Yair rand (talk) 21:10, 21 September 2010 (UTC)

Oh, I misread. My apologies. I'll let Jimbo answer that one, then, as he has the links to WMF, not I. CycloneGU (talk) 22:24, 21 September 2010 (UTC)

Jimbo, I didn't closely follow events when the trial was agreed, so maybe I don't have the full story. But what I'm seeing now is (a) a lot of voices asking why PC is still active when consensus was only for a two-month trial, and (b) the fact that you invited us here to perfect the wording of a poll about what to do between now and November when the new version will be implemented. I was not aware you were offering us a choice about whether the November implementation goes ahead. Yet you say above, "Let me be clear: I do not intend to force the new version on the English Wikipedia regardless of consensus." Please help me to understand what I'm missing regarding both (a) and (b) above. PL290 (talk) 10:22, 22 September 2010 (UTC)

Stopping watching page now poll has ended; please direct any reply to my talk page. PL290 (talk) 17:29, 29 September 2010 (UTC)

• Corrected to say "whether pending changes should be utilized". Risker (talk) 19:48, 21 September 2010 (UTC)

FWIW, I use Monobook skin with Dutch as an interface language, and there isn't a watchlist notice of any kind. Usually I get the English notices if no one has customized a Dutch one.—Kww(talk) 19:54, 21 September 2010 (UTC)

I'm also puzzled as to why you and the WMF don't just add the feature by fiat. While I appreciate the degree to which you've allowed the community to have a say in its future, it seems that avoiding this entire mess of polls and consensus building would have minimized the drama; the vociferous debate here has only added fuel to the fire, so to speak. Of course, ending the poll now will lead to cries of disenfranchisement, so... Archaeo (talk) 23:09, 21 September 2010 (UTC)

Well what actual evidence do you have that the WMF really supports PC?©Geni 23:30, 21 September 2010 (UTC)

It's still running, when by consensus, it was but a test which should not be running now. It's ok to try things, it's ok to fail, it's ok to try again, it's not ok to break consensus without saying something like, "This website is the private property of WmF and Wmf wants PC to keep running, whatever the consensus may have been." Nothing untowards about the latter at all. Gwen Gale (talk) 00:17, 22 September 2010 (UTC)

True enough: it's WmF property, and they can do with it what they will. I'm just infuriated by the constant mouthing of "consensus" when that doesn't seem to be really of much concern.—Kww(talk) 00:48, 22 September 2010 (UTC)

Consensus is a primary concern.--Jimbo Wales (talk) 01:26, 22 September 2010 (UTC)

That is extremely hard to reconcile with a trial that is being continued despite a lack of consensus to do so, being replaced by a trial that there is no consensus to start, where that trial has a feature set that there is no consensus is adequate or desirable.—Kww(talk) 03:26, 22 September 2010 (UTC)

Consensus can't be *that* primary a concern if the "trial", which now looks more like a case of gamemaster fiat, is still running, now, can it? —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 16:23, 22 September 2010 (UTC)

I'm also puzzled as to why you and the WMF don't just add the feature by fiat. - I'm a complete nobody in the grand scheme of things...and I'm fine with that. I'm probably done with the site if consensus says that we have to have caretakers approving edits anyway. I'm done today if Jimbo decides by fiat that this is the future of the supposed site that anyone can edit. --Onorem♠Dil 03:29, 22 September 2010 (UTC)

I should've chosen my words more carefully, especially since I'm a nobody too. I think Jimbo's caution here is admirable; look at redesigns like the new Digg, which alienated a huge portion of the user base. Regardless of what happens, I'm pleased by the fact that people were given a chance to weigh in. On the other hand, I also think that the way this was set up was needlessly divisive. The trial period should've ended as scheduled, which would've prevented all this brouhaha. Adding it by fiat would've prevented all this stuff, but at the cost of possibly driving people away from the project. It's complicated, is what I'm saying, and getting consensus for such a major change, as Jimbo seems to want, will take a lot of time and drama. Archaeo (talk) 06:54, 24 September 2010 (UTC)

Simple majority is not consensus. I agree that the WMF should just make a decision and forget "consensus" because it's not happening either way. Polls like this are just more of a dramafest; executive decisions are final and everyone will shut up faster. /ƒETCHCOMMS/ 15:27, 22 September 2010 (UTC)

Irrepairable damage ?

I've been following PC for a long time and have particpated in the discussions, polling and attenpted to use it during the trial period. What I havent seen is where November 9 comes into the discussion, where is the community consensus for the new trial or implementation which will occur from Nov 9. What I'm seeing here is a red herring that has been successfully thrown out there such that while editors question why the trial hasnt stopped we are being conditioned to accept that from Nov 9 PC will no longer be in a trial phase but become perminently active. Enjoy the continuation of this because irreguardless of the result (with or without canvassing) its all irrelevant to future of PC come November 9 PC is here to stay. Already consensus that was a corner stone of the foundation and success of Wikipedia appears to have been abandoned, following close behind is WP:AGF and trust because editors(formally known as the community) have another prime example of how consensus was ignored, trust was broken, WP:AGF is now damaged goods. Everbody should hang their heads in shame and ask themselves was PC worth the damage it has done, can it be fixed. Jimbo still has time to redress some of the damage by honouring the conditions of the trial and shutting it down now. Gnangarra 23:58, 21 September 2010 (UTC)

Gangarra, you are wrong about the facts. It is absolutely false to say that "from Nov 9 PC will no longer be in a trial phase but become permanently active". I don't mind people dissenting and debating, but please get the facts before you say things like that. Consensus is not being ignored, and consensus will be required in order for PC to be kept permanently.--Jimbo Wales (talk) 01:25, 22 September 2010 (UTC)

Jimbo, I think much of the community is under the impression that a new trial beginning on November 9 will happen regardless of consensus ("It has been announced that a new version is slated for release on November 9. The community should now decide if the current implementation should still be used between now and the release of the new version." seems to imply that). Some people may be voting based on this assumption. Regardless, the unexplained continuation of this trial is dividing the community. I agree with Gnangarra, there is a lot of damage being done by this. Closing the poll and trial now would save the situation. I'd suggest simply closing the poll now, stopping the trial, and letting things cool off before continuing. The community needs to pull together and find solutions to problems. Some of the issues with PC can't be fixed with just developer time. We need to analyze the results of the trial, explore possibilities like the PC compromise, and have a simple discussion without continual support/oppose tearing the community apart. We don't need to rush through this, we need a result everyone can agree on, even if it takes a few months longer to get it. --Yair rand (talk) 01:44, 22 September 2010 (UTC)

(edit conflict)But consensus was for a 2 month trail with a defualt of discontinuing, the original terms of the trail are being ignored as I said trust has been lost and AGF appears to be damaged, the current situation appears that PC is a perminent addition and these polls are just lip service to appease those that have concerns. The real sad thing is I think that PC has merit but needed work, but I dont think its worth is greater than the damage this process has done. Gnangarra 01:56, 22 September 2010 (UTC)

Jimbo, while I firmly believe you have setup this poll in good faith, some of the criticism in this section is valid and highlights how unintentional bias may have slipped into the poll wording. You are absolutely entitled to your opinion, but it's incredibly difficult for anyone to not let their opinion ever slip through. A prime example is the wording for the close option: "Remove Pending Changes from all article pages until a new trial after the November release." This says that a new trial is definitely coming and I think it has directly led to a number of keep votes with the rationale of "it's not worth it to turn it off for just a short time." Perhaps you have a great plan to conduct an RfC on the new trial between now and November 9. And it's fine if you would like to see that RfC succeed. But this wording does make it sound like a done deal. While I voted for close, I'm in favor of a new well structured trial that avoids the pitfalls of the first trial. I'll also reiterate my poll comment that the new trial may use a different set of articles, which shows that leaving it on may be not only useless to the new trial, but risks biasing the configuration of the new trial to the same set of articles when a different set may provide more meaningful results as was expressed in numerous comments from the first trial. If we want to follow a scientific approach with meaningful stats, shutting down the first failed trial to prevent bleed through bias to the second trial is the best way to go. And this is from someone who supports a second trial. I'm sorry that I didn't think more about this in my original comments on your talk page to the setup of the new poll; I got hung-up on opposing the poll due to the high tensions and distrust. I wish I had gone to the next level of the technical reasons favoring closure. —UncleDouggie (talk) 08:33, 22 September 2010 (UTC)

To further clarify my support for a second trial: I don't blindly favor a repeat of the first trial with whatever new UI the developers are able to give us by November 9. I would like to see some debate about what the next trial should look at to give us meaningful guidance for a final implementation, if there is one at all. I have proposed integrating PC with some new automated tools. While we may not be able to implement all that for a second trial, we might still be able to collect good data. For example, we could construct a new edit filter that would only fire on PC protected articles and assign a tag in place of modifying the current PC workflow. The tag would be used for data analysis only to compare actual vandalism between edits with and without the tag. We could also do a similar thing with a bot. —UncleDouggie (talk) 11:10, 22 September 2010 (UTC)

Closing the poll early

Yair, please explain to me how a discussion between 600 or more editors could even be possible, let alone "simple"? If this poll and trial were closed now, a significant majority of Wikipedia's editing community would object. There is nothing "simple" here. Ronk01 talk 03:13, 22 September 2010 (UTC)

The point is that this poll isnt a discussion it never has been and never pretended to be, closing now would upset many editors but the polls integrity is under question and not just because of alleged canvassing. As said the damage is done, the validity of WP policies & principles are compromised so is this process worth that cost? Gnangarra 04:00, 22 September 2010 (UTC)

There is a reason why this poll isn't a discussion: Communities this large can never achieve consensus. The only way to look at this is to see what the community supports, and currently, that appears to be keeping PC turned on (61% now, without the "canvassing" it would be about 60.8%.) Ronk01 talk 04:08, 22 September 2010 (UTC)

Thank you for that perfect display of one of the dangers of this process. Would you care to demonstrate a few more? --Yair rand (talk) 04:14, 22 September 2010 (UTC)

Your blatant Assumption of Bad Faith is not appreciated, I would not do the same to you, so please show others the courtesy of AGF. I still believe in the value of consensus in policy discussion, RfA, talk pages, etc. But here, in poll about extending a trial a few weeks? By the time the community had reached anything even resembling consensus, it would be November 9th, and the second trial would be starting. I will repeat myself once more: In discussion among more than five hundred individuals, most votes will not achieve more than a 55% majority due to the awkwardness of discussion, and the varied opinions of the group members. That's why Obama had a landslide with only 56% of the popular vote. Ronk01 talk 04:25, 22 September 2010 (UTC)

One of the more dangerous aspects of this poll is that people will start believing that. We have no deadline, no need to do things faster at the cost of quality of the result. Accepting a decision based on anything less than clear consensus is dangerous. Extending the trial past two months is dishonest. Discussion is what leads to actual helpful results. --Yair rand (talk) 04:40, 22 September 2010 (UTC)

Again, I ask, how can a discussion between 600 editors take place? Ronk01 talk 04:44, 22 September 2010 (UTC)

The same way it often takes place at RFA with a couple of hundred, which still manages to hold a discussion, and arrive at an outcome judged by a neutral party, with votes given more or less weight according to their strength, and the putative "range" for success being a guide for the closer not an immovable red line. You seem to be implying that the more responses a discussion has, the less effort we need to take to properly assess consensus. There are several things wrong with that, most of which should be obvious, but one interesting result of that line of thought is that you need to invent yet another new arbitrary formula that defines exactly how much consensus should be sacrificed according to the number of participants. You could create "bands" according to total participants, or a graph like relationship to mathematically decide how much less care is taken to assess consensus for big decisions, but since what you would be doing is exactly that, I respectfully suggest that you'd be better off just accepting that big decisions are big decisions, and need to take longer when consensus is difficult - there is no deadline, except those we arbitrarily impose on ourselves, and reducing our standards for more important discussions flies in the face of reason. Incidentally, comparison to a "landslide" as an election result being 56% is artificial, since the final result is a composite of many binary sub results for states. You could theoretically win every state with exactly one more vote than your opponent and achieve 100% "landslide" with 50 votes more than your opponent out of many millions. If that's what you'd like to see happen here, then sorry, I would wish no part in it.  Begoon•talk 05:37, 22 September 2010 (UTC)

(61% now, without the "canvassing" it would be about 60.8%) the messaging of 400 supporters is about equal to the total number of voters we currently have, and much greater than the ~230 odd supports currently we have. Before the messages there were ~100(ish) supports, I'd love to see the maths that converts that to 0.2%. Quoting made up statistics is 93% less effective than building logical arguments Jebus989^✰ 07:35, 22 September 2010 (UTC)

Ah, yes, but since this is an odd numbered month, and only has 30 days, you need to factor in "stuff", don't you? So I did a quick calculation on a beer mat, and I think your 93% should have been 92.8%, or possibly even some entirely different number? I'll see if it comes out differently when I make it up again later. :-) Seriously - yes, it's silly, and adds to the whole "numbers game" here that devalues consensus.  Begoon•talk 01:01, 23 September 2010 (UTC)

Ronk01: There aren't 600 different opinions. There are usually a couple of key points that make people go one way or another and then a bunch of additional comments for future improvement/direction. Occasionally, someone will come along and frame the problem in a different way that helps guide the process. Something comes up like the difficulty of turning it off and the ones in favor of shutdown need to respond with a detailed picture of what shutdown will require. It's all about acknowledging the concerns of others instead of fighting over them because the decision has already been made. —UncleDouggie (talk) 09:16, 22 September 2010 (UTC)

This poll could be compared with the institution of the calendar year; we decided it would run from Jan. 1st to Dec. 31st. We can't just go and extend a couple of weeks or months even because that'd suit some folks better. You either stick to something that has been agreed upon or you don't negotiate at all.Qwrk (talk) 09:58, 22 September 2010 (UTC)

Why, then, is the trial being extended past the term that was originally stated? Or was the two months not the result of an agreement, but just somebody’s guess at the time it would take for sufficient testing? Odysseus1479 (talk) 00:15, 23 September 2010 (UTC)

Perhaps a PC workgroup? Basically, 20 or so of the most involved editors (10 from each side) who are willing to compromise (ie. willing to give up on screaming "PC Must Die!!!" or "If we don't get PC, I'm leaving") Who would discuss possible compromises for implementation of PC. I'm sure almost all of us here in some way believe that that we are arguing is what's best for Wikipedia, if you aren't, then please leave, because this is not about personal vendettas, it is about implementation of a tool that the developers spent years working on. 11:52, 22 September 2010 (UTC)— Preceding unsigned comment added by Ronk01 (talk • contribs)

It doesn't matter how long the developers spent working on it. DuncanHill (talk) 12:04, 22 September 2010 (UTC)

I do believe that everyone is arguing for what they feel is best for Wikipedia. Unfortunately, a hand-picked group of moderate folks won't come up with the best solution. Many of the "extremists" have the most in-depth experience and can provide the best feedback. And excluding people will only further aggravate the feelings of being left out of the process. If the trial had ended after 2 months, most people would be working together constructively. —UncleDouggie (talk) 12:15, 22 September 2010 (UTC)

Typically these discussions attract the most involved people. You don't have to fix something that ain't broken, and you don't have to engineer the reaching of consensus. The process has been working just fine for years. When given an opportunity to work toward consensus, the community (surprise) reached one. Revcasy (talk) 12:33, 22 September 2010 (UTC)

That wasn't a community compromise that was about ten editors who were still hanging around when the community had moved along. Off2riorob (talk) 12:48, 22 September 2010 (UTC)

I agree that the proposal needed to be more widely discussed (and I said as much several times on that talk page). I also think that the final outcome would have been similar to what we arrived at there had the process been allowed to carry on. I merely hold that discussion up as an example. Unfortunately, many people seem to value speed over consensus. Revcasy (talk) 12:59, 22 September 2010 (UTC)

The exact same thing can be said for the discussion which created this poll, Rob. _EricLeb⁰¹ (Page | Talk) 19:30, 22 September 2010 (UTC)

Yes agreed, but we don't need a community consensus to open a straw poll do we? We will be going around in circles forever. That discussion was similar but that is all it was, a small group of editors discussing the issue, that was being presented as a compromise closure which it really wasn't, it may have been the basis for yet another straw poll to see if there was support for it but there had only days before been a huge poll of almost 700 editors. Off2riorob (talk) 19:36, 22 September 2010 (UTC)

small group of editors discussing the issue, that was being presented as a compromise closure which it really wasn't, Shorter Off2riorob: Anything that doesn't give me exactly what I want is no compromise. It is your shameful strawpoll which got us into this situation; why don't you sit down and let those who are willing to compromise discuss the matter? The result of your actions has been to drive those of us who would like PC to continue to be tested and improved but don't think it is a magic wand into opposition. Septentrionalis PMAnderson 20:46, 22 September 2010 (UTC)

"... let those who are willing to compromise discuss the matter"

Where in the heck are those people?? I have seen absolutely ZERO suggestions from the "oppose" side, which do not begin at "Shut PC down! RAWR!" I've asked about a dozen times already... aside from 'shut it down, and then maybe I'll have an idea or two', what are the compromises that at least one person in opposition to PC would support?? BigK HeX (talk) 23:36, 22 September 2010 (UTC)

I would support keeping it switched on on the understanding that it would be used only on BLPs that aren't already semi-protected, or would be added to BLPs along with semi-protection. I think this is an option lots of people would support, but it was not asked in the first poll. I tried to add it to the second poll during the discussion on Jimbo's page, but I got nowhere. SlimVirgin ^{talk|contribs} 23:47, 22 September 2010 (UTC)

Anything that doesn't give me exactly what I want is no compromise.

Um...I agree with the other comment that the messed up poll started this entire mess, and thus will not touch on that at all. However, I do disagree with the above statement. Every user has a different opinion of what they want. When two users want different things and will accept no other alternative, there is no way to compromise between the two. Now, for 600 people...it would take an Act of God to find such a compromise. I'm not saying Jimbo is God, since he or the WMF will have to do something in the end, but it still would take a miracle to turn this ship around. Now, if I am misreading and you are quoting something else, please clarify, because the statement I am discussing is not an attitude that should be around here. No offense. CycloneGU (talk) 04:13, 23 September 2010 (UTC)

Section Break

BigK, now that is a bit disappointing, considering that you were a party to the good-faith discussions after the previous poll. Simply because our discussion was stepped all over and ignored by others does not mean that you should suddenly re-start negotiations from a new, fraudulent premise, claiming that the opposition made no efforts to compromise. Our compromise was devalued and our position undermined, why should we re-enter discussion in good-faith when we have been treated that way? Revcasy (talk) 23:51, 22 September 2010 (UTC)

comment - this user never used the tool at all and has no idea as to its workings and therefore was not even involved in the trial. Off2riorob (talk) 11:42, 23 September 2010 (UTC)

If you are purposely trying to make me angry Off2riorob, you have succeeded. I have never misrepresented my feelings about PC. I have said here on this page, on Jimbo's talk page, on the previous poll's talk page, and in my votes in both of the polls that my issue is with the way in which this process is taking place, and that I have no strong feelings about PC itself. But even putting aside the idea that the process on its own is a valid reason to object, I am a part of the community which must live with this tool. I was once an IP editor, I encounter articles every day (even with the current limited trial implementation) which are PC protected, and I am familiar with the workings of PC. To claim that my opinion is somehow not valid because I never actually approved a change is patently false, and against the spirit not only of consensus and community involvement, but even against the spirit of this farcical poll. Furthermore, I strongly object to being singled out by you, as I am not even involved in a conversation with you here. Your actions smack of personal vendetta, intentional or not, and that is not appropriate behavior, whether or not I happen to disagree with you, or your absurd polls. Revcasy (talk) 12:24, 23 September 2010 (UTC)

Your anger belongs to you alone, I am not responsible for it. Your comments are worth what they are worth according to your experience in using the tool of which you have none at all. I assure you I am in no way singling you out, I feel the same about all users that didn't trial the tool and yet are vocal objectors. Off2riorob (talk) 13:27, 23 September 2010 (UTC)

Indeed. Have you contacted those other users repeatedly on their talk page as you have me? Have you made a special effort to point those other users out here in this discussion as you have me? As you say, if you look around I believe you will find several users who have similar objections to the process underway here, and little or no objections to PC itself. I do appreciate that you find all of those user's opinions invalid, and their concerns about process unimportant. However, I (and I am sure they) feel similarly about your opinion. Now, if we could move on, I have no interest in continuing this personal confrontation. Revcasy (talk) 13:56, 23 September 2010 (UTC)

I feel the same about all users that didn't trial the tool and yet are vocal objectors - but not those who didn't trial the tool and are supporters? Interesting Jebus989^✰ 14:02, 23 September 2010 (UTC)

To Revcasy .. You were the first user with whom the issue arose, other users that also didn't actually contribute to the actual tools trial are likely aware of who they are. This to me is about the tool, all users have opinions about process but those personal opinions should not imo allow those users to seek the closure of this simple tool without even trialling it. Process discussions would be better elsewhere. This is actually about the tool. Off2riorob (talk) 14:10, 23 September 2010 (UTC)

Off2: I breathlessly await your further efforts in bringing to light these offenders. Also, I have raised objections to the process in other places. However, I absolutely reserve the right to protest a poll on the poll's discussion page. Revcasy (talk) 14:22, 23 September 2010 (UTC)

To Jebus989 .. Users that didn't trial the tool but are supporters .. I find that less of an issue. It is possible to accept with good faith that the tool is being developed to the simple benefit of the wikipedia and getting in the way of that when the tool is up and running and in further development without the wiki wheels dropping off is simply unnecessary and detrimental to the development of the project.Off2riorob (talk) 14:10, 23 September 2010 (UTC)

Well, to be fair, the poll is about leaving it dormant until the next trial is implemented. I've made many posts which allude to policy on this page to support my belief that to avoid long term damage to trust and support it would be preferable to go dormant, and move to a new trial when the software addresses the current issues. As to PC as a tool, I support it strongly. I'm interested in whether you think my posts would also "be better elsewhere", along, of course with any other posts on this page discussing process.  Begoon•talk 14:22, 23 September 2010 (UTC)

Issues regarding policy are discussable here but meaningless as regards the tool trial, if users have not used the tool and only object to some aspect of the process they should make that clear. I was not involved in the trial so I have no idea about the tool but I don't like the process of straw polling, or whatever they object to. This to me is about the tool. I have trialled it and used it as much as possible and when people that have not used it at all want to switch it off I think I have an issue with that. Off2riorob (talk) 16:09, 23 September 2010 (UTC)

(far too many colons, now, but I can't sensibly outdent in the middle here :-)) Thanks, that clarifies your point a bit. I still don't agree with any sort of limiting of input, because it looks to me like a pretty mad place at the inevitable end of that line, but I understand a little better. Incidentally, there is a very interesting section on this page where the use of the tool trial is discussed almost exclusively in the context of process and policy. It relates to editors' responsibility and duties as a reviewer and whether it carries any liabilities - tied in with the implications of not clarifying that before a new trial begins or the tool is rolled out any further. I don't wish to say that is a direct comparison to what you believe is "wrong", but it at least demonstrates that there is no bright line between policy, process and discussion of the tool trial.  Begoon•talk 16:26, 23 September 2010 (UTC)

Thanks yes, I commented below in the The insanity must stop section about the tool qualities and benefits and here about the legal liabilities of reviewing. Off2riorob (talk) 17:55, 23 September 2010 (UTC)

Ah, yes - so you did - I hadn't seen the post here, I agree entirely that it's almost certainly a negligible chance that liability could be, or would even attempt to be enforced. I do, though for the peace of mind of editors concerned about this, think it's important to clarify it somewhere, preferably with something more concrete than my, or your opinion. More important, though, imo, is clarifying the role of the reviewer - does he just check for obvious vandalism, with the inherent risk that this will let bad edits "sneak in" on a page he may never have visited before, but found through the outstanding PC pages, or should he review the edits exactly as though found through Recent Changes or his watchlist. This seems to be an interesting balance to be weighed between making reviewers less willing to review, and dealing with bad content introduced in this way. I think both of these points need to be clarified preferably before a new trial (because they may require minor UI changes/messages etc), and certainly before any kind of mass rollout. It's basically the headlong rush with no dotting of i's or crossing of t's that concerns me. It's unnecessary and risky - guess what I'm going to say next. There is no deadline that means we need to not do this properly, except the ones we choose to impose on ourselves.  Begoon•talk 18:20, 23 September 2010 (UTC)

Interesting comments Begoon. I have only ever reviewed pages from my watchlist, although some of them I am not knowledgeable about at all, currently I don't support random reviewing and with the current number of articles under pending there is no need for it either. Personally I review the edit quite thoroughly and if I find an uncited addition is perhaps correct I go to google it and if I find it is correct I accept it and then add the cite I found that supports it and if needed refine it, for example add any internal links that maybe the unconfirmed user has not added. If after reviewing I assess the edit is a false addition or similar I template the user as I would any edit and then I check his previous contributions as if he added a false addition to that article he may have added a false addition to other articles, as users would do if they found a false addition when reverting vandal additions. Off2riorob (talk) 19:27, 23 September 2010 (UTC)

Indeed. I do the same, but with the addition that I do look at the PC outstanding page. I suspect that the majority (if that word won't get me lynched here) of reviewers do as we do. However, there are also a good many who are of the opinion that it should not require validating the content, but just be a "quick vandalism" check, and they will be less likely to review if more work is required on each review. There's at least one discussion on this page along those lines, and I've seen many other similar discussions. As you say, whilst PC is on the < 2000 articles it is now, that's manageable, but once it expands there become some serious potential issues which need to be dealt with now, so that we head them off. If reviewers are going to do a simple vandalism check, then there is a real danger poor edits will "slip through" when the next reviewer assumes it was properly checked, and allows the subsequent change. This potential obviously just gets higher the more widely we use the feature. The answer, imo, is to clearly define what is expected in a review, and consider what effect this will have on available willing reviewer resources. Aside from the need to make sure the revised software addresses anything it needs to here, and the many other identified problems before we start another trial, clarifying points of procedure/policy like this all need to be clear before we embark on a new trial. To do this properly takes time. Add that to the obvious widespread feeling that this is being rushed through without proper process, and, to me at least, the need to honour the original conditions of the trial becomes paramount. Without that, I'm seriously concerned that irreparable damage might be done to long term trust and confidence not only in PC itself, but in the project as a whole regarding integrity in process management. For the sake of all those risks I really don't see what the problem is with a proper pause between trials to assess the results. Any slippage of proper process management to adhere to arbitrarily defind, self imposed "deadlines" is just too risky to make any long term sense, imo.  Begoon&#149;talk 23:52, 23 September 2010 (UTC)

Break

Exactly what is supposed to be disappointing about my comment that opposers continually demand "shut it down"? I don't see how a proposed closing statement which offered precious little more than "Shut it down", is somehow evidence that runs contrary to my premise that opposers have offered zero compromises that don't start with "Shut it down"....

Even though I've long said that the support was easily significant enough to make a trial extension a reasonable and fair outcome, I'm still quite willing to support "Shut it down" as a compromise measure, but that's only if opposers similarly respect that they have significant ground to cover on a compromise. I've looked, but it's been difficult to find much willingness among many of the vocal opposers. BigK HeX (talk) 06:18, 23 September 2010 (UTC)

BigK, I would also remind you that "I believe PC is a fundamentally flawed idea, and I don't support any implementation of it" (which you portrayed a bit more sarcastically) is an entirely valid position. It's not my own, and I'd be willing to consider another trial (if and only if it's understood that the end of the trial is an unequivocal kill date, and that true consensus must exist afterward to turn it back on), but only if a lot of things get fixed. Seraphimblade ^{Talk to me} 05:29, 23 September 2010 (UTC)

I don't think I ever disputed the validity of those who oppose the concept of PC. I'm a bit unsure why you directed that comment at me. BigK HeX (talk) 06:18, 23 September 2010 (UTC)

In any case, it is refreshing that I've finally come across a compromise that isn't premised on "Shut it down". Thanks, User:SlimVirgin. BigK HeX (talk) 06:20, 23 September 2010 (UTC)

I'm willing to listen, but I have a hard time seeing how a compromise will address the concerns I have raised below. Even SlimVirgin's proposal wouldn't have helped here because my example case is a BLP. To me, it's a simple matter that the drawbacks outweigh the benefits at this time. I do acknowledge that there are some benefits, and I know that some of the "keep" voters acknowledge the drawbacks. Even if we all agree on these points, how will that get us anywhere? —UncleDouggie (talk) 06:42, 23 September 2010 (UTC)

Actually, it's not so much the details of SlimVirgin's proposal, but rather his willingness. He's evidencing a willingness to meet closer to the middle. That attitude is far more conducive to finding a consensus than most of the other vocal opposers who have basically stuck by the idea, "screw the significant majority, we still want it OUR way and won't even entertain discussion until that's done." We had a trial and then got community input. For all of the "we demand it be shut down", I haven't come across much of an explanation as to why the option to temporarily extend the trial is somehow unfathomable, as opposed to being one of the possible fair/reasonable outcomes from the first straw poll discussion. Seems that some opposers attempted to force the outcome of the first straw poll into a rigid binary decision, with little more rationale than that it selfishly furthered their own aims. I hope to see more open-mindedness than that. BigK HeX (talk) 14:29, 23 September 2010 (UTC)

I agree, much of the drama resulted from people who constantly demanded "SHUT IT DOWN NOW OR I WILL LEAVE OMG LOLZ" and so on. It was their not bringing up legitimate reasons to shut it down. That is because it works. However, even though it works, it doesn't work everywhere, and naming even one single situation isn't enough to warrant removing a tool that has worked elsewhere. That is just my view.

I will applaud SlimVirgin for trying to come closer to the middle as well. I saw the remark last nite and was too tired to post a reply. CycloneGU (talk) 21:49, 23 September 2010 (UTC)

The insanity must stop

And I mean Pending Changes v1, not this poll (although that's another matter entirely). The pending changes UI is a disaster that is causing more long-term damage to Wikipedia than it is preventing. Even after all this discussion, just take a look at the last 5 days of history on the Bam Margera BLP, including the PC log. First, an IP vandal tries to do some damage. Auntof6 tries to undo it but accidentally accepts it. An IP does the right thing in taking out the vandalism, while unknowingly zapping the original good text in the process. Wolfnix appears to make the same PC UI mistake as Auntof6 in accepting the removal, reverses themselves, and then realizes it was originally vandalisim rather than section blanking and finally takes it out! The good portion of the text doesn't get restored until 4 days later when I find the fiasco while doing some research on PC. Just before that, I messed up just as badly as everyone else in accepting an insertion that I was really trying to take out, which of course I fixed. I don't want to think about what would have happened if I unaccepted instead of editing. The discussion here has made me treat the unaccept feature as radioactive. We can't let this disaster continue. Shut it down before even more damage is done and we all go nuts. The folks that think it's working may be just as confused as I was in thinking that I was preventing a bad insertion through my review when I was actually allowing it. Before anyone tries to say I made a bozo move in my accept, as a simple read of the log would seem to show, may I point out that this article is 0 for 3 in correct PC reviews for the last 5 days. If everyone is messing it up, something is horribly broken. —UncleDouggie (talk) 01:41, 23 September 2010 (UTC)

Yes when I made the changes, I accidentally accepted, because well the UI for pending changes is awful. If you put in a reason for why your denying, It blocks out the unaccepted option period. This is unacceptable. I reverted the mistake because the section about Bam's death was unconfirmed, and the article cited did not list this information, period. So, I considered it vandalism or some sort of Personal Attack against the subject. Since then, and because of this specific incident, I hardly do any pending changes, and if I do, I make sure, that it is the proper option. --Wolfnix • Talk • 02:55, 23 September 2010 (UTC)

Lets get that compromise going! I'm tired of the PC this PC that!!--Novus Orator 03:08, 23 September 2010 (UTC)

My edit filter compromise wouldn't have helped us here, unfortunately. Any decent edit filter would have flagged an insertion to a BLP with the word "death" as requiring review. To make the compromise work, we need a good reviewing infrastructure to deal with the high risk changes. We don't know yet just how well PCv2 will address the deficiencies in this area in a real-wiki environment. —UncleDouggie (talk) 03:21, 23 September 2010 (UTC)

I agree that this can get very confusing and complicated. To me, it boils down to just what a reviewer's responsibilities are. Are they supposed to:

• Verify accuracy of changes?
• Ensure that all additions are sourced?
• Just make sure there's no profanity in a change?

My understanding was that reviewers are only responsible for making sure there is no "obvious vandalism", but what's obvious to one may not be to another. I didn't "accidentally accept" the change mentioned, I consciously accepted it because it didn't look like vandalism to me. I'd do so again under my current understanding of the rules. If reviewers are directed to verify changes, I for one won't be reviewing as many articles. --Auntof6 (talk) 03:09, 23 September 2010 (UTC)

I neglected to explain the reason for my mistake described above. I now see that Auntof6 and Wolfnix made different mistakes from me, so I should elaborate. I thought I was looking at the diff from the last accepted revision to the latest revision, when I was actually looking at the diff from the latest revision back in time to the last accepted revision. It's possible that my use of pop-ups and Twinkle gave me hyperlinks that a vanilla reviewer wouldn't have had available and contributed to the confusion. What I do know is I was convinced that the IP edit was trying to take out the bad ref rather than add it. If I had to do it again, of course I would only use the hyperlinks from the article history page and ignore all the bogus links on the reviewing screen. But this would also slow down the process and others might not be so careful. And even once I did all that, I would still be constantly second guessing hitting the accept button. If I did lots of reviewing, I suspect I would get fed up and just adopt the workflow to always accept and then go sort it out on the history page, rolling back as needed. I know that I'm not alone. Will the stats collected in v2 account for this workflow? —UncleDouggie (talk) 03:36, 23 September 2010 (UTC)

I would like to also add that while there has been a lot of vandalism previously prevented on Bam Margera, almost all of it has been through Huggle or Twinkle rather than PC reviews. A simple delay in public visibility of edits would have accomplished the same benefit without the complexity of PC. —UncleDouggie (talk) 05:05, 23 September 2010 (UTC)

[Outdent]

I would support having two separate pending changes queues: one for BLP that gets more rigorous attention from those active in those areas (admins. especially), and another for non-BLPs which are not as serious, but still vandalism.

As for the incident starting this section, if Auntof6 had checked the sources in the original version or even visited the source, it would have been found that invalid information was added, the edit was vandalism, and the undo or rollback buttom required clicking. I've reviewed the odd BLP edit, I think, and I was always careful to check sources for any new information. No source, I reverted. I think the problem is less that the later edits were accepted; it is more about the incorrect accepting of the first edit, resulting in text that people assumed was accurate in later edits, and the eventual good fortune that the mistake WAS found. If I came along in the second phase of that, I may have also assumed Bam was dead as it was a previously accepted edit, and I would have thought Auntof6 had reviewed it properly. Proper procedures for reviewing edits will make pending changes much better. (And getting rid of the unaccept button and providing proper instructions to click Undo to deny the edit...or Rollback...or etc...) CycloneGU (talk) 04:04, 23 September 2010 (UTC)

Personally, I won't accept any edit, or series of edits through review that I wouldn't accept at recent changes, or a watchlisted article, because I view accepting an edit as confirmation that it is valid in all ways. But that's what I personally feel, and I completely agree that the guidelines need clarification. I'd prefer them to be that way, because, although it means review sometimes takes more time, it avoids the situation where a poor edit is "reviewed" but not checked. On the other hand, if the side effect is that reviewers are less willing to review, we need to consider the guidelines carefully. We'd need to weigh up the pros and cons of that undesirable side effect against the danger of an edit never being checked because it was assumed to be checked when reviewed. Of course, the answer is not simple, and I don't have it - except to endorse the fact that it is an issue that needs addressing properly before any kind of mass rollout of PC. If that isn't done, then the "few" errors or misunderstandings of this kind will potentially become unmanageable, another potential outcry for getting rid of PC. Again, my position is that there is no deadline here, except arbitrary self imposed ones, and the better option for a long term PC acceptance and support is to iron these things out before the next trial. I would like PC to be a success, not something a high number of people feel was rushed through too quickly.  Begoon&#149;talk 04:26, 23 September 2010 (UTC)

I do not propose to spend the time it would take to give a detailed commentary on this, but I will just make the following remarks. When pending changes started I tried using it, and very soon found that it was so confusing and error prone that I gave up, and have never touched it since. It may be that some editors who have taken the trouble to learn how to use it effectively can do so, but I would be prepared to bet that there are far more who either mess up with it or avoid it like the plague. JamesBWatson (talk) 09:37, 23 September 2010 (UTC)

Have to agree with JamesBWatson gave it a try but found the UI confusing now I just ignore changes that are flagged by PC. Mo ainm~Talk 11:17, 23 September 2010 (UTC)

It does take a little bit of trouble to get used to it but when you do it is quite simple and works fine. I have even been involved in a couple of edit wars with it and it works well, the edits build up and rather than having to continue to revert, users stopped accepting the disputed edits, the article became stable and Admin action and discussion ensued, articles where the regular vandals got the message that they were wasting their time calmed down and imo when new issues arose that would have attracted a lot of vandalism, there was less unconfirmed editing imo because of pending and this was a big improvement on allowing the free disruption at the article to result in long term semi protection. Off2riorob (talk) 12:24, 23 September 2010 (UTC)

So then, will you be doing all reviews between now and November 9? I think it is foolish to force the continued use of a tool that is so delicate and error prone that it can only be safely operated by a select group of users. This situation also truly makes it vulnerable to POV pushing by the elite reviewing experts. —UncleDouggie (talk) 21:51, 23 September 2010 (UTC)

I wil do whatever comes up on my watchlist if I am online. It is not a difficult task to review an edit, you do not need to be a select user at all. There is also no evidence at all of your allegation of POV pushing by elite reviewing experts as you call them. Also there is nothing delicate and error prone about the tool either. Have you got any diffs of complaints about these POV pushing reviewing experts as you call them? Off2riorob (talk) 21:57, 23 September 2010 (UTC)

I said it is vulnerable. As more reviewers dropout from frustration with the tool that only you seem to know how to use properly, the reviewing pool will shrink until we have only professional reviewers left. The concept of using the tool for POV pushing was first raised by others on the last poll. I had dismissed it at the time, but now I'm starting to think that maybe they were right. Exactly how many editors have to say that the tool sucks before you will listen to us? —UncleDouggie (talk) 22:13, 23 September 2010 (UTC)

246/161 So Far

That is 60.4% as of right now, not counting abstain or poll opposition (some of whom voted one of the other options). This begs the question what will happen from this result, since it's not 66% or even 65%. As I type, there are four days left in the poll. CycloneGU (talk) 04:18, 23 September 2010 (UTC)

It's been up for only three of the seven days. The poll could easily change by 6% by then, or more, in either direction. Don't be too premature on this. Silverseren^C 04:25, 23 September 2010 (UTC)

What makes 65 or 66% a number to watch for? As far as I can tell, this never-should-have-happened poll only requires 1 vote more than half for PC in its current form to continue being used. --Onorem♠Dil 04:56, 23 September 2010 (UTC)

I used that number because it's been floating around several discussions, and I do not know what number Jimbo is looking for in regards to consensus. If this can even be called that. CycloneGU (talk) 05:14, 23 September 2010 (UTC)

They aren't looking for consensus. That target was too difficult to achieve so they simplified the process. --Onorem♠Dil 05:21, 23 September 2010 (UTC)

This poll was going to be 50/50 from what I remember hearing.Doc James (talk · contribs · email) 06:20, 23 September 2010 (UTC)

It was clear from Jimbo's talk page that this poll has a 50% threshold. I don't understand why that isn't made clear on the poll page. It shouldn't change anyone's vote, but it could lead to hard feelings afterward. —UncleDouggie (talk) 06:32, 23 September 2010 (UTC)

WP:CCC makes it clear that consensus can actively change, so an arcane discussion on Jimbo's talkpage does not necessarily have precedent over novel debate right here. But whose original idea was a simple majority, anyways?   — C M B J   06:56, 23 September 2010 (UTC)

Jimbo decided that on his own. Revcasy (talk) 13:00, 23 September 2010 (UTC)

• The poll used to include the words "The poll will be closed strictly according to majority vote." but they got removed at some point in the last couple of days. DuncanHill (talk) 11:59, 23 September 2010 (UTC)

The comment was removed in this diff by user NW. Off2riorob (talk) 12:13, 23 September 2010 (UTC)

Thanks for finding that. I have undone the removal, as the wording was there from the very beginning and its absence was confusing some editors. DuncanHill (talk) 12:53, 23 September 2010 (UTC)

Proposal to address such actions

I've seen a poll fall apart on Commons because the Commons:Sexual content page kept being edited during the poll. I'm thinking of proposing something along this line:

• All polls should be conducted in a portion of the WP: namespace beginning "Polls/".
• A poll should reference only templates in WP:Polls/ that are protected during the poll, and static history versions of any other relevant pages outside "WP:Polls/". The former would explain the start, end, and methodology of the poll.
• A class of users, which I dub "Pollmeisters", should administer polls according to (some set of policies; will fill in if formally proposing this). They should be granted the admin power to protect and unprotect pages within WP:Polls/ (to allow protection of the poll parameters and, when it closes, the results) but no other admin power (except as obtained by other mechanisms).

Obviously some details need to be filled in here, but I'm thinking that with a system like this, we could save whole rainforests of virtual trees on the talk pages. I don't mean to suggest that opposition to an atypical poll like this one should be dismissed - only that we should keep it outside of the poll in progress. Wnt (talk) 21:41, 23 September 2010 (UTC)

Not that it matters one bit because this is Jimbo's poll to do with as he likes, but one stat I'm curious about for our history crunchers is: How many of the voters for each option have performed at least one review? —UncleDouggie (talk) 22:27, 23 September 2010 (UTC)

I would very much like to know that. It might affect the determination of consensus. Ronk01 talk 02:39, 24 September 2010 (UTC)

It certainly shouldn't. I have made many, many reviews, but the concept that my opinion should thus matter more than someone else's solely on that basis is wrong. If my opinion is logical, factual, and relevant then it should be considered of no more importance than that of an editor who has never used the tool, but provides an equally valid opinion. I have never piloted an aircraft, but I would be most upset if a comment I made in a discussion on, say, Air Traffic Control was viewed with less weight than that of someone who has piloted, rather than just assessing the relevance and strengths of the two opinions. It's perfectly valid for someone who hasn't used the tool to offer an opinion, after all, it could be to give the reason why they have not used it, which is surely useful information, or to support quicker development because they will be willing to begin to use it when the software has addressed issues which have been reported. I truly see a very dangerous place at the end of the road if we accept anything which considers factors other than strength and relevance of an opinion as a determiner. Now, having said all that, sure, there may be cases where an opinion does have less strength, and this is because of lack of use of the tool, but that can only be assessed for individual opinions, not by some blanket application of what is, in effect, a veto. To illustrate the point, in certain circumstances, it could be appropriate to give an opinion less weight because it appears to be from someone who has used the tool a great deal and either loves or hates it to such a degree that they seem unable to perceive anything outside that love or hate - a kind of "wood for the trees" scenario. Basically - consider opinions individually, on their merits. Now, to be clear, this response is to the premise that we should judge consensus in this way. Historical statistics based on this, or any number of different demographics are a different matter, and could, indeed, be very useful and interesting.  Begoon&#149;talk 07:02, 24 September 2010 (UTC)

Consensus isn't an issue in this poll. Jimbo has made it clear that it's strictly a numerical vote and majority rules. How could it be anything else when discussion is basically banned? If we did care about consensus, I largely agree with your points about the value of all opinions. I asked the question out of curiosity as to which option is more favored by actual users of the tool. This would be only one data point in evaluating the discussion and help answer the question about how difficult users have found the tool to use. All opinions would still need to be considered. The trouble is that many votes are of the form "it works" or "roll it out to everything now", without any supporting rationale. Knowing whether such voters have used the tool would let us understand just what they are complimenting. I would prefer that the voters just explain themselves more fully. However, the poll instructions discourage this, so looking at edit history would be the only way to get additional information. —UncleDouggie (talk) 09:02, 24 September 2010 (UTC)

Oh, sure. I was basically replying to Ronk's comment on this method being used to determine consensus, which is why I added the last sentence. Since I'm utterly opposed to not using consensus here, I didn't even consider whether it could be helpful in this scenario, since it didn't seem useful to imagine ways to make a (to me) utterly wrong process slightly less awful, when far more acceptable alternatives already exist, and are usually used effectively. I suspect I would still be uncomfortable with it, if I did lend it much thought, on the basis that if the process is failing to gather enough information, then it's the process that needs altering, not the rules for attributing worth to votes. I would certainly be very interested, after the event, to see the answer, it might indeed carry valuable clues as to how we better gather and assess voters' opinions to fine tune the construction of future votes - I just don't think it could, in itself, ever be used in isolation or in a "blanket" fashion to assess a vote. As I said above, there are occasions where it might be a factor in judging a particular individual's vote, but also, there are situations where it would be utterly irrelevant.  Begoon&#149;talk 10:18, 24 September 2010 (UTC)

Begoon, are you saying that comments by an individual who has never used the tool, but has heard hearsay about it are just as valid as someone like us, who has used the tool, and actually know how the system works? I think comments should be judged on their worth, that is, is this a comment by an editor who knows something about PC, or is it just someone who has heard Jeské talking about PC like it is going to make the wheels fall off and screams SHUT IT OFF!!!! incessantly? I know this sounds Citizendiumish, but we need to consider experience. And UncleDouggie, the poll instructions make it very clear that comments are welcomed. Ronk01 talk 18:21, 24 September 2010 (UTC)

I tried to very carefully explain what I was saying - sorry if it wasn't clear. I'm saying that any comment is as strong as its actual points and arguments. Once you start discounting or enhancing it with blanket parameters you are on a road that ends in places I wouldn't wish to be. Judge each contribution on its own merits. If the particular parameter you have in mind is relevant to that particular vote, then yes, apply it. If it's not - then don't. But be very careful that it really is a valid judgement for each vote you apply it to. This is all a part of what the neutral closing party needs to do. If the strength and subject matter of the vote has no need to have used the tool - such as the examples I give above, then don't apply the parameter. I'm struggling to explain it any better than that. Oh, and to answer your last point, I agree, and I would give equally low consideration to a vote that came with the rationale "SHUT IT OFF!!!!" or similar, as I would to one that says "Keep it going - makes sense", or similar  Begoon&#149;talk 19:22, 24 September 2010 (UTC)

I don't need to have pushed the button myself to have watched along and considered whether or not I would accept a revision in certain situations during the trial. I chose not to actively participate, but that doesn't limit my ability to observe for myself. Please don't assume that everyone that didn't accept a revision is basing their vote entirely on hearsay. --Onorem♠Dil 18:28, 24 September 2010 (UTC)

Unfortunately, many users one both sides are voting based on hearsay, not facts. Onorem, I agree that every user cannot be assumed to be voting on hearsay, but quite a few are. Ronk01 talk 18:54, 24 September 2010 (UTC)

Users that did not involve themselves actively in the trial have no idea how the tool actually works and their spitting the dummy process issues griping is doing the wikipedia a disservice. I suggest the opposes that oppose for such issues simply vote with their feet and stop reviewing (oh, they haven't been reviewing anyway). Oh the trust issues, the wiki will collapse, yea, drama much, simple tool, why not get over it and try it. Off2riorob (talk) 18:41, 24 September 2010 (UTC)

I've followed along. I don't need to push the button to understand how the tool works. <horrible argument>Are you a fan of rape and murder? Please don't say no unless you've tried it for yourself.</horrible argument> I understand the concept. I understand how the tool works. I didn't use it because I disagree with the concept, and don't need to push the buttons to watch an article that has PC imposed on it and think to myself whether or not a specific edit should be approved. [several statements deleted to avoid npa templates being placed on my talk page...] --Onorem♠Dil 18:54, 24 September 2010 (UTC)

Firstly, I am glad that I didn't have to template your talk page. Secondly, your example is a perfect example of the hyperbole that this page is filled with. This is an editing tool, not a capital offense! Experience with the tool is valuable in making this type of situation. Ronk01 talk 22:04, 24 September 2010 (UTC)

It was obvious hyperbole...but with a purpose that still hasn't seen a response. Please explain why I need to have actually used the tool to have an opinion. I'm not arguing against the tool from a technical perspective. How the tool works doesn't matter if you don't think the tool should exist in the first place. I understand how it works. I've simply watched it closely without approving an edit.--Onorem♠Dil 22:16, 24 September 2010 (UTC)

The answer is surely very clear; those in favour of railroading the present buggy implementation through are desperate to invalidate the opinions of those who are not. Malleus Fatuorum 22:20, 24 September 2010 (UTC)

I have just spent some time reviewing some PCs, and Lo! the scales have fallen from my eyes! I see the light...no, actually my opinion has not changed the slightest bit. Is it suddenly valid now because I have used the tool? PC works exactly the way I thought it did, I still don't have strong feelings about it, and I still think these polls have been an inappropriate manner of discussing (or not discussing) it. Revcasy (talk) 22:45, 24 September 2010 (UTC)

You have only had the tool for a couple of hours, thanks for your efforts but as a review and understanding of the tool they have little to no value at all. Off2riorob (talk) 22:52, 24 September 2010 (UTC)

I'm not saying that we discount the editors who have done research or know anything about the tool, I'm talking about people who have little to no knowledge of the tool. You wouldn't want an uninformed plumber to vote on the fate of a new surgical procedure, would you? Similarly, you wouldn't want someone who knows nothing of the tool to vote on either side, until they have done some research. That might explain some of the rather short yes votes, and similar opposes. And for those who advocate discussion, I will ask again, how do you plan to coordinate a discussion between hundreds of editors? I still haven't gotten a valid response for that question. Malleus, remember, this is a temporary implementation, not a permanent one, no one is "railroading" anything. And for those who don't think PC should exist at all, please give an argument or two. Ronk01 talk 23:32, 24 September 2010 (UTC)

An uninformed plumber

"You wouldn't want an uninformed plumber to vote on the fate of a new surgical procedure, would you?" - I think his opinion would be very likely to carry far less weight than that of qualified medical practitioners if his opinion was on a technical matter - but I certainly wouldn't automatically discount it without assessing its relevance, strength and worth. "I still haven't gotten a valid response for that question." - I gave you what I considered a valid response last time you asked the question. Now, if instead, you mean an answer you agree with, I probably can't help. But not being in agreement with an answer does not invalidate it, it just leaves you unconvinced.  Begoon&#149;talk 06:23, 25 September 2010 (UTC)

Did you miss the word "buggy" in my posting, or did you just blank it? Malleus Fatuorum 23:18, 24 September 2010 (UTC)

I absolutely agree that the current version of PC is buggy, but that isn't a reason to oppose the idea of PC. Ronk01 talk 23:32, 24 September 2010 (UTC)

And I utterly agree. But in this poll we are voting on continuing the trial until the new software is implemented - that's all - and I see it as perfectly valid in the context of this poll to hold the opinion that it is better to stop here and reassess the position since at some point it needs to get back on the "discuss and define trial, approve trial, have trial, analyse and discuss results of trial, implement feature or return to go" cycle, and since it would restore a good deal of faith and trust to do that right now, and it needs to be done at some point, I can see no time like the present. The bugginess of the version is entirely relevant to that position.  Begoon&#149;talk 06:33, 25 September 2010 (UTC)

Actually, Off2riorob, since you apparently have no respect for process or consensus, I would say that it is your opinion that is of little to no value. We are all well aware that you like PC. We are also aware that you don't care that it is being implemented despite a lack of consensus, and that you consider it to be a technicality that a trial described as being two-months long has been made indefinite without any consensus to do so. If you have so little respect for such things, why should your opinion on the tool be considered?—Kww(talk) 23:05, 24 September 2010 (UTC)

Hi, your comments are close to my position but I do have some care and consideration to some of the issues you mentioned. I ask only for my opinion as regards the tool and especially its actual usage and related benefits and negatives to be considered and respected through the total involvement I have had with using it over the last ten or eleven weeks. As an addition I can tell you also that I don't give a f***, I can use semi protection just as well as the next user. If the tool is rejected by the community I will simply move on and protect BLP articles and the Wikipedia in general as best I can with the remaining tools. Off2riorob (talk) 23:42, 24 September 2010 (UTC)

Off2:Ok, so I guess I will spend the next few weeks reviewing and then meet you back on (I am sure) yet another inappropriate poll discussion. At that time my objections will not have changed because the principle will not have changed and you can go right on ignoring it, but at least I will be entitled to an opinion then. Or is my opinion just invalid until I agree with you? That seems closer to the actual situation at this point. You started with an indefensible position, which I called you on, and now when it becomes clearer that you are absolutely wrong, you stick your fingers in your ears and go on as before. Oh well, we can continue indefinitely. Revcasy (talk) 23:54, 24 September 2010 (UTC)

Your comments are about something I am not involved in, and are as such hilarious. You have missed the actual issue.. it is not about me, regards and thanks. Off2riorob (talk) 23:58, 24 September 2010 (UTC)

No actually my comments are about your insistence that I should not be involved in this discussion because I have not used PC. Well, now I have. You were wrong. You still are. You can pretend you do not know what I am talking about, but it won't make you any less wrong. Revcasy (talk) 00:04, 25 September 2010 (UTC)

Haha, while I appreciate your assertions that I am , wrong, wrong , wrong, at least I am still in a majority. Off2riorob (talk) 00:08, 25 September 2010 (UTC)

The wrong I refer to is clearly your idea that those who have not used PC should not be allowed a valid opinion on it. I don't think that the majority of users want to disenfranchise those who disagree with you? *shrug* Your insistence on calling me out personally still has yet to be explained. I have not seen you make a list of all the other editors who's opinions you feel should be disregarded. Also, it is almost magical how it is alright according to you for those with no experience to support PC, but not the other way around. Revcasy (talk) 00:25, 25 September 2010 (UTC)

"Haha, while I appreciate your assertions that I am , wrong, wrong , wrong, at least I am still in a majority." - a happy day for consensus :) - Kingpin¹³ (talk) 05:39, 25 September 2010 (UTC)

Gentlemen, gentlemen. We need to relax here. Personally accusing one party of being wrong is not going to get anything solved. I admit I like PC as well, but there are places it doesn't work. The WMF has raised concerns about disabling PC in its current context just to switch it on for a trial of a newer version. I am therefore in support of getting additional data keeping it on until the new version. But I take offense to Revcasy's assertion that Off2riorob is wrong. That is IMO a direct attack against his beliefs, and consequently against mine to a certain extent as well. We all have our own opinions on the tool, but no one opinion is more right or wrong than the other - the exception being that PC will kill users inexplicably in its current context, THEN it would be wrong to keep it. CycloneGU (talk) 17:47, 25 September 2010 (UTC)

Off2riorob is outright wrong with statements such as: You have only had the tool for a couple of hours, thanks for your efforts but as a review and understanding of the tool they have little to no value at all. I believe that the WMF objection was to turning off the PC extension itself, not to removing articles from PC protection, which is done quite frequently for legitimate reasons. —UncleDouggie (talk) 19:37, 25 September 2010 (UTC)

That is what I was saying, that WMF didn't want to do the technical work of removing it, then readd it in two months' time. Jimbo is trying to work with that concern as well as the concerns of the community. I trust that the best interests of the community are being considered here. Many of the Keep votes may actually turn to Close after the new version is seen - I don't even know yet whether I will like the new version. I am eager to see it tho. once it's done and continue testing. CycloneGU (talk) 21:41, 25 September 2010 (UTC)

I'm confused. I think the majority of the close voters understand the benefit of keeping the extension enabled for testing purposes in advance of any trial for PCv2. It is clear in the poll wording that closing the trial means only removing the PC protection status of articles. You stated above that your are in favor of keeping PC because of the WMF's objection to turning off the extension, which isn't what is being proposed. Please clarify. Thanks. —UncleDouggie (talk) 23:03, 25 September 2010 (UTC)

I am not much of a source of information on technical wiki topics, but I understand words. The trial for PC version 1 was the time that PC1 was being observed to prior to discussing and voting on keeping or dropping PC1. That is over. Now we are in a temporary extension of running PC1 and this straw poll is to determine if the temporary extension is a good thing to keep until PC2 is ready to test on the 9th of November or the temporary extension is a mistake and PC should be shut down completely until it is restarted for the PC2 trial. I hope that helps. --Fartherred (talk) 19:46, 26 September 2010 (UTC)

Also, I never said Off2riorob was right either. I found the comment "You are wrong" out of place for a public talk page and suggestive of the beginnings of an argument that I'd like to see not take place. Both parties were attacking each other and I took offense to the one particular statement personally. Frankly, this discussion shows that perhaps there may be some people who just aren't on the same common ground and of whom neither will budge to the other ground an inch. I will say, however, that a couple of hours is generally enough time to figure out how something works; however, continued usage will give a better feel for using it, and thus a better opinion. I think Off2riorob meant that. CycloneGU (talk) 21:41, 25 September 2010 (UTC)

Fair enough. —UncleDouggie (talk) 23:03, 25 September 2010 (UTC)

Indeed, I was overly emotional, and confrontational. I apologize to Off2riorob, and the entire community taking part in this discussion. Taking things personally does nothing to raise the level of discourse or reach consensus. I would like to clarify again though, that the things I was saying--including my 'wrong's--were directed at the position I perceive Off2 to have taken in regards to the value of input from inexperienced editors and my input in particular. On the subject of Pending Changes I acknowledge (again) that I do not feel strongly one way or the other, and I understand and respect why many support the use of the tool (while also understanding the reservations of the opposition). My concern is now, and has always been, the use of straw polls in place of consensus. In any case, I am sorry. Revcasy (talk) 00:30, 26 September 2010 (UTC)

Hi, I had taken this page of my watchlist but noticed this.. I do appreciate your comments and I am sorry also especially that you felt singled out , that was never my intension, you don't really have anything to apologize to me for, It did get a bit heated on both sides, and a little emotionally involved, but I thought it was never uncivil or nasty. I am off again now, back to articles, I have contributed more than enough to this sticky issue, best regards to all. Off2riorob (talk) 06:46, 26 September 2010 (UTC)

Archiving

If there are no objections, I will set up auto archiving for this page, with threads 2 days or older being archived. Ronk01 talk 14:40, 23 September 2010 (UTC)

It is rather long, isn't it? I guess my only concerns would be that 2 days is fairly short, and anything else would be fairly pointless, and additionally that archiving may result in users coming here to discuss an issue already discussed, and, not finding it here, restart the whole discussion. I suppose that adds up to "I don't know", really :-)  Begoon&#149;talk 14:55, 23 September 2010 (UTC)

I suppose the best thing we can do is wait and see what other people think. Perhaps 3 days, with a notice to inform editors of the archiving, so they don't repeat topics? Ronk01 talk 15:45, 23 September 2010 (UTC)

This poll will close on Sunday or Monday (see the poll page). After that, discussion on this page should cease. Archiving is not necessary IMO unless you want to archive a few discussions manually, then later reinsert them on this page after the poll is closed. CycloneGU (talk) 16:02, 23 September 2010 (UTC)

No need to archive. It is important that voters can easily see all these discussions. This poll has a limited life anyway. —UncleDouggie (talk) 21:43, 23 September 2010 (UTC)

Remove "closed by majority vote" from poll

I'd like to reinstate an edit made by NW remove the "closed by majority vote" from the poll as WP:NOTDEM makes it pretty clear we don't operate that way. Thoughts? Hobit (talk) 15:20, 23 September 2010 (UTC)

Unfortunately, in this case Jimbo has decided that this poll will be decided by a simple majority. I think the wording is necessary to make it even clearer to people how this poll is being conducted in a manner which contravenes established policy. Revcasy (talk) 15:28, 23 September 2010 (UTC)

(edit conflict)While we traditionally don't operate that way, unless something changes, that's exactly the way that "consensus" is going to be determined in this case. I think it's a plus to let people know that as they cast their votes. (Not !votes, but votes) It doesn't do any good to remove the statement when that's how the decision is planned to be made. --Onorem♠Dil 15:33, 23 September 2010 (UTC)

Depends which part of Jimbo's talk page you want to quote, e.g. during earlier discussions he said:

"The goal is to get to 80%+ eventually of either support or oppose. Along the way, the vote results will determine what to do...

50% - we turn it off and go back to the drawing board as to whether or not PC is what we want to be doing at all anyway
50%-66% - we turn it off temporarily and ask the Foundation to go back to the drawing board
66%-80% - we keep it on but ask for further revisions and a v3 poll

80%+ - we keep it as a stable permanent feature but also welcome improvements over time, as with anything"

Quote from here Jebus989^✰ 15:48, 23 September 2010 (UTC)

I'd honestly be against removal of that - the comment you quote is prior to his announcement that this poll would be simple majority. I think he views this poll as outside of his voting bands - but it has become rather hard to follow  Begoon&#149;talk 16:42, 23 September 2010 (UTC)

The numbers really don't matter. There's always room for another "poll" if the results aren't what they're looking for...but since they'll obviously hit 51% here, we should be able to ignore the farce for a couple months. --Onorem♠Dil 16:53, 23 September 2010 (UTC)

• In short, Jimbo wants these few pages under PC for a few weeks so much that he is willing to ignore our policy on what consensus is and break the promises with which this test began. A few people on this talk page support this attitude. For my part: fool me once, shame on you; fool me twice, shame on me. Septentrionalis PMAnderson 17:00, 23 September 2010 (UTC)

Um, no, those percentages listed above were for the original poll about PC. Jimbo considered it close enough to 66% to leave it on, but felt that (since some changes were going to be made by the time of its more full introduction in November) this straw poll should be made to determine whether PC should be left on in the interim. Silverseren^C 17:03, 23 September 2010 (UTC)

For ther record they were made at 21:33, 12 September 2010 (UTC), 8 days after the close of the first poll Jebus989^✰ 17:54, 23 September 2010 (UTC)

Yes. Those percentages were not something that Jimbo has stated before the poll, but he said that they were what he was going by and looking for in the results of the poll. The time of when he stated them doesn't really matter. Silverseren^C 18:00, 23 September 2010 (UTC)

I was trying to give you a gentle prod to actually read the section I'm quoting, if you so you'll notice it starts with Jimbo suggesting for the next poll, Why not simply have two voting options? support and oppose. The goal is to get to 80%+ eventually of either support or oppose. Along the way, the vote results will determine what to do... edit: I do obviously realise he later went on to ask for a majority decision, I'm just showing how things have changed in a fairly short time period and so this query is not all that unreasonable Jebus989^✰ 18:07, 23 September 2010 (UTC)

His meaning was quite clear to me. He is doing this poll as a majority decision, but over time through all of the testing, he eventually wants to get to an 80% decision either way regarding keep or scrap. That 80% has nothing to do with this poll. CycloneGU (talk) 00:48, 24 September 2010 (UTC)

Cyclone has it right. The quote of me above, talking about voting ranges, is a discussion strictly with respect to the vote at the end of the test of the second version. My desire is that we make absolutely certain that, unlike the previous vote in which people felt there were "promises" made (though not by me), and some assumptions made (about whether and when the system would be turned off) that we go into the next real vote after testing the next version with easy and absolute confidence of what is going to happen next.--Jimbo Wales (talk) 12:22, 25 September 2010 (UTC)

Replacing 50% with consensus

OK, let me instead propose replacing that with "This poll will be closed by an uninvolved admin who will determine WP:CONSENSUS as normal." I realize Jimbo put the 50% bar there, but I think we can replace it given consensus on doing so. Thoughts? Hobit (talk) 17:04, 23 September 2010 (UTC)

Consensus doesn't mean 50%, unless you are going to say that the closing admin has to choose "Yes or No" and that there's no such thing as No Consensus in this case. But then, how is that any different from just deciding by a 50% minimum? Silverseren^C 17:07, 23 September 2010 (UTC)

What I'm shooting for is that without consensus we'll go with the last thing that had consensus--the 2 month trial which is now over. Suggestions on wording? Hobit (talk) 17:12, 23 September 2010 (UTC)

So...if your "No Consensus" is, essentially, a "No", then isn't this proposition going directly against what Jimbo wanted for this straw poll (and how he made it)? Silverseren^C 17:16, 23 September 2010 (UTC)

That was my suggestion. I think it's an action the community could take, but clearly there isn't support for it. Hobit (talk) 20:18, 23 September 2010 (UTC)

I honestly don't think it's a good thing to do, Hobit. Jimbo said that would be the criterion, and the message warns/informs people of that. That's how it will be closed, as far as I can tell, and much as we may wish it were different, adding what is likely to be incorrect information will do nothing but further confuse, imo.  Begoon&#149;talk 17:26, 23 September 2010 (UTC)

Agree with Begoon, it would only confuse voters. Ronk01 talk 19:02, 23 September 2010 (UTC)

Noted. Consider it withdrawn. Hobit (talk) 20:18, 23 September 2010 (UTC)

I think this was unnecessary. He wasn't 'attempting to change' it, just proposed a change and opened a discussion. Besides, we all know Jimbo is following this poll and discussion as and when time allows Jebus989^✰ 20:57, 23 September 2010 (UTC)

Perhaps my wordsmithing was a bit faulty, but, as you can see, he hasn't been here in awhile. I determined that he needed to know, since it would have been a major change in the polling process if executed. Ronk01 talk 21:58, 23 September 2010 (UTC)

I'm sorry I wasn't online much the past couple of days, due to meetings in New York. But I am watching everything very closely.--Jimbo Wales (talk) 12:22, 25 September 2010 (UTC)

Well, if you really want to say to everyone who has voted "sorry, we lied, the poll won't be closed in the way it said right from the start, we've decided to move the goalposts" go ahead. I am sick and fucking tired of the downright dishonesty that passes for "consensus building" around here. The poll was opened as strict majority, why fuck around with it now we are over half way through? Are people actually insane or just deliberstely trying to piss people off? DuncanHill (talk) 00:54, 24 September 2010 (UTC)

• I think it was fair to see if others felt it should be changed. If something is wrong, being wrong for a long time is hardly a reason not to change it.... Hobit (talk) 01:51, 24 September 2010 (UTC)
• I'm of mixed feelings. It is a goalpost change, but it seems to be a direct response to those that say "sorry, we lied, the trial won't be closed in the way it said right from the start, we've decided to move the goalposts." This whole thing is a clusterfuck of unprecedented proportions.—Kww(talk) 14:50, 24 September 2010 (UTC)

The original trial was closed with enough support found to continue PC on a temporary basis. The original proposal NEVER established the only options to be just "turn it off" and "keep it permanently". BigK HeX (talk) 13:40, 25 September 2010 (UTC)

Excuse me, but that is just blatantly false: the original trial was never closed, and that was, is, and always will be the basis of my fury.—Kww(talk) 15:54, 25 September 2010 (UTC)

There's nothing blatantly false about it. You may not have liked the conclusion of the trial and the messy straw poll dsicussion, but Jimbo very definitely came to a conclusion about it, and that conclusion was for a temporary continuance of PC. You may not agree with the result of a "temporary continuance" as being one of the valid outcomes of the previous trial/straw poll/discussion, but that was the result we got. I haven't heard many reasoned objections as to why "temporary continuance" is treated as such an unfathomable option, nor why it would be unreasonable given the significant majority support for PC. BigK HeX (talk) 16:21, 25 September 2010 (UTC)

The trial was never stopped, which means it was never closed. That's a very simple concept, and I have to believe your failure to acknowledge it is willful.—Kww(talk) 16:27, 25 September 2010 (UTC)

Alternatively, another very simple concept is that Pending Changes does NOT have to be removed from every article for a trial to have made it to a conclusion. BigK HeX (talk) 16:32, 25 September 2010 (UTC)

That simple concept is a simple falsehood. Trials don't come to a conclusion by being indefinitely continued without consensus. That's called indefinite extension without consensus. The trial was never brought to a conclusion.—Kww(talk) 16:41, 25 September 2010 (UTC)

You're free to characterize a decision made for temporary continuance based on a discussion evidencing significant majority support and set to be reviewed at an explicitly specified date as an "indefinite extension without consensus". Cheers. BigK HeX (talk) 16:50, 25 September 2010 (UTC)

That was one man's decision made after the trial was not stopped at the appointed time. Jimbo has the right to force things, but that doesn't make them consensual. Regardless, it was made well after the trial was supposed to have stopped. You really should take the time to describe things truthfully.—Kww(talk) 17:25, 25 September 2010 (UTC)

Seems like there are more productive lines of discourse than to insinuate that someone engages in "willful" falsehood merely because they do not share your continued conflation between "concluding a trial" and "removing PC from all articles". It appears that you refuse to acknowledge that there was a trial period which ended in a discussion, and a discussion that ended in significant majority support, and a reading of the discussion that determined a temporary continuance of PC to be the reasonable way to respect the points made in the discussion of the original trial. It doesn't seem fruitful to ignore that there WAS indeed a discussion following the trial period, and that a decision WAS indeed drawn from the discussion on how to move forward from the original trial. However, if the complete removal of PC from all articles is the only event that you can recognize, then there's not much left to be said. Good day to you. BigK HeX (talk) 17:49, 25 September 2010 (UTC)

• I don't think it would have been correct to alter it, as it would be misleading, and much as we might like to do something like that to make a point, it wouldn't really be appropriate. I also see nothing at all wrong with Hobit bringing it here for an opinion, and then accepting the consensus - seems to me exactly the correct way to do things...  Begoon&#149;talk 19:40, 24 September 2010 (UTC)
• To be clear, my primary goal is that we not have another "clusterfuck of unprecedented proportions", in Kww's colorful phrase. We will know with absolute certainty, to the maximum degree humanly possible, what will happen and when, at the end of the trial of the next version. The goal is to get to very strong consensus in the end, one way or the other.--Jimbo Wales (talk) 12:22, 25 September 2010 (UTC)

  And what if there is none? Shutting 30 to 40% of the community up to chase libel ghosts? Artem Karimov (talk | edits) 13:08, 27 September 2010 (UTC)

November trial

Will there be a discussion on holding that trial? If so, the wording of the poll is highly misleading (indicating a trial will happen) if not, why not? Hobit (talk) 15:26, 23 September 2010 (UTC)

from what i can tell the November trial will happen no matter what. as for a discussion, that would be after the trial.--intelati1^(Call) 15:28, 23 September 2010 (UTC)

I would prefer that there was consensus (real consensus) for a second trial, but that's unlikely. Ronk01 talk 15:43, 23 September 2010 (UTC)

I think there are way too many possible benefits that can be provided by pending changes to NOT have a trial. Also, I think the WMF is eager to show those currently opposed that their concerns ARE being addressed; therefore, having a poll and having those people just vote against the second trial and derailing it would waste the WMF's time and work. So I think having the new trial is just a matter of course at this time, even though discussion would definitely be beneficial as long as it didn't become anything like this last month. CycloneGU (talk) 16:05, 23 September 2010 (UTC)

I'll also assume there will be an ongoing conversation during the two months between this poll and the trial. For suggestions and (uhh) such.--intelati^(Call) 16:35, 23 September 2010 (UTC)

Yes, there will be a trial of the new version, and then a poll at the end of the trial, with the results of that poll determining what will happen next, depending on level of support. We have time to work on the wording of that poll to make sure that it is a reasonable middle path that most people are comfortable with.--Jimbo Wales (talk) 12:24, 25 September 2010 (UTC)

Related to this, this is something I've just been discussing with Begoon on my talkpage. Our concerns are broadly that we need to nail down, not just the poll, but what it is we hope to achieve with the trial. It's a lengthy discussion, but I summarised it as: "We need to sort out what we're measuring, how we measure it, how we report it, how we discuss it, and how we decide what to do with it. And we need to do this before the next trial." Note that Begoon and I took different sides in this poll, though we're both broadly in support of PC. TFOWR 12:38, 25 September 2010 (UTC)

Well, firstly, thanks to TFOWR for the discussion you linked. A very pleasant experience in this climate to have a productive discourse with someone who (god forbid) voted the opposite way. I stole a bit of that discussion to post above, as a summary of my position: I see it as perfectly valid in the context of this poll to hold the opinion that it is better to stop here and reassess the position since at some point it needs to get back on the "discuss and define trial, approve trial, have trial, analyse and discuss results of trial, implement feature or return to go" cycle, and since it would restore a good deal of faith and trust to do that right now, and it needs to be done at some point, I can see no time like the present. There are many usability and policy concerns, plus unexplored scalability issues that it is essential to trial with the next version, and to start the next trial before there is a concrete plan for it seems unwise. There is no deadline except for those we arbitrarily impose on ourselves, and if we arbitrarily impose such deadlines in such a way that the process of implementation is damaged - well, that's just silly. I strongly support PC, but let's do it right.  Begoon&#149;talk 13:54, 25 September 2010 (UTC)

Support —UncleDouggie (talk) 20:09, 25 September 2010 (UTC)

Support Let's get this right. Ronk01 talk 21:00, 25 September 2010 (UTC)

Thanks guys, very much appreciated. It's a cause of huge frustration to me that much of our time here seems to be wasted with discussions on pro/con entrenched positions when in fact our priority needs to be getting the whole thing back on track properly, with the minimum anger/frustration or loss of long term trust, and this just seems the best chance and time for doing that, to me at least.  Begoon&#149;talk 01:20, 26 September 2010 (UTC)

Thank goodness I am not in charge

It seems that our behavior on Wikipedia cannot be completely prescribed by a set of rules. Let us be nice to one another and accept that we will not get our own way in every case. If one is determined to be fussy enough, one can be dissatisfied continuously. I commend the WMF for allowing people to have their say. I know that the great many words cannot all amount to much because in the end only one thing or another will be done. This straw poll may have been a mistake considering the amount of acrimony revealed, but it was done in good faith. Let us not think too much of our own opinions and graciously accept what good has been offered us in the more than three million articles available on Wikipedia. We can ignore the unavoidable inconsistencies in policy, the bureaucratic mistakes and the articles we do not enjoy. Let all of the unimportant Wikipedians like myself just wait and see what happens and not complain too much. --Fartherred (talk) 17:34, 23 September 2010 (UTC)

Support :D--intelati^(Call) 17:35, 23 September 2010 (UTC)

Support also, though I disagree with the words "unimportant Wikipedians". Anyone who contributes positively is an important Wikipedian. CycloneGU (talk) 21:42, 23 September 2010 (UTC)

I believe "unimportant" is used in a sarcastic tone.--intelati^(Call) 21:55, 23 September 2010 (UTC)

Point taken. CycloneGU (talk) 22:48, 23 September 2010 (UTC)

Support. AGF, relax, feel special to be part of this amazing endeavour. Geometry guy 22:30, 23 September 2010 (UTC)

Is that supposed to be sarcasm? Almost impossible to tell the difference between the best sarcasm and someone just honestly speaking their mind when all you get to go by are written words. In case it is not, I feel compelled to point out that three million articles were not "offered to us". Neither did they come into existance by everyone adopting "wait and see". Every editor who ever did take action and added one good edit to an article helped create these three million. --Xeeron (talk) 23:29, 23 September 2010 (UTC)

It is my belief that sarcasm was not intended, but I'll let the user defend himself as only he can confirm. CycloneGU (talk) 23:49, 23 September 2010 (UTC)

The word unimportant is ambiguous in this case. I was thinking of my importance in determining administrative and policy directions at Wikipedia. A few of us have special importance in determining policy and the technical methods that make Wikipedia work. Most of us must share our importance with 13,099,589 other registered users. Even if I would rank myself a little higher than the absolute lowest level of importance in policy determination, and I do, that still leaves me with a rather small share of importance. --Fartherred (talk) 09:09, 24 September 2010 (UTC)

Wikipedia offers millions of articles to anyone with a computer and an internet connection, registered and unregistered users alike, contributors and noncontributors. The articles are not forced upon us. We must seek them out, search through the labyrinth of categories and links and find those jewels of articles that are precious to us. Without the Wikipedia organization and other wikis, such articles would not be available to be found. If we participate in producing this offering, we are offered some of the fruit of our own labor, but any one person's contributions are such a small fraction of the whole that it is difficult to measure.

As awe inspiring as Wikipedia is, it still does not put bread on the table or clothes on the back of most Wikipedians. In this sense Wikipedia is unimportant to me. I do not need to feel threatened if things go awry at Wikipedia. Actually, my financial situation might improve if I stop spending so much time here. So I do not get upset if Wikipedia follows irrational and counterproductive policies. If it collapses under its own weight, that is someone else's worry. --Fartherred (talk) 10:02, 24 September 2010 (UTC)

What rubbish. Certainly, Wikipedia doesn't put bread on the table or clothes on the back of those who contribute to it. What those of us who do try to write content for it are at least hoping for, is that by providing a reliable and freely available source of information, we can have a material positive impact on peoples' lives. ("I’m doing this for the child in Africa who is going to use free textbooks and reference works produced by our community and find a solution to the crushing poverty that surrounds him. But for this child, a website on the Internet is not enough; we need to find ways to get our work to people in a form they can actually use. And I’m doing this for my own daughter, who I hope will grow up in a world where culture is free, not proprietary, where control of knowledge is in the hands of people everywhere, with basic works they can adopt, modify, and share freely without asking permission from anyone.") If you're not concerned that the unique culture that's made this possible could collapse, then frankly you shouldn't be here. (BTW, you're actually "sharing your importance" with 3,868 other users, not 13,000,000—"number of registered accounts"≠"number of active users".) – iridescent 10:18, 24 September 2010 (UTC)

My contributions to this thread were called rubbish. I hope to show that there are nuggets of wisdom in the heap. After the poll on whether to keep PC revealed complaints as well as support, the decision was made to try to correct faults. Since there would be difficulties in changing editing practices to stop PC then start it again in a short time and associated difficulties with starting and stopping software, It was decided to use a poll ask the community's indulgence for an unplanned continuation until the software bug fixes were ready for trial. This was jumped on as a broken promise by some people. It does not seem to me to rate claims of a broken promise. Just vote down the continuation if you do not like it and it seems to me the PC will be stopped and restarted. If there is a consensus to indulge a continuation, that is just the way things are done at Wikipedia. We should accept it. Why make the claims in advance that it is expected that consensus will not be followed? Wait and see if there is a good reason for complaining before complaining. For myself I will wait and see and not complain either way except to complain about complaining. Luckily I do not have a stick that I can beat Wikipedians with. So I am not tempted to say be nice or I will beat you with a stick. I did not write that I do not care if Wikipedia collapses only that it is not my worry. It is like my death. I try to hold off death with reasonable effort, but having made that effort I do not worry about it. It is out of my hands.

For a comment on the importance of our own opinions see: Six Blind Men & The Elephant. It was written about religious arguments but it seems to fit.

User A1 and iridescent have made a conspicuous display of asking for revokation of reviewer status and granting it. It is not surprising after User A1's comments that he or she is unwilling to review articles. I do not know if this will lead many Wikipedians to follow User A1's example. I do not know if WMF will be able to make a version of PC work. I will wait. Knowledge will come. --Fartherred (talk) 06:14, 25 September 2010 (UTC)

November US elections

Guys, I really think that one of the reasons for keeping the interim usage may well be to prevent a lot of articles relating to politicians running for office and certain related issues from getting regularly vandalized by partisans. In all honesty, I personally think that instituting a pending changes function around the time of major elections in any English-speaking democratic country is probably a good thing, because it reduces the likelihood of a lot of people having to look over dubious sources to determine if the possibly inflammatory material pending inclusion is well enough sourced to be included, particularly when the material might well have an impact on the elections. I personally very much think that doing whatever we can to help ensure that we do not get misused by partisans for political purposes right before any major contented or contentious election is probably a good thing. John Carter (talk) 15:06, 24 September 2010 (UTC)

I think you're missing the point; the problem is that, regardless of the pros and cons in theory, the current implementation does not work on high-traffic pages. (See the Bam Margera discussion above.) In any case, around the time of major elections in any English-speaking democratic country translates as "always"—do you realise just how many English-speaking countries there are? – iridescent 15:12, 24 September 2010 (UTC)

Yes, I do realize how many there are, and, in all honesty, if we were to place pending changes protection on articles about national elections in any of those countries for a few months before those countries hold contentious national elections, that may well be reason enough for a lot of less-than-honest political partisans to attempt vandalism. And I was primarily thinking of countries like the US, UK, Ireland, Canada, Australia, and New Zealand, where there are a lot of people willing to engage in such vandalism. And I definitely know that I would hate to find out that false or poorly-sourced information we included in an article might change the outcome of an election, particularly if it winds up being an important one. John Carter (talk) 15:21, 24 September 2010 (UTC)

(ec) I agree. I am a bit concerned about a possible side effect, though. You can imagine that IPs who see that their edits are not geting through, will sign up for an account when without the PC system, they wouldn't have done that. So, we could get a larger influx of new Wiki-editors with a profile that makes them less likely to become good Wikipedians.

So, perhaps we also need to make other changes in policies, like restricting all new editors to 1RR for their first 1000 edits. If they violate 1 RR or are otherwise engaged in edit warring, they should be given a warning and the edit count for lifting the 1RR will be reset to zero. Count Iblis (talk) 15:22, 24 September 2010 (UTC)

No objections to adding other changes as well, or in changing the nature of the existing system elsewhere, particularly regarding whether newly registered users automatically get "grandfathered in". However, I still think that, with the very contentious US election which is upcoming, and the likelihood that some political operatives would be tempted to see us as a platform for less-than-ethical behavior, we are at least temporarily better off making that harder for them. Like I said elsewhere some time ago, some years ago I was a bit of a political operative myself, and was involved in an election which was pretty filthy from both sides. I've known for some time this site would be more effective than what we did then, and I hope we don't give people a chance to prove me right. John Carter (talk) 16:00, 24 September 2010 (UTC)

I thought that the point of "pending changes" was to encourage newer editors to participate under supervision. If you're going to set up new rules to restrict their editing, perhaps its purpose is superseded. Wnt (talk) 00:49, 25 September 2010 (UTC)

No, this is a problem for which PC is demonstrably not a solution. Barack Obama was moved back to semi-protection while the trial was still underway; the backlog became unmanageable - and his election is two years off. Other elected officials will have the same problem when elections roll around (there may be fewer vandals - or there may not; but there will also be fewer watchers and confident reviewers. Septentrionalis PMAnderson 16:57, 25 September 2010 (UTC)

It has nothing to do with elections; rather it has to do with their controversies and notoriety. George W. Bush was also put under PC and also had to be returned to semi-protection because of extreme vandalism. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:39, 25 September 2010 (UTC)

And does anybody seriously suppose that Barbara Boxer (and in all likelihood Carly Fiorina) won't have the same problem in a month's time? Septentrionalis PMAnderson 22:39, 25 September 2010 (UTC)

Request revocation of reviewer status

I wish to request that my "reviewer" status be revoked; who can do this? User A1 (talk) 20:20, 24 September 2010 (UTC)

✓ Done. If you want them back, let me know. – iridescent 20:23, 24 September 2010 (UTC)

Thankyou. User A1 (talk) 20:29, 24 September 2010 (UTC)

2:1

Some people voting seem to think that 283/192 is 2:1. It is not, it is closer to 3:2. SpinningSpark 10:26, 26 September 2010 (UTC)

No it is not, it is closer to 28:19. Geometry guy 11:47, 26 September 2010 (UTC)

LOL. Arteyu ? _{Blame it on me !} 12:09, 26 September 2010 (UTC)

No, it's much, MUCH closer to...wait, come again? =P I do know the percentage is now about 59.58%. Not a firm consensus either way. =/ CycloneGU (talk) 19:58, 26 September 2010 (UTC)

For an encore, I suggest a new straw poll on "What does no consensus mean at the straw poll on interim usage?" (as this is disputed). Infinite regress is cool, as are continued fractions. Geometry guy 23:41, 26 September 2010 (UTC)

• Support :D--intelati^(Call) 23:42, 26 September 2010 (UTC)

Vandalism Detection Competition

I've written up a brief overview of the recent vandalism detection competition on the page with my proposed compromise for pending changes.This may be useful in plotting the future course for PC. I would appreciate for any comments to be left on the proposal page to keep things together. Thanks. —UncleDouggie (talk) 00:22, 27 September 2010 (UTC)

Ok

289-199, we'll keep using and testing PC in a very very limited way as described in the poll, until such time as the Foundation rolls out the next version around November 9th. If they delay until December 31st, we'll hold a new poll to see whether to continue using it at that time (but let's not debate about this right now, let's wait until then, so we can do so with some actual information).

The things to be figured out now will be details about how we will use it for now. And we have plenty of time now for a reasoned and detailed discussion of the parameters for the closing of the next trial. I most particularly want to hear from those who voted "Oppose the poll" - now is your chance to get things back on track by helping to specify a reasonable set of parameters for the next trial. Let's do so in good faith and with all of us working towards a harmonious and sensible conclusion, please.--Jimbo Wales (talk) 13:16, 27 September 2010 (UTC)

I agree that a fresh start is in order. Perhaps a change of venue as well? Community participation in discussion is bound to be higher elsewhere than on a closed poll's talk page. Moving would also allow starting with a blank slate (perhaps improving moods), without all of that history hovering up there. Revcasy (talk) 13:28, 27 September 2010 (UTC)

You have my full support in that. There are two questions, I guess, so probably there should be two discussions, on different pages. The one about what to do in the interim should be clearly marked with absolutely clear warnings that this is ONLY about what to do in the interim, and that it sets NO precedent of any kind for future use. The one about the next poll should make it clear that there is a strong desire on the part of many supporters and many opponents of PC that the next poll have absolute clarity around what happens next, and that I will endorse a specific wording of the poll based on a thoughtful consensus-seeking discussion about that.--Jimbo Wales (talk) 13:35, 27 September 2010 (UTC)

• The first thing is that a clear end date must be set, and that no pending changes continuation be permitted to outlive the trial. I don't care if it's a bot, or if it's a patch that says "if the ending time of the protection is greater than the end of the trial, set the ending time to the end date of the trial", but there can't be any more of these occurrences of people refusing to honor the end of a trial.—Kww(talk) 14:44, 27 September 2010 (UTC)

• How very unfortunate that the fact that there has been a 10% drop in support for continuing this trial in just two weeks has been glossed over, or that a very significant portion of the support votes are philosophical supports for a trial that is not taking place and has not been proposed (i.e., applying this to BLPs), or that the number of articles currently enrolled in the trial is down more than 30% from its peak, at only about 950 articles.

  In the next six weeks, it is very important that the community re-examine the criteria for any future trial. As well, the revised software must be available for testing on the test wiki at least a week prior to its introduction in exactly the configuration proposed for the second trial. That is, if there is sufficient interest to do so; very few administrators have added any articles in the past month, while many articles have been removed from the trial because pending changes has not been helpful.[2] It would really have been better to withdraw the software now, with a complete rethinking and proper redesign uploaded on 15 January 2011 (giving the developers enough time to do things right and test it, and allowing the community to make an informed and considered decision on what exactly should be included in any trial). And the bot to convert articles back to semi-protection, which would have taken less than 24 hours to write, be approved, and complete the task, should be lined up for the end-date of the next trial, which should be no longer than two months. Risker (talk) 14:58, 27 September 2010 (UTC)

I'd support for a broader use. Anonymous IP's seem to pick up articles on PC to show everybody that they r stronger. We need a better statistic: The difference between before n during PC. How many constructive edits of anonymous IP's, before n during PC. How much vandalism of anonymous IP's, before n during PC. If all vandalism goes to articles on PC protection, then do the other articles get less vandalism ??? --Chris.urs-o (talk) 15:02, 27 September 2010 (UTC)

Nobody knows, because nobody has analysed the results of the first trial. (There's been a tiny bit of analysis but not any significant review,seen here.) Further, the software requires acceptance of an edit before it can be reverted or removed, so the data will report lower than actual reversion rates if another edit is accepted before an IP's revision is removed. Risker (talk) 15:14, 27 September 2010 (UTC)

Oh, and it is not possible to determine how many constructive IP edits were made before the initiation of the trial, because the trial was almost completely focused on articles that were under long-term semi-protection. Some others were added in along the way, mainly those involving current events which had insufficient past history to determine any change in pattern. Risker (talk) 15:26, 27 September 2010 (UTC)

That was dishonest on your side, Jimbo. 2 months is 2 months. And you also forgot about WP:POLL. Artem Karimov (talk | edits) 15:58, 27 September 2010 (UTC)

If the anonymous IP vandals decided to destroy Wikipedia, and PC protection in on their way, then, if we do nothing, there is no Wikipedia anymore. Formalisms do not help to defend Wikipedia. --Chris.urs-o (talk) 16:13, 27 September 2010 (UTC)

I do not think it is a bad thing for PC to be removed from articles for which it has not been helpful. Having a small number of articles running under PC keeps a small number of editors in practice using it.

Should we emphasize that people who do not like PC are not required to be reviewers as a condition for editing? Where PC does not work well it can be removed from that article. I fail to understand the reason for the intensity of the opposition to PC. The whole thing might be scrapped as unhelpful after we see the new version. We should not let a few editors vaguely define principals after the fact that have been violated. Vague claims of dishonesty are disurptive. We should strive for clear communication. We should discuss what we intend beforehand. Then if our actions match our words, people who complain can be told it would have been better to make such complaints in the discussion. Let us agree that we want to be a community that can operate by consensus. --Fartherred (talk) 16:19, 27 September 2010 (UTC) -- altered --Fartherred (talk) 16:21, 27 September 2010 (UTC)

It's because the user interface is awful, the software facilitates the recording of vandalism and BLP violations in page histories because the edit is already part of the article before it can be removed, and the trial was poorly considered and implemented (not to knock those who did it, but they only had a few days' notice to set things up). This was intended as a proof-of-concept trial, there have been some significant holes in the concept identified, and keeping bad software running when there are other alternatives is generally considered to be a pretty poor systems engineering decision. Risker (talk) 16:32, 27 September 2010 (UTC)

You shouldn't emphasize that when it's bordering on misinformation. The PC feature is planned to exclude autoconfirmed users from making direct edits to the most protected pages. They will only be able to make pending changes. This contrasts with the current state where autoconfirmed users can edit semiprotected pages freely. The prospect of making it harder for autoconfirmed users to edit articles is one of the things that makes me oppose the feature in general. The other is the demonstrated disregard for consensus. Sakkura (talk) 16:33, 27 September 2010 (UTC)

It is not about PC only. What it is about is the undermining of Wikipedia basic principles like:

a) The encyclopaedia that everyone can edit

b) Wikipedia is not a popularity contest and

c) Wikipedia is not a bureaucracy. Artem Karimov (talk | edits) 17:08, 27 September 2010 (UTC)

Yeah, yeah, the fortress under siege. Remember that Wikipedia was built by anons and a large share of edits belongs to the unregistered. Turning this share away is a grave net negative. Also never forget to assume the good faith of the anonymous contributors. Artem Karimov (talk | edits) 17:08, 27 September 2010 (UTC)

So semi (which totally excludes anons) should be used instead? A bit of a logical fallacy there. Ronk01 talk 19:15, 27 September 2010 (UTC)

You forget that PC (and FR in general) increases access for all IPs, good or ill. PC isn't the magic bullet it's being made out to be by supporters, since anything that improves the ability for good IPs to edit does the same for bum IPs. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:50, 27 September 2010 (UTC)

I agree, but it is important that good IP's are able to edit semi-protected pages (from past discussions with you I am fairly sure we agree here) The question is "how?" PC can work for backwater pages, but that really isn't enough. A user has suggested a combination of PC with an aggressive edit filter to prevent some of the issues that PC has on more prominent pages. Ronk01 talk 21:20, 27 September 2010 (UTC)

An edit filter is only aggressive if it's manned 24/7. Once the details of an edit filter become known (whether thru brute-force attacks or a leak) then it becomes worthless as an antivandalism measure until altered. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 01:48, 28 September 2010 (UTC)

It was a time when a few programers did beautiful voluntary work for a crazy idea. Now Wikipedia is a source of neutral and serious information. A kind of bad anonymous IP vandals do not like this at all. --Chris.urs-o (talk) 19:19, 27 September 2010 (UTC)

• As one who opposed & someone with reviewer right, i will do my very to avoid articles under PC leaving the burden to others who supported PC until the end of the interim period. For the next trial, we need clear rules & clear way to assess whatever this new trial is a success or not. Personally until i will only feel satisfied with PC when i will cease to have that unease feeling i had when dealing with reviewing PC edits. Now that said, every one should be also aware that there are editors to "re-convince" out there and it's has been proved that it will be way harder than just convince them. --KrebMarkt (talk) 21:46, 27 September 2010 (UTC)

• I asked above about liability issues. Editors deserve to hear an official Wikipedia position regarding whether reviewers should expect to be held legally liable for approving a libelous edit. This position should be independent of any internal Wikipedia policy issues regarding (for example) what happens to reviewers who approve clear vandalism, unless such facts are actually relevant to the legal outcome. We should also hear what response is expected (legally and/or in terms of Wikipedia policy) when reviewers encounter damaging facts in BLPs cited to print or paywalled sources. Wnt (talk) 06:41, 28 September 2010 (UTC)

Let's all BLP be semi-protected and finish this issue. --Chris.urs-o (talk) 09:05, 28 September 2010 (UTC)

That's not actually an answer. We've just had a poll to use PC, so the question now is what happens if we do. Wnt (talk) 17:47, 28 September 2010 (UTC)

I believe Artem Karimov above expresses complaints based upon conflicting definitions of words. Artem thinks a trial is not over until the software being tried is turned off. The Jimbo et al think the trial is over when the discussion and voting are complete and that turning off the software is another matter. Two editors disagreeing on the meaning of words does not mean one or both of them are dishonest. As for [[WP:POLL]] linked above, it states: "Although such polls are occasionally used and sometimes helpful, their use is often controversial and never binding." What is binding here is the decision of WMF and the developers. They have to make some decision and have decided to be guided by a poll. As for changing plans to deal with poorly thought out situations, all people need the ability to be able to change plans once in a while. We would be crippled without it. This has been done openly. It is not dishonest.

Sakkura writes that emphasizing that people who do not like PC are not required to be reviewers as a condition for editing is "bordering on misinformation". Again there is a disagreement between editors. I am not trying to misinform. I went to another computer terminal and signed on as User:156.99.55.125 and edited the Venus article which is under PC. The edit at 16:59:44 Wikipedia time was not difficult. I added a link. Making the edit as a PC delayed it a little. Sakkura, you seem concerned about being able to only make PC edits instead of editing freely in some situation. Please explain why this worries you. It will be taken into account. As for the "demonstrated disregard for consensus," how would following the will of the minority show more regard for consensus than following the will of the majority? In fact WMW cannot always even take into account the will of the majority or even inform the majority about sensitive legal issues. Having everyone discuss a violation of privacy would violate that privacy even more. WMF and the developers have not done a perfect job in every case, but certainly they have done better than I could have done even with a thousand hired administrators to help me. --Fartherred (talk) 20:50, 28 September 2010 (UTC)

The purpose of WP:Consensus is avoid having decisions divide us into a majority and minority; the way to consensus is for the discussion to head for a compromise position on which an overwhelming majority of both sides can agree. In this case, a proposal which seemed likely to do that was outlined on the talk page of rthe original poll.

We do not operate by majority vote; the majority is not entitled to its way. We often do nothing instead - and remain with the status quo until a consensus postion is found and adopted. That's policy; those who disagree with it should organize elsewhere and see if their method of writing an encyclopedia works better. It will be more decisive - and it will hemorrhage volunteers like water. Septentrionalis PMAnderson 21:10, 28 September 2010 (UTC)

WMF certainly did paint itself into a corner as far as finding a way to move beyond the trial without violating consensus. I think they just assumed that keeping PC running until the new version is ready for trial was such an obviously good idea that the Wikipedia community would just rubber stamp that. Then, after the second trial we would have a consensus for PC. I look at how it would be like to be in charge myself and see myself painting myself into a new corner every couple of months. The only difference is that I am so lovable that everyone would instantly forgive me. We could go on to discussing What's Next which just happens to be the next thread. --Fartherred (talk) 02:15, 29 September 2010 (UTC)

What's Next?

A lot of the replies here focus on past events, or on personal perception of PC. All of that has its place, but right now, it is more important to discuss the future, namely:

1. How long should the second trial run (2 months? X months? Till there is a majority to close it?)
2. At what point is consensus needed (To start the trial? To keep the trial running? Never?)

The second question is really the key one, especially if we get a repeat of the current situation (no consensus for PC, no majority against PC) at the end of the second trial. --Xeeron (talk) 17:00, 27 September 2010 (UTC)

The consensus for the PC should be no less than 75%. Otherwise we have a significant minority. Turning on the PC with 51/49 goes afoul of WP:NOTDEMOCRACY. Artem Karimov (talk | edits) 17:08, 27 September 2010 (UTC)

1. (edit conflict)Same, Two months. Then keep it on for a one week Poll (like this one. YES, or NO) just a simple majority.
2. To keep the trial running, and to figure out what to do next. Keep it?, Scrap it?...--intelati^(Call) 17:10, 27 September 2010 (UTC)

Writing strictly from what can be done on the non-developer end of things:

• First, design the trial, identifying objectives, criteria, and what will constitute success/failure. The present trial is focused on long-term semi-protected articles, but a significant number of participants in this poll clearly identified that they want to see a different type of articles targeted, so selection criteria must be developed. This is so critical that I believe the next trial should not begin until this is put in place.
• Second, the trial must have a clear end date, with understanding that it ends when it ends absent a strong consensus that the trial should continue. The community should select a specific group of editors to assess the consensus; it should not be anyone associated with the WMF or the Board of Directors, who have a vested interest.
• Third, ensure that the proposed upgrade is thoroughly tested in the configuration to be used onwiki before it is installed on this project. This will require a significant number of editors, particularly those with limited technical background, to participate.
• Fourth, ensure that data is collected and analysed throughout the trial, and that this data and analysis be easily available onwiki. This may require developer resources for data analysis, and may be outside the scope of our own project.
• Finally, give thought to the idea that those who have actually participated in the trial should have their votes more heavily weighted in whether or not it continues. That would include administrators who have added or removed PC, reviewers who have actually reviewed edits, editors who have edited articles that are on pending changes, and those who have participated in analysis and review of the results of the trial. The number of voters in this poll who are commenting strictly from a philosophical point of view and have no actual knowledge or experience has rendered this current poll largely useless.
• My two cents. Risker (talk) 17:25, 27 September 2010 (UTC)

I'm not sure if this is obvious ... but for trial purposes, we should actually be preparing to suffer additional vandalism. If we want real numbers on vandalism, then we'll have to actually have articles of different types go through various phases of unprotection and PC-protection (and possibly semi-prot as well). For statistics-gathering purposes, having articles straight PC protected for the entire duration of the trial is less helpful than for PC to not run continuously on an article for maybe more than 1 week. For the trial, it'd really help if a bot were created that could database PC protected articles (and the duration of PC assigned by an admin) and then rotate those articles through periods of unprotected status and semi-prot status.

For any future trials, it should be made explicitly clear whether temporary continuation of PC-protection on articles is to be an acceptable outcome of the trial discussion. Though realisitically, if support for PC based on personal feelings is a valid point of discussion, then temporary continuance will again be the likely outcome. Those opposed to PC are sure to see the inevitability of this (and likely already have ... which would explain much of the vehemence). Those in opposition to PC will feel trivialized if there are not criteria which could give a larger chance for all PC usage to be suspended. To respect the opposition noted in the first straw poll, at a minimum, a period of PC suspension should just be built-in to any future trial from the outset. I'd suggest that the best time for suspension would be following the 2-month experiment with active use of PC, so ... during the analysis and discussion time.

If given ideal support, I'd recommend a progression as follows:

1. 2-month experimentation with rotating articles through an unprotected state, PC-protection, and semi-prot;
2. on the 61st day, a bot removes all articles from PC-protection, and applies semi-prot of the same duration;
3. 15 days are given for analysis of the impact from PC;
4. only after time for analysis is a discussion held;
5. the community is asked where the data indicates that PC has been successful and what should be done for the future
6. discussion is reviewed
7. it is decided if there is consensus for a permanent decision on acceptance or rejection of PC
8. failing consensus, the discussion should be reviewed to decide whether PC should be suspended for X months pending further work from developers, or whether PC usage should be temporarily reinstated where analysis shows success.

I think this is as fair as can be hoped for, but if temporary continuance is an option though ... I expect that many of the vocal opponents are still going to be hesitant to go along willingly. BigK HeX (talk) 21:13, 27 September 2010 (UTC)

A small note on gathering the data: Rotating is possible, but the much better option, when aiming for a clean evaluation is a randomized experiment. That is, identify a number of articles that could fit PC, and (if PC, semi and none are the options) randomly put 1/3 to PC, 1/3 to semi and 1/3 to no protection. Compare how the articles developed in the end. --Xeeron (talk) 21:26, 27 September 2010 (UTC)

That'd work on paper, but I'd guess we can't reliably get a set of comparable articles, since there probably aren't even any decent theories on how to classify an article's susceptibility to vandalism. Given a choice between a variability in time-of-year, and variability in articles' susceptibility to vandalism, I think we have little choice. BigK HeX (talk) 22:08, 27 September 2010 (UTC)

Huh? If there is no way to choose, how did admins select the pages currently on PC? Simply let them make that selection again and then divide 1/3-1/3-1/3. --Xeeron (talk) 22:34, 27 September 2010 (UTC)

I think a consensus is needed to start any new trial. Any "trial" needs to be of a fixed duration, and the system needs to be configured to automatically shut down at the end of the trial. There can't be any repeats of the refusals to shut a trial down.—Kww(talk) 21:16, 27 September 2010 (UTC)

With the number of voters in this poll, I am afraid that I must point out a fault in your plan, There is no way an organized discussion can be coordinated between this many people. This is not a talk page proposal, this a Wiki-wide change, more than a few people will participate. Ronk01 talk 21:23, 27 September 2010 (UTC)

In my experience, there is no maximum number of people who can participate in a discussion. --Yair rand (talk) 21:31, 27 September 2010 (UTC)

But have you ever coordinated a discussion among the hundreds of people who would likely comment on a PC discussion? Discussion has its limits. Ronk01 talk 21:48, 27 September 2010 (UTC)

Ah, see, now we're back to a discussion we had earlier. If you reduce the value of discussion in relation to the number of participants, then you erode consensus the more important the discussion is. Seems a bit arse-backwards to me. RFA discussions with 200 participants are judged by consensus quite regularly in a week, so if 800 participants mean it will take 4 weeks, hey that's just how it is. Reducing the need for consensus on more important decisions feels like a unique kind of solution that ends up in very silly places to me, big decisions take longer - dropping the "push it through to an artificial WP:DEADLINE" approach is a far better plan than devaluing the most important decisions, don't you think?  Begoon&#149;talk 22:07, 27 September 2010 (UTC)

• I like Risker's points. A new trial must be a trial, with a definite end and a result that can be analyzed to determine effectiveness, broadly construed.

• We must have a target group of articles to test on. Preferably a good spread and variety, to see where it is effective.
• We must set an end date, for which there will be no extension of any sort. Two months sounds fine; anything over that seems a bit excessive and anything less seems ineffective.
• We must have a set group of users determine community consensus on how this new trial will have gone, after those two months (or however long). This group will ideally include editors from various backgrounds: technical, vandalism patrolling, admins, non-admins, etc. and should be determined by the community. I am not opposed to WMF staff being part of this group of the community chooses so. The group should be chosen before the trial, so that there is no dispute over its membership during the trial.
• We must have perfected the new version of PC before its rollout. Technical issues and concerns over performance must not stand in the way of its effectiveness; if such issues continue to plague PC, it should not be used until said issues are resolved.
• Data must be collected an analyzed. A system of doing so must be determined before the trial, so that there will be no dispute over how the data is analyzed during and after the trial. A full analysis should include views from both a technical and non-technical viewpoint, and must be completed before PC is fully implemented (if ever) and before the community once again votes on whether to keep and/or expand the use of PC on this project. The data and full analysis must be easily accessible by any user of this project.

• This is what I find necessary to begin a new, efficient, and conclusive trial over the use of PC on the English Wikipedia. Preparing all this before starting the trial should alleviate the arguments and concerns raised during the previous trial. /ƒETCHCOMMS/ 21:33, 27 September 2010 (UTC)

I agree with nearly all of this, with the addition that I think that the preset "end date" should be set for a time after the vote, and should happen based on what happens in the vote. I am a big fan as well of Risker's point that we should run this trial carefully on a set of articles based on an informed selection of articles based on what we have learned so far. I am also a strong proponent of requiring a one week running configuration of *exactly* what will be trialed, on the test server, before the launch of the trial proper. This would mean that if the Foundation delivers the software on November 9th, then the trial would start November 16th. I think 2 months is a fine time for a trial, same as before, so that would run until January 16th. Then a one week vote would take place until the 23rd. Then allow up to 48 hours for analysis of the vote (this should be super super super easy if the poll is designed properly to give simple answers and then on the 25th would be either shut down, or shift into interim use mode while we wait for the Foundation to release a new version - depending on the overall outcome of the poll.

Additionally, as of the start of the vote, January 16th, the developers should give us an update on what further revisions of the software would be available and by when so that all voting can take place under the maximum possible set of information.--Jimbo Wales (talk) 22:10, 27 September 2010 (UTC)

If the discussion is to start at the end of the 60 days, then actual analysis of whatever data is available would likely have to begin a fair while before the end of the 60 days. Personally, I'd prefer that editors be given maybe 14 days after the experimental period in order to collate and present their analyses of the data. And, then hold the community discussion on the trial. BigK HeX (talk) 22:16, 27 September 2010 (UTC)

Jimbo, I genuinely believe that November 9th is way too soon to start the next trial. That timeframe only allowed for the developers to *attempt* to address the most serious concerns about the software, and even if they are completely successful in meeting that deadline for those specific concerns, there were still many other very important ones that will have been left by the wayside. It is much better to give them more time to solve a greater number of problems. The community, as well, needs more time to prepare for the next trial, which will take more than six weeks to do. One of my major criticisms of the initial trial was that it was rushed into existence before the community had decided how to carry out the trial, without any forethought to how to tell what effect the tool was actually having. A second poorly designed trial with a second generation of software that still doesn't come close to meeting community expectations is a recipe for failure. Give people the time to do it right. Risker (talk) 22:21, 27 September 2010 (UTC)

Re: "I think that the preset "end date" should be set for a time after the vote" Absolutely not. Trials have a fixed start, a fixed stop, and then a return to the status quo. They do not linger on forever while people analyze votes and shift the passing thresholds around. The best I can do in terms of assuming good faith is that people took an easy path instead of an intentionally deceitful one, but that cannot be permitted to happen again. When the trial is over, it's over, until there is a consensus to restart it in the same or modified form.—Kww(talk) 22:37, 27 September 2010 (UTC)

Setting the end date after the vote sounds as though the same thing that happened here is going to be attempted again. Dispute over what "trial" means, whether consensus is needed for or against it, and an unending trial delay. Even if there is an absolute "turn it off no matter what" date, I have no doubt that at least a few people will try to argue that the feature should be kept on anyway. I have serious doubts over whether the PC issue will be decided by community consensus in the end, and I doubt that PC is actually going to be turned off on November 9 as the poll said. If we're actually going ahead with having yet another trial, can we please keep it as clean and honest as possible? --Yair rand (talk) 22:47, 27 September 2010 (UTC)

"Clean and honest" went by the board some time ago. You're going to have PC whether you like it or not. Malleus Fatuorum 23:54, 27 September 2010 (UTC)

• There is one more thing that I think should be done now. The administrators who entered articles into this trial should probably start reviewing those individual articles to see if the effect has been positive or negative and, if appropriate, return the articles to semi-protection or, if there has been no significant vandalism or other problems, consider lifting pending changes as unneeded. As the administrator who placed the largest number of articles into the trial[3], I know that my work is cut out for me; however, since there has been no other effective review mechanism, I think this is important. I'll try to figure out how to log my decisions in an effective way (suggestions welcome--post them on my talk page) and make them available to the community. Risker (talk) 22:03, 27 September 2010 (UTC)
  • I recommend everyone participating here, including myself, to give serious thought to all the points made by Risker here. There's an awful lot of good sense seems to be coming from that direction.  Begoon&#149;talk 22:12, 27 September 2010 (UTC)
    • I have been going back and forth in my mind about this, but I think we should give some serious thought to developing a clearer and more complete policy for reviewers about how PC should be used. In an ideal world, we could wait until we had finally decided on whether to implement PC permanently before working seriously on policy, but I think that for a trial to work properly, and give us the best chance of collecting good data, all of the reviewers need to be on the same page about how to use the tool. I know that everybody here has been in situations where they made a decision about whether to accept a change or not and had doubts or second thoughts before and after the fact. PC is supposed to be used to stop vandalism, but the distinction between vandalism and other types of problematic edits is not always clear. Also, is there a different standard for BLPs than for other articles? Should unverified (and possibly, though not certainly, libelous) information be rejected as vandalism in a BLP? What about in non-biographical articles? Is the standard simple vandalism, or something else? We need to set out guidelines for reviewers in official policy space, and preferably before the next trial starts. Revcasy (talk) 22:36, 27 September 2010 (UTC)
      • Yes, I've given exactly that opinion myself. There are many things that need to be clarified before starting another trial, to prevent it failing to address concerns. To quote myself: "it is better to stop here and reassess the position since at some point it needs to get back on the "discuss and define trial, approve trial, have trial, analyse and discuss results of trial, implement feature or return to go" cycle, and since it would restore a good deal of faith and trust to do that right now, and it needs to be done at some point, I can see no time like the present. There are many usability and policy concerns, plus unexplored scalability issues that it is essential to trial with the next version, and to start the next trial before there is a concrete plan for it seems unwise. There is no deadline except for those we arbitrarily impose on ourselves, and if we arbitrarily impose such deadlines in such a way that the process of implementation is damaged - well, that's just silly. I strongly support PC, but let's do it right." Now that it has been decreed that we won't do the "stop here" part, I don't believe that should preclude the rest of the "getting back on track" part  Begoon&#149;talk 22:46, 27 September 2010 (UTC)

A few suggestions:

• Two months is a good length for the trial, but one week is insufficient to analyze data and prepare for a meaningful poll. At least two weeks is needed for analysis and at least a week for a poll. My plan would be to have pending changes 'running' in total for three months, though data for analysis would only come only from the first two months. Then two weeks for discussion, one week for the poll, and one week to close the poll and implement the decision. Three months total (unless consensus was to continue).
• If even that timeframe has any chance of being effective, we need to determine the metrics upfront. We need to know what data we are collecting, how we will analyze it, and what the goals are. This has started but needs refinement.
• An agreed upon notion of consensus, as mentioned above. Could the two week discussion try to establish consensus and the poll itself just be a straight vote?
• The 1/3 - 1/3 - 1/3 proposal (pending-semi-none) is a good model, but we have already seen pending changes not work on pages with a lot of vandalism rates and be unnecessary on very little. We also know that there is desire to see pending changes on pages which are 'highly sensitive' rather than just 'highly vandalized' and low-watchlisted BLPs and medical articles are top on that list. In addition to using pending changes on the existing set, we could try it on a few hundred BLP articles, and a distinct set of medical article as suggested by Doc James. These results should be analyzed separately, since they serve different functions.
• There is a need to clarify policy on when to accept an edit. As long as it's not blatant vandalism? Or can reviewers use more discretion...
• Innovative ideas like integrating pending changes with the edit-filter, or with existing recent changes patrol software like Lupin's, Huggle, and STiki should be explored. All of this highbrow grunt work is a waste of time if an efficient algorithm can handle the bulk of the screening. Ocaasi 11:52, 28 September 2010 (UTC)

• Support "Then two weeks for discussion, one week for the poll, and one week to close the poll and implement the decision." Two weeks for discussion before the poll is a big step in the right direction as far as allowing consensus to develop vs. simply voting. It may not be everything I'd like to see, but it would help. Revcasy (talk) 12:03, 28 September 2010 (UTC)
• Oppose There's no reason to let it run for an extra month. Run it for two months, shut it down, and then analyze. The tool needs to be configured to automatically shut down at the end of the trial so that there can be no repeats of the problems we've had this time.—Kww(talk) 13:31, 28 September 2010 (UTC)

Well, that is a valid point. I see no reason why the trial should not be ended before discussions. If nothing else, it would allow the analysis and decision-making to take place in a calmer atmosphere than has happened here. Revcasy (talk) 13:41, 28 September 2010 (UTC)

Either way, as long as it's clear and decided ahead of time. PC supporters will find it silly to shut it down while the poll is going on. PC opposition will be skeptical if it's not. I don't particularly mind either, but if it is going to be left on for that extra month, it should be made clear up front so it is part of the plan. Ocaasi 14:49, 28 September 2010 (UTC)

I wont push 1/3-1/3-1/3 anymore after this, but I want to point out one last time that it does not matter at all for the method whether there are pages that have been previously shown to not work with PC or whether there are new pages that people want to test with PC. All that is needed is a list of pages that people think PC should be applied on in the second trial. Then randomly split that list in 3 groups, and voila, you got a randomized experiment. --Xeeron (talk) 16:13, 28 September 2010 (UTC)

• Support – Running for an extra month is valuable to the discussion. Availability of the analyzed data and insights from the discussion will certainly make users want to check out the real impacts of the pros and cons being discussed. The best way to do this is to still have PC running for the specific purpose of supporting the discussion. This is what I did with Bam Margera above. —UncleDouggie (talk) 08:58, 29 September 2010 (UTC)

One feature necessary to the test is that it be possible to shut PC down and then turn it on again, without leaving the intervening edits on any article on which it is active as a large pending change (which Tim Sterling appears to be saying is the case now). This can be worked around with a PC-removal bot, but that's a non-trivial nuisance too. Septentrionalis PMAnderson 20:32, 28 September 2010 (UTC)

Who remembers this blast from the past: "4. Any changes to the software must be gradual and reversible." - Jimbo Wales (original emphasis) The trial should not restart until a method has been devised to close down PC if required Jebus989^✰ 21:43, 28 September 2010 (UTC)

Discussion on Wiki-en-L Foundation-L

For the record: there is an extensive discussion on Wiki-en-L Foundation-L with respect to Pending Changes in which I have been quite active; unfortunately, the archives aren't quite up to date yet so it may be a few hours before those who aren't subscribed have a chance to read it. However, this is from Erik Moeller of the WMF:

How and under what conditions Pending Changes is used is up to the enwiki community. All we're doing is leaving the feature in place: the community can decide to defer its continued usage, to narrow it, to broaden it, to restrict it in the scope of a trial, or to discontinue it. We're going to base our resource allocation for future development as much as possible on the emerging consensus in the enwiki community, and we'll try to support that continuing process with data as much as possible. It's evident that these discussions are still very much in flux.

It's my strong impression from that and the remainder of the discussion on that thread that there's no intention to *ever* turn off pending changes on this project, and that there never was an intention to stick to the original requirement of consensus to *continue* the tool here. With that in mind, I'm not sure that there is much point in developing a second trial. Risker (talk) 00:05, 29 September 2010 (UTC)

Apologies to all: the email at the start of the thread was sent to multiple mailing lists and came through only on my Wiki-en-L tab, so I thought I was responding there. Instead, I was responding on Foundation-L. The archives, for those who do not subscribe, are here[4] (thread near the bottom). Risker (talk) 00:43, 29 September 2010 (UTC)

I share in Risker's disappointment, but the possibility of discontinuing it sounds very like the proposed consensus settlement of the original straw poll. Perhaps a settlement on that basis would be workable. Septentrionalis PMAnderson 00:20, 29 September 2010 (UTC)

This is the purest example of fiat I have ever seen. Whatever respect I had for the WMF has been lost. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 00:39, 29 September 2010 (UTC)

I think that's exactly right Risker; there's no intention to ever switch PC off; these interminable polls and trials are just so much political spin. Malleus Fatuorum 00:42, 29 September 2010 (UTC)

Would ArbCom be open to an request on the matter? Certainly this has the flavor of intervention by an involved admin. Septentrionalis PMAnderson 00:46, 29 September 2010 (UTC)

I couldn't dream of speaking for the Committee on this point; clearly, as the administrator who put the largest number of articles into the trial (20% of the total), I'm way too involved to even consider raising it with my colleagues. Realistically, though, I'm hard pressed to see what action the Committee would be in the position of taking. Risker (talk) 01:30, 29 September 2010 (UTC)

I think there are many people who are going to graduate from high school and find their expectations of life, especially of the professional community, shattered. Wikipedia is a business. It's filed as a not-for-profit corporation. The WMF since its inception has had the intent of gaining a legitimate and internationally recognized academic standard. The academic community, also since the inception of the WMF, have claimed that Wikipedia is flawed because its susceptible to vandalism and intentional misinformation. Pending changes, like semi-protection (also met with the same fervor by its opposition to a 'free wiki that anyone can edit') has been a step in that direction. If PC's can increase the WMF's credibility in the eyes of the academic community 1% they will go with it. To be honest, the 'I was promised and feel cheated' cries mean very little to the WMF in retrospect to the number of visitors to the site, and to their other projects in their entirety. While I'm sure they'll lose a few hundred editors over this, the literally millions of visitors to the site and the ever increasing popularity of the site I doubt it will hurt their public image and most editors will continue to contribute with in the wiki project. Mkdw^talk 00:58, 29 September 2010 (UTC)

I suggest that you think again. Your idea that businesses are somehow permitted to behave immorally simply beggars belief. Wikipedia may or may not lose a few hundred editors over this debacle, but what is perhaps more certain is that many editors like myself will just ignore any and all articles that PC is applied to. Malleus Fatuorum 01:10, 29 September 2010 (UTC)

Businesses are not permitted to behave immorally insofar as they exist under the threat of an enforcement of morality, by the regulatory authorities, lawsuit, or public protest (including internal protest). Absent that - well, consider the conduct of British Petroleum in the Gulf of Mexico. We are the public protest, which is just as well - since nobody wants the first two options, and they would have even less traction than ArbCom. Septentrionalis PMAnderson 16:35, 29 September 2010 (UTC)

I don't know what has ever given you the idea that pending changes has anything to do with being able to improve the reputation of wikipedia in the eyes of the academe, and its presence or absence won't have any effect on the number of visitors to the site. No, the real question is whether enough administrators can be persuaded to use pending changes on sufficient articles for it to ever be anything more than a very expensive trinket. We couldn't even manage to get up to 2000 articles on PC during the trial, and it's now down below 950. Risker (talk) 01:26, 29 September 2010 (UTC)

I'm going to ignore Mkdw's above comment since it is, in my opinion, flatly ridiculous and completely incorrect in every respect. Further comments on the thread seem to indicate that the WMF does not intend to force the community into using PC. (Jimbo's intentions might be different entirely, but I don't see that as being particularly relevant.) I suggest we go ahead in constructing a possible trial, and try to make it as honest and clear as possible. Clear end date upon which PC is removed, with no possibility for debate. Unambiguous guidelines. Actual discussion of the issues. And working on possible alternatives and compromises. Per the recent poll, PC will be removed on November 9, and since the new version is not to be launched until November 16 and there will be a week of having the version testable in labs, that leaves plenty of time to make sure that all points are covered and that there will be no further deception. The community is in charge entirely, we have to remember that. --Yair rand (talk) 01:28, 29 September 2010 (UTC)

I think that the Foundation comments to the effect that they won't force enwiki into using Pending Changes are best evaluated in light of the Foundation comments to the effect that they would turn off Pending Changes after two months if there wasn't consensus to continue - just a face-saving delaying tactic to be used until the situation is a fait accompli; just a way to make it so by fiat with enough wiggle room to claim it wasn't implemented by fiat. I don't think that means that Erik Möller is going to turn on PC on a bunch of pages here by fiat, but I bet that by the time we get around to the start of the "consensus" second trial, there will have been a press release from the WMF claiming that we are now actively using PC. Very disappointing. — Gavia immer (talk) 01:55, 29 September 2010 (UTC)

For the record, I'm reachable here as well. Please see my response to Risker. Where, how and to what extent PC is used on the English Wikipedia is entirely up to its community: the fact that there's a software extension running on our server cluster doesn't change that; the community can decide to narrow, broaden, or eliminate the scope of its use as it sees fit through the normal process by which it makes such decisions. That process includes Jimmy (not as a WMF official but in his historical role as a community leader), who is, as far as I can tell, trying his best to facilitate a consensus outcome.

No matter what sequence of events had played out here, there would be considerable voices of objection, concern, and frustration about the process: it's a divisive and disruptive change that's captured a lot of mindshare (arguably far greater than is justifiable by its currently very limited impact and use). The reality is that we're trying to help the community make sense of whether this particular tool works for its intended purposes (subjecting selected pages to more intense and more efficient patrolling, receiving more constructive edits on pages that are otherwise locked down). If it doesn't work and is a dead end, we need to know, so we can redirect development resources to more worthwhile endeavors. If it does work, we'll need to know, so we can make it better. That's a hard and complex question that I'd love to help find answers to, instead of having this process conversation ad nauseam. Wearing my WMF hat, obviously,-Eloquence* 02:48, 29 September 2010 (UTC)

Well, it's captured a lot of mindshare because of the manner in which it's been handled from Day One, Erik. More effort was put into the press release about this change than was put into communication with the project itself beforehand. (One questions why a press release would be required for what was intended to be a trial of software that had no guarantee of succeeding - would the WMF have issued another press release at the end of the trial if PC was resoundingly tossed out?) We barely limped through the first trial, and never even met the targeted 2000 article mark, which everyone expected would be expanded after a few weeks. Most of the changes now on bugzilla were known in the first few days of the trial (some even before that), so it's disheartening to be left with the impression that nothing was done about them in the past three months. Sure, put the next generation up on the test wiki for November 9th or 16th or whatever (and listen and act immediately on any concerns expressed there), but give us enough time to construct a proper trial that can genuinely explore this tool, and help us to line up the analytical resources needed to make an informed decision about the tool's value. Seven weeks is not enough time to do this, nor to heal some of the (very deep) wounds that have come with this. This isn't Vector, where each individual gets to decide what skin they wish to use. An article on PC is on PC for every editor, new or old. (As an aside to Yair rand, one week on the test wiki will be insufficient; none of the issues identified in the week before the trial started was corrected last time, except for turning on reviewer permission, which they'd forgotten to do, so it had gone completely untested.) Risker (talk) 03:19, 29 September 2010 (UTC)

There was no press release; there was a blog post precisely to avoid the kind of misreporting that we've seen in the past, which worked quite well. Again, it's up to you all how and when you want to use the code; all we're committing to is updating and deploying it. If you can establish consensus here for a timetable and use scenario that's more conservative, then by all means, go ahead and do it. My own take is that at this point this is much ado about very little; the wiki-world won't suffer terrible damage if there's a production test on a few pages with improved software, and it won't suffer terrible damage if there isn't.--Eloquence* 03:57, 29 September 2010 (UTC)

I am also extremely disappointed by the message this is sending. Jimbo has basically promised that the new version would be much improved with major existing bugs fixed and such, not that it would stay on forever, used or not. Now, that November (or whenever) release had better be much improved. In fact, if most of the issues haven't been fixed and the whole system and configuration thoroughly tested by the day before the launch, the launch date must be postponed until the testing and fixing is done. We're not wasting another two months with defective software and we're certainly not spending the rest of our lives with PC left on for no reason. Either fix it, and people will vote to keep it, or don't bother with a rush job and actually take it off sometime. /ƒETCHCOMMS/ 03:54, 29 September 2010 (UTC)

My disappointment stems form the fact that the trial with the buggy PCv1 is still running. We know that the next trial isn't starting on November 9, and if someone forces that date it will be another disaster. The supporters of PC should be able to see that it's in their best interest to turn off PCv1 and get their act together to do PCv2 the best they can. Risker pointed out above the need to select different articles for the next trial, which was supported by several users. I made the same point during the poll. We need to disable PC on the current set of articles anyway, so do it now and make everyone happy. The only people who will be confused for a bit are the "keep" voters who drank the Kool-Aid. Hopefully someone can mix them up another batch. I second the request that all admins should remove their articles from PC now. —UncleDouggie (talk) 09:30, 29 September 2010 (UTC)

The supporters of PC are either too far blinded by the shine coming from PC's baroque features or are Foundation shills. Much like I would not trust myself to review PC stats (as I am heavily biased against what amounts to a gang-unit-like gestapo)I would not trust them to do the same (because they are heavily biased for it. Before you even try a second trial, study the data that has already been gleaned. Shut off PC in the meanwhile so that you can get a cleanly-done sample. If you have to hire a university prof, do it and make sure he's got no rooster in the ring. We should not be coddling either side here, and the actions and words of the WMF and Jimbo suggest to me that they're attempting to sell us a bill of goods. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:48, 29 September 2010 (UTC)

While I agree people with strong feelings either way aren't the people who should be reviewing any stats, I think the comments above "The supporters of PC are either too far blinded by the shine coming from PC's baroque features or are Foundation shills" are a great example of why having a rational discussion on this, let alone achieving any sort of consensus is extremely difficult. Sadly, it often seems to be the case on big issues that people get way too emotionally involved and WP:NPA goes out the window Nil Einne (talk) 03:24, 2 October 2010 (UTC)

Clarifying PC and statistics for future tests

Pending changes is an alternative to semi-protection that

1. Increases user participation (i.e. edits)
2. And protects against vandalism

That means we need numbers for unprotected pages, semi-protected pages and PC protected pages

• Edit numbers before and after PC is applied to an article
• Vandalism numbers before and after PC is applied to an article

Can we get a statistician from WP:WPMATH to help design a statistically valid test so that we don't have questions about the numbers? This should help determine if PC was successful and to what degree. Please correct me if I have gotten something wrong here or left something out. Thanks. -      Hydroxonium (talk) 13:13, 29 September 2010 (UTC)

How large a test needs to be to be statistically valid depends on what you are testing for. What is the present rate of vandalism on unprotected and semi-protected pages? what improvement is worth testing for? (If PC pages have only 99% of the vandalism they did before, it still may be worth doing; but the test required to demonstrate that would be on the order of 50,000 articles.) Do we want separate results for different categories (BLPs, heavily trafficked BLPs, unwatched articles...) and if so, which ones? Septentrionalis PMAnderson 16:42, 29 September 2010 (UTC)

Heavily trafficked articles of any type are unsuitable if they are already under semi-protection, it's the one thing we did learn on this trial. They receive significant vandalism (which doesn't occur if the article is semi-protected), receive almost no useful edits, and are a time sink. Back to Hydroxonium's question, perhaps a mathematician or statistician can help with some of the "number" type facts, but any data gathering needs someone with API or Toolserver access, and the time, energy, ability and willingness to write the scripts to gather the data. Risker (talk) 17:18, 29 September 2010 (UTC)

Articles being actively targeted by off-site campaigns are likewise unsuitable, since, if the campaign is sustained and aggressive enough, they also become a time sink, and "usefulness" gets turned into a more subjective measure than it needs to be. It feeds an "us vs. them" mentality. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:54, 29 September 2010 (UTC)

I have a degree in mathematics and have done statistics; the mathematical questions here are fairly straightforward. The data - and deciding what we want to know - are far more difficult. Septentrionalis PMAnderson 20:46, 29 September 2010 (UTC)

If a PC protected article gets targeted by anonymous IP vandals to demonstrate the unusability of PC protection, the vandalism numbers are useless. We can only check then the constructive/nonreverted edits of anonymous IPs. --Chris.urs-o (talk) 04:20, 30 September 2010 (UTC)

Good point. Is there a page somewhere that is organizing the next test and how it will be conducted? In not, it might be helpful to set one up early and get this kind of input there to help design a proper test. Thanks. -      Hydroxonium (talk) 07:16, 30 September 2010 (UTC)

The proper way to conduct the test is to not let the users know which articles are under test and which aren't. That means no press releases or blog posts. Notify reviewers to test out the UI in advance of any articles being enrolled. Then setup the article selection criteria and randomly split the selected articles into three groups as outlined above. Protect the PC backlog page so only reviewers can see it. Let anyone edit and give them all the same message that their change is subject to review. For the articles randomly assigned to no protection, let the edit go live after 5 to 12 minutes (delay selected randomly for each edit), unless of course there has been a normal Huggle/Twinkle revert. For the PC group, obviously wait for a review. For the semi-protection group, dump the edit in the bit bucket. This would require dropping all the PC adornments that the developers have sprinkled over the article and history pages, except possibly for reviewers. All those do is make it impossible to run a blind test. Even with all this, we can never hope for a truly double-blind test procedure as is standard in the medical field, but we could at least stop begging the entire world to come bias our results. —UncleDouggie (talk) 08:21, 30 September 2010 (UTC)

That is not possible, UncleDouggie, since FraggedRevisions application shows up in the prot log and on the top of the page. (Ya know, that section that reads "Last accepted revision"?) You will not be able to stealth-test it. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 19:00, 30 September 2010 (UTC)

That's why I said that "This would require dropping all the PC adornments that the developers have sprinkled over the article and history pages." If nothing can be changed, there is no point in conducting a trial. It's not our fault that this software was developed without any regard to how to test it. —UncleDouggie (talk) 08:40, 1 October 2010 (UTC)

I heavily doubt the devs will ditch the prot log notation, if for no other reason than transparency and accountability (on the chance that an admin will clap a dragnet on an article with an extremely crappy reason for doing so). I also heavily doubt they'll ditch the notes in history because this could have severe consequences for areas that have traditionally suffered from what I like to call "Meadowlands Syndrome" - two entrenched sides who support two sides of an extremely polemic issue. In most of these areas, content arguments are usually won by getting one side sanctioned. —Jeremy ^{(v^_^v PC/SP is a show-trial!)} 04:27, 2 October 2010 (UTC)

Straw poll on interim usage is there a summary some where?

Just asking? The Resident Anthropologist (talk) 17:31, 29 September 2010 (UTC)

289 For, 199 against, "We'll keep using and testing PC in a very very limited way as described in the poll, until such time as the Foundation rolls out the next version around November 9th."--Talktome^(Intelati) 17:33, 29 September 2010 (UTC)

Ok sounds good. If some one could put a summary with the closing it would be very helpful The Resident Anthropologist (talk) 17:35, 29 September 2010 (UTC)

If someone could do so, and make it make any kind of sense, that would be good. I'm not sure how we can. The "Summary info for editors" says, Trial – The trial has ended - yet, of course, it is still in use. The intro also says, "The two month trial of Pending Changes ended". None of it makes any sense, because the whole concept of a finite trial, which was promised [5] has been disregarded. Chzz  ►  19:32, 29 September 2010 (UTC)

You know what? I think it's time to take PC off all articles until November (or when the next version of PC releases) OhanaUnited^{Talk page} 03:14, 30 September 2010 (UTC)

It seems to me like all those who !voted neutral were just upset there was another poll after the first one, sort of inherently voting against it, but meh. NativeForeigner ^Talk/_Contribs 03:36, 30 September 2010 (UTC)

It wasn't a !vote, it was a vote - as decreed by Mr. Wales. There was no neutral option.  Chzz  ►  16:57, 30 September 2010 (UTC)

Where do we go from here?

Alright, given the poll has been closed for 3 days now, what approach is going to be taken? ~100 more people voted for, I don't think that that can be interpreted as consensus since that's barely over 50%. I support PC but these straw polls have gotten us nowhere and have been an utter waste of time because it seems like the poll was started as a means to get an outright majority of support votes. —Ғяіᴆaз'§Đøøм • Champagne? • 4:08pm • 06:08, 30 September 2010 (UTC)

We go one step closer to a champagne supernova and we enjoy the ride.
A poll on where to go from here? Three more polls and it could be moot with the new version then available. ...But Seriously with so much fuss to turn one version on and then about turning it off as we count the hours to a forthcoming (poll on whether to turn on a) new version there has come a point where i don't think i can continue to care. I love polls. If anything around here shows the core collective indecision it is this issue. If it bugs you then ignore it; someone will be along shortly to review the edit(s).
More people want it on now than want it off. More votes were cast about turning it on in the first place.
This is either the closest thing to crazy or the closest thing to closure. delirious & lost ☯ ^~hugs~ 00:28, 1 October 2010 (UTC)

PC is still in effect as of now, so I'm assuming they're keeping it until the updated version comes out. Sw♠rm^Talk 05:40, 3 October 2010 (UTC)