User talk:Jimbo Wales

Create tool for random-jury selection

Introduction

I agree with Lila Tretikov that software development is the key to massive improvement of Wikipedia and Wikimedia content. Besides using better AI techniques to detect/remove copyvio text which has been copy/pasted into pages, other smart tools could be developed. A major advance would be a "random-jury selection tool" to wikimail perhaps 50 people to debate an issue from pools of hundreds/thousands of interested regular users.

For years, WP issues have been decided by self-selected groups of intense users, due to inability to wp:CANVAS potential !voters in a random manner, without seeming bias in notifications. The result has been many wp:TAGTEAM gang decisions which have acted as a "jury of sneers" to denigrate editors such as myself, with 2 degrees in computer science, having completely written several macro text editors and graphics editors, as if I do not know what templates or software products would be useful. The decision to mandate wp:DASHes into hyphenated titles was also forced in a relatively small group of self-selected users, later claimed as too much trouble to reverse.

Instead, if whatever-language wikipedia had a random-jury selection tool, then perhaps a wider cross-section of their editing community could be involved in decisions. After a few years, then policies such as wp:CANVAS might be reduced to allow broader, randomized consensus on major issues, without hoping enough people would sacrifice months of time to obsess about a wp:RfC which relatively few people knew was pending. As evidenced by websites such as Google Search or YouTube, software can change the world, make an unknown person famous, or discover a little-known talent to become a cultural superstar. Some software tools can become "killer apps" in their impact to transform the workplace. -Wikid77 (talk) 11:09, 24 February 2016 (UTC)

Yes, I totally agree that we need jury selection (which I mentioned above in this comment) instead of the horrible state of affairs of self-selected users making the decisions and there being no differentiation between interested and disinterested parties in the decisions. I did not realize there were others who have been contemplating similar solutions. That's good to hear. Smallbones seemed to agree on the need for some kind of reform here. Can we start a project or task force to address this? It is something that bothered me from early on when I started working on Wikipedia and it seems as bad as ever. I thought that trying to participate more in the noticeboards as a neutral third party to break up petty disputes was the best way I could contribute to address the problem I saw that you describe above, but the gangs and their sympathetic admins immediately chased me off. And that's why I too have been thinking jury duty is the way to go. --David Tornheim (talk) 11:28, 24 February 2016 (UTC)

@Wikid77 - This is an excellent idea. Carrite (talk) 12:16, 24 February 2016 (UTC)

I've long favored this idea. There are some details I would suggest though. There's too much room for distrust when an issue has widespread political significance, and so the procedure needs to be designed to dispel all doubt.

• The random numbers have to be genuinely, provably random - i.e. we say in advance how the generation is going to be done, and base it on a public lottery drawing, or preferably two from different countries. Otherwise he who controls the pseudorandom number generator controls Wikipedia - whether anyone at Wikipedia knows it or not!
• Notification of selection should be as a public on-wiki comment, because emails sent by some individual out of public view conceivably could be selectively mislaid.
• The random selection should be per edit rather than per user. We never really know how many usernames someone has, and more to the point, I think that the people who contribute more should have more say. (Possibly, we might agree on some namespaces to be excluded from consideration, like User talk...) Selecting random recent edits also ensures we get active users.
• Users with certain administrative statuses at the time of selection subject to blocks or certain kinds of topic bans might be excluded from the pool, but we must agree not to allow any future administrative action - even blocking! - to disqualify the juror from voting. Otherwise, jury selection could come to be dreaded, as rogue admins watch the conversation and start digging into the background of any juror who makes a comment they don't like.

Wnt (talk) 13:38, 24 February 2016 (UTC)

I like the idea. I'll respond to just a few points.

• I don't think worrying about the provenance of the random number generator is a very sensible place to worry, other than that it be a typical random number generator in Linux. The vectors of attack on such a system to game Wikipedia surely lie elsewhere, so preventing implementation based on a fear that would involve requiring random numbers being generated from different countries is not really optimal.
• We should absolutely be aware that putting a call out for 50 random volunteers will not result in 50 people responding - active wikipedians will be much more likely to respond, for example, but often times even they will decide that they aren't sufficiently knowledgable or interested to weigh in on an issue.
• Rather than excluding admins, I think it will often make sense to include them more often. They tend to be highly experienced users trusted by the community. There are sometimes complaints that the admins are tyrants, etc., but these generally have little merit. Rogue admins are seldom the real issue.
• It would be good to limit the use of the tool somehow, in terms of sheer numbers. If people start getting bombarded with 20 requests a day, it will just be a nuisance and people will tend to opt out of it. It should be reserved for use such that most active users only receive 1 request per month max.--Jimbo Wales (talk) 14:50, 24 February 2016 (UTC)

• random # generator I agree with Jimbo that the nature of the random # generator will not likely be an issue, and that a typical pseudo-random # generator found in most programming languages should be sufficient and can be made publicly visible. The choice of seed would be key, such as date and time as received on a particular WMF server, to make it hard to know or control what seed would be used. Also easy to verify the results after the seed was chosen. Perhaps adding a layer of public key encryption between the editor requesting a jury and the jury administration bot would further make it impossible for anyone to know or control what the seed is before it is chosen. Is there a way for two people to flip a coin using public key encryption? I do not think there is such a thing as a "truly random # generator" or being able to prove one if you had it. I think you can prove something isn't random or how well it stands to randomized tests, but not the reverse.
• participation I agree that editors may not want to do jury duty. To address this I believe it should be made mandatory (just like if you want to vote in real life), if you want to continue editing. Once the system is implemented, each editor would be required to submit their name to the jury pool before making any more further edits. If they fail to show up for jury duty, they would lose editing privileges until they participate. There might be some liberal provisions to allow unavailable for jury duty, such as going on vacation, or deferring jury duty if selected at a bad time. Or asking to be taken off the case because of COI, or feeling lack of competence, etc., just like in real life, and having an admin. on duty required to rule on the request. And like in real life, the editor(s) of the dispute might be allowed a small # of pre-emptive disqualifications of jurors without a showing of cause. Yes, that could be pretty complex, but if done correctly most of it could be done by automatically under clear timelines.
• complexity/frequency What I described in the above paragraph could be fairly sophisticated and have a time line as long as an ArbCom case. For major issues, it might be worth it. There could be different levels, with different timings. One might ask for a simple jury, where only a few editors need show up, or a full blown case, like ArbCom. And maybe limit the # of requests for jury decision per editor.
• seniority/# of edits in selection I do not think selection should be based on # of edits or length of service. I disagree with Jimbo about putting admins. on--they have plenty of power/authority as it is. We need fresh neutral eyes for content disputes, not people with entrenched unshakable views to have extra power. Witnesses can always speak in the witness box area, and that can include anyone, regardless of experience--as we have now, but the jury would be making the call, not the involved parties and self-selected participants in a case. For policy issues, a jury is probably not the best way to handle things. Then I agree that experience would be more important.
• testing I agree such a model would need to be tested in a very limited venue, and some general agreement on those to be the guinea pigs would be advisable. With feedback, the system could be expanded. I think if we found an appropriate notice board. Or took every Nth case for that noticeboard, or let requesters opt for a jury would be a way to start.

--David Tornheim (talk) 00:37, 25 February 2016 (UTC)

This seems on the face of it like a good idea, provided there is some basic gatekeeping to stop it being used frivolously. Guy (Help!) 15:01, 24 February 2016 (UTC)

I will emphasize the points of both Jimbo and Guy that there should be some limitations. This ought to be natural as any rollout of any new idea ought to be tested in phases, and gradually ramped up. (Unfortunately, we haven’t always followed such rules, but we can this time.)--S Philbrick(Talk) 15:49, 24 February 2016 (UTC)

@Jimbo Wales: I don't think a random number generator attack is out of the question, especially if the jury mechanism ever gets used to decide things like whether to have a site blackout over some surveillance bill an order of magnitude worse than anything proposed before. Even ordinary hackers might have both the means and some obscure but compelling reason to influence how the site is maintained. Wnt (talk) 16:17, 24 February 2016 (UTC)

I think juries should primarily deal with behavioral and content disputes, not policy matters. --David Tornheim (talk) 00:53, 25 February 2016 (UTC)

Also, I apologize for being unclear above - I wasn't suggesting that admins be denied a chance to vote, but that they not be allowed to remove voters by finding something to block them about. (I'm assuming that these "votes" will be prolonged discussions where jurors get to make the ultimate decision, but will listen to one another and any parties involved first, thereby giving others some time to see which way they are leaning) Wnt (talk) 16:51, 24 February 2016 (UTC)

I am okay with limiting jurors to lay editors--admins. have a enough to do as it is. But I agree that there is reason to be concerned about "jury tampering" as you described. Blocks after jury selection should probably not apply to jury participation unless the admin. overseeing the jury finds them to be disrupting the jury process. One of the things I really don't like is that admins. are not assigned duties randomly but just chose to do whatever they feel like, in the same way that noticeboards are self-selected. A recent decision that I described above I found rather outrageous, where a major decision was made by an admin. that many--including myself--believed was involved and therefore had COI. I much prefer the American court system where docket #'s are assigned to cases and the selection of judge is neither up to the judge nor the person filing the case. That would really help here. --David Tornheim (talk) 00:53, 25 February 2016 (UTC)

While this may look like a good idea, it's also a way of leaving the decisions to people with no particular knowledge nor interest in a topic. Is that really what we want? In a "real" jury, people are forced to sit through a a long and complex legal process, guided by professional lawyers and judges, confronted with expert witnesses, and instructed by the the judge. And the results are still not very good. --Stephan Schulz (talk) 01:15, 25 February 2016 (UTC)

As someone who works in and studies law, I respectfully disagree. I think the legal system (dispute resolution system) we have in the U.S. is far better and fairer than its counterpart on Wikipedia, which looks far more like the Wild West. In the U.S. too many people think that what they see on TV has a reasonable connection with how the legal system actually works. (See [1] and [2]). My legal courses quickly showed me that the legal system is far different and more reasonable than what I saw on TV. I quickly realized the representations made about the McDonald coffee cup case could not be accurate based on what I was learning. With minimal research I quickly found that what the media said about it and what actually happened had little in common. The media made it sound crazy that someone would sue for >$1M because coffee is too hot--seems reasonable. Until you find out the feeble elderly woman, sitting in the passenger seat of a parked car, sustained 3rd degree burns, 2 weeks hospital stay and skin grafting, and that when she asked McDonald's to help her with the bills, they told her to piss off. And that it was not her, but the jury that was so upset with McDonald's callous attitude about it, that they wanted a punishment McDonald's with 1 day's sales of coffee. Should we be surprised that McDonald's and corporations whose advertisements fund the media are more than happy if the media makes the plaintiffs and their attorneys look unbelievably selfish and opportunistic and leave out the plaintiff's version when they can make it sound more outrageous? So, no the court system works far better than the general public realizes. But our dispute resolution on Wikipedia undoubtedly have serious problems. At least our coffee case article is far better than what I heard on TV! --David Tornheim (talk) 02:09, 25 February 2016 (UTC)

I don't think we'd get far with the McD coffee case. But what about O. J. Simpson vs. Albert Woodfox? The best justice money can buy? Yes, our system is not perfect. But I'm not convinced that leaving decisions to less informed "juries" will be an improvement. --Stephan Schulz (talk) 02:30, 25 February 2016 (UTC)

^The two cases you give are just typical media cases. Most cases are not like this. Please review the articles I provided above.

I'm not convinced the gangs of self-proclaimed "informed" editors are always as expert as they want us to believe. With anonymity and no provision for qualification of expertise, an anonymous editor can make up any b.s. they want about how "expert" they are (or that they have no COI) and it is unverifiable. These self-appointed experts also think it is okay to say others they disagree with on content are "incompetent", including people who are actual experts. In contentious cases, I trust a person who has never seen the RS before to read it and summarize it over the kinds of behavior I have seen by the "informed" and highly invested editors who have decided a priori what is in the WP:RS and throw a temper tantrum of numerous bogus allegations and distractions if you point out they are wrong or lying. Wikipedia is supposed to be verifiable, so in most cases lay people should be able to read and verify what is in the RS matches what is in the article.

Now, I will admit, material such as Maxwell's equations or Schrödinger equation, few people with a basic high school education would be able to read any of the RS for the article, or understand much of the article itself. I do wonder to what extent Wikipedia was designed to require such specialized knowledge for editing or reading articles such as these. Compare these with the treatment of Britanica: [3] and [4] which permits the high school reader to walk away with a minimal understanding of these extremely important equations. ---David Tornheim (talk) 13:39, 25 February 2016 (UTC)

@Stephan Schultz: Fair point, and this does also raise another risk. One of the advantages of current policy is that it doesn't matter how popular something is, if it is objectively wrong, we portray it as such. A "jury" of randomly selected Wikipedians may be inclined to vote on what they feel about the subject, rather than the actual evidence. Take GMOs: what would Wikipedia be like if, say, a "jury" voted to allow Séralini's findings to be represented as fact, despite the evidence that they are scientifically unsupportable? We know that things like homeopathy, reiki, chiropractic and so on are popular - we also know they are laden with bullshit. Currently Wikipedia does not do votes and does not allow numerical superiority to override verifiability. Would that change? Several of the people agitating for this kind of reform are known to support fringe views, and their problem with Wikipedia is that they perceive our reality-based bias as unacceptable - they characterise the reaction to their POV-pushing as bullying, uncivil, non-neutral and so on - it's everybody else who is the problem. There are definite risks in this proposal unless it is implemented in a way that ensures policy, not popularity, always wins. Guy (Help!) 14:23, 25 February 2016 (UTC)

Wow. Your true spirit is showing. You obviously don't trust the ordinary randomly selected editor to review the evidence (WP:RS) and make the right decision. They apparently can't be bothered to follow our rules. They obviously don't know the Truth™ and won't find it even if you show them the RS. You on the other hand do have access to the Truth™, and you have admin. tools and the threat of using them to prove it. Those who disagree with you are just pushing WP:Fringe views and are a bunch of advocates who don't get it, pretty much like the other average ordinary users who can't be trusted either. It would probably be better if none of these clueless people were here and you and those who think like you do took over. In fact, the entire project of Wikipedia is in question with the presumption that "anyone" can edit, because after all, as you said, they can't be trusted to look at the RS: they will just favor their beliefs instead. This looks to me like the classic fallacy of Argument from authority, where you are the self-identified authority and everyone else who doesn't think like you do, well, we can't be trusted... --David Tornheim (talk) 04:58, 26 February 2016 (UTC)

Above we have a good expression of Guy's McCarthyism. The belief that you known what's correct about everything, and others can't be trusted. That we need the benevolent dictator, or otherwise we'll be drowning in bad content. That we need good hearted bullies to maintain the content. I say no. We have policies and we have editors who understand the policies. We represent reliable sources correctly, and voila, we have good content. We don't need philosopher kings. We need fair application of policy. Guy's approach sounds superficially convincing, but it's deeply flawed and harmful. In other words, we don't need to be protected from the Communists. We much moreso need to be protected from those who say they're protecting us from Communists. We are community with principles and policy that will suffice, and I certainly support any means to getting better peer review of edits and avoiding canvassing, rigging, and gang-based POV enforcement. SageRad (talk) 14:56, 25 February 2016 (UTC)

Where did I say I know what's best? The issue is as I stated. Consider an analogy: state leigslatures may pass laws banning gay marriage, but the Supreme Court, which has no jury, has ruled that this is unconstitutional. So if you are going to have juries to rule on content, you'd have to have an overarching system which examines whether the decision of the jury is actually in line with policy. You can't qualify the jury by credentials, because (a) credentialled editors are every bit as susceptible to bias as lay editors - hence Citizendium nuking a big chunk of its expert-generated content a while back, because the experts in question turned out to be fringe POV-pushers rebuffed from Wikipedia - and (b) lay editing is a foundational principle.

The accusation of McCarthyism is hyperbolic and frankly irrelevant. Wikipedia is, by its very nature, inclusive. A few people get sanctioned or banned, but only a few, and mainly in areas where there is a well-documented real-world collision between fact and ideology (climate change, creationism, homeopathy, psychic phenomena and the like). Wikipedia is a reality-based project. If it ever ceases to be a reality-based project, it will lose its credibility and it won't be Wikipedia any more. If Wikipedia said that reiki works, rather than it being so obviously nonsense that a nine-year-old student could devise a test that shows it doesn't work, then we would have failed. Being neutral and accurate does not mean giving parity to belief and empirical fact. Guy (Help!) 23:26, 25 February 2016 (UTC)

JzG/Guy: Your claim that "only a few" people get banned is a bit disingenuous. At the GMO ArbCom Case, a single editor Tryptofish proposed having 7 editors site-banned [5], including me, SageRad, DrChrissy who have spoken here, Prokaryotes (whose successful topic-ban you later supported, resulting in a voluntary 6-month site ban.) You did not oppose any of the 7 proposals for site bans, and indeed spoke up in support of the proposed site ban of Petrarchan47. To say site bans are uncommon is not consistent with your willingness to let 7 of them move forward in just that single proceeding. --David Tornheim (talk) 04:33, 26 February 2016 (UTC)

• Ruthless decisions and deletions are rampant: Evidentally many users are unaware of the ruthless decisions being made. Now non-admin closures of XfD discussions are commonplace, concluding clear "consensus" to delete templates or pages used for years, just waiting for an admin to delete the pages not wanted by non-admin decisions. The deletions have been horrific; just look at TfD of Template:Convert/flip2, which deleted 3 complex templates in use for 5 years, based on 1 "Delete" !vote while another user urged to save those templates for use on other MediaWiki websites not using Lua script versions of templates. What some people do is to systematically remove valuable templates, from hundreds or thousands of pages, then claim, "template unused" as justification to delete a page used for years to provide data not possible in any equivalent way. -Wikid77 (talk) 12:12, 25 February 2016 (UTC)

Perhaps have a quorum of noninvolved users

It might seem great to poll only 3rd-party, non-involved users, but that seems unrealistic because the basis of the not-vote (!vote) strategy is to debate on merits of issues, rather than sheer numbers, and the involved users typically know those details well; however, too often the count of Oppose vs. Support is used to determine consensus, and hence a minimum count of non-involved users could be set, as an outsider quorum, to ensure enough outside 3rd-party opinions to deter an insiders-only majority rule. -Wikid77 (talk) 20:42, 24 February 2016 (UTC)

Yes, I agree, that a quorum of non-involved editors' decisions should carry the weight. And their discussion should be separated from the involved parties--like at ArbCom. They can listen to involved parties comments as witnesses and say they agree with Witness X's comments, etc., ask questions of witnesses/parties, and anyone else who feels like showing up, etc. Their decisions can be based on the old consensus model, but the self-selected parties and witnesses do not get an !vote in that.

I'm not sure what is to be done if the 3rd party uninvolved can't seem to agree. Like a real jury, they could be required to come up with something that all or at least, say 2/3rd of them can live with before they are released from duty.

I think it essential that the 3rd party non-involved are selected randomly from editor space. Editors may claim not to be involved, and just show up. But they may have an interest, or a strong opinion on the subject matter, even if they have never edited the talk page or article space before. In fact, the editors who show up could have had their accounts created specifically for that purpose--to be able to show up on key decisions and to be able to claim neutrality, since they have up until then spent all their editing time on other subjects. I read a disturbing Wikipediocracy article that described such methods a group of COI editors could use to hijack a particular Wikipedia article(s) which sadly could easily be in place right now and very hard to detect or prove, even if suspected. The random jury selection process would make the methods described in that article ineffective. --David Tornheim (talk) 01:17, 25 February 2016 (UTC)

Jury tool as part of user-decision toolset

In terms of WMF budget allocations, I would recommend a 2-year minimum to develop a broader toolset with the "random-jury selection tool" plus some related tools, as a broader toolset, to help assure the wp:developers a longer-term employment, in case of fears of project cancellation would deter programmers from sticking with the user-decision tools project. -Wikid77 (talk) 20:42, 24 February 2016 (UTC)

What if you gave a trial and nobody came?

I think the proposal is well meant but impractical. For example, in the U.S. if you're selected for jury duty you have to serve under pain of legal sanction including fine or imprisonment, while Wikipedia has no such powers of compulsion. On the other hand there's much to be said for giving more weight to those who haven't previously been involved in the matter at hand -- or at the very least, giving them space to be heard above the din of partisan sniping. Shock Brigade Harvester Boris (talk) 01:27, 25 February 2016 (UTC)

Simple, make it mandatory if you want to continue editing (as I described above under "participation"). How many people you know who went to jail or were fined for not attending jury duty?  :) Jury duty is only required if you want to vote. --David Tornheim (talk) 02:13, 25 February 2016 (UTC)

I've known people who were fined (though not jailed) for blowing off jury duty. The link between voting and jury duty is a misconception that holds over from the days when jury lists were taken from voter registration rolls. Nowadays the jury pool is often supplemented with other lists, such as licensed drivers.

As for the possibility that one could be blocked for not serving on the Wiki-jury, this has so many potential undesirable outcomes that the chance of it being accepted is zero to negative.

Outside-the-box thinking is good, but this idea just isn't going to fly. Shock Brigade Harvester Boris (talk) 02:38, 25 February 2016 (UTC)

I'm not seeing any evidence for "so many potential undesirable outcomes" resulting from mandatory jury duty, any more than we have in real life. It's simply a civic responsibility to create a fair system of justice by a jury of one's peers who are neutral rather than having the disputes boil over to different forums with all the same self-selected editors continuing to dominate the discussion/decision. As it is, admins. have wide latitude to block or ban editors for anything they consider to be a behavioral issue if a POV gang has it in for that editor; likewise, admins. have wide latitude to do nothing, when, in fact, action is needed. I have seen admins. say they were "afraid" to take a stand. I have seen countless double-standards, even lying tolerated, but with no venue to address the problem. I really believe having juries would address some of these most serious problems that demoralize editors.

Failure to participate with jury duty could be enforced by a bot. Although editors might try to trick the bot with bogus edits to "prove" participation, the other jurors who took it seriously would complain. I don't see much potential for abuse even close to what we already have. --David Tornheim (talk) 12:45, 25 February 2016 (UTC)

I very much wish that this could work. I would be relieved to see some way to get truly random and not self-selected participation in reviewing contested issues about content. I've participated in many RfC call from Legobot out of a sense of duty, the same sense that would compel me to go to jury duty. However, i have seen too many RfC's gamed by the same few editors with a point of view to push. I've seen them canvassed for votes, and don't trust the RfC process anymore. I've seen the same things that David Tornheim writes about: editors blatantly lying, distorting things in strawman arguments, misrepresenting sources on talk pages, derailing through fallacious arguments intentionally, filibustering, etc.... the whole gamut of "Civil POV Railroading" techniques, and many times not so civil -- and nothing ever done to remedy it through AN/I or any other process. In other words, an ecology where a few editors with an axe to grind can take over articles and force a POV to be represented as reality, in Wikivoice. I've seen this ugly process too much here. I've pretty much given up on actually editing anywhere if a few bully editors show up, because i know it's their way or the highway, and they'll stop at nothing, whatever the reliable sources may say, whatever reason may say, until they get their way. They have an agenda and they force it upon Wikipedia nonstop. Wikipedia has a serious infection of establishment trolls. Do not reply and say that i am promoting "pseudoscience' or "woo" or whatever. I am not. It's political. It's not about science versus pseudoscience. It's about forcing one version of selective representation of science into many articles. SageRad (talk) 19:43, 25 February 2016 (UTC)

You are promoting pseudoscience and woo. Pseudoscience-pushers have always said exactly the same things you are saying ("it's conspiracy and corrupt politics"), they do now, and they always will. However, this phrase, "establishment troll" is a new one for me. It is great, in its bizareness! And it could not make your stance outside the mainstream more clear. Jytdog (talk) 21:36, 25 February 2016 (UTC)

No i am not promoting "pseudoscience" and "woo", and stop the god damn personal attacks, man. There are dynamics going on, some of which serve establishment ends. That's observable. There are indeed real conspiracies in the world, too -- there are warranted conspiracy theories and unwarranted one. The idea of "the mainstream" is a term of control, and conformity to "the mainstream" is used to promote ideological conformity. That is not what Wikipedia is about. It's about getting things right. If there is a solid source that says a certain chemical causes a certain harm in humans or in animals, with due weight, then it belongs in an article about that chemical. Those who push it out of the article do seem to be "establishment trolls" when they do that systematically and round-the-clock as if their life depended on it. Stop your vile personal attacks and speak with integrity, Jytdog. SageRad (talk) 13:15, 26 February 2016 (UTC)

I support what SageRad says. There is a group of "establishment trolls" who have a strong POV and impose their POV on Wikipedia. Any opposition to their POV is labelled "pseudoscience" and any references which do not support their POV are labelled "unreliable". The result is that Wikipedia is heavily biassed in one direction and is losing credibility. Biscuittin (talk) 13:59, 26 February 2016 (UTC)

I've always recognized that our jury duty would need to be voluntary. I don't think this is a problem for us though, because think about the choices: with IRL jury duty, you go sit in a room and probably not get called, or go to work and get paid. With Wikipedia jury duty, we could assign you straight to a case that might interest you, and the alternative is doing other volunteer work on the site. Additionally, of course, we can simply notify more people - with (in)justice systems, people may worry about self-selection if you did that, but here, we'd be moving from an entirely self-selected system to one less so, and that's still progress. Wnt (talk) 17:46, 25 February 2016 (UTC)

Like I've suggested this a couple times. I did this: I used a jury of three for my second RfA, and it worked great -- 2-1 decision. Granted the jurors were picked by me, but I picked two of the essentially at random. Worked out great, and I always hoped it'd be a test case. Can we test this? Let's pick a not-terribly-important RfC or two and try it?

As to random numbers, back in the day of play-by-mail games (and I mean play by mail, we didn't have email yet) we'd write "Alcoa, shares traded, July 17" (July 17 being in the future, and "last digit" being assumed) to give the equivalent of a ten-sided-die roll (also works for even/odd). This gives a common number than anyone can check (due to an oddity of statistics, it actually only works with numbers not generated by human activity -- temperature (last digit) would do.) Herostratus (talk) 17:59, 25 February 2016 (UTC)

• Support Fantastic idea. I think something along the lines of Legobot that picks random editors for RFC's would work. I also think that Admins should be included, and perhaps a separate pool with random admins chosen to close noticeboard sections on AN, AN/I, and AE would be a good thing when dealing with behavior issues. AlbinoFerret 21:31, 25 February 2016 (UTC)

• Comment I don't think this is a practical idea and I think it is problematic make demands of volunteers who freely giving their time and attention to work on areas of the project that they find rewarding. Pulling them away from improving articles or fighting vandalism and insisting they serve on a jury if their number comes up will likely find some editors simply disappearing from Wikipedia for a few weeks or months. Since I think that the very active editor pool on Wikipedia is around 3500 individuals, I think you are overestimating the number of editors who regularly edit and would want to participate in this process.

But that's just my opinion. As in all major changes on Wikipedia, try an RfC rather than a talk page discussion and see what kind of support this has. In general, editors don't like major changes but the RfC on RfA reform succeeded and maybe this one would too. Liz ^{Read! Talk!} 21:52, 25 February 2016 (UTC)

• Comment A very interesting idea. Being selected as a jury member should not be related to number of edits. My own style of editing is multiple, brief edits, meaning I have a high frequency. I would not really like to think this makes me more likely to be selected, although I would be willing to complete my community service. If editors were to avoid jury service, perhaps a sanction that makes editing problematic e.g. 3 edits/day until engagement in jury service, rather than an outright block might be an encouragement. Randomly-selected jury service for matters of behaviour is fine - we should all know what is appropriate behaviour on here and what is not. Matters of content is slightly different. What if editors were requested to state their areas of interest (limited to, e.g. 5) - many of us already do this on our user page - and the jury members were selected from this group of knowledgeable (expert?) editors?DrChrissy ^(talk) 22:56, 25 February 2016 (UTC)

Thanks for the brainstorming! Yes, reduced editing fuctionality might be better than a complete block. I think reduced function would be preferable to some of the black/white punishments we have for other issues as well. It could also be used like plea bargain. --David Tornheim (talk) 23:52, 25 February 2016 (UTC)

Drawing people from the pool of those who identify as interested in a subject is the exact opposite of what this proposal is trying to achieve. You would pretty much guarantee that the "jury" would be comprised of partisans. I am sure many of those topic-banned in the GMO arbitration case consider themselves experts, or at least knowledgeable and interested, but it turned out that the more "expert" they were, the more they were pushing a POV. Hence bans for partisans in both camps. Imagine trying to adjudicate on the validity of claims in an article on a subject rife with pseudoscience - acupuncture, homeopathy, whatever - with a jury of people who self-identify as experts in that field. You'd end up with the situation they had at Citizendium, where they had to nuke great swathes of citable articles because they had been hijacked by "experts" like Dana Ullman. Guy (Help!) 23:12, 25 February 2016 (UTC)

It would actually depend on the size of the pool. If the pool is big enough, the selection would be random (within the confines of the subject matter) and this would prevent the sharks in the pool from chomping on the others to POV, or playing tag with other sharks. That is why I suggested a limited number of categories - so that editors would have to think about where they wanted to have influence and this would be limited.DrChrissy ^(talk) 23:23, 25 February 2016 (UTC)

The idea of users selecting categories is interesting. I agree with JzG that allowing self-professed experts to limit categories of jury participation might allow continued advocacy in that topic area. In particular, when it comes to health articles, we have an embarrassing problem with the influence of "sketpic" partisan activists promoting Skeptic magazine and Quackwatch through their Wikipedia editing as a way of proselytizing and recruiting to their their radical ideology regarding science and "truth" as witnessed by this video of Susan Gerbic. A skeptic I met in real life had a poor understanding of science and very little training. Science was more like some "fad diet", that the person had jumped on and thought would save the world. I notice that many skeptics are people who turned away from Christianity [6], [7], [8], [9] from one dogma to another, using all the same methods of promoting the dogma.

But I also think it would be unfair to force someone to participate in something they really feel they have no knowledge of the RS that may be fairly complicated and have no interest in learning it. That's why I would propose that editors be permitted to opt-out of certain discussions if they feel they are not up to it, rather than the other way around. I would rather users be randomly pushed into topics that are new, but interesting to them rather than having a big say in choosing the topic to participate in.

That said, I have close family members who are professors in science, and if they ever were to edit, they might prefer their entire editing experience was only related to science and/or their area of true expertise. I'm not sure that would be an unreasonable thing to ask, and hence, asking to have their jury service restricted to science (or other subject they feel somewhat knowledgeable on the RS) does not seem entirely crazy. So I'm a bit on the fence with your idea on self-selecting categories. Keep in mind that establishing a category will require the person requesting a jury to chose one, and there could be a big disagreement about what category the dispute fit into best. --David Tornheim (talk) 01:08, 26 February 2016 (UTC)

It is a mistake to characterise skeptics as "partisans" in this way: skepticism is the default in the scientific method, and it is skeptics who have led the way in exposing the fraudulent nature of homeopathy, the lack of credible evidence for acupuncture, the entirely bogus claims of reiki and other "energy medicine" modalities, the fatalities caused by chiropractic manipulation of the neck (a procedure with no demonstrable benefit to offset the risk), the activities of anti-vaccination activists and so on. Skeptics played a prominent role in Kitzmiller, too. Skeptics have exposed the reality behind several paranormal claims, and exposed fraudulent claims around bigfoot and other cryptids, including tracing the origin of the chupacabra.

You seem to think that all skeptics are like the newbies Susan Gerbic trains. That's not true. My skeptic friends in real life include several professors, many medical doctors, a ton of PhDs - I am ofetn well outclassed with my humble BEng. Sure, soe skeptics are clueless. But most aren't. The thought leaders of the movement included people like Martin Gardner, Carl Sagan, Robert L. Park and others. Oh, and of course Randi, who, as a stage conjurer, is very adept at spotting sleight of hand, so nailed Peter Popoff and Uri Geller to name but two out of many frauds and charlatans he has exposed.

Skeptics are basically just science advocates. Read Voodoo science some time. Sometimes science says things that people don't want to hear (the earth is warming, GMOs are safe, vaccines are safe and effective and so on). Wikipedia does not work like the news media and sees no need to give parity of esteem to science and nonsense. This is, as far as I can tell, a feature, not a bug, and any attempt to change it through the back door, by allowing votes to override policy-based arguments, would degrade the project. Guy (Help!) 23:38, 26 February 2016 (UTC)

Regarding juries, I'm not sure why you keep talking about majority. Juries often require unanimous verdicts [10]. If you have not seen the famous movie 12_Angry_Men_(1957_film), it would be well worth a view. It makes a strong case for why it might be okay to allow one juror to block a questionable decision of the other 11. My guess is that if juries are implemented, they would continue to follow the long established Wiki-consensus rules/principles, but I assume that the consensus would be of just the jurors rather than including those in the dispute--just as is done in the courts. I would prefer a super-majority or close to unanimity, but I would probably not want a single defiant juror the ability to stonewall the decision.

Thank you for the long explanation of your thoughts on Skeptics. If you got a BEng, then I'll bet you had to take some liberal arts classes and some in the humanities. Most engineers in my program (and that I have known since) hated liberal arts classes, thinking the courses were too "touchy-feely", too mooshy. They strongly preferred math, science, logic and other left brain thinking which they saw as grounded in verifiable "fact", formula and rules of nature, and did not see much need or value for the other wishy-washy foolishness. But if you take a little philosophy (especially regarding Epistemology), the solidity of science in finding all "truth" is not all that it is made out to be by science advocates who hold it on a pedestal. Empiricism has always had its limits pointed out by Plato in his Allegory of the Cave [11], David Hume in An Enquiry Concerning Human Understanding, Descartes' Meditations on First Philosophy and Immanuel Kant's novel metaphysics and the unknowability of the thing-in-itself. Science may be great for designing motors and predicting the movement of planets, understanding DNA, evolution, measuring and predicting temperatures, understanding many things about living beings, etc., especially the concerns that come up with pure materialism. But if you want to understand abstract concepts, history, the nature of morality and beauty, culture, good writing, power, etc., science and its reliance on the scientific method and use of induction have limits in explaining power and experts in many academic fields would wisely not rely solely on science for all "truth". And those were many of the well established fields the engineers who loved science tried to avoid. Imagine an architecture class based solely on science. --David Tornheim (talk) 04:05, 27 February 2016 (UTC)

David I am a humanities and a science person and the way you are mixing up po-mo and science is garbage that will never get traction in WP. Jytdog (talk) 05:23, 27 February 2016 (UTC)

You sure had me fooled by your edit history that is only about pharma, medicine, science, chemicals, etc. Of the many posts of yours I have seen in different venues, never once have I seen even the slightest interest in humanities, philosophy or epistemology. You might take a look at our article space just mentioned here. Science is just a wee percentage of the articles we have. It's funny how you brought up Postmodernism. All of the well respected philosophers I mentioned died centuries before the concept came into being. After all, science is just a branch of philosophy (see Natural philosophy) with a fairly recent and dramatic rise in interest during and since the Renaissance and Industrial Revolution (History_of_science). --David Tornheim (talk) 13:31, 27 February 2016 (UTC)

Perhaps contact 200 users to get 50 respondents

As Jimbo noted above, contacting 50 users for jury dury would likely not get 50 replies, and we would not know all of their editing-habits. So by wikimailing to 200 users then perhaps 50 would reply in time to debate the issues. -Wikid77 (talk) 17:03, 28 February 2016 (UTC)

I'm not sure why we are talking about trying to obtain 50 jurors, unless the decision was of grand importance, perhaps even more significant than an ArbCom proceeding. My thought on juries is 12 or fewer jurors, perhaps as few as 4 or 5 for ordinary decisions. There could be a provision that if one or both sides of the dispute request it (or the number of parties in the dispute is high enough), then a larger jury would be seated prior to the beginning of action. But more than 12 starts to seem unmanageable, especially if the goal is for a close to unanimous result.

I am thinking about ordinary things I see on noticeboards where there is a heated dispute about something that to others who don't edit in the topic area might see as relatively minor, but for those in the dispute may be heavily invested and ready to argue ad nauseum. Neutral 3rd parties are unlikely to participate in these posts, because it is too complicated and they don't want to get dragged into the drama, perhaps afraid of alienating one or both sides, especially if they see it as no big deal. And that basically keeps the fire burning, rather than creating an appropriate compromise that might be reasonably satisfying to both sides--something the parties themselves may be incapable of creating on their own. I don't think these little (and common) flair-ups should require a large jury.

Are you thinking it might be more like a single grand jury that listens to multiple disputes? I think it would take a very high level of commitment--like that of ArbCom--to be on a grand jury. I personally think it would be more like a specific jury for each action. I am going to make a separate post of any idea I have. --David Tornheim (talk) 09:24, 29 February 2016 (UTC)

Idea: Standing Juries

Perhaps we could have a SET of standing juries (perhaps 10 or more of these). Each standing jury would initially be staffed by a random selection of jurors who have been called for service and have also agreed to participate for the time period in question. When a jury is needed, one of these juries is selected randomly to hear the dispute. This has the advantage that the jurors would get to know each other and might find some common rules they like to operate with. And they would be available at all times to act quickly, rather than trying to assemble a bunch of strangers who do not know how to work together, as is the case in juries in courts.

Term: Editors would serve for some specific amount of time on the jury they have been assigned with. Terms would be staggered for continuity (such as 1 week or 1 month terms). Jurors could stay longer if they want (perhaps some maximum # of terms), which would create some stability and experience in that jury. Once a juror leaves, a new eligible (and willing juror) would be randomly selected to replace them.

Removal: By unanimous vote of the other jurors, anyone could be kicked off of the jury for future proceedings for that jury if they are too disruptive.

Size: I see some interesting options for size of the individual standing juries:

• Small: 4-12 members: Ordinary disputes. Entire jury always participates, except for COI jurors.
• Medium: 20 members: With this variation, not every member would be required to participate in every decision. This would be quite helpful if a juror needs to be off-Wiki for a few days. COI editors would need to opt-out. Jurors might be permitted to opt-out voluntarily for matters they feel unqualified for.
• Large: 50 members: Jurors could be to assigned to a set of subject-matter specific committees based on expertise and interest. When that standing jury is called, the appropriate committee would serve as the jury for any proceeding involving that subject matter--somewhat like how it is done in Washington D.C. in Congress and other governments. However, there is a good chance this would duplicate all the same problems we already have if the jurors are self-selecting their subject matter interest. It would also require a huge pool of jurors and overhead, and so is also impractical when combined with the idea of multiple standing juries. Nonetheless, since this is how elected governments work, some element of the committee idea might be useful with respect to juries.

--David Tornheim (talk) 09:59, 29 February 2016 (UTC)

Jan 2016 edit-counts rebound from December

See: https://stats.wikimedia.org/EN/TablesWikipediansEditsGt5.htm

January 2016 S M T W T F S 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

As could be expected, the January 2016 edit-counts have rebounded right back up as usual. Even with the reduced edits of the Friday Xmas & New Years Day weekends, the broad edit-count activity (5 or more edits) is only about 3% lower than January 2015 (the prior Friday New Year was 2009). Typical pageviews were also low during the January 1-3 weekend, so I figure low pageviews predicts low edit-counts that weekend. The next common year starting on Friday is 2021, then 2027 and 2039, as likely lower January edit-counts.

Meanwhile, the Turkish Wikipedia is on fire, with now 800 active editors, as the highest in forever (previously 739). A low spot is the Vietnamese Wikipedia, down worse than Chinese, so perhaps China has targeted the Vietnamese site as well. That reminds me of President Lyndon Johnson with the quagmire in Vietnam, where supposedly his potential win in North Vietnam was stopped by China's threat to nuke a U.S. takeover and millions of people (hence later Nixon in China & detente). Anyway, the strong January-2016 edit-counts seem unaffected by any rumors about the WMF staff, or whatever complaints here. -Wikid77 (talk) 10:55, 25 February 2016 (UTC)

• It is good to see that whateverthehell it was that happened in December was a temporary blip... The count for Very Active Editors (100 edits/mo.) across all projects topped the 11,000 mark and was the highest since Jan. 2012. On En-WP the count of 3,492 was up 5.8% from the same month of the previous year and topped the figures for January 2015, 2014, 2013, and 2012. Reports of Wikipedia's demise have been greatly exaggerated. I know that there is a way to check the percentage of edits attributable to VE, which had been running about 4% of total edits last I heard. Can anyone provide that figure for Jan. 2015 from the available data? (That would be a good number series for WMF to generate in tabular form, just saying...) Carrite (talk) 18:12, 26 February 2016 (UTC)

• Actually that is not right, there has been "a steady overall decline in active editors" up to today's date (26 February 2016), according to this article in The Register by Andrew Orlowski. MPS1992 (talk) 22:05, 26 February 2016 (UTC)

The Register is a secondary source, if you go to the primary sources and things appeared to decline from 2007 to 2014 then bottomed out and have recovered a little. One measure is the time between 10 million edits; that peaked at 37 days in 2007, dropped to 73 days at the slowest point in 2014 and is now running at about 66 days. Another measure at the other end of the scale is the editors doing over 100 edits per month in mainspace, that seems to have peaked in 2007 but rallied in 2015. So if you look at raw figures the English Wikipedia is below the 2007 peak, but that ignores the effect of the abuse filters, faster vandalism reversion and blocking and the migration of intrawiki links to Wikidata. All of those improvements have reduced raw edit count and we don't know what it would now be if they hadn't happened. ϢereSpielChequers 17:47, 27 February 2016 (UTC)

The operative phrase is "Active Wikipedians," which means a very specific thing — the group of those making 5 or more edits a month. LINK This number has not rebounded as robustly as the core volunteers number, the "Very Active Wikipedians" making 100 or more editors a month. In fact the January 2016 figure of "Active Wikipedians" on En-WP (30,993) continues what may be characterized as a "steady overall decline" from the peak year of 2007, when WP was expanding fast and was something of an internet fad. So Orlowski and The Register — who are consistent critics of WP — aren't exactly wrong, they are just phrasing the metrics in a very tendentious way. For all practical intent, the pivotally important declining count of core editors attenuated in 2012-2013, took another significant drop in 2014, and recovered to the 2012 level over the past year. There are somewhat fewer casual editors, but that seems to me natural with a maturing project, in which footnoting becomes mandatory and the typos are steadily being fixed. Carrite (talk) 17:09, 28 February 2016 (UTC)

Ah good point about editors making five edits in a month. But remember how many edits it takes for a vandal to get four warnings and a block. The edit filters have lost us a lot of the vandals who were in those stats in 2007, though as far as I'm aware by now they must be getting diminished returns, in fact the first half of 2015 showed month on month increases. Since mid last year that has dipped again. As for whether we would expect to lose some of the editors who make >=5 edits a month, I'd agree that readers who fix the odd typo will be finding fewer typos as they get rarer, so some of those editors will drop below the 5 edit a month threshold. But more generally I'd say that whatever happened to new autoconfirmed editors mid year seems to have also hit the "active editor" count, though not as sharply. ϢereSpielChequers 21:54, 28 February 2016 (UTC)

Yes there was a processing glitch re the December figures. Also there does seem to be a problem with editors reaching their tenth edit, that has stopped rising and started dropping again since July, especially recently. I'm in discussion with a number of people at WMF and elsewhere to work out why. It could be just a glitch. ϢereSpielChequers 17:50, 27 February 2016 (UTC)

And would a glitch in the December edit-counts also be affecting the January counts as well? -Wikid77 (talk) 23:42, 28 February 2016 (UTC)

There was a glitch in the December count that was fixed in the January run. One of the theories for the Dec-Jan dip is that it could be another as yet unidentified processing glitch. ϢereSpielChequers 09:23, 29 February 2016 (UTC)

WMF mission - the unstated part about self-preservation

Just a quick thought. I think one of the key challenges for the WMF itself, is how it can stay relevant. Lila's long statement includes this:

As we matured, we encountered two fundamental, existential challenges. <snip> The other is external and is emerging from our own value of freely licensed content: Many companies copy our knowledge into their own databases and present it inside their interfaces. While this supports wider dissemination, it also separates our readers from our community. Wikipedia is more than the raw content, repurposed by anyone as they like. It is a platform for knowledge and learning, but if we don't meet the needs of users, we will lose them and ultimately fail in our mission.

I read that, and thought, "Really? Which part of the WMF mission is she talking about failing there?" I realized that it is the unstated part - that the WMF itself should persist, grow, remain relevant. Most every organization wants that. It is an unstated thing, generally.

And I thought - "existential challenge" to WMF? As in "live or die" crisischallenge? What is that? What Lila doesn't say, is maybe that the way things are going, people get the content we create elsewhere, readers don't come to the WM/WP sites so much any more, donations fall, WMF itself cannot grow, maybe even shrinks.

I think that is a big deal to the WMF board and ED.

When there are existential threats, people think about self-preservation. So - how do they keep the platform relevant, is #1. I think that is what the "knowledge engine" is/was all about - making the wikipedia.org portal, and search results produced by it, why people come to Wikipedia. They want that, so the WMF itself can persist and remain relevant.

I think, that maybe that has nothing much to do, with supporting what we content-creators do. It is one reason I have been saying that I have this strong sense, that the WMF seems to be ready to throw the editing community under the bus (or push us to the back of the bus? bad metaphors) - to "hijack the platform", as I have written. (whether this is "paranoid" crazy talk or not, I don't know, as the WMF is not sharing its actual strategy with us in any clear way)

Does the editing community overall agree that the WMF is an "existential crisis?challenge"? (outside of the issue's with Lila's leadership - I mean bigger picture stuff that hiring Lila was meant to address in part)

Is what the WMF sees as a life-threat a bad thing to us at all? I don't know.

Personally I have just wanted the servers and software to work so we can publish our content and get it out there, and haven't even thought about the WMF much at all before James got dismissed. Not really part of the world I see, even - just something in the background. I am even unaware if donations are falling and the WMF is shrinking (outside of people leaving because they are so unhappy now) .

This is one place where I think there are not shared values or priorities, or even a shared sense of reality, between the editing community and the WMF leadership.

This may be obvious to tons of people here. It is just something I have been getting clarity on. Jytdog (talk) 18:26, 25 February 2016 (UTC) (striking, valid point. i can acknowledge and change when people point out i made a mistake Jytdog (talk) 18:52, 28 February 2016 (UTC))

I recommend being really careful about accuracy. Don't carelessly slide from "challenge" to "crisis". There are many challenges which are not crises. I recently joined the board of The Guardian. They face a big challenge - revenues are much less than expenses, and the trust fund which covers the difference won't last forever. By some estimates, there are 8-10 years of money left. That's a big existential challenge, but it isn't (yet) a crisis.

Our movement - dedicated originally to creating a free encyclopedia for everyone, in their own language, and dedicated more broadly to "a world in which every single human being can freely share in the sum of all knowledge" faces real existential challenges. They are not (yet) crises, so we should not run around in panic. But we should be serious and aware of changes in the world which will impact, negatively in some cases, our mission.--Jimbo Wales (talk) 19:13, 26 February 2016 (UTC)

• The Knowledge Engine by Wikipedia™ idea was all about creating an app or a function that would drive readers to the Wikipedia website, where they would be exposed to donation banners saying that unless people donated money there would be engineers going without coffee and the servers would stop running, or whatever the A/B tested winning nonsense line of the year was. And then the money would continue to flow and the engineering staff would continue to grow, and the sun would never set on the San Francisco empire. Of course, we the core volunteers care little about empire building. We want the software to continue working, to be improved in an evolutionary manner without disruptive bugs, for our backs to be covered if we are sued for trying to maintain NPOV, for research resources to be made available for us, and for filters and bots to be developed to help us perform our work. This is a big element of the fallout over the Knowledge Engire — we have no intrinsic need as encyclopedists to take on Google's commercial search engine just because they've figured out how to monetize our non-commercial work in a way that perhaps endangers the multi-millions that WMF rakes in each year. Carrite (talk) 15:26, 26 February 2016 (UTC)

@Carrite: I wouldn't belittle it like that - bypassing our site means that there are fewer new editors, not just less in donations. Still, I think we should have pinned our hopes on hooking, not crooking - only a solid web of quality content ensures that search engines continue to find it necessary to deliver readers here. I'd rather Wikipedia have fewer restrictions, cover less notable events, allow more social media like use of the site by individuals to encourage their interest - compete more with Yelp than Google, even. I mean, if we wanted, we could throw out WP:NOT, literally have a map of your hometown with pictures of all the local businesses - while that of course invites unwholesome influences, they would be on a smaller scale than the ones that have "notable" products. We could have a hierarchy of content, some of which is more personal experience and unreliable reference than the hardcore encyclopedia, with proper disclaimers and warnings. There's no reason why you should have to be born a robot to be allowed to snarf up content on the web that isn't suitable for a proper encyclopedia article! Wnt (talk) 17:26, 26 February 2016 (UTC)

Speaking for myself, I don't think of improving internal search and discovery to be primarily about revenue and page views. Those things are important, to be sure, but they are never the most important things. As Wnt says, the visibility of our work, the usability of our site, the usefulness of our navigation, impacts our work in lots of different ways direct and indirect, including attracting quality editors and retaining quality editors. (I'm not saying I agree with all of Wnt's suggestions - I don't, although I encourage people to brainstorm openly like this.)--Jimbo Wales (talk) 19:13, 26 February 2016 (UTC)

Actually, one could make a statistical argument that the growth of Google's "Knowledge Boxes" or whatever they call them is positively correlated to editor growth, not negatively correlated. There is no statistical evidence whatsoever for a negative correlation. The very early thinking about "Knowledge Engine by Wikipedia" (the actual first name!) revolved almost entirely around needing to come up with a compelling feature that brought readers to the WMF site (implicitly, so they could eventually be solicited to donate money). I'm not being cynical, this is historically accurate. Carrite (talk) 17:53, 26 February 2016 (UTC)

I don't agree. I think it there is very clear and compelling evidence that improvements in the quality of Google's search results has resulted in decreasing traffic for us. 10 years ago if you typed "How old is Tom Cruise?" (or variants on that) Google didn't know the answer, but they sent people to Wikipedia. Now, they just tell people the answer - because their machines have read and digested Wikipedia. On one level, that's totally great - that's part of why our work is freely licensed - we want people to use it to make the world better in ways big and small. We aren't a commercial business trying to maximize page views. On another level, we need to be objective and acknowledge what it means for attracting editors and - yes - donors.--Jimbo Wales (talk) 19:13, 26 February 2016 (UTC)

@JW. It's not really an opinion I assert, it is a demonstrable statistical relationship (albeit a counterintuitive one). To wit: Knowledge boxes are maybe 3 years old. The count of core volunteers has improved, not declined over those three years. Therefore the statistical correlation is positive rather than negative. Proving that is just a matter of showing the statistical arithmetic. Now page views, those I will give you are way off — which raises the interesting hypothesis that there is little connection between page view numbers and volunteer numbers. Lest you worry too much, there has never been a decline in WMF donation levels, Google knowledge boxes or not... Carrite (talk) 17:18, 28 February 2016 (UTC)

Speaking only as a sample of one, I click through from Google to Wikipedia much less often now because of their knowledge box and Google Graph info box. (I missed a flight from London to SF because Google's "knowledge" box sent me to the wrong terminal.)

Yes, if our work gets pushed so far back behind re-users' portals that readers almost never land on wikipedia.org sites, Wikipedia as such will become a failed experiment. But that won't happen. People will always want thorough coverage of topics and issues. It seems inevitable, though, that our traffic will decline significantly. The best counter to that is to make our articles even more awesome - by that, I mean improve the quality and rigour of our offering. I don't support Wnt's notion of becoming more like Yelp.

By the way, Wikidata is taking information from Wikipedia and feeding it to Google, Bing etc. without requiring attribution. At least now, when I see Wikipedia content in Google's Knowledge Graph box, I can click the link to Wikipedia for more. With Wikidata just raiding our data and pouring it into commercial portals via the back door with no attribution required, they seem to be accelerating the process you describe as a "real existential challenge" to Wikipedia and the movement. Good for Denny's employer, not so good for the WMF that relies on the rivers of gold flowing to it through Wikipedia banners. --Anthonyhcole (talk · contribs · email) 07:26, 27 February 2016 (UTC)

As with JW above, you are observing the true phenomenon that WP's page-view numbers are way off due to the Google info box things and postulating a problem which may not even exist based upon that. Page-view numbers are down, but donations are way up. These two things, which intuitively seem to have a relationship apparently do not have such a relationship at all. WMF should start surveying its donors at the time of their donations as to why they are donating — although I doubt they will do that because it is a set of answers they probably do not want to hear. But if one wants to learn who is donating and why, that's how one would find out... Carrite (talk) 17:25, 28 February 2016 (UTC)

• This is a great discussion. I would just like to focus here on the gap in missions. What editors here are saying, is that from their perspective - if WMF's goal was primarily to support and grow its flagship (the encyclopedia) - and keep it relevant, the WMF would have made different choices. "WMF will remain relevant" is not part of the mission of the movement. It is an unstated part of the WMF's mission that I think moved to the fore.

This comes down to the board itself. Every entity eventually starts working on self-preservation, regardless of why it was originally set up. Its why we can't get rid of government agencies once they are set up. I think this is also what happens when you get significant voices on the board not coming from or not in touch with the movement. Folks who serve on for-profit boards are wired to make that organization grow and flourish. They don't even question if that is a good thing or the right thing - it is just what they are do.

I think this is part of what went wrong with Lila; nobody articulated that a big part of her mission seems to have been to "remake the WMF so that the WMF as an entity remains relevant and flourishes" - to make it a relevant and important tech company (and freedom to "break eggs" along the lines of what George Herbert wrote here). She apparently didn't execute on that well, but part of the problem was, I think, unarticulated mission clash.

Now, post-Lila, might be a good time for the board and the community, to look at the board and what its mission is going forward. Jytdog (talk) 19:44, 28 February 2016 (UTC)

Lila resigns

Within the last 20 minutes Lila Tretikov has announced her resignation as ED on wikimedia-l. BethNaught (talk) 19:04, 25 February 2016 (UTC)

Preliminary followup from Patricio:[12] --Guy Macon (talk) 20:04, 25 February 2016 (UTC)

Whew. Now we really have to try to reset and start over, somehow. A good starting point would be the WMF Board telling us the story of what has gone on, in a way that respects everybody's experiences, without further obfuscation. that would be amazing. Jytdog (talk) 19:30, 25 February 2016 (UTC)

I continue to advocate for that.--Jimbo Wales (talk) 19:18, 26 February 2016 (UTC)

And this validates all the "discussions" held on this and other pages in what way? I fear I do not see how any of this arm-waving and angry discourse actually benefits either Wikipedia (as an encyclopedia) or the WMF (as the legal entity behind Wikipedia), and it decidedly seems to make Wikipedia into the "tabloid capital of the Internet" with regard to what is, and what is not, actually important. Collect (talk) 19:37, 25 February 2016 (UTC)

Well, I partly agree. The drama has been sadly much higher than normal, and that's unfortunate. But it is also really important that we stay wildly open. Democratic institutions are often noisy, and not all the rhetoric is always helpful. But that's a lot better than a closed approach where dissent is discouraged or eliminated.--Jimbo Wales (talk) 19:18, 26 February 2016 (UTC)

The WMF has lost a lot of great people - see [13] - most recently User:Ironholds, as of the 23rd. Unfortunately, few of them are really explaining what's wrong well enough for some of us in the audience to understand exactly what has gone wrong. I am suspicious though this goes all the way back to when we lost Mike Godwin, seemingly over a minor spat involving the article about the FBI - and if that's true, the problem predates Tretikov, and will very likely continue under her successor. Wnt (talk) 19:55, 25 February 2016 (UTC)

I don't think Mike Godwin's departure had anything at all to do with the article about the FBI! I've never even heard such an idea.--Jimbo Wales (talk) 19:18, 26 February 2016 (UTC)

@Jimbo Wales: I remembered he took a lot of heat in this discussion and related offline commentary in August 2010, and when he left in October 2010 I think there was some speculation along that line, neither more nor less convincing to me than a lot of the speculation we're hearing now about why various WMF employees left recently. I apologize if this impression was mistaken. Wnt (talk) 01:54, 27 February 2016 (UTC)

• I hereby call for not commenting on the accuracy of the claims made in the resignation message, and for those who agree with me on this to completely ignore any such comments (any response just encourages further discussion) It would not be fair to do anything other than allowing Lila her final say and moving on to a discussion about how to move forward. Who is with me on this? --Guy Macon (talk) 19:58, 25 February 2016 (UTC)

In general, I am, subject to the caveat that I don't think we should try to shut down discussions that people want and need to have. I do encourage dignity and that people be forward looking. The fundamental question is "How can we improve things?" - This necessarily involves some element of looking back on "What went wrong?" But we shouldn't get bogged down in only "What went wrong?"--Jimbo Wales (talk) 19:18, 26 February 2016 (UTC)

While one could wish that the director of a foundation of this size and importance could use the word "equitably" correctly, and that her personal staff would in any case detect the mistake in a document of such importance to their director and to the foundation, I agree with Guy Macon that we ought to avoid a long wrangle here about the substantive claims. MarkBernstein (talk) 20:03, 25 February 2016 (UTC)

Agreed. There is no need to kick someone when they're down, or to shoot them from behind as they walk away. (Don't bite: metaphors.) The record can be left to speak for itself. BethNaught (talk) 20:07, 25 February 2016 (UTC)

This response seems to set the right tone:

. . . I know many people disagreed with your strategy and your decisions, but as I had the privilege to knew your personally during the last 2 years, I know each one of them were truly in the favor of the movement in your personal point of view. Thank you for all the works during the last two years. For being available to speak personally with each one of us. I learned a lot for each meeting or conversation with you. Thank you Lila, and good luck whatever will the roads will take you. -- Regards,Itzik Edri , Chairperson, Wikimedia Israel [14]

-- Alanscottwalker (talk) 20:19, 25 February 2016 (UTC)

From what I've gleaned over the stories and messages I've read, much of the clash seemed to be putting a business-oriented person in charge of a large non-profit where many of the employees work because they believe in the mission, not for the size of their paychecks and so demanding results (mentioned in an earlier email from Lila) will garner a different response than it would at a startup. I think she and the Board underestimated how much they need the support of the larger Wikipedia community and how much a non-profit needs to care about process as much as the achieving results. Both are possible, of course, but the repeated calls for more transparency indicates that concerns about the process of governance and leadership might have suffered. Again, some of these incidents wouldn't have been an issue in a for profit company but in an organization that relies so much on volunteer labor, it's essential. Liz ^{Read! Talk!} 21:00, 25 February 2016 (UTC)

I think Liz's comment hits the mark. It all comes down to whether Wikimedia is a movement or an entity. Process is paramount with a movement and results is paramount with an entity. Most movements seem to evolve or devolve into entities within a decade or two, or else evaporate. But a few, a very few, last long enough to give evidence that a movement can possibly continue indefinitely. In this tiny category I think of movements which have an indefinite "live the movement" or, to be dramatic, "live the revolution" mantra and ever present commitment among the participants, with the goal of continuing ad infinitum as opposed to say, working towards IPO day. Right now the only examples I can think of are major religions and one or two long lasting political movements, but I see no reason why there could not be a "movement", dedicated to something as simple as spreading free knowledge, that could continue ad infinitum. If that is the purpose, then perhaps its best to accept that process must rule, and quantitative results should be seen as icing, or not, on the cake. Nocturnalnow (talk) 05:54, 26 February 2016 (UTC)

Let us put in perspective the assertion "to be putting a business-oriented person in charge of a large non-profit where many of the employees work because they believe in the mission, not for the size of their paychecks and so demanding results will garner a different response than it would at a startup". From the Knight Grant document, the average paycheck of the WMF technical people is 125K$ per year. Asking for some results doesn't seem so cruel. Pldx1 (talk) 19:29, 26 February 2016 (UTC)

Sticking with perspective, employee retention in San Francisco is a huge priority because of a supply shortage in the tech sector. Also, San Francisco/San Jose are beyond expensive, like average house prices of a million. 125 there is like 15 in Waco. Context, Context, Context. Nocturnalnow (talk) 04:53, 27 February 2016 (UTC)

Results and process are both important. Lila was there because the previous regime were not producing technical results. Failure after failure. Do I need to list them? Sue Gardner narrowed her focus on getting technical results, and sensibly passed as many non-technical functions as possible across to the community, like the North America education program and the Funds Dissemination Committee. She should have replaced the C-level technical team, and a number of others, before she left. Leaving that to the new ED was a poison chalice in my opinion.

Presently, on wikimedia-l and Andrew Lih's Wikipedia Weekly Facebook group, the discredited, discarded tech team, who treated us with such contempt and whom Sue failed to flush away, are rewriting the narrative. It seems there were no serious tech issues. Expecting them to treat us with respect is unreasonable - we're too rude and luddite to deserve it. Expecting them to generate and monitor key performance indicators is too much to ask of underpaid missionaries. It must have been a shock to these people, who had a job for life under Sue, to realise their jobs depend on them actually producing desirable results.

Perhaps Lila did the right thing in resigning, given the transparency issues around the Knowledge Engine and staff complaints that she was not communicating with them enough. Only a rigorous independent enquiry can clarify that.

Regarding the other frequent complaint, the persistent rethinking on the strategy, I think it was a good thing, though obviously very uncomfortable for the staff. Under her, the WMF was moving closer and closer to a responsive (to the community and readership) strategic planning process - including the 2015 Strategy Consultation, a project that collected input from the wider volunteer community (not just the most vocal few) and, most importantly, the readership. (Off topic but: surprisingly, although there were some different priorities in those two groups, there was a lot of congruence. If you define reliability as neutrality and accuracy, for readers and the wider volunteer community alike, the biggest concern is the unreliability of our offering.) --Anthonyhcole (talk · contribs · email) 06:26, 27 February 2016 (UTC)

Well, it is funny in a sad way, that one heard so many criticisms of Sue, and then she steps down, in part in response to that criticism and it's de ja vu, someone new to beat-up on - Sue, at least, had the benefit of hiring, hiring, hiring, and it necessarily became 'now what' with all these people. Alanscottwalker (talk) 15:23, 27 February 2016 (UTC)

@Nocturnalnow: According to [15], maintaining the WMF at SF rather than moving to LA is yet another way for bullying the poor underpaid staffers. Wouldn't they be more productive when being paid half an house each year, with a boat anchored at the rear of the garden ? Pldx1 (talk) 15:13, 27 February 2016 (UTC)

I'm late as usual but am disappointed to hear this. Anthony makes some very good points. I also had the impression that Lila was actually willing to listen to and engage with the community, in stark contrast to Sue. Shock Brigade Harvester Boris (talk) 15:35, 27 February 2016 (UTC)

Gee, you mean Lila Tretikov never demeaned the community as in Sue Gardner's comment of "filthy T-shirt eating Doritos" (CBSnews.com); when I heard Sue say that comment, I was shocked at first, and then it made my blood run cold, as no wonder the MediaWiki software had major bugs unfixed for years because not only did she not care about Wikipedia, she evidentally deeply despised the editing community, even if she was lying to herself or others to claim otherwise. It makes you wonder what kind of mean-spirited leader would stereotype users in a biased way by saying, "in a filthy T-shirt eating Doritos". That was a giant "RED FLAG" and thank God she said it to reveal her true mindset, but she even had many WMF staffers fooled to think she was nice (to this day?). -Wikid77 (talk) 18:06, 27 February 2016 (UTC)

Looks like Sue Gardner was talking about the hacker stereotype: "Sue Gardner: Women are less likely to kinda geek out at their computer for 10, 20, 40 hours. I mean, there's a reason that the stereotype of the hacker is a guy in a filthy T-shirt eating Doritos, right?" Nocturnalnow (talk) 00:41, 29 February 2016 (UTC)

Boris, the wider volunteer community has not really been aware of this drama (as opposed to the staff, the activists and the inner circle of volunteers who engage regularly with staff). A lot of it has played out on wikimedia-l and the Wikipedia Weekly Facebook group. Yes, Lila did give a damn about the communities, and also about the reliability of our content. But she lost a very important constituency - her staff. I really hope the WMF gets someone independent, with the right qualifications and experience, to undertake a rigorous, deep inquiry into how this all happened - including the Doc James dismissal and the hiring of Arnnon Geshuri. If they don't, I see more than just Lila's head rolling here. Those afore-mentioned players are fairly united in their desire for such an inquiry. --Anthonyhcole (talk · contribs · email) 16:23, 27 February 2016 (UTC)

An inquiry. That's a great, no downside, idea.... a rigorous independent inquiry might be really helpful going forward. Just brainstorming here and thinking that maybe the most experienced and respected volunteer community members, perhaps administrators and past ArbCom members, can make up the Board of Inquiry? Nocturnalnow (talk) 18:05, 27 February 2016 (UTC)

I nominate Risker as chair of the inquiry. She's pretty awesome with investigative projects. Liz ^{Read! Talk!} 01:18, 28 February 2016 (UTC)

She is awesome. I'd personally like to see her on the board, but that's up to her and the voters of course. :-) --Jimbo Wales (talk) 17:02, 28 February 2016 (UTC)

This is an interesting idea. But I doubt the WMF would give a group of volunteers the unfettered access to confidential information needed for a meaningful inquiry. On the other hand it couldn't hurt to ask. Shock Brigade Harvester Boris (talk) 03:30, 28 February 2016 (UTC)

We have to fix the software, nobody else will

As Jimbo has welcomed various software talks on his talk-page, and now, after 5 years, we need to take more-direct action to fix bugs in the MediaWiki software. In particular, the Community Tech team is ready to fix diff problems (an added blank line seen as the next line totally changed, or no specific differences shown when a line is several hundred thousand characters long). -Wikid77 (talk) 21:27, 25 February 2016 (UTC)

Yes, I have been noticing the diff problem, and it is really a pain to figure out what the change is when it does that! --David Tornheim (talk) 23:24, 25 February 2016 (UTC)

There have been considerable advances in computing minimal edit differences since the days of "diff". See Edit distance. Wikipedia should be using them. Wikimedia diffs displayed often indicate much larger changes than were actually made. John Nagle (talk) 23:29, 25 February 2016 (UTC)

I suggest that we wait a bit until the WMF management issues sort themselves out, then start with a small, easy-to-implement, and noncontentious change. If we can actually open up a line of communication that results in the change either being implemented or rejected, then we can move on to larger things. Something really small, like changing the CR+LF to LF in the HTML we output. See Newline. --Guy Macon (talk) 00:21, 26 February 2016 (UTC)

Diff must span newlines & use multi-line resync

Naturally, the Community Tech team could become overwhelmed by diff problems because the current tactic of matching line-by-line will abandon phrase comparisons when a line is split with more newlines as manual line breaks. Instead, the diff program must be radically altered (eventually) to compare text across multi-line breaks, to show how one line was split into parts and then text altered in those separated parts. Also, the resync tactic will need to match partial (modified) lines, plus multiple lines as a look-ahead tactic for wp:wikitables where lines seem repeated for "|-" plus column styles "|valign=top|" as otherwise the diff would mismatch rows of a table as if all subsequent rows had changed, rather than just the fewer actual changes. As a phase-one improvement, just increase current diff to show changes between 10,000-character lines, which currently highlights the whole line as completely new text. Because of the complexity involved, the WP community needs to help explain the long-term solutions so the wp:developers don't give-up in frustration and go work some easier issues to avoid the workload pressures of fixing the vital diff operations. -Wikid77 (talk) 15:16, 26 February 2016, revised 11:52/12:31, 28 February 2016 (UTC)

Example diff of line over 10,000 characters

The megadiff problem was noted in pop singer's article "Tishma" of 28 Sep 2015 (later rewritten to have shorter lines), as an example diff of a line over 10,000 characters long, which could not pinpoint the phrases changed but instead showed the entire massive altered line as if totally new text (see: dif6701). To handle such long lines, the diff utility might need to be expanded with much larger line buffers to compare all 10,000 characters across the line. Meanwhile, splitting long lines can help diff of future revisions; here I split that long line in User:Wikid77/Tishma_test, as 7 parts in 2 revisions, to reveal the changed text was removal of a Bot-generated cite parameter: "|unused_data=@ http:/ /www.bangladesh-pratidin.com" (see: dif7822) hidden in the original long line. -Wikid77 (talk) 09:41, 27 February 2016, revised 11:39/12:31, 28 February 2016 (UTC)

Have submitted a phabricator bug for this? Most diff programs work on a line by line basis so this is quite expected output. There is a bug T123110 which seems to be for this exact problem. Adding this example to that bug might help gain some traction on the bug. --Salix alba (talk): 12:18, 28 February 2016 (UTC)

The WMF Community Tech team is working to fix diff, as a high-priority project for Spring 2016. -Wikid77 (talk) 12:31, 28 February 2016 (UTC)

I've mentioned this thread in that bug.--Salix alba (talk): 14:22, 28 February 2016 (UTC)

Where we left the discussion

copied out of archive, as I just posted this today. will let this go if it gets archived again Jytdog (talk) 03:31, 26 February 2016 (UTC)
Jimmy, in my professional life I generally wait three days before following up on an ongoing conversation. So I have here. Here is where we are.Jytdog (talk) 19:26, 24 February 2016 (UTC)

Hello Jytdog, I can't speak for Jimmy, but since he has some quite pressing things to deal with at the moment, and I appreciate your concern, I thought I would respond to points that I can. I was not on the board when this grant was submitted, but was until July. – SJ + 05:19, 26 February 2016 (UTC)

Sj That is super nice of you. Yes I imagine Jimmy has his hands very full in SF and I hope he is bringing comfort, and learning. Thanks! Jytdog (talk) 07:35, 26 February 2016 (UTC)

• About James dismissal, I explained how the gap between what the sides are saying is tearing up the community, asked if you understood that, and asked if you would commit to work toward a joint statement, perhaps with a mediator. you wrote, I understand that. I'm open to trying, but I don't see how it is possible. Perceptions seem radically different.. I asked if you would actually commit to trying. That is unanswered. Jytdog (talk) 19:26, 24 February 2016 (UTC)

Sure, I will if he will, but I continue to think that this isn't really going to be the most effective way forward. What I continue to advocate for at the board level is the freedom to talk openly about why we voted James off the board. I think I've been the most open amongst the board members, and that isn't a criticism of other board members - we've been advised to be cautious about what we say, and I'm just a little more empowered by my position to be blunt. I've advocated for something that is probably a bad idea - publishing a whole whack of internal board emails. Ok, so that's probably extreme, so I am not pushing hard for it. But the current situation is not where I would like it to be.--Jimbo Wales (talk) 19:33, 26 February 2016 (UTC)

Doc James, would you please reply here, to the "I will (commit to trying) if he will" from Jimmy? Thanks, Jimmy, and thanks in advance, Doc James. Again, I encourage use of a mediator due to everything that has transpired. Jytdog (talk) 23:19, 26 February 2016 (UTC)

I am supportive of an independent review of these matters as I have stated a number of times already. I am also happy to go over the issues and documents publically following the publishing of "a whole whack of internal board emails". Doc James (talk · contribs · email) 04:58, 28 February 2016 (UTC)

Thanks James. Yes, independent people being involved is important if the community is going to something that makes sense of all this. Jytdog (talk) 05:02, 28 February 2016 (UTC)

• I've asked in several different ways for you to tell the story of how the vision for technology development has evolved over the past year - the vision that produced Discovery's three year plan and the long-term plans discussed in Knight grant including its board-approved $2,445,873 budget for Stage 1. I've also asked where this vision stands now. You have not provided that. Jytdog (talk) 19:26, 24 February 2016 (UTC)

I think I've answered with all that there is to know, as I've said. Sj's comment, next, is quite good. I think you're looking for something much more grand. To really understand all this, you have to really grasp that there was never a grand plan to fundamentally change everything. There was what I think is great - a discussion of a challenging but achievable project to improve search and discovery on the site.--Jimbo Wales (talk) 19:33, 26 February 2016 (UTC)

• First, the Board did not specially approve a $2.5M budget for 'stage 1' of the grant proposal. Financial approvals by the board happen in bulk every May-June when it reviews the annual plan. The board approved the 2015-16 annual plan, with funding for the entire staff, in June. This included funding for the discovery team. The board had not seen any version of this grant at that point; the team's budget was funded based on its public projects and targets. The budget mentioned in the grant proposal seems to be the annual budget for that team. [Trying to fit the realities of an annual planning cycle into the requirements of a funding proposal is not unusual in the grants world.]

  Second, the vision for the discovery team was only part of the total technical vision (and technical budget). The overall vision was most concisely defined earlier in 2015 by the Call to Action. Portions of that call that this grant might touch on [I am speculating here]: a simplified user experience, new models for knowledge creation, and strengthening partnerships with orgs that use or contribute free knowledge. The only discussion the board had about the development of a Discovery roadmap and vision, before the grant was submitted, was around a presentation at the July meeting. It seemed a piece of the scaffolding in the call to action, facilitating curation. It did not seem like a dramatic shift for the projects or tech infrastructure overall. – SJ + 05:19, 26 February 2016 (UTC)

I hear all that. This is more detail that what Jimmy provided in the first go round (now in the archive). You have focused tightly on the Knight grant, and that was not really what I was asking about. It is more about the the vision/long term strategy that produced the Knight grant - the story that would have been told while the three-year Discovery deck was being delivered - the arc that would WMF from where it is now to being a really innovative tech company, centered around delivering value in the form non-commercial knowledge that readers could get only through the wikipedia.org portal. That thing. The thing that Doc James said he was told had a multi-year, multi-tens-of-millions of dollar cost. Was any of that vision solidifying while you were on the board? Jytdog (talk) 07:35, 26 February 2016 (UTC)

Ah, I see. "Arc to a really innovative tech company" – interesting question. I think there are some great innovations now, in individual teams. An overall arc is an open question; every plan and strategy thinks about / touches on that, but I didn't sense a grand arc tied to "changing the tech essence of the foundation" tied to Discovery. (That said: it is a new team, with a new lead, and tackling fairly new use cases, so as with all such teams, a chance for something different to develop.)

I don't recall discussing a three-year Discovery plan; just the ideas around 'Year 0' with some general thoughts about how this could have excellent feedback w/ wikidata and maps. The greater detail seems interesting, and might have spurred useful board discussion - but it didn't come up while I was on the board.

Even in the three-year plan, I may have missed something, but I don't see anyone aiming for 'things readers could only get through wp.org' – the project seems committed to connecting many different free knowledge sources and making them all more visible. I hope it is possible to learn more about these ideas this year, in the public strategic brainstorming. I think it's essential for the projects and the foundation to look outwards, form partnerships, and think+work at this scale, and hope this series of communication failures doesn't discourage such brainstorming in the future.

Finally: I may draw the line at three-indent discussions :) I don't want to belabor these points & trying to support the current changes in other ways. But thanks for these patient questions. – SJ + 22:34, 27 February 2016 (UTC)

I hear you. Thanks very much. Jytdog (talk) 22:50, 27 February 2016 (UTC)

• I've asked you to let me know if the WMF board considers that vision and its evolution to be confidential and said that this appears to be the case. You have not responded to that. Jytdog (talk) 19:26, 24 February 2016 (UTC)

I don't think long term vision should ever be confidential. My point is that you're looking for something that, in my opinion, doesn't exist and didn't exist. Indeed, one of the major staff criticisms of Lila and one thing she acknowledged as problematic, is that formalizing a long term strategy was taking much too long.--Jimbo Wales (talk) 19:33, 26 February 2016 (UTC)

This is hard to hear. Saying "vague" and not defining further, is not helping. I feel like you are relying on a narrow interpretation of "strategy" here and above to not tell the story. There is a story. Things happened with regard to planning. I can only conclude that the board considers the evolving story confidential. I'm sorry. Jytdog (talk) 23:40, 26 February 2016 (UTC)

No, I'm telling you everything. SJ's comments are pretty thorough.--Jimbo Wales (talk) 10:06, 27 February 2016 (UTC)

Gorillawarfare's timeline makes it clear that things were first pitched to Knight in April and the KE was actually presented to the board at the end of June. The board sat there and heard that, and there was a story that was told as the slides were presented. We don't know what was in either of those pitches, nor if that was in the main stream of what Lila had been talking about or if she and her team went rogue. Please tell the story. If you actually don't know the story, please say that. If you consider the story confidential, please say that. Jytdog (talk) 17:49, 27 February 2016 (UTC)

• This was never the case when I was on the board; I do not think that this has changed. It is more likely that the vision was simply unclear, than that it evolved into something confidential that could not be shared. – SJ + 05:19, 26 February 2016 (UTC)

I hear that and this is also what Jimmy said - about it being vague. This is hard to understand because things like the 3-year discovery plan was actually presented and the Knight foundation people were told something about the bigger story that the grant application is the first stage of. In my professional life I don't pitch things that are not well thought out and that I have buy in from my management for. The WMF - the entity - seems to have had some story definitive enough to tell. How can I hear that story - and the variations on it as it changed? That is what I am after. I know I am nobody and nobody owes me anything. But not getting an answer (or being told - hey go ask over there) is making it ~feel~ like this is stuff that is considered confidential. That's all. Thanks again! Jytdog (talk) 07:35, 26 February 2016 (UTC)

"How can I hear that story - and the variations on it as it changed? That is what I am after. I know I am nobody and nobody owes me anything. But not getting an answer (or being told - hey go ask over there) is making it ~feel~ like this is stuff that is considered confidential." Well, if this is all about wanties not needies, and all based on feelies, why not back off and end it here? AnonNep (talk) 10:41, 26 February 2016 (UTC)

Because it is important to me at least, and I think to others. And I like Sj and want to explain myself. I don't know who you are, and based on what you wrote here, I am good with that. Jytdog (talk) 11:06, 26 February 2016 (UTC)

"Because it is important to me" Hence my point - ego based discussions are rarely constructive. "I don't know who you are, and based on what you wrote here, I am good with that." Ditto. Meow, girlfriend. AnonNep (talk) 11:16, 26 February 2016 (UTC)

You may consider it weak, but I am trying to keep this conversation human, and own what i know and don't know, and be clear that I am only speaking for me, and not making Grand Statements like people tend to make here. I am not climbing the reichstag; I am trying to have a conversation. Now shoo kitty. Jytdog (talk) 23:48, 26 February 2016 (UTC)

You wanted respectful discussion & answers? Look at your own behaviour & don't howl about how you didn't get it (in more ways than one) *whoosh* (sound of conversation sliding above your head). AnonNep (talk) 08:36, 27 February 2016 (UTC)

• As Lila has acknowledged that not engaging the community sooner was a mistake, I've asked you about the board's role in that, as the body to which she is accountable. You have not responded. Jytdog (talk) 19:26, 24 February 2016 (UTC)

As far as I know, there was never any pressure from the board nor any individual board members to not communicate with the community about long term vision and strategy. Indeed, the sentiment from the board - both community-selected members and other members - has always been broadly to encourage open discussion and disclosure.--Jimbo Wales (talk) 19:33, 26 February 2016 (UTC)

Yes this is what you said before and it is still high level, as it was before. Please be aware that all the information in GorillaWarfare's timeline is there for everyone to see. The lack of transparency, and the criticism given directly to the board over the lack of transparency from many quarters (and not from flakes) is very very clear. Please address the question. Thanks. Jytdog (talk) 23:40, 26 February 2016 (UTC)

Yes, I've read the timeline. What I'm telling you is that as far as I know, there was never any pressure from the board or any individual board members to not communicate with the community about long term vision and strategy. I'm not sure what you think is wrong with that statement. I think I've asked the question - you asked me about Lila acknowledging that not engaging the community sooner being a mistake. You asked me what the board's role in that was. I'm telling you: the board has broadly encouraged open discussion and disclosure, and I'm unaware of anyone individually giving her advice to hide anything about long term strategy about Discovery from the community. How is that not addressing the question - it seems like a very direct answer to me.--Jimbo Wales (talk) 10:06, 27 February 2016 (UTC)

[16] interesting read--Ozzie10aaaa (talk) 13:29, 27 February 2016 (UTC)

it is not answering the question because it leaves a perfect hole where a direct response would be; my sense is you are asking me to read between the lines of "broadly encouraged open discussion and disclosure" and "no one told her not to" and the only things I can put there are things like: "Lila refused to bring this to the community although the board told her to do so, and we didn't make it clear enough that her job depended on her doing that and doing it well" or "we actually didn't know the kind of plans that were being pitched in the WMF's name; and if we had, we would have driven community engagement sooner". Please don't ask me to read between the lines. I am asking you to please tell me where the board was, on overseeing Lila as she made the ongoing mistake of not engaging the community. Jytdog (talk) 17:23, 27 February 2016 (UTC)

The board has broadly encouraged open discussion and disclosure, and I'm unaware of anyone individually giving her advice to hide anything about long term strategy. Going into slightly more depth than that, I didn't see anything particularly unusual or controversial about the concepts being presented to us about the evolving ideas around improving search and discover, and I simply assumed that there was community discussion and consultation about it. The grander concept which, as I now understand, Damon was pitching via cloak-and-dagger PGP encrypted files (one employee told me that he had to give his PGP key on a USB stick because Damon didn't trust the public keyservers), didn't really get traction and was quickly abandoned. By the time of the board meeting in Mexico City, we specifically discussed that this would not be anything like a "Google competitor". As to the exact details of every single discussion with funders, obviously the board is not privy to those as a practical matter. Certainly had we understood that a disconnect was going on, and that the community was not being consulted, we absolutely would have pushed harder for community engagement sooner. As it is, I think most likely other board members, like me, simply assumed that it was being talked about and not treated as some kind of super top secret thing. Is that helpful?--Jimbo Wales (talk) 23:25, 27 February 2016 (UTC)

Thanks for replying. With all of the sturm und drang that is detailed in the Timeline - much of which concerns transparency (which includes engaging the community of course) - and it is all right there, from the funds dissemination committee in May, to James email in October, to Asaf's response to Lila in November, especially the Funds Dissemination Committee note on November 23 ("They state that they are "appalled by the closed way that the WMF has undertaken both strategic and annual planning, and the WMF's approach to budget transparency (or lack thereof)" ..... all the evidence says that Board was well aware that Lila was not engaging the community. I started this in the hope that this could be an authentic engagement and you could move toward regaining the community's trust. You. are. lying. You are displaying the arrogance of power and demonstrating the reality that you and the Board are not accountable to the community. And that is where we are. Jytdog (talk) 03:23, 28 February 2016 (UTC)

Jytdog re: "You. are. lying." You need to take a break. Please come back when you can discuss matters politely. Smallbones_(smalltalk) 05:34, 28 February 2016 (UTC)

Calling a lie a lie, is not rude. It is completely disrespectful to pretend to have a conversation, which is what Jimmy has been doing. This will all come out eventually. I was trying to open the door wide for him to walk through. I am not taking a break. I am done. Jytdog (talk) 08:50, 28 February 2016 (UTC)

Good, Jytdog, you never came here to have a conversation. You came to accuse from your very first mistake laden post. So by your standard: you. are. lying. - Alanscottwalker (talk) 19:33, 28 February 2016 (UTC)

Alanscottwalker, I'll ask that we just leave this alone now. Jytdog says he is done, and I'll enforce that if needed. Smallbones_(smalltalk) 23:01, 28 February 2016 (UTC)

Uhm . . . What? What an odd thing to ping me for. Your last statement is worse than silly, Smallbones. Alanscottwalker (talk) 23:19, 28 February 2016 (UTC)

That is where we are, from my perspective. Jytdog (talk) 19:26, 24 February 2016 (UTC)

I truly hope that Jimbo Wales will not try to wriggle away from addressing in detail the damning sequence of events described so clearly in the timeline written by GorillaWarfare, and thanks to her for taking the time to compile it. There are many things that could be said, but haven't yet been said, though the slanderous "utter fucking bullshit" originated right here on this talk page, from Jimbo himself. The time to apologize and explain is right here right now, but board members say they need to consult with lawyers first, and the lawyers are just too busy. Why not do the right thing, and resign from the board now instead? Cullen³²⁸ Let's discuss it 07:51, 27 February 2016 (UTC)

I fully stand by that comment. I continue to urge the board towards full publication of the details.--Jimbo Wales (talk) 10:01, 27 February 2016 (UTC)

Tons of weasling here. Why, for fuck sake, didn't the board or yourself come up with this on your own account but hide behind such nonsense like urging towards full publication, encouraging open discussion and other non-content sentences, when you could do this stuff all by yourself? You simply ditched the responsibility you have as a board by staying non-open and thus encouraged the secrecy. Come out in the open with everything, nobody needs to be asked, as openness is one of the fundamental principles here, you have to have very good reasons not to be completely open, and up to now not a single one was given. Grüße vom Sänger ♫ ^(talk) 10:40, 28 February 2016 (UTC)

How is "demonstrating the reality that you and the Board are not accountable to the community" "displaying the arrogance of power" or even "lying"? It's just the way the system works. 81.149.218.171 (talk) 15:48, 28 February 2016 (UTC)

Take time to extend your life & telomeres

Jimbo, I am beginning to think this talk-page is likely to shorten your life, so beyond drinking plenty of fluids, perhaps also take more time here to read about health issues, such as:

• "Regenerative medicine" - as with organ transplants from cloned tissues
• "Life extension" - the general topic about living longer
• "telomeres" - the "null padding" after genes to survive DNA truncation

Please take more time, for yourself, to benefit from (re-)reading various acticles here to improve your life. There have been so many scientific advances in the past 15 years, the progress has been amazing, and you can check each article "History" tab to track the updates. As you know, many other users will add replies in other threads on this page, so you never need worry if other people can't get enough key answers. Just sayin'. -Wikid77 (talk) 01:41, 27 February 2016 (UTC)

Thanks, that's very kind of you. Actually, since December 1st I have been on a health kick. I lost 5 lbs in December and 5 lbs in January - February has not been as productive, but I have maintained my weight. I still need to lose a fair amount of weight to get to what is considered medically optimal, but I already feel much better. A big part of that has been watching my food intake, but another big part is a dedication to trying really hard to average 10,000 steps a day - I haven't quite hit the target for February I'm afraid but I'm damned close and hope to get there in March.

Also, in large part because my 50th birthday is coming up, I've had a fair number of health tests since October and I'm happy to report that I'm in decent shape. Not quite in need of organ transplant from cloned tissues just yet. But I'll read the articles you mention anyway - it's an interesting topic.--Jimbo Wales (talk) 10:00, 27 February 2016 (UTC)

It might interest you to know (if you don't) that the "medically optimal" weight might not actually be optimal. If weight is classified using body mass index (the usual way), then statistics show that people in the "overweight" range actually have slightly better life expectancy than people in the "normal" range -- people in the "obese" range have substantially reduced life expectancy, and even more so for people in the "underweight" range. (Citations can be found in Body mass index#Variation in relationship to health.) Anyway it's still a very good idea to be aerobically fit -- fitness is probably a more important predictor of heath than weight, overall. Looie496 (talk) 22:54, 27 February 2016 (UTC)

Thanks. Yes, I've read that sort of thing. I'm currently at 190, down from 200 (actually down from like 208 last summer). Certain charts say I should be 158-159 based on my height - I think that's pretty unlikely! But I think 175 (my first long term goal) is a good starting point, and increasing aerobic fitness is important too. My sister has challenged me to run a half-marathon and so I've been easing into aerobic fitness.--Jimbo Wales (talk) 00:37, 28 February 2016 (UTC)

Some friends of mine have switched to using vegetable soups to stay full, or so-called "negative calorie foods" (celery, apples) which burn calories to chew, but that can be difficult when dining out. I've heard waiters in the U.S. will serve "half-portion" meals when asked, otherwise dining out can derail the meal plan. -Wikid77 (talk) 11:11, 28 February 2016 (UTC)

Jimbo, according to the London Cycling Guide, on the last Friday of each month (so you just missed one) there is a "Critical Mass" ride around town. It starts off from the National Film Theatre beside Waterloo Bridge and sounds like a swell way to keep fit. All the best, 81.149.218.171 (talk) 15:54, 28 February 2016 (UTC)

• I mentioned Critical Mass to Jimbo the other day, when he was telling us that he used Boris bikes. The London ride last Friday was really quite cold but I stuck with it for about two hours. The route is unpredictable and this time included Paternoster Square, the Millenium Bridge, the South Bank and Oxford Circus. For a taste, here's the last Halloween ride. Andrew D. (talk) 20:45, 28 February 2016 (UTC)

Interesting video. Did they ride on Christmas Day? 86.155.149.208 (talk) 13:41, 29 February 2016 (UTC)

Genius: Wikipedia's competitor?

To continue from a comment I made above under "WMF mission", here is an example of the sort of thing that Wikipedia could have been doing, which we've instead left to a "web 2.0" type private competitor. According to a recent article, ProPublica is trying to track down records about what 700 U.S. Navy ships were doing in Vietnam, primarily to assist soldiers exposed to Agent Orange to pursue medical claims. Dating back to Wikipedia:September 11 victims and the WP:NOT restrictions, Wikipedia doesn't provide a place for this kind of data collection, and so the people at ProPublica have turned to Genius.com to track the data via annotations on their article page (using various scripts such as at genius.codes). Despite being featured on Google News, it doesn't look like they are getting a lot of response - it's a pretty arcane thing to request help for, especially in a transient news article. Still, it seems like a wake-up call, because as you can see at the genius.com site, the company is trying to get crowdsourced annotations on historical documents in a way that competes directly with WMF projects.

My feeling is that Genius is deeply handicapped by programmer sensibilities - it looks like what Wikipedia might look like if the people who designed Flow had had a chance to write and carry out a complete wish list. Just try reading "Heart of Darkness Part I" scrolling up and down one little column of text riddled with white space. More to the point, the text is filled with contiguous blocks of text that seem to be claimed by the first person to post, apparently, who puts up the wrong kind of picture for the ship for example, and then you read the most-upvoted response for why he's wrong. All this depends on a huge amount of slow scripting from aites all over the web, including one I remember mostly for the time it was hacked and readers of major American newspapers were being sent to the Syrian Electronic Army.^[17] It is not exactly a beguiling site, but the rule of crowdsourced projects is that whoever actually does something is the best person for the job.

I'll indulge a moment to gripe about the ill fate of my attempt to do a similar annotation project for Charter 08 at Wikisource (s:Charter 08). I simply linked various terms to what articles I could find about the relevant Chinese history terms, nothing fancy, but I thought it was a good way to add value to something that to me was not very easy to understand. Well, you'll see it's a dead link now because someone over there decided that the charter - which people went to prison to disseminate to the world - was not "public domain", even though the translation I cited referenced itself as GFDL. There are people who say that deletionists are not a conspiracy seeking to destroy Wikipedia, but they would be wrong. Technically, we have a vastly better medium to do this kind of annotation, but the question is, can we harden our resolve to use it? Wnt (talk) 19:24, 27 February 2016 (UTC)

With regard to Charter 08, the deletion log states: "Note that translation has GFDL release, if/when the orginal comes back into PD, the translation can be used." The status of the translation is irrelevant if the original is copyrighted and (presumably, unless you can show us evidence) unlicensed, because it is a derivative work. Are you advocating that Wikimedia projects should host more unlicensed copyrighted material? (This is a more general question.) That would destroy us, because we are about free knowledge, which is why we use free licenses. BethNaught (talk) 19:39, 27 February 2016 (UTC) And before you ask, yes, I do think fair use material is overused on enwiki. BethNaught (talk) 19:37, 27 February 2016 (UTC)

@BethNaught: It's a request for more information than I should need to have. When you find a respectable website that has a copy or adaptation of a document and says the copy has a specific free license, you should be able to take that at its word. It's not feasible to investigate and second-guess further. For example, Wikimedia Commons will host content found free-licensed on Flickr, without trying to track down each and every person who put the photo on Flickr and verify where they got it etc. Especially when it is so implausible that its distribution would be restricted in any way, and many people - such as Marco Rubio, currently a leading candidate for U.S. president - currently host copies of this document.^[18] — Preceding unsigned comment added by Wnt (talk • contribs) —Preceding undated comment added 22:51, 27 February 2016 (UTC)

I didn't think that anybody was blaming you. But if it has emerged that the source document is non-free, no amount of reasonable assumptions can change the truth. If someone uploaded a Flickr file to Commons in good faith, and it was later found to be a copyvio, nobody should blame them (I can't say if this happens in practice) but it should still be deleted on copyright grounds. As for other people hosting it, I do not see how that is relevant for our purposes. BethNaught (talk) 22:56, 27 February 2016 (UTC)

This is kind of a tail wagging the dog here, but to go into it a little further, AFAIK nobody ever provided any reason to think the original was not free-licensed. It was discussed here - I think (can't see myself) that after I uploaded the translation with its license, someone added a tag saying it was public domain in China because it was censored - then China changed its law in 2010 to say that censored works didn't lose copyright - then in 2012 someone noticed and deleted the whole thing because under URAA it might not be PD in the U.S. even if it remained PD in China because there was no retroactive change there. If that seems a little crazy, well, the impression I have is that Wikisource will do just about anything to get rid of manifestos, Dylann Storm Roof, that shooter from Finland and so forth - you should see the veritable river of nonsense they tapped into to suppress Ted Kaczinski's manifesto based on a claim that a government auction of stuff in his shack years afterward retroactively took it out of the public domain. There's really no ideological or practical distinction in this kind of thing; anyone who decides to censor the Unabomber's manifesto is bound to censor for the government of China, once they've had a few years to work it through in their head. Just ask the leading Internet corporations... Wnt (talk) 02:56, 28 February 2016 (UTC)

Wikipedia censorship

Any mention of censorship, as in Wnt's last 2 sentences, grabs my attention, because, I was brought up in a maddenly stupid systemically censored society evidenced by when when I let an old black maid sit in the front seat of my parents' car when taking her home, my mom was called by several friends who saw us driving through town and one warned my mom I might be a "niggar lover"; a name I was also called by friends a few times when a child for exhibiting just casual friendliness towards a black boy who had the misfortune to live near our neighborhood. Jimbo, may have a slight knowledge of what I'm talking about, albeit, hopefully not by the time he was a child. I know, the word is racism, not censorship, but it sure felt like censorship, telling me what I can say and do, and so I can argue that one of the perhaps lesser effects of systemic racism is a type of censorship in what people can say and write.

To the point, I was trying to educate myself about the points Wnt and BethNaught bring up here so I went to our serch bar and typed in "wikipedia censorship" which sent me to [19] a subtopic in child porn and sex stuff. So, I will chime in on the discussion about public domain, copyright, licensing etc. to give my opinion that somebody has to stand up against the Chinese Government pricks and if not us, who? Nocturnalnow (talk) 16:10, 28 February 2016 (UTC)

What an interesting story. May I ask where you grew up? And you are right that by the time I was a child, much of that sort of overt talk had passed, but you did still hear a bit of it around here and there, and of course in many cases it was simply replaced by less overt forms. One topic I'm interested in is how societies can quickly flip from "you aren't allowed to say this is wrong" to "you aren't allowed to say this is right", much faster than the actual practices change.

And yes, I think it is important that we be uncompromising worldwide on government censorship, and China is the prime example.--Jimbo Wales (talk) 16:58, 28 February 2016 (UTC)

Macon, Georgia. Nocturnalnow (talk) 00:17, 29 February 2016 (UTC)

Hey I travelled through western Russia for 2 weeks; saw thousands of people but one black person; almost got arrested for jaywalking in middle of block rather than at crosswalk, but my Russian tourguide, she claimed my passport was locked at hotel safe, and the policeman did not search for our passports; instead I paid him a small fine in rubles. -Wikid77 (talk) 00:11, 29 February 2016 (UTC)

Important

You have email from me that you should read. — Ched :  ?  12:40, 28 February 2016 (UTC)

Problems with Selecting Correct Link due to Shifting Content

Please excuse the section title but I'm not sure of the technical term for this and the pun may help attract attention.

I just experienced a common aggravation with the Wikipedia browser interface. What happens is that I loaded a Wikipedia page into a browser. The page happened to be a stub but it can happen with any page and is more likely with long pages. As the page started loading, I spotted a link that I wanted to click on and so stabbed at it with my mouse pointer. But, as the page completed loading, the alignment of the visible portion changed, jerking the position of the link to a different place on the screen and so I actually clicked on a different link. I then had to back up and do over. This happens frequently and it's quite annoying.

My impression is that happens because there's a layer of secondary content presented by javascript or templates and this is only parsed and formatted after the primary page contentshifts as would include banners, menu options and other material which appears at the top of the page display.

My device or browser may be a factor. In this case, it was a Chromebook which has a slow processor and so takes time to digest a complex page. I also observe similar behaviour on my tablet which likewise has a comparatively slow processor and so takes time to finish loading a page.

The WMF developers in SF may not notice such effects if they are using powerful workstations with high-speed internet connections. When Jimbo is over there, please can he find out what is done to check that Wikipedia works smoothly on a variety of common devices – both fast and slow.

Andrew D. (talk) 13:08, 28 February 2016 (UTC)

Reverted the edit on this. Content is useful feedback & title is 'roll the dice' on that feedback. No need for an over-zealous delete. AnonNep (talk) 13:23, 28 February 2016 (UTC)

Happens to me frequently. But usually on Watchlist, where I end up reverting or rolling back useful edits (obviously with no edit summary) or even thanking unknown editors for vandalism (even though it may be really good vandalism, of course) Martinevans123 (talk) 13:31, 28 February 2016 (UTC)

I'm fascinated. What is "really good vandalism"? 81.149.218.171 (talk) 15:41, 28 February 2016 (UTC)

You know, the sort of stuff that one day might be worth a fortune, if you live in Bristol. Martinevans123 (talk) 17:17, 28 February 2016 (UTC)

• Screen shifted by Javascript: See wp:custom to re-customize the screen. The jerk-around screen also happens in some categories with long blurbs, and a click-on a page could trigger a nearby page instead. You might learn to "pre-wrong" click the opposite wrong word so it will become the correct click once the screen shifts afterward, but you need to pace your responses to the speed of the current device. -Wikid77 (talk) 16:41, 28 February 2016 (UTC)
• I notice it on the tabs across the top of the page, because I have options selected to reformat them - they show up one way first, and then shift around a little, and I end up clicking the wrong tab and doing things like watchlisting instead of viewing history, etc. I'm pretty sure it's nothing to do with your computer power or internet bandwidth - I'm using a reasonably powerful Mac Pro and 100mbps broadband. Boing! said Zebedee (talk) 16:55, 28 February 2016 (UTC)

Indeed. On one occasion Banksy painted a picture on a wall and council workmen, unknowing, painted it out. The council chided their employees, who confessed that they had thought it was just vandalism. Our local council is not enlightened - a Banksy drawing on a wall was eliminated soon after it appeared. In the same street, the famous Dalston peace mural (a work by local artists) got the same treatment. 86.155.149.208 (talk) 13:36, 29 February 2016 (UTC)