Wikipedia talk:Wikipedia Signpost/2010-09-13/News and notes

← Back to News and notes

Discuss this story

Pending changes

• Wales has been one of the most outspoken cheerleaders for Pending Changes (or whatever it's being called at the moment), it is impossible for him to act as a neutral mouthpiece for the community in this matter. His dismissive comments on his talk-page about the need for consensus suggest that he would have found some way to turn the two-month trial into something more permanent whatever the numbers were. It's time the Foundation realised that Wales does not represent the community in any way. DuncanHill (talk) 19:38, 13 September 2010 (UTC)

Duncan is completely wrong about that. I am not dismissive of community consensus at all. I think consensus is necessary to make this a permanent feature. I have at no time said or suggested in any way anything like what he is claiming. Full stop. 65% of the votes were in favor of it, and there was significant dissent requiring further attention before we can hold a final vote.--Jimbo Wales (talk) 02:46, 17 September 2010 (UTC)

You've previously been very dismissive of voting, and you yourself have said that 65% is not enough... it seems your principles change to suit the result you've always wanted. You were entirely the wrong person to interpret the poll to the Foundation. You've fiddled a two-month trial into an indefinite continuation of the changes you have been trying to impose for ages. DuncanHill (talk) 14:38, 17 September 2010 (UTC)

I strongly support that we vote on features like this, and I think that 65% is not enough to make it permanent. I have not fiddle a two-month trail into an indefinite continuation of changes, nor have I been trying to "impose" any changes at all, not now, and not "for ages". It would be best, Duncan, if you pay attention to the facts here, rather than simply making up things out of thin air.--Jimbo Wales (talk) 15:23, 17 September 2010 (UTC)

So why not ask for a vote specifically on continuing the trial? As for trying to impose PC, you kicked the whole thing off by announcing (on your talk page, instead of in any of the more proper fora, but then that's par for the course for you) that you were going to ask the Foundation (of which you are a trustee) to turn it on, without any attempt at the time to gain consensus. You have been the main driving force behind this proposal from the start, you cannot possibly act as an unbiased interpreter of a confusing and malformed poll. DuncanHill (talk) 15:32, 17 September 2010 (UTC)

DuncanHill should realise that that he does not represent the community in any way. Joy. —Preceding unsigned comment added by 217.44.191.121 (talk) 01:18, 15 September 2010 (UTC)

I don't recall ever claiming that I did. DuncanHill (talk) 20:12, 15 September 2010 (UTC)

Yeah, but you keep saying you care about consensus, and then you immediately say that the "65% support" means that there is consensus to keep it. Which isn't what I think consensus is about at all. Just because some users were in the minority doesn't mean they can be ignored. To quote myself from below, "there seems to be the idea here (and Jimbo seems to think it too) that there are two parties, and you have to end up upsetting one, so why not upset the smaller one? However, this is not how consensus should work. I think both "parties" need to be much more willing to compromise, so we can reach a solution which encompasses everybody's ideas." - Kingpin¹³ (talk) 05:31, 17 September 2010 (UTC)

But it is patent nonsense to ignore a substantial majority, is it not? Ronk01 talk 05:49, 17 September 2010 (UTC)

I'm not saying we should ignore what you claim is a majority, I'm simply saying that you shouldn't ignore what appears to be the minority, just on the basis that it's a minority. We need to try to address points raised by both both sides, rather than just making sure the majority is happy. I don't even know why we have sides in the first place, it would be easier to make sure everyone is listening to each other, if you didn't dump them into two different sections which just increases hostility/competition, rather than promoting discussion/co-operation. - Kingpin¹³ (talk) 08:23, 17 September 2010 (UTC)

Not sure how you can argue that 65% isn't a majority, it might be 1% away from consensus, but it is a majority. Besides that, it is impossible to make any moves until later today, when the timetables fro release are posted. Ronk01 talk 16:36, 17 September 2010 (UTC)

• At this point, the best thing to do with PC (until the second straw poll) is to leave everything as it is, to turn it off would mean semi-protecting half a million articles, and thus making Wikipedia less accessible. Ronk01 talk 21:06, 13 September 2010 (UTC)

Because the poll was biased... and as I said before, no single option was voted for by 50% of the voters, even allowing editors to vote for multiple options - Kingpin¹³ (talk) 16:40, 17 September 2010 (UTC)

Your analysis was shaky at best. It was decided that the support sections would be lumped into one option for simplicity. And before you say that there were editors who said that they would vote for option one over any option but theirs, I checked, and those votes, if moved to oppose, do little to affect the 65% majority that was established. Ronk01 talk 17:23, 17 September 2010 (UTC)

Yes but the editors who we are actually looking at, are those who haven't stated their opinion on PC, and it's being assumed that they support it in any form. As I said way down below " it's apparently okay to assume that anybody in the support column thinks we should have PC in any form, despite some clearly stating that they preferred option 1 over some other options, but it's not okay to assume that they oppose the other options"? - Kingpin¹³ (talk) 21:18, 17 September 2010 (UTC)

Well the point of a vote is to determine the views of a cross section of a population, however, only fifty two thousandths of a percent of active Wikipedians cast !votes, all we can say is that of 52/1000 of a percent of active Wikipedians, 65% want PC in some form, and only 35% want it off in some form. We can not go around saying that there is no consensus for PC, since a tiny fraction of a fraction of a fraction voted! In reality, this resembles one of the petty content disputes that I mediate, not a logical discussion of a new feature that has not caused any major disturbances that were not fixed within seconds. We need more editors involved, and we need to have a community discussion, not a squabble among a few diehard advocates of their position. Ronk01 talk 23:17, 17 September 2010 (UTC)

• IMO, calling the 65/35% Support/Oppose statistic in favor of implementing pending changes is bullsh*t; that poll was clearly stacked in favor of the support camp since they had multiple options for support whereas the oppose camp had only one choice. Why no one wants to go back and actually tally the 'support on option x' figures is beyond me, but I would be willing to bet then when you look at those numbers no single option has more than 50% support, which poses a problem since any attempt to create unity by ignoring options will result in shift by the supporters in a neglected camp either a different camp, or to the oppose camp. Should the opposes then start to gain momentum its entirely possible that the opposition will end up outnumbering the support camp. Just an observation. TomStar81 (Talk) 02:56, 14 September 2010 (UTC)

It has been decided that a simple majority will be sufficient to keep PC running until PC 2.0 has been released, assuming tha tthe timetable for release is reasonable. Ronk01 talk 03:35, 14 September 2010 (UTC)

Not when the poll is purposely designed to skew in favour of implementing pending changes. You may be able to fool some people, but not the rest of us. Plus, who says simple majority is consensus? Wikipedia doesn't work by going with simple majority votes. OhanaUnited^{Talk page} 04:00, 14 September 2010 (UTC)

• That those who oppose PC seem to think that they can discount the opinion of 65% of people who weighted in I find a little strange. I thank Jimbo for stepping in and supporting the opinion of those in the majority.Doc James (talk · contribs · email) 04:54, 14 September 2010 (UTC)

I'm rather disgusted by the whole way pending changes is being implemented, the supporters seem to care only for the number of users in the support and oppose sections, and not what those users are actually saying. I think it's ignoring consensus and based only on numbers. Also, the "majority support" seems to get used as a justification for almost completely ignoring the opposes, despite it being clear that we don't use majority support at Wikipedia (take a look at AfD, RfA, RfB etc.). That pending changes wasn't turned of after the trial expired shows that the "people at the top" had already decided that this was here to stay, and Jimbo's "evaluation" reads more as a personal opinion then a proper representation of the community. Anyway, I did a bit of my own research into the poll, if we take every vote, and unless there is an explicit support for an option, we assume it's an oppose to that option (e.g. Support 2 and 4 = oppose 3). Then we get these results (bear in mind that this is something of an estimation, since, despite careful examination of the supports, I may have missed some explicit supports of other options. I won't vouch for this being completely accurate, and if you want to verify it you'll need to look at the poll yourself):

Total votes: 624
If option two is implemented:
Explicit supports: 117
Opposes: 507
If option three is implemented:
Explicit supports: 308
Opposes: 316
If option four is implemented:
Explicit supports: 162
Opposes: 462

I used a .NET C# program to store the data ("voter" class with four booleans for each option) anyone who would like to use it just ask me, and I can try to format it for you. - Kingpin¹³ (talk) 09:03, 14 September 2010 (UTC)

Firstly your "count" is an extremely biased example of why individuals with a strong POV should never count votes. (Look what happened in Florida in 2000)I would remind you that the straw poll has been reinterpreted as a decision to keep PC running until a second straw poll which would be a simple yes or no vote over whether to keep PC turned on until PC 2.0 is released. Ronk01 talk 14:01, 14 September 2010 (UTC)

Erm... care to explain how a simple count can be biased? All I'm trying to show is that no single option at the straw poll got >50% support, so simply claiming a majority is in support is misleading. Also, I don't see how this poll could be interpreted as a decision to keep PC running until a second trial, since it didn't even have the option of turning it off until a second trial is organised. Maybe because there is a false belief among the supporters of PC that the opposition only want PC turned off completely, never to return, and are unwilling to compromise? When in truth plenty of people want it turned off for now, while problems are fixed, and then on again once it actually works - Kingpin¹³ (talk)

I knew it: independent of the true total no option did have 50% support. For my part I would demand an rfc on the outcome and a recount so we can get an actual, factual read out of the PC system rather than an a rigged poll. TomStar81 (Talk) 16:17, 14 September 2010 (UTC)

What has me concerned -- & is pushing me towards opposing pending changes entirely -- is that no one in favor of leaving PC on has provided a convincing reasoning -- or frankly, any reasoning -- for implementing it. We're not talking hands here; supposedly consensus on Wikipedia is not determined by number of votes. What we're talking about here are persuasive arguments for or against making changes. About the only analysis of this experiment that has been done is Wikipedia:Pending changes/Metrics/Preliminary Analysis, whose primary point -- whether intended or not -- is that no one who started this test had any idea of how to evaluate it as a success or failure. Providing a rationale for the change would make it easier for those of us -- like me, until now -- who have no strong opinion on the matter to accept the change, instead of suspecting this to be little more than a public relations stunt which will not benefit Wikipedia in the long run. -- llywrch (talk) 20:25, 14 September 2010 (UTC)

What needs to be considered is the fact that all of the individuals who voted for the keep options want PC kept, while they might not agree on exactly what form, they do agree that it needs to stay. Unfortunately, the vote came very, very close to 2/3, which means that either a significant majority gets angry, or a minority gets angry. PC was released too early, it has flaws. But flaws can be fixed (in fact take a look at this.) PC was meant to change the way we protect articles in such a manner that Wikipedia is more open to IP contributors, and to protect minor articles (like low traffic BLPs) that IP's make major contributions to, and are thus unsuitable for Semi-protection. Two months was not a long enough period to test PC, thus the lack of and conclusive research, the time period simply was not statistically significant. Reviewers had to learn how to use PC, Admins needed to learn which pages PC worked on, thus the initial troubles (along with the technical problems that PC 2.0 should fix). PC is a tool, that like semi-protection can be used to protect pages, but it doesn't stop IPs from editing like semi-protection does. Remember, IP's make 20% of Good Faith edits, and only 30% of all IP edits are vandalism. To address problems with the poll itself, the polling period was far too short, more editors needed to be involved, because with the way it was, only editors who care one way or another about PC voted (with several exceptions). That is not an acceptable way to poll, neutral editors are needed to ensure that the interests of all Wikipedians are represented. Of course, Straw polls do not establish consensus, but how do we have a discussion between thousands of editors? Ronk01 talk 21:53, 14 September 2010 (UTC)

That's easy: we open an rfc and let the community speak for itself. As to your conjecture that the PC trial came too soon, I agree. It should have been implemented with strict parameters and on predetermined articles to see how things would work. Releasing this to us in its current form was like asking us to understand and apply a concept that has no form or function. How were we supposed to do that. Moreover, the results on individual level are striking: no single PC option on support has a majority. I am not for pending changes, I admit that, but in the interest of fairness I will point out that for this to work on here in any capacity we need to disable pending changes, sit down, have a long conversation with the community about what went right and what went wrong, tweak the PC option, then conduct another trial on a controlled group of articles to see if the improved version of PC if properly applied would actually help the articles it was being applied to. For the time being, and RFC would help with sit down and have a long conversation with the community part of my plan outlined above. If nothing else, it would be a sign of good faith between disagreeing parties here. TomStar81 (Talk) 03:42, 15 September 2010 (UTC)

Ronk01, keep in mind that there is an even larger group whose opinion needs to be considered: the rest of the Wikipedia community who did not vote. While it would be simple to dismiss them for being "apathetic", I believe many of them are, in fact, interested but hanging back to see just how this process will end. If it appears that the process was rigged & Pending changes would be adopted, willy-nilly, then they will be very angry -- & there are far more of them than participated in this "straw vote" for or against it. It appeared that everyone involved was moving towards a do-over of this step of the process (it would not be hard to do better than a "straw poll" with four competing options), until Wales' impudent claim that everyone was for Pending changes -- well, everyone whose opinion matters to him. He obviously wants it, whether or not it works. Or whether anyone understands how to use it. -- llywrch (talk) 05:49, 15 September 2010 (UTC)

Also in reply to Ronk01 ;). Not all the people in the support column supported PC in any form, many specifically stated they preferred option 1 over some of the other options. Also, there seems to be the idea here (and Jimbo seems to think it too) that there are two parties, and you have to end up upsetting one, so why not upset the smaller one? However, this is not how consensus should work. I think both "parties" need to be much more willing to compromise, so we can reach a solution which encompasses everybody's ideas. - Kingpin¹³ (talk) 07:51, 15 September 2010 (UTC)

I agree that an RFC is needed, along with a drive to get more community participation. I would however contend that, like any debate of this size, there are users who are absolutely unwilling to compromise on PC, on both sides. Ronk01 talk 15:10, 15 September 2010 (UTC)

I find it very depressing to see that PC is going to be continued despite what seems to be no consensus for it to continue at the current time. My rationale for not supporting its continuation was its rather broken state-- I felt that the interface was confusing, and it didn't do what it set out to do (let IPs edit high traffic articles). I feverishly object to this being forced down on us, and I'd welcome an RfC which didn't have a flawed straw poll method to it. Nomader ^(Talk) 07:14, 16 September 2010 (UTC)

The topic was also discussed in yesterday's IRC office hours with Sue Gardner, where she suggested holding another office hour specifically about Pending Changes. Regards, HaeB (talk) 08:56, 17 September 2010 (UTC)

I haven't seen a single attempt at compromise other than "Shut it down, RAWR!". Any of the folks opposed to PC want to suggest an actual compromise (instead of repeating their desire to trample the wishes of the significant majority)? BigK HeX (talk) 11:43, 17 September 2010 (UTC)

That's rich. A poll is set up and the only option for people who don't like PC is "shut it down completely", then you complain about users opposing it only saying they want it shut down? If you'd actually take a look through the poll you'd see that opposes have still managed to voice some of the problems they have with PC anyway, for example, it's slow, complicated for new comers, has poor review guidelines etc., all of which are things which could be fixed, and are all things the majority of the supporters would seem to support fixing as well. Sure, there are some editors who are opposed to it on principal, and that wouldn't be "fixable" as such. But the bulk of the opposition (including some opposed to it in principal) are not saying "Shut it down, RAWR!", and I don't think you should be representing the opposition like that, when many of them have given valid reasons for opposing. I also think that while we reach a compromise and improve PC, it should not be used in articles. This whole PC thing seems to have been a headlong charge, without actually stopping to think. There has been no proper centralised discussion on it, the trial was implemented through a poll (<not consensus), then kept in because the devs then decide to tell us they can't take it out, another arguably rigged poll is then used as justification to keep it in (again not consensus, and gives the impression this is a competition, and since the supporters have "won", the opposition is not important). I think by turning it off, and spending some time reviewing, will make the opposition feel like they're not being ignored for once. Also, you claim a majority are in support of PC, and yet out of all the users who voted on this poll, no single option was voted for by more than 50% of them, again showing that this poll was bias. - Kingpin¹³ (talk) 12:11, 17 September 2010 (UTC)

So ... you are here pretty strongly requesting PC to be shut down, and you're also objecting to my characterization of the opposition's demands to "Shut it down RAWR"....?

Also, "no single option was voted for by more than 50% of them, again showing that this poll was bias"

Yes there was an option with majority support... that being the option of "keep PC in some fashion" received 65% support. Using the breakout of support levels against supporters is asanine, as it is almost certain that most of the people supporting one of the options would find other "Support" options acceptable, even if their preference was for one. I can personally attest to that being the case for me. I, personally, noted my support for what I believed to be most appealing option, since it was pretty clear that Supports and Opposes would be grouped, as any other interpretation of the poll would be nonsensical, given that a simple 2-option poll was the obvious alternative. BigK HeX (talk) 12:17, 17 September 2010 (UTC)

Thanks for your reply, to respond to some of your points. I don't think I'm strongly requesting that PC be shut down here, I'm requesting that in a two month trial, the trial is ended after two months, yes. I also believe I'm displaying reasons as to why I think this, not just saying "Shut it down RAWR" (which implies I don't care about trying to reason, and just have my own opinion and don't care to discuss it, clearly (I hope) not true).
Okay, it's not my fault that this poll was poorly set up. You want to count it up assuming that anyone supporting option 2, 3 or 4, also supports the other options. I have counted it up as assuming that they don't (although I've also actually READ through the poll, to see if users have explicitly stated they also support the other options. Something which you have not, as if you had you would see that some users have also explicitly stated they are opposed to the other options). I'm not saying that I think that my count should be used as the final tally, I'm just trying to show that we shouldn't have a messed up poll (like this one) and then ASSUME that people are in support of certain options. Because the problem with assuming is that you can assume either way (as I have shown). I've tried to make it very clear how I counted up the supports/opposes for each separate option, something which has not been made clear for the final tally. I think there is a bit of a "blindfold" being pulled in front of the eyes of the community here, and I'm just trying to present a bit more data, a bit more clearly. - Kingpin¹³ (talk) 12:42, 17 September 2010 (UTC)

I reviewed the poll discussion in-depth. From my review, the overwhelming support for PC is clear, even despite your .... interesting view on the segregated results. As for the "end of the trial" ... we had a two-month trial, and found significant support for PC in some fashion. The trial of PC has been extended on that strong support. Seems pretty reasonable. Jimbo asked multiple times for examples which would indicate that retaining PC-protection is causing problems on some article or another. I don't believe there was ever a direct response to that. If there's insignificant evidence of PC causing problems, then it seems pretty respectful of the strong support for PC to extend the trial. I've asked many, many, many times for compromise proposals from opposers, and just about every single response I got was based on "shut it down". Seems like we could be far more focused on specific problems.... BigK HeX (talk) 13:22, 17 September 2010 (UTC)

If you would post your review maybe we could examine it, rather than listening to you repeatedly claiming that there is overwhelming support for PC, without any evidence. Yes, I think it's very interesting too, that it's apparently okay to assume that anybody in the support column thinks we should have PC in any form, despite some clearly stating that they preferred option 1 over some other options, but it's not okay to assume that they oppose the other options (even after I read through the entire support column and made sure I wasn't counting anyone who explicitly stated they supported the other options too), I'm glad you agree. The original trial was a two-month trial, it was made clear it would be switched off after two months to allow for review, not that there would be a poll to extend it. This is a POLL, I think it's sad that one of the biggest technical changes to the way Wikipedia works in possible the last couple of years, is being rushed through with a poll. You can't judge community support like that, and the trial shouldn't be extended based on that. Especially when that poll is rigged to show a larger support for PC then there actually is. I just gave you three specific reasons PC is harmful, picked out from the comments at the poll, displaying large problems with PC, and yet you seem to have suddenly gone deaf, and still insist that everybody's response is "based on 'shut it down'" and I'm not focused on specific issues? Try actually responding to me message please (see message at 12:11). - Kingpin¹³ (talk) 14:13, 17 September 2010 (UTC)

It's a bit odd that you believe people "have suddenly gone deaf", when your responses didn't seem very relevant to my actual statements.

I said, "Jimbo asked multiple times for examples which would indicate that retaining PC-protection is causing problems on some article"

To which you referenced your comment that, "it's slow, complicated for new comers, has poor review guidelines". NONE of these points us to an article where we can find and view for ourselves indications of these problems. Moreover, it's been beaten to death that Pending Change is FAR SUPERIOR to the {{editprotected}} process that newcomers are forced to endure when the lack of PC-protection forces admins to use Page Protection.

Also, I said, "I've asked many, many, many times for compromise proposals from opposers, and just about every single response I got was based on "shut it down""

And your response was that you gave a lot of rationales for shutting it down ... which only affirms my statement that all of the opposers keep proposing that we shut it down. Certainly there have been rationales given for shutting it down, but regardless of the rationales, the proposal has invariably been to shut it down. Do opposers not see any other routes???? BigK HeX (talk) 14:48, 17 September 2010 (UTC)

What I gave were specific problems with PC on the whole, which were causing problems on ALL the aritcles where it has been implemented. I beleive the Barack Obama article had a few problems with PC, since you want an article. At least some of the opposes seem to think that {{editprotected}} is a better system. My responses where, as I pointed out earlier, "all [...] things which could be fixed", not neccasairly reasons for shutting pending changes down. Again, as I said earlier, the opposition were not given any option except from to support the proposal to shut it down, due to the poor poll - Kingpin¹³ (talk) 15:00, 17 September 2010 (UTC)

Well ... the poll clearly had its problems. However, "shut it down" is the proposed move that I get even outside of the poll. I'm still interested to see whether any of the opposers have any other possibilities in mind. I came to this thread hoping that this would be a possible place to find such alternatives. BigK HeX (talk) 15:22, 17 September 2010 (UTC)

page blanking bot

Hi, I'm involved in that cleanup project and some of the statistics in the Signpost summary got mixed up. The correct picture seems to be:

• Darius did about 163k edits but a lot of them were minor
• Darius created slightly under 10,000 articles
• Darius made significant (non-minor) edits to about 23,000 articles including the 10,000 described above
• Darius made edits (minor or otherwise) to a total of about 40,000 articles
• Almost all the articles were about athletes or athletic events
• A lot of the articles don't seem to be copyright problems because they contain very little text (in some cases none at all), just lists of names and events and other such statistics. Darius wasn't big on writing prose, but he was quite skillful at table formatting, so many of his articles were in pure tabular form.
• The bot's initial run will blank only the 10,000 articles Darius created. A subsequent run to get the other 13,000 of the significantly edited articles will take more complex programming and is still under discussion. The remaining 17,000 won't be affected unless something changes.
• Manual review of some the affected articles indicate that about 10% (depending on who you ask) of the articles contain copyvio text, not 10% of Darius's total edits. It's not clear if that's based on only the 10k created (which would mean about 1000 vios), or 23k significantly edited (which would mean around 2300 vios). I don't think anyone expects the minor edits to contain vios, but checking this stuff is still a work in progress.

The main info/discussion page is Wikipedia:Administrators'_noticeboard/Incidents/CCI though it is getting sort of unwieldy and maybe it should be digested/reorganized.

User:Moonriddengirl and User:Uncle G are the main principals in this, more or less.

75.62.4.206 (talk) 23:25, 13 September 2010 (UTC)

This could be a problem for many other projects, too. Many smaller projects just copy/translate articles from the English WP, without checking for copyvios (or errors). So the copyvios could have already been copied to dozens of other projects. But unlike the German Wikipedia, those projects don't import the version histories of translated articles, so there's no easy way to find these articles. --Kam Solusar (talk) 13:05, 14 September 2010 (UTC)

A few post publication corrections are in order: thanks. ResMar 17:39, 25 September 2010 (UTC)