Wikipedia:Village pump (policy)

The policy section of the village pump is used to discuss proposed policies and guidelines and changes to existing policies and guidelines.
If you want to propose something new that is not a policy or guideline, use Village pump (proposals).
If you have a question about how to apply an existing policy or guideline, try one of the many Wikipedia:Noticeboards.
This is not the place to resolve disputes over how a policy should be implemented. Please see Wikipedia:Dispute resolution for how to proceed in such cases.

Please see this FAQ page for a list of frequently rejected or ignored proposals.

« Archives, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192

The criteria of WP:NSPORT here are too inclusive

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

When closing a discussion Closers are cautioned to: to judge the consensus of the community, after discarding irrelevant arguments: those that flatly contradict established policy, those based on personal opinion only, those that are logically fallacious, and those that show no understanding of the matter of issue. There are a number of statements about inclusion of sports figures that essentially come down to either WP:ILIKEIT or WP:IDONTLIKEIT, none of which are especially useful in evaluating consensus. These statements were partially or fully discounted in our evaluation.

This RFC discussed two related themes. The first theme was whether the subject notability guidelines given at WP:NSPORT are being used and interpreted correctly at AFD. The second theme was whether some of the subguidelines within WP:NSPORT are too inclusive.

The first theme developed a strong consensus that the GNG is the controlling guideline, while the criteria at NSPORT are useful tools to try to quickly determine the likelihood of an article meeting the GNG.

The second theme was the question of whether some of the NSPORTS criteria are too inclusive. Significant discussion only occurred for the cycling criteria. No consensus emerged on any specific adjustment to the cycling criteria, and no other criteria were discussed in sufficient depth for anything resembling a consensus to emerge.

The conclusions that can be drawn at this point are:

There is clear consensus that no subject-specific notability guideline, including Notability (sports) is a replacement for or supercedes the General Notability Guideline. Arguments must be more refined than simply citing compliance with a subguideline of WP:NSPORTS in the context of an Articles for Deletion discussion.
There is a general unhappiness with the permissiveness of some NSPORTS criteria, but no consensus in this discussion on any specific changes to any of them.
There is rough consensus that sources on older athletes are concentrated in print media. Because it is impossible to prove the negative that the sources do not exist to support an article, some intermediate standard is required for determining when an article on these athletes should be deleted due to lack of notability.

Due to the tangential discussion here on some points that have been raised below, there are issues that remain unsettled by this discussion. Follow-up discussion should be held to determine (in no particular order):

The appropriate result for an article at AFD about a sportsperson where they verifiably meet a NSPORTS criterion, but various levels of effort have not yielded significant sourcing (especially for older athletes or athletes from non-Anglophone locations)
The possible cut-offs for "older" or "non-Anglophone" in connection with the above question
Proposed specific changes for the criteria in each sports-specific subguideline

As with the RfC on secondary school notability, this should not be an invitation to flood AfD with indiscriminate or excessive nominations. (non-admin closure)

Eggishorn (talk) (contrib) 19:26, 6 June 2017 (UTC)[reply]

Tazerdadog (talk) 23:25, 6 June 2017 (UTC)[reply]

Why is a substub like Magdalena Zamolska encyclopedic? Because WP:NCYCLING. I don't agree that participating in some competitions, without winning, or generating any coverage but a mention in the list of participants should suffice. Who decided those and similar exceptions to WP:BIO? Current research shows that 30-40% of biographies in Wikiepdia are of sportspeople. Most of them are substubs like this, people who did nothing except participate in events. Participation should not be sufficient to be in an encyclopedia; we are turning into a sport stat site. I am hoping the discussion here would be more representative of our community's thoughts than one that could happen at the sports notability talk page (where I'll of course leave a ping). --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 07:55, 18 April 2017 (UTC)[reply]

You didn't notify the cycling project about this (which I have done for you). The answer is that this individual represented her nation at an event that's the highest level for that sport. It's a long established consensus for sporting individuals that it'll pass the threshold for inclusion, along with single appearances at the Olympics, in cricket matches, football matches, baseball games, etc, etc. There's a lot of areas on WP that you could say are "too inclusive". Such as all populated places. How is this stub any better? Lugnuts ^{Fire Walk with Me} 08:06, 18 April 2017 (UTC)[reply]

Leaving aside WP:OTHERSTUFFEXISTS (you are welcome to start your own discussion about possible problem with WP:NCGN), the problem with claiming we have consensus is that it is difficult to prove (citation needed). And when traced, we often find it was a discussion of five users... Anyway, I see some merit in saying that a sportsperson who "represented her nation at an event that's the highest level for that sport" is notable, but speaking as a someone who knows very little about sports - wouldn't that be Cycling at the Summer Olympics? Cycling, to stick to a single example (that is however representative of the problem with NSPORT) lists, in addition to Olympics, five more events: UCI World Championships, UCI World Cups, Grand Tour (cycling), UCI World Tour and Classic_cycle_races#The_.27Monuments.27, for most of those they are in fact several competitions, most of which occur each year. I repeat: it sands to logic to say that competitors at a single top level event are notable. When we get to a dozen of two top level events, we have a problem (of inclusionist creep). --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 11:11, 18 April 2017 (UTC)[reply]

The individual in question did compete in the 2008 women's road race, the very top level race for cyclists, organized by the UCI (the cycling governing body). The stub also addresses WP:BIAS, in this case a female athlete from Poland. Lugnuts ^{Fire Walk with Me} 11:40, 18 April 2017 (UTC)[reply]

And to answer your first question - encyclopedic means comprehensive, which is a different question to the one you're really asking (I assume about notability). Lugnuts ^{Fire Walk with Me} 14:17, 18 April 2017 (UTC)[reply]

@Lugnuts: Who decided that 2008 UCI Road World Championships – Women's road race is "the very top level race for cyclists"? Again, I don't understand the criteria here. Olympics, yes, but what else and why? You are ignoring most of my arguments - I asked why do we have the 20 or so top level races? Bias is irrelevant, we don't create article about women or other minorities because they are underrepresented, we create them if they are important. Oh, and the subject seems to have placed 89 out of 91 in said competition. I could understand the top 3 medal winners could be notable. Anyone else - I still don't see why. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 04:41, 19 April 2017 (UTC)[reply]

The community decided. Just like everything else around here. UCI events are the top level for the sport of cycling. You've said yourself that you are "someone who knows very little about sports", but you can find out more here. Lugnuts ^{Fire Walk with Me} 06:42, 19 April 2017 (UTC)[reply]

Agree, and not just for cycling. All of sports suffers from this junk. Many of these contentless microstubs are so meager that they'd barely fill out a few cells in an excel spreadsheet. The "sourcing" tends to come from websites that indiscriminately aggregate statistics, often with a dubious record of accuracy. Not a word of prose to be found anywhere. In fact, many of these so-called articles rest on sources so poor that it's hard to distinguish two similarly-named people from each other or to determine whether two stat entries with similar names are actually the same person. Of course, these terrible articles get inflated to large sizes with endless infoboxes and templates to mask the lack of content. Wikipedia readers would be better served if this extremely useless clutter was trimmed, or merged into list articles. But no, it's super duper mega important to erect a shiny little shrine to every farmer who once hit a cricket ball around a church backyard in 1834. Reyk _YO! 08:15, 18 April 2017 (UTC)[reply]

Said the person who created this. And this. Chuckle. Lugnuts ^{Fire Walk with Me} 08:19, 18 April 2017 (UTC)[reply]

Off-topic baiting and snapping collapsed
The following discussion has been closed. Please do not modify it.
I know better now. Fuckwit. Reyk _YO! 08:24, 18 April 2017 (UTC)[reply] I strongly recommend you apologise and retract that last comment. Lugnuts ^{Fire Walk with Me} 08:25, 18 April 2017 (UTC)[reply] No and no. Reyk _YO! 08:27, 18 April 2017 (UTC)[reply] Hi, got here from Special:RecentChanges. Please, not in front of the children.MikeTango (talk) 08:32, 18 April 2017 (UTC)[reply]

Yes, I wrote a handful of articles that don't meet the standards I now consider important. So what? It dawned on me what gibberish I was writing back then, so I stopped. Same reason I don't participate at DYK anymore; it was encouraging me to write trash in exchange for baubles. This tactic of digging through ancient edits hunting for a "gotcha" is neither helpful nor particularly honest. And if in future you want to try that again in hopes of getting me to lose my temper, save your energy. I know what the game is now, and it won't work a second time. Reyk _YO! 09:09, 18 April 2017 (UTC)[reply]

Everyone made mistakes in their early wiki career, I also created a lot of entries on non-notable fictional topic that I am not often nominating for deletion, because geek fancruft is not notable. If anything, this only drives home the point that sport geeks need to raise the threshold; currently it is really hard to be a non-notable sportsperson on Wikipedia. We should not be a catalog of sport stats. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 11:14, 18 April 2017 (UTC)[reply]

What makes you so sure the early stubs created were "mistakes" rather than the deletionist current perspective being espoused? Carrite (talk) 03:59, 30 April 2017 (UTC)[reply]

Because I am saying it was a mistake to create that kind of article in that state. Reyk _YO! 22:57, 13 May 2017 (UTC)[reply]

Agree. "The answer is that this individual represented her nation at an event that's the highest level for that sport. It's a long established consensus for sporting individuals that it'll pass the threshold for inclusion" simply isn't true. Many sports are so minor that even the world championships themselves don't get an article here (or much media attention), never mind every athlete appearing in them being notable enough for an enwiki article. For example, Inline speed skating. I have created articles on a few inline speed skaters, I'm not biased against the sport in any way, but most competitors at the world championships are not notable. And then we have sports where NSPORTS explicitly sets the bar lower than competing at world championships. This is in itself defensible, e.g. everyone who plays in the Premier League is notable, fine. But this has been taken to extremes in some cases, like cricket, or cycling (especially women's cycling, where the bar for inclusion is lower even though the sport gets less media attention now and used to get massively less media attention in the past), or soccer (playing for a few minutes in an official game with the national team of a micronation is sufficient to be notable according to WP:NFOOTY). Often the mistake has been made that because people in sport X here and now are notable, this should be included in the guideline for then and there as well. This ignores the massive changes in populatity / notability one can have per sport over periods, countries, and gender. Even for the Olympics, one can reasonably argue that competing in the first games (up to at least 1920) often was not important and has in some cases not even been registered. Fram (talk) 12:01, 18 April 2017 (UTC)[reply]
- @Fram: Wow. You know, I totally missed the FAQ. You are right, NSPORT clearly states that the subject still has to meet the GNG. But in my experience, the problem is that AfDs are often closed as keep solely based on NSPORT, with arguments that the subject doesn't meet GNG falling on deaf ears. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:15, 19 April 2017 (UTC)[reply]
  - Closers of AfD discussions for sports figures need to be reminded that if they ignore the context of the sports-specific guidelines, which is clearly laid out in the lead section and the FAQ, they are ignoring the community consensus that formed the guidelines, and the multiple subsequent reaffirmations on the talk page of the guidelines. isaacl (talk) 05:59, 19 April 2017 (UTC)[reply]

Comment Part of the problem is that many of the sport-specific guidelines were created by editors from sport-specific wikiprojects; and while not everybody is inclusionist about their favorite sport, very few editors are deletionist about their favorite sport. So instead of a nice balance between inclusionists and deletionists, there's just inclusionists and moderates, which tends to lead to an inclusionist bias.
That said, many of the sports stubs (and substubs) are about genuinely notable people (as in, people who meet GNG); given that the number of articles about sportspeople is so huge, it's hardly surprising that sports editors simply haven't found the time to flesh all of them out; and NSPORT guidelines, as they're intended to do, discourage AfDs in cases like that.
Part of the problem, too, is that many editors at AfD apply NSPORT overenthusiastically; as in "keep - meets NSPORT" without any thought as to whether the athlete actually meets GNG. "Keep - meets NSPORT" can be a good and valid argument, but AfD !voters often forget that GNG must also be met; and if the case is made that an athlete meets NSPORT but fails GNG, "keep - meets NSPORT" is too simplistic to be of any value as a !vote or a reply. Sideways713 (talk) 12:41, 18 April 2017 (UTC)[reply]

As a further note: the OP asked "Who decided those and similar exceptions to WP:BIO?" NSPORT doesn't create any exceptions to WP:BIO or WP:GNG; on the contrary, it explicitly notes (in the FAQ) that all sports biographies must meet WP:GNG to be kept. Sportspeople do get a freebie compared to non-sportspeople in that WP:BIO1E is practically never applied to a sportsperson, but that's not codified in NSPORT. Sideways713 (talk) 14:58, 18 April 2017 (UTC)[reply]

Disagree. I think there is a problem with there being too many athlete articles as a proportion of all biographical articles on Wikipedia, for sure. But the problem is not that we're allowing too many articles on athletes, but rather we don't have enough articles on non-athletes. -- Earl Andrew - talk 13:26, 18 April 2017 (UTC)[reply]

It would be great if we could include more articles on other professions (academics for example are ones perennially under-represented) but then the issue is a matter of WP:V; it is very difficult to support articles on these other professions, whereas even for minor sports, websites with statistics exist allowing individual players to at least meet WP:V. --MASEM (t) 13:39, 18 April 2017 (UTC)[reply]

Agree The ONLY principle for deciding if an article should exist is "could I fill it with enough well-referenced text to make it worthwhile". Any criteria that encourages the creation of articles which will never have enough text is simply ill-advised. --Jayron 32 13:41, 18 April 2017 (UTC)[reply]

Which is a much bigger project-wide issue that's not just related to sportspeople. Lugnuts ^{Fire Walk with Me} 13:52, 18 April 2017 (UTC)[reply]

How do you eat an elephant? One bite at a time... --Jayron 32 17:50, 18 April 2017 (UTC)[reply]

Indeed. Let's not derail the discussion by saying "there are other problematic areas". This one is a very good place to start. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:13, 19 April 2017 (UTC)[reply]

Note that NSPORT is, in theory, supposed to simply be a quick way to judge if it's likely that an article can be filled with enough well-referenced text to make it worthwhile; it's not supposed to create any backdoors whereby athletes who fail that basic criterion can be declared "notable", though admittedly that does sometimes happen in practice. Sideways713 (talk) 14:02, 18 April 2017 (UTC)[reply]

Intent means nothing, all that means anything is what actually happens. Every notability guideline beyond WP:42 is an invitation to create articles which will never be developed. --Jayron 32 17:51, 18 April 2017 (UTC)[reply]

And how do you decide what is "enough text"? You are proposing that we introduce subjectivity instead of objectivity. "Enough text" could be 95% waffle and duplication of info from other articles. NSPORT generally uses objective criteria such as "must have played at the highest level of domestic competition". Any stub can be enlarged but would the additional text be useful? Jack | ^{talk page} 16:18, 18 April 2017 (UTC)[reply]

WP:42 is a good start. WP:GNG, the expanded version, is pretty solid as well. Not sure we need more than that. But yeah, if an article can never be developed, it should not exist in the first place. If all we have to say about a person is that they lived, why even have an article? If we can't write about their childhood, their careers in some sense of completeness, analysis of their lives by others, etc. then what are we putting in an article? In general, the text of a Wikipedia article should be less than the text which exists in the rest of the world outside of Wikipedia, because Wikipedia should be a summary. When Wikipedia articles are longer than reliable source material about a person, and consist solely of three lines of text and a table, that's a problem. --Jayron 32 17:50, 18 April 2017 (UTC)[reply]

I've always thought a good rule of thumb is whether or not the article shows a reasonable potential of being developed into a Featured Article. Even a lot of perennial stubs show that potential. The article that sparked this thread might have that potential, if and only if someone is willing to do the research to show how that athlete fits into the larger context of the history of cycling in Poland. (And if the material to be researched doesn't exist, well, merging or deletion are the only viable alternatives. -- llywrch (talk) 19:39, 19 April 2017 (UTC)[reply]

Potential article quality has little relation to suitability for inclusion in an encyclopedia. If we're covering someone as a sportsperson, the only essential is that we have sufficient sourced content about their significance as a sportsperson --Michig (talk) 19:51, 19 April 2017 (UTC)[reply]

Your answer shows you haven't thought much about the issue. I have, because it's not unique to sports biographies. There are many articles -- both biographical & non-biographical -- currently existing in Wikipedia that will never be expanded beyond a stub because there isn't any further information about the subject. Yet one can argue the subject is notable.

So what to do? Avoid creating it in the first place? Delete them because they will never be developed beyond a stub? Or merge them into a list, much as tv episodes or fictional characters are? (You could call that the "Pokemon solution", in memory of the first time this issue came up.) Or perhaps there is another solution that this old Wikipedian is blind to. But a solution needs to be found because the issue of permanent stubs in Wikipedia is one that will only become more apparent as time goes on. -- llywrch (talk) 15:56, 20 April 2017 (UTC)[reply]

"sufficient sourced content about their significance as a sportsperson" is merely the standard we need to mention the person at Wikipedia as a whole. The issue is more about how to include text about their lives at Wikipedia: As part of a larger page or as a stand alone page. We aren't saying "Should this person be removed from Wikipedia entirely". What we are saying is "Is there enough text about this person's life to write to justify putting it on a separate page" If all we have to say is "Jane Smith competed in fencing at the 2012 Olympics" and we literally know nothing else about them, we can note that fact just fine in an article titled "Fencing at the 2012 Olympics". There's no need to ALSO create another page whose sole purpose is to contain that line of text. Jane Smith is at Wikipedia, all the information we have is already in the encyclopedia, and unless and until we can justify creating an entire page to house her biography (rather than just one or two facts about her participation in a sporting event) then the extra page isn't needed. No one is saying to remove the information from Wikipedia, just that creating an entire new page to house that information is not worthwhile. --Jayron 32 16:09, 20 April 2017 (UTC)[reply]

"Your answer shows you haven't thought much about the issue" - could you possibly get any more arrogant? I don't get your obsession with articles having to go beyond stubs - have you ever looked at the size of many of the articles in other works such as Encyclopaedia Britannica? I have no objection to merging really short articles to wider topics, but that's an editorial decision on how we organize content and a separate issue to notability. --Michig (talk) 17:38, 20 April 2017 (UTC)[reply]
I've never said such a thing, and I would thank you for not accusing me of doing so. --Jayron 32 01:51, 21 April 2017 (UTC)[reply]
Agree As a compromise, can we just have a guideline that makes such entries as the example in the OP presumed ok for a list but not for an article? Alanscottwalker (talk) 14:17, 18 April 2017 (UTC)[reply]
Comment: Unfortunately for the majority here, you're writing for an encyclopedia where the fundamental underpinning of notability criteria isn't "We think your profession is important," but "The world has heard of you." Of course there will be many more athletes and entertainers than academics represented here, because our culture pays attention to athletes and entertainers, and a third-line forward for the Boston Bruins or a minor actress who gets her kit off for three scenes in Game of Thrones has vastly more name recognition than your average Nobel Prize winner. I'm just curious as to what some of you think should be done about that aside from jettisoning WP:V. Ravenswing 15:03, 18 April 2017 (UTC)[reply]
- Subject-specific notability guidelines can be more restrictive than the GNG (They can't be less restrictive though), if that helps to prevent excessive coverage in a topic area where imbalanced sourcing can exist. But the case argued here is different: it is that criteria in NSPORT have been selected that allow for the presumption of notability based on weak criteria that can be easily met with WP:V (like the OP substub) but show no strong indication of eventually meeting GNG. Just because an athlete can be documented doesn't mean they are necessarily notable. --MASEM (t) 15:14, 18 April 2017 (UTC)[reply]
- While I agree with the rough sentiment, and I would not suggest any quotas on sportspeople, nor would I disagree that they are a highly visible type of a profession with high notability compared to many others, we have to draw a line a bit higher; consider the example I linked in the OP: that cyclist has very little world recognition, and I doubt her name is recognized by anyone except die-hard cyclist fans in Poland. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:20, 19 April 2017 (UTC)[reply]
Agree that its too inclusive, not sure what you can do at this point. I think you would probably get a majority of people in a site wide RfC to agree on that, but I don't think you could get agreement on what the content specific guideline should be. On a practical level, any drastic change would probably be more messy than even the schools issue in terms of flooding AfD. I think for any changes you'd need to have a very carefully crafted RfC. TonyBallioni (talk) 15:24, 18 April 2017 (UTC)[reply]
- I would definitely agree that an RFC regarding this must pre-emptively set up some type of grandfathering approach to existing articles, and a means to address fait accompli at AFD. We learned from the SCHOOLOUTCOMES RFC that if these aren't set in stone to start, a lot of editors will reject the change due to their fears that we'll flood AFD. Preemptive processes for this will help alleviate those matters. --MASEM (t) 15:50, 18 April 2017 (UTC)[reply]
  - Yes. In my mind the proposed grandfather clause would need to be set as the date the RfC would be opened to prevent gaming the system, which would happen just as surely as a flood on AfD. Sports aren't my particular area of interest or expertise, but if someone decides to set up a draft RfC, feel free to ping me. I was involved with setting up the schools RfC, so I have thoughts on both the positives and negatives that came from that which might be helpful. TonyBallioni (talk) 15:54, 18 April 2017 (UTC)[reply]
Disagree. On the whole, WP:NSPORT is fine, but fine tuning is always possible and sometimes desirable. Having read the arguments raised by User:Piotrus, his issue is not about NSPORT but specifically about WP:NCYCLING and he needs to raise it with the cycling project. While he's there, he might ask them to stop obstructing the traffic on public roads. Jack | ^{talk page} 15:49, 18 April 2017 (UTC)[reply]
- No, you are wrong. My issue ws with NSPORT, I just used NCYCLING as an example. Half if not more of the NSPORT guidelines are super-inclusive. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:27, 19 April 2017 (UTC)[reply]
  - @Piotrus: Perhaps, but not WP:CRIN of which I am the main author (it is summarised by WP:NCRIC within NSPORT). CRIN is actually exclusive because we insist that all cricketers must have performed at the highest domestic level to comply with WP:Notability and the article must clearly demonstrate, with verification, that the player has done so. For the most part, it means he or she has played for a team that is ranked "first-class". At WP:CRIC, we take any article that does not comply with CRIN straight to AfD. I don't know about the other sports projects but CRIC has set a high standard for inclusion and we strictly apply it. We have even taken cricket articles to AfD because there was no indication of first-class standard, even though subsequent investigation revealed that the subject did comply with CRIN but the author had been incompetent by not stating and verifying the notability. Maybe what is needed is for other sports projects to follow our example. Jack | ^{talk page} 18:14, 19 April 2017 (UTC)[reply]
Comment - I tend to agree that NSPORTS is far too inclusive. At the top of NSPORTS there is a FAQ. The first item in the FAQ refers to the GNG. The answer says:

The topic-specific notability guidelines described on this page do not replace the general notability guideline. They are intended only to stop an article from being quickly deleted when there is very strong reason to believe that significant, independent, non-routine, non-promotional secondary coverage from reliable sources are available, given sufficient time to locate them.[1][2][3][4] Wikipedia's standard for including an article about a given person is not based on whether or not he/she has attained certain achievements, but on whether or not the person has received appropriate coverage in reliable sources, in accordance with the general notability guideline.

My opinion, which I've expressed at several AfD, is that this makes NSPORTS and all of it's sport specific criteria, secondary to the GNG. It helps clarify who might meet the GNG but I tend to think there are cases where a person who's done something specific in the sports related field is unlikely to meet the GNG in itself.

Perhaps that's a way forward? Yes, they've entered this race/played in this match/been part of this competition. So find sources to satisfy the GNG for them then. Blue Square Thing (talk) 16:55, 18 April 2017 (UTC)[reply]

Yes, it's not just your opinion: this is the consensus that was agreed upon when the sports-related notability criteria were initially created, and reaffirmed multiple times since. isaacl (talk) 18:27, 18 April 2017 (UTC)[reply]

And to confirm further, most editors that were involved in NSPORT recognize that it sets a presumption of notability in lieu of meeting the GNG that can be challenged if an editor does the necessary legwork to confirm that the GNG cannot otherwise be met (eg following the steps at WP:BEFORE, particularly scouring print sources for athletes prior to 2000). If you can reasonable show no sourcing exists beyond stats, then it is acceptable to seek AFD for that athlete. Most of the editors involved in NSPORT that I'm aware don't think it overrides the GNG. --MASEM (t) 18:34, 18 April 2017 (UTC)[reply]

Disagree. The guideline is, on the whole, a pragmatic approach that avoids spending time discussing every single article in depth to get consensus each time. So we have lots of stubs - so what? So we have a small proportion of sports articles on people who would likely be considered by most not significant enough to have an encyclopedia article - better than having a debate every time, and we have to consider the difficulty in finding coverage that undoubtedly exists on subjects who were competing more than 20 years ago. Our starting point should be to ask at what level in each sport do competitors become worthy of inclusion in a wide-reaching and comprehensive encyclopedia, not to ask how much content we can find about someone right now, which is all the GNG gives you. In that respect, while not perfect, WP:NSPORT is doing the right thing (although personally I would drop the nonsense about having to satisfy the GNG from it). --Michig (talk) 17:07, 18 April 2017 (UTC)[reply]
The problem is not the existence of stubs. Stubs are fine, so long as they could eventually be expanded with existing source material. The problem is the existence of stubs which can never grow beyond a stub because no source material exists. If we have a person for whom all we can say is "They competed in this one event." and have zero biographical material, well, we're creating an article for nothing. We already have documented their having competed in that one event. We don't have to do it twice. We really only need an article if we can write a biography about them. If we can't write a biography, why have an article? --Jayron 32 17:54, 18 April 2017 (UTC)[reply]

But who's to say that x stub can never be expanded? Lugnuts ^{Fire Walk with Me} 18:08, 18 April 2017 (UTC)[reply]

"You can't prove sources DON'T exist" is a bullshit rationale and you know it. Demanding that people prove the non-existence of sources before allowing an article to be deleted is so stupid, I won't even insult you by explaining why. --Jayron 32 01:26, 19 April 2017 (UTC)[reply]

Typical non-answer and a great way to dodge the question, clearly demonstrating you can't actually answer the question. Lugnuts ^{Fire Walk with Me} 06:45, 19 April 2017 (UTC)[reply]

OK: Here is your answer: You cannot justify anything by demanding that those that oppose you prove the non-existence of proof. That's called Argument from ignorance and is a core informal fallacy because it is ultimately impossible to do. If you demand that unicorns exist because no one has ever produced evidence of a "lack of unicorns" in the world, that's baldly silly: You can't demand that we accept the existence of the unobservable merely because no one has produced a lack of observation of the unobservable. You simply can't demand that the world accept the existence of unicorns on the basis that no one has produced a "lack of unicorns" as a proof. It stretches credibility. In the same way, you can't demand that all articles be allowed to exist forever once created merely because no one has produced proof that sources don't exist. If THAT was the standard for deletion at Wikipedia, we couldn't delete anything because no one can actually produce proof of non-existence. That's why in LIFE (that is, in the entirety of human experience, including outside of Wikipedia) concepts like null hypothesis and Philosophical burden of proof exist. You can't just go around making random assertions and then demand people assume your assertions are true without any positive evidence! Because (and this apparently includes you) some people are unaware of the way this reality of human existence works, Wikipedia has a policies in place like WP:BURDEN which establishes that the burden of proof that content should exist at Wikipedia lies solely with the person who wishes to keep the content at Wikipedia. I had assumed you understood how basic logic and proof of existence of things worked. I will not assume that level of intelligence in you any further. --Jayron 32 01:43, 20 April 2017 (UTC)[reply]

I didn't "demand" anything. I stopped reading your pointless rant at that point. Lugnuts ^{Fire Walk with Me} 08:36, 21 April 2017 (UTC)[reply]

You are seriously trying to pull the n00b arguments of WP:THEREMUSTBESOURCES/WP:ATA#CRYSTAL on us? Really? --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:30, 19 April 2017 (UTC)[reply]

Stubs are absolutely fine if they are properly sourced and tell our readers the key facts about a subject. There is no need for them to be expanded beyond that level to become worthwhile articles. --Michig (talk) 17:52, 19 April 2017 (UTC)[reply]

Well, it is awfully inclusive. But one big advantage is that it sets an objective guideline, and this helps reduce arguments, probably. Like WP:BASEBALL/N... if you had one at-bat in a major league, you are in, period; and if not, you're not (except if you meet WP:BIO on other grounds, which is a different question). There's no back-and-forth of "Well, but he didn't play much... define 'much', where's your cutoff?", with a random decision based on who happens happens to be populating the AfD discussion for that particular page. This is a net positive for the project IMO. And course these articles don't harm anyone. UNLESS... one editor above says "The 'sourcing' tends to come from websites that indiscriminately aggregate statistics, often with a dubious record of accuracy". Well if that's true it's very different. Is it? That's the only question that excites me (For baseball players, the sourcing is usually Baseball-Reference.com, which is comprehensive and reliable. But I don't know about other sports.) Herostratus (talk) 17:43, 18 April 2017 (UTC)[reply]
- The general consensus for discussions involving baseball and hockey players is that sites which are compendiums of stats are considered to be routine coverage, and thus not sources indicative of the player meeting the general notability guideline. (They can be fine sources for the actual stats, as is the case for the Baseball Reference site, but they aren't used to show that English Wikipedia's standards for having an article have been met.) isaacl (talk) 18:23, 18 April 2017 (UTC)[reply]

Really? My understanding is that you have one at-bat in the National League in 1904, you get an article. Are you saying that the "general consensus for discussions involving baseball... players" is that at AfD many major league player articles are being deleted? Yeesh, that's a big change. Herostratus (talk) 19:47, 19 April 2017 (UTC)[reply]

A player with an at bat meets the baseball-specific notability guidelines, but as described in the lead section of WP:NSPORTS, as well as the FAQ, the article can still get deleted for not meeting the general notability guideline. To determine if this guideline is met, a reference in a stats site is not considered to be sufficient. I do believe there are some corresponding examples of baseball player articles that were deleted, and there have been examples of professional European hockey player articles that were deleted where an entry on a stats site was insufficient to keep the article. isaacl (talk) 20:15, 19 April 2017 (UTC)[reply]

A fully inclusive approach via NCRIC and others is the only way to achieve a proper, complete, NPOV encyclopedia. Why is it that not a single person who says "this is wrong" is prepared to offer a viable alternative solution? Let's take NCRIC for an example. One first-class appearance. Want to make it two? Three? Fifty? Fine. Offer that suggestion as an alternative. State NPOV reasons why. But doing so is more POV than any existing guideline.

How easy is it to invent similar "guidelines" for other sports? A single professional appearance? Soccer

, MLB

, NFL

, NHL

, NBA

... already done. No further need for discussion. Any other need will encourage violation of NPOV. This whole discussion is happening about ten years too late. Bobo . 18:36, 18 April 2017 (UTC)[reply]

We could just as easily set it at "played one complete season"

for team sports, and won an international championship

for individuals. Both would give us a much more realistic expectation of actually establishing notability through sources. Blueboar (talk) 18:55, 18 April 2017 (UTC)[reply]

How do you define "playing in one complete season"? What if a player misses one game? Or two? Or ten? Obviously it depends on how many games a season includes (MLB teams play 162 games in their regular season, whereas NFL teams play only 16), but your criteria are still arbitrary. The idea that winning an international competition is the only thing that makes a player notable is laughable; some of the greatest soccer players of all time never won an international competition, and some sports (read American football) don't even have an international competition to speak of, or at least the highest level of their sport is contained within one nation. – Pee Jay 19:58, 18 April 2017 (UTC)[reply]

I was considering that, Peejay. What is a "complete season"? The example I use is Robin van Persie. The season before he signed for Manchester United was the first season in which he played all domestic league games, and therefore the first in which he could be fully judged as a player in the English league. He then played another season before SAF retired, in which he once again played every single league fixture. After that, boom. Anyway, that's just the sports geek in me talking! For "amateur" players, who may have made only a single appearance, we probably don't know what they did for the rest of the season. Did they stay with the team or did they go back to working in their day job in the local pie shop? Bobo . 20:21, 18 April 2017 (UTC)[reply]

It's extremely rare to see MLB players play all 162 games of a season anymore. And what if someone's hurt and misses one game? Or suspended? A player could conceivably play for 20 years, be one of the elites in his sport, and never play every single game in a given season, especially in baseball. That criteria doesn't really work. Smartyllama (talk) 17:51, 28 April 2017 (UTC)[reply]

Disagree For the sports that I'm familiar with (football and rugby league) the guideline is probably about right in terms of identifying which players are definitely notable. If anything, WP:NFOOTY is used more to prevent articles being created rather than to allow them, as it's easy to create a well-sourced article on footballers playing well below the fully-professional leagues and we have a constant stream of articles on semi-pro players being deleted at AfD (there is also a small group of editors who have persistently attempted to relax the guideline further to allow semi-professional non-internationals). There may well be some individual guidelines that are too lax, but let's not throw the baby out with the bathwater here. Number 5 7 20:28, 18 April 2017 (UTC)[reply]
This is true. I remember one of the first things I did when I started trying to be "inclusionist" with footballers was to add articles on footballers for Alloa Athletic, who at the time were playing in the third tier of the Scottish pyramid. I now know the reasons why this was a bit excessive - and I understand these reasons fully, though I didn't at the time and it seemed inconsistent.

That and no two sources appeared to agree on how to spell Michael Bolochoweckyj's name! Bobo . 20:35, 18 April 2017 (UTC)[reply]

Disagree For American sports, the ones I am most familiar with, the guidelines appear appropriate. Anyone playing in even one game in the National Football League, Major League Baseball, National Basketball Association, and National Hockey League are bound to have sources from at least two areas. The first is their college/junior careers. These sports, in the U.S. are well covered and therefore generate a lot of coverage. The second is that the teams are very well covered. Minor roster moves are covered in detail, including the drafts. For non-team sports, the events in question gain a lot of coverage. Boxing title fights, golf majors, the Ironman triathlon, etc. are all significantly covered. Would be nice if other professions got that coverage, but they don't. There is a reason kids, at least in the U.S., collect baseball cards with baseball statistics and not scientist cards that list scientific discoveries. Its because of how well followed baseball is and not the sciences or other areas (see [1]). RonSigPi (talk) 21:33, 18 April 2017 (UTC)[reply]
But if the sources already exist, why do you need the guideline in the first place? Doesn't WP:GNG already allow us to create the article? If the only reason we have the secondary guideline is to repeat exactly what WP:GNG says, then it is redundant. If the secondary guideline allows us to include material which can't be sourced, it's a bad idea. Which is it? --Jayron 32 01:29, 19 April 2017 (UTC)[reply]
The secondary guideline is there to help editors evaluate whether it's likely that an athlete will meet GNG when sources in the article (if any) fail to demonstrate that he does, but sources not in the article (including hard-to-access offline sources and non-English sources, but also English online sources no one's gone to the trouble of adding) might show otherwise. As the OP noted, Wikipedia has countless stubs about sportspeople; some of them can never be expanded and some can, and NSPORT helps editors make a quick educated guess about whether a particular stub can (given enough time and effort to locate the sources) be expanded and shown to meet GNG. If that guess is "yes" more often than it should in a particular sport, then the sport-specific guidelines for that sport are too lenient; which is almost certainly the case for many sports, but the existence of thousands of poorly sourced stubs doesn't in itself prove that, and is indeed almost inevitable considering the vast number of articles to be written and the limited number of dedicated sports editors.
The guess is always just a guess, and can be challenged at AfD if you think the real answer is "no" - though you'll need to be strong of heart to try that with a cricket article... Sideways713 (talk) 10:27, 19 April 2017 (UTC)[reply]
People keep saying that as though if by mere repetition, they would make it true. You may WISH the secondary guidelines exist to "help editors evaluate whether it's likely", but in the decade or more history of Wikipedia using such guidelines, we have ample evidence of their real use to Wikipedians: To cite in deletion discussions to stiffle discussion of source text and retain otherwise unexpandable articles. That's the raison d'etre of all secondary notability criteria: to provide a loop-hole to allow articles to exist which could otherwise never contain any substantive, referenced prose. You can keep asserting that they only exist to "help editors evaluate" but the bulk of evidence from actual use of these guidelines is quite not that. --Jayron 32 01:49, 20 April 2017 (UTC)[reply]
There's no need to take such an adversarial tone, as if I or other NSPORT editors were in disagreement about that. On the contrary, many of us are just as frustrated with NSPORT being misused that way as you are; and yes, sometimes it's not just misapplied by AfD editors, but used as a tool by inclusionist sports editors to justify keeping their pet perma-stub. But that's not the only reason NSPORT exists; it has a legitimate purpose, and not all articles it helps keep are unexpandable.
And articles that meet NSPORT do sometimes get deleted in real life for failing to meet GNG; less often nowadays, unfortunately, but it does still happen. This AfD is a recent case. Sideways713 (talk) 09:43, 20 April 2017 (UTC)[reply]
If you're frustrated with the use of the guideline, the solution is to nuke the guideline. Leave people with the GNG to work with. That would solve the problem quite well. --Jayron 32 15:13, 20 April 2017 (UTC)[reply]
I'm frustrated with the current use of the guideline and with shallow "keep - meets NSPORT" !votes; but I'd be even more frustrated if AfDs for track and field athletes who do meet GNG suddenly spiked tenfold and I had to run around looking for sources to prove the keep separately every single time, which is what might happen if NSPORT was nuked. There are thousands of track stubs with minimal sourcing, not because the sources don't exist (true in some cases, but not most) or even because the sources are hard to locate (true in many cases, but not all) but because WikiProject Athletics doesn't have enough manpower. The people we do have have better things to do than run from AfD to AfD, or to be at loggerheads with each other about notability. I'm not entirely happy with NSPORT or the track-specific guideline (which is a mishmash of strict criteria, lenient criteria, strange criteria and some misplaced stuff no one's ever got around to removing); but they give us more time to concentrate on actually improving articles. Sideways713 (talk) 22:42, 20 April 2017 (UTC)[reply]
Agree, and as to the above about what the standards should be, we already have that standard. "If a topic has received significant coverage in reliable sources that are independent of the subject, it is presumed to be suitable for a stand-alone article or list." If an athlete meets that standard, they are notable. If not, they are not. But a brief blurb just mentioning that they happened to appear is not significant coverage. Seraphimblade ^{Talk to me} 21:52, 18 April 2017 (UTC)[reply]
Agree. Seraphimblade says it well, Alanscottwalker offers a possible compromise. When push comes to shove, I can't think of any topic notability guideline that works well and all of those I have seen acknowledge the pre-eminence of GNG. They are at best well-intentioned attempts by aficionados to carve out some sort of standard but they are used in effect to bludgeon through things that would never apparently meet GNG. - Sitush (talk) 23:28, 18 April 2017 (UTC)[reply]

The GNG is one of the worst standards ever adopted by this project. Any other sensible encyclopedia (as WP originally did) would take objective criteria much along the lines of the SNGs as their starting point rather than judging which subjects to include based solely on how much coverage they can find elsewhere. --Michig (talk) 18:00, 19 April 2017 (UTC)[reply]

So, where are you going to find source texts to add information to articles at Wikipedia, if the GNG wasn't there? The GNG only says that source text needs to exist before creating an article. Why is that so wrong? Are you really proposing we do away with that standard, and allow articles to be created on subjects even when no source text exists about that subject? --Jayron 32 15:13, 20 April 2017 (UTC)[reply]

What you have just stated makes no sense. We have the policy WP:V which would prevent articles existing even when no source text exists, whatever guidelines we might have. We can often glean sufficient sourced information for a properly sourced short article from several reliable sources, none of which would be considered to contain 'significant coverage', but GNG demands that the coverage in individual sources is significant. Plenty of unencyclopedic topics have 'significant coverage', and at the end of the day all it gives is the ability to create more detailed articles. --Michig (talk) 17:38, 20 April 2017 (UTC)[reply]

Disagree Like RonSigPi, I haven't had any trouble finding material on Australian athletes. The abundance of articles reflects the abundance of material available. That in turn reflects the degree interest of our reader base. Per Seraphimblade, GNG is our standard, and that means athletes get bios. Our guidelines merely reflect our accumulated experience of GNG, which is wider and more encompassing. An teenage athlete once asked me why a teammate warranted a Wikipedia article. I told her that it was because her teammate was a Paralympic medalist, and hence was notable. She said she thought you had to be a national hero. I replied that a Paralympian is a national hero. Hawkeye7 (talk) 23:57, 18 April 2017 (UTC)[reply]
Comment It's objective standard? No, it is not, it's a made-up guideline, and it thus can be re-made-up. The arguments based on V and on NPOV, are just irrelevant, when what we are talking about is Notability - what we cover in a separate article, a separate biography, in fact - a biography covers birth to death - but what some appear to keep arguing is we need no reliable sources for their biography, just vital statistics and sport statistics, and the made-up guideline. And then there is the argument that there are a ton of reliable sources on these people because they are popular, well fine, but obviously they pass the GNG, then. Come now, and compromise over perpetual stubs (move them to Draft, put them in a list), now, what have you got to lose. -- Alanscottwalker (talk) 00:00, 19 April 2017 (UTC)[reply]
Disagree. There is no reason to change a standard that is working fairly well. We have a bright line of notability that makes sense and is easy to follow. Anything else and you get bogged down in perpetual afd debates about what quantifies as enough coverage. These arguments come up every so often by people who don't think athletes deserve the coverage they get... but thats rubish... more people follow sports than many other professions and will want comprehensive coverage of the major sports. Spanneraol (talk) 00:09, 19 April 2017 (UTC)[reply]

No one here is begrudging whatever coverage they get, but why is that apparently mountainous coverage not in these biographies. Alanscottwalker (talk) 00:16, 19 April 2017 (UTC)[reply]

And they can get that coverage from, in many cases, the single sports ref website that we cite. Simplistically, if we are aggregating information then we are sort of adding value to the reader experience but otherwise we are not. - Sitush (talk) 00:18, 19 April 2017 (UTC)[reply]

So we are saying that people who are looking to research their favorite team should go somewhere else to find information about the players who arent covered here? We should offer complete coverage of major sports leagues as we do. And Yes many of the bios can be better with more information in them... but that isnt any reason to change the guidelines... diligent editors are constantly improving the stub articles to make them better and more informative...deleting the stubs doesnt help them do that. Spanneraol (talk) 00:32, 19 April 2017 (UTC)[reply]

Like I say, if the article has been uninformative for perhaps 6 months, send it to Draft. And send new ones (after a date certain) to draft immediately.Alanscottwalker (talk) 00:38, 19 April 2017 (UTC)[reply]

Why? People are more likely to find and improve an article if it is in main space than if it is hidden on some draft page somewhere. Why the arbitrary timeline to improve an article? Spanneraol (talk) 00:42, 19 April 2017 (UTC)[reply]

Because they are not getting improved, obviously. As for finding them anyone who wants to know and write about them will find them in Draft space. Alanscottwalker (talk) 00:47, 19 April 2017 (UTC)[reply]

We are not the article quality police.... and no it is much harder to find stuff in draft space. Spanneraol (talk) 00:49, 19 April 2017 (UTC)[reply]

Police? What? We are not here to not write actual encyclopedic biographies. And I'll add, that those articles that basically just regurgitate that single sports stat site, does make Wikipedia terrible in the eyes of readers. The least one can do is put all those sources about their life you say exist in a further reading section, so the reader might actually learn something. Alanscottwalker (talk) 00:55, 19 April 2017 (UTC)[reply]

Did you run some survey that led you to understand that to be the case or is it your personal opinion? You are essentially saying articles should be deleted if they don't measure up to your personal standard of what a good article should be. Thats contrary to the mission of wikipedia. Spanneraol (talk) 16:04, 19 April 2017 (UTC)[reply]

No. Per Wikipedia policy as in WP:PAGEDECIDE, I am suggesting putting these non-articles in Lists and moving them to Draft pages. The only thing contrary to the mission of Wikipedia is the continued refusal, or inability to substantiate that these subjects have the sources. As for the rest, it takes no survey to know an article that rips off another website by just regurgitating its contents is bad. -- Alanscottwalker (talk) 20:39, 19 April 2017 (UTC)[reply]

Part of the problem is that many editors think that a "presumption" of notability means something actually is notable. Our SNGs could clarify this by using "likelihood" instead of "presumption". It is true that the more an athlete has done in his/her sport, the more likely it is that they will meet our GNG. Blueboar (talk) 00:31, 19 April 2017 (UTC)[reply]
- The first sentence of the sports-specific notability guideline is the following: This guideline is used to help evaluate whether or not a sports person or sports league/organization (amateur or professional) is likely to meet the general notability guideline, and thus merit an article in Wikipedia. The third sentence uses "likely" again: If the article does meet the criteria set forth below, then it is likely that sufficient sources exist to satisfy the inclusion criteria for a stand-alone article. The third paragraph starts as follows: Please note that the failure to meet these criteria does not mean an article must be deleted; conversely, the meeting of any of these criteria does not mean that an article must be kept. I'm not sure how many more times this can be emphasized in the lead section of the guideline. isaacl (talk) 00:43, 19 April 2017 (UTC)[reply]

Too often, the only part of a SNG that our fellow actually bother to read is the bullet pointed list of things that create a "presumption" of notability. They (mistakenly) think meeting one of these criteria makes the subject/topic notable... when all meeting the criteria indicates is that the subject/topic is likely to be notable (i.e. likely to pass GNG). Blueboar (talk) 00:57, 19 April 2017 (UTC)[reply]

I appreciate that many editors don't bother reading the guidelines they quote, but closers of AfD discussions really must read the guidelines referenced within the discussions, otherwise, they are overriding the consensus that created the guidelines. Short of inserting the sentences I quoted above into every subsection, I don't know what more to do to publicize the underlying context of the sports-specific guidelines. I wrote the FAQ, with references to the original discussion that created WP:NSPORTS, to make its relationship to the general notability guideline crystal clear. isaacl (talk) 01:10, 19 April 2017 (UTC)[reply]

A topic is generally considered to be notable if it has received significant coverage in reliable sources that are independent of the subject. While the exisatence of such sources is possible, and frequently easy, to prove, the lack of them is impssible - especially given the fact that these sources need not be in English, on the internet, or freely accessable to the general public. In stead, we have specific guidelines which tend to indicate when topics are likely to be notable. עוד מישהו Od Mishehu 03:46, 19 April 2017 (UTC)[reply]

The burden can't be on proving a negative. At some point, people who create the article, or argue to retain it, need to actually come up with references, not just handwave that they've got to be out there somewhere or another. Seraphimblade ^{Talk to me} 07:43, 19 April 2017 (UTC)[reply]

Agree... But that begs the question: when do we reach "at some point"? A month? A year? Blueboar (talk) 10:05, 19 April 2017 (UTC)[reply]

@Blueboar: Well, to me, especially for an area where the vast majority of the articles will be BLPs, that standard is analogous to WP:BEFORE. Before you create the article, ensure that you have found sufficient source material to sustain a full article, not just a blurb or mention that would be sufficient to stave off BLPPROD. If you failed to, the article will be deleted, but if and when you turn up those sources, you can recreate it from them. Seraphimblade ^{Talk to me} 15:29, 19 April 2017 (UTC)[reply]

The majority of the articles that are actually stubs are not BLPs because they deal with athletes from the 19th century or early 20th century for which online sources are difficult if not impossible to find. Current living internet-era athletes in major sports are fairly easy to source and reference. The guidelines are meant to account for these people for whom sources are only available in newspaper microfiche at libraries. Spanneraol (talk) 16:02, 19 April 2017 (UTC)[reply]

Well, then, same standard, really. Once you go to the library, read the microfiches, and determine that the source material really does exist, you can create the article, citing those sources. But just "Well I think it's at a library...on something...somewhere!" doesn't do. Create the article once you actually have the references in hand, not just with a wave of a hand that they must be out there somewhere. Seraphimblade ^{Talk to me} 18:42, 19 April 2017 (UTC)[reply]

Agree that our standards for athletes' notability are far too lax and allow far too much cruft into the project. Take, for example, this guy: an 18-year-old kid who kicks around a football for a living. Essentially no in-depth coverage exists, but he's "presumed notable" per WP:NFOOTY, which apparently trumps WP:BASIC. (Certainly it does so at every relevant AfD I've witnessed.) By contrast, none of the faculty at the local university have articles, because WP:PROF is quite stringent, and rightly so. Nevertheless, these are people who've put in time and effort in obtaining doctorates, writing publications and teaching students. Similarly, none of the physicians at the local hospital is considered notable, although these individuals have studied medicine and are caring for patients every day. Again, I don't object, but what I do find objectionable, indeed downright bizarre, is that out of the economist, the oncologist and the basically random adolescent, it's the last we presume notable.
In terms of what should be done, I agree it's a more complicated problem now than it was ten years ago, but I think the focus should certainly be on having articles that comply with WP:BASIC: "people are presumed notable if they have received significant coverage in multiple published secondary sources". Presumption of notability should be granted for significant coverage, not simply participation and routine mention. - Biruitorul ^Talk 16:20, 19 April 2017 (UTC)[reply]
What's actually being proposed here? One note to keep in mind: how much coverage is out there that isn't online or in English. I've skimmed this thread and seen people challenge the notability of a Romanian and a Polish athlete. Has anyone checked the local papers where they come from? – Muboshgu (talk) 17:33, 19 April 2017 (UTC)[reply]
- I don't think that's such a salient issue for contemporary athletes - in this day and age, newspapers if anything have more of an online than a print presence, and we have editors who know the major European languages. The heart of the problem lies with the too-permissive nature of WP:ATHLETE, which presumes notability for sub-stubs even in the absence of in-depth coverage. - Biruitorul ^Talk 19:05, 19 April 2017 (UTC)[reply]
  - I don't think the presumption is the problem as much as editor's application of it at AfD. In some cases, old articles that have shown no sign of being able to satisfy GNG are deleted despite meeting ATHLETE, but too often editors keep articles based on ATHLETE alone. Jogurney (talk) 19:11, 19 April 2017 (UTC)[reply]
    - While I agree the problem is the editor's application, I don't think the old article example is valid. The policy is good when a subject is relatively old. The older a subject is (or for an athlete the older their prime sporting years), the harder it is to actually find sources. While certainly this applies to pre-Internet players, I think that can go back as few as five years as many sources are not online anymore. Let's face it, most editors are sitting at their computer and doing online research. No one is going through hard newspapers finding sources. However, based on various factors, such as attendance and general understanding of newspaper, radio, etc. it can be viewed that era would have been covered - its just difficult and a lot to ask a volunteer website to go check those. Its a presumption the community that has knowledge on the topic has created and it should be respected. To be more honest, we have to make a judgement. Who can say that a 1900 Football League player was notable? None of us were alive then and no one is actually checking the sources. Its an unreasonable proposition to force the keep position people to find sources. If that is the case, then we might as well just say this is a project from 1997 on. Another time presumptions are good are foreign language. How many editors we have that know both English and Thai, are knowledgeable on the Thai League T1, and are in Thailand to access and evaluate sources (reminder - not everything is on the Internet even today). I assume very few if any. We have to make judgment calls and the sports projects (and other projects) are in the best position to make those calls. While I was a commenter, I think this whole idea was well reasoned by a number of editors in Wikipedia:Articles for deletion/Andy Ainscow. I think the solution is to have admins decide debates using critical thinking. Its not a vote, so a bunch of "Keep - meets WP:NFOOTY" should be discounted. For articles that are likely to have sources found (e.g., a 2017 National Football League player), make editors find sources since they should be able to be found. For a 1930s era racer at the 24 Hours of Le Mans, let's keep since no one is going to go though 80 year old French newspapers, but we know generally how well covered the race and its competitors were. Admins should use sound judgment, I think that solves about 95% of the problems. RonSigPi (talk) 22:55, 19 April 2017 (UTC)[reply]

One thing that leave me cold in these discussions is that nobody is making the case "Having these articles degrades the reader's experience, and/or is a burden on our system not warranted by the article's small value, because _________". Well what goes in the blank? If you don't have something exciting to put in the blank, you can count me off the boat, and a lot of other editors too: you are not going to get anywhere.

I'll tell you what doesn't go in the blank: "It it too permissive" or "these are not notable people" or "it doesn't meet our notability rules which were handed down to Moses on Mount Ararat" or "our printing costs are too high" or "it annoys me to see these articles" or "too many of these come up if I use the Random Article function" or "it is overloading the servers with data" or "these articles have small value" (true, but what I want to know is net value), and stuff of that nature

You know what I haven't seen? I haven't seen anybody say "I don't come here anymore, instead I use Britannica or just Google, because there are too many articles here about cyclists, and this degrades my ability to use the Wikipedia".

Now, there are arguments against having articles on unnotable subjects, and they mostly have to do with our own internal dynamics, and the one good argument that's raised for all obscure articles (not just athletes) is that is that there some overhead cost to any article -- which basically means, it has to be watchlisted or checked in the Recent Changes feed.

I wouldn't worry about that too much myself. Obscure athletes are pretty popular, relative to all obscure subjects. People like sports! Just as some people like to watch over maintain articles on lichen species, Parliamentary elections of the 17th century, and towns in Ecuador, so there are people who like to watch over articles on Portuguese cyclists. This enthusiasm is evidenced, in part, by the very existence of the articles.

And if you take those articles away, those editors are not necessarily going to be like "OK, they've cleared out most of Category:Portuguese cyclists, guess I'll focus on Category:South Carolina State Senators instead". It doesn't necessarily work like that.

On that score, there're a lot of subjects before athletes where I'd worry that nobody is watching them. I just found vandalism from 2014 in a (non-athlete) article. Is that happening a lot with athlete articles in particular? Nobody in the discussion above has claimed that it is. Is it?

Another good argument could be "no one -- quite literally, virtually no person -- is interested in this class of articles, as shown by the page views which are consistently at or near zero in the my sample of these kinds of articles, so the value is not just small but is zero". Of course, that's only a good argument if it is true. Is it?

In all the above discussion, I have seen one, and really only one, good argument against these articles: that they are frequently inaccurate -- the statistics and other data in them are wrong, to an unacceptable level of frequency. However, this was presented as an offhand aside and I suspect it is just an uninformed opinion. It if was true, then we would have a problem. Is it?

Go get some data, or even anecdotes, showing some of these things to be true. Absent that, you haven't convinced me that it is a problem. Herostratus (talk) 10:31, 20 April 2017 (UTC)[reply]

Any stub has a higher chance of being if not inaccurate then at least misleading, because a stub will likely be based on a single source or a very small number of sources. Sources, even good reliable sources, are wrong sometimes, omit important information or concentrate on only one aspect of the subject (potentially creating WP:UNDUE or WP:BALASP problems); which is a problem with only one cure, namely introducing more sources. This is of course true for any stub, not just athlete stubs; but it does make it understandable that some people aren't happy to see so many athlete stubs, particularly given that many of them are BLPs.
A very concrete case of athlete articles being inaccurate unacceptably often was the recent Sander.v.Ginkel debacle, which led to the mass deletion of thousands of athlete BLPs created by Sander.v.Ginkel. The problem there was caused by a bad editor, not by too lenient sport-specific notability guidelines, but the stubby nature of the articles was still a factor; articles that had been fleshed out by editors other than Sander were less likely to have serious problems, though they were still included in the cleanup effort. Sideways713 (talk) 11:21, 20 April 2017 (UTC)[reply]

We've had that argument before: it's the fiction purges of the late 2000s. --Izno (talk) 12:39, 20 April 2017 (UTC)[reply]

OK, but "argument made previously" is not the same as "wrong argument"; so that's not really a refutation. @User:Sideways713, OK, fair point. But a lot of the material in these articles, I think, is "She finished third in the 1999 such-and-such race with a time of 27:13" and so forth. Well, its either true or isn't; its a statement of objective measurable fact. So if we have a site that is regularly giving times of 27:13 when the actual time was 26:43, then we need to blacklist that site. Most sites giving statistics are probably getting them from the official source, or are the official source, I would suppose. Although granted typos occur, yes.

However, most statements in the encyclopedia have a single source (if any). And multiple refs don't necessarily help that much, because errors replicate. I get that with birthates sometimes: eight sources that have it wrong, because they're copying from the same flawed source or each other. So I dunno. Possibly the single-source angle is an argument against a lot of lower-level articles, from politicians to villages to lichen species. Whether articles on athletes in particular are a problem on this account I don't know. Herostratus (talk) 13:12, 20 April 2017 (UTC)[reply]

"She finished third in the 1999 such-and-such race with a time of 27:13" is a sentence not an article, and for an Article the pedia needs multiple qualified RS. Moreover, NPOV would suggest that most statements in the Pedia have multiple reliable sources, and common sense would seem to back-up that the tons of innocuous statements like "Paris is the capital of France" have multiple reliable sources, and for statements 'of note' per NPOV/OR/BLP, multiple sources would exist, as we rely on noted received knowledge. Sure, only one source is cited but it has 'weight' because of the multiple other sources that reconfirm it, contextualize it, and qualify it. Alanscottwalker (talk) 15:45, 20 April 2017 (UTC)[reply]

Beyond what Alanscottwalker said, there's a difference between statements that have only one source and whole articles that have only one source. Bad sources, typos and the like can cause a single statement in any article to be wrong; but if that single statement is the entire article, the consequences are more dramatic. For example, the article may be named incorrectly; obviously if there's a typo/mistake in the source, but also if the source uses a name that isn't the WP:COMMONNAME. With multiple sources it's usually easy to tell what the COMMONNAME is, or at least discard very uncommon names as definitely not being the COMMONNAME.
An even worse problem would be a poorly supported controversial statement in a BLP.
Often, though, the problem with an article that only has one source isn't so much what is in the article as what is not in the article. Say:

The source correctly notes she finished third in such-and-such race in 1999, but doesn't note that she won the same race in 2001, so that more noteworthy fact won't be mentioned in the article.
The source, and consequently the article, concentrates on her running career and doesn't mention that she's equally or more notable as a coach, a triathlete, a local politician, etc.
The source omits important controversies surrounding the athlete.

In these cases what's in the stub is perfectly true, yet still gives the reader a misleading overview of the topic. This kind of thing can happen very easily with athletes, maybe more easily than with non-athletes, because databases (commonly used as the single source in athlete stubs) sometimes give an incomplete view of a subject's athletic career and almost always give an incomplete view of their life as a whole. But the other reason why it may be a bigger issue with athletes is simply that there are so many of them. Sideways713 (talk) 15:53, 20 April 2017 (UTC)[reply]

Just as a general comment for brainstorming: WP:N does not require a standalone article should the GNG be met; WP:N only requires that the presumption that the GNG can be met to support a stand-alone article. What this means in another way is that we are not required to include every athlete that possibly may meet the GNG. Wikiprojects are allowed to specify SNGs purposely more stricter than the GNG if they believe that will improve their project's coverage. (As an example, over at the video games project, we developed what is now WP:NOT#GAMEGUIDE to avoid extensive strategy guide material, most which can meet the GNG but fails NOT. I will point out that WP:NOT suggests we are not simply a "Who's Who" type work, so while having the ability to document every athlete that might have ever played doesn't mean that is a feature we want. That said, finding a bar here that is above the GNG and still reasonably inclusive rather than deletionist is not a simple task at all.
With that in mind, one of the things I do find about the current NSPORT that I've pointed out in the past is that they make presumptions on importance that while can be documented, really beg the question of whether they meet the GNG. A common one is for any professional NFL player that at least steps onto the field once. The argument, fairly enough, is that for someone to be pro NFL, they had to have had a college football career, and that college career can be documented. However, to some extent, that type of coverage could be considered rote or routine for most college players; there's definitely a few that quickly bubble to the top but with so many teams, the only type of coverage that gets every single players starts to become more localized (to the school), and more on stats than actual "content". In other words, if we took a stub for an NFL player than had an average college career, and only one game appearance before retiring, and fill out the article with their college career, we're still looking at a very weak article, content-wise, even if it is meeting WP:V, one that will likely never reach a GA or FA (which should be an ultimate goal for every article). Making this NSPORT presumption far too inclusive. Not all of NSPORT is like that, but I really do think there needs to be a re-evaluation that is going to avoid unexpandable permastubs that even if fully sourced will go nowhere.
I don't think ditching NSPORT is right as there's far too many issues with print-only sources that the presumption of notability is needed to allow articles to expand, but there does need tightening of the loosest inclusion metrics for players. --MASEM (t) 18:14, 20 April 2017 (UTC)[reply]
- Perhaps a better place to start is expanding on the meaning of "significant coverage" in English Wikipedia's basic criteria for the notability of persons: the independent, non-promotional, non-routine, reliable secondary coverage of the subject should be sufficiently detailed that a reasonably complete overview of the person's life can be written. A reasonably complete overview includes major life events, accomplishments, and other significant influences spanning most of the person's lifetime. isaacl (talk) 18:40, 20 April 2017 (UTC)[reply]
  - That's opening a huge can of worms - such a change could be used to drive massive deletions, not just of sportspersons, and not just of minimal content stubs - "major life events, accomplishments, and other significant influences spanning most of the person's lifetime" means that the majority of biographies, even of major figures and featured articles, could be deleted because we don't have references of where the person went to school, or what they did in retirement. For example - see Robert Millar - the article of this notoriously private person contains nothing of his life post 1998. Does that mean it should be deleted? Because by strict application of this rule, it would be liable for deletion.Nigel Ish (talk) 19:06, 20 April 2017 (UTC)[reply]
    - A better line, but this would require a lot of thought, is the idea of avoiding people that are doing their profession in a routine manner, while while may be very visible to the world via sources, is just not necessarily an encyclopedic topic. This for NSPORTS would require recognizing that every professional player is not necessarily unique. Furthermore, this at least sets an apples-to-apples comparison between players within the same sport, or if outside NSPORTS, within the same profession (for example, not all CEOs of Fortune 500 companies are necessarily notable, some are just doing their job to keep their business running). This needs a lot more thought before even going that way, but it is a possibility. --MASEM (t) 19:11, 20 April 2017 (UTC)[reply]
      - There appears to be a presumption in this discussion that a large proportion of articles on sportspeople MUST be deleted and that Wikipedia should have much less coverage of sport. Some people here are proposing much stricter application of GNG than would be applied to say 10th century royalty or to Politicians. This risks increasing systemic bias - by removing coverage of older (pre-internet) sportspersons, participants in less fashionable sports, participants in non-English speaking nations or of sportswomen, all of whom may suffer from any attempt to artificially ratchet up standards for sportspersons.Nigel Ish (talk) 19:31, 20 April 2017 (UTC)[reply]
        I'm not making any presumptions either way; I rather agree that sports figures and other persons should be treated in the same manner, as much as is feasible, which is why I think it is better to look at potential changes to Wikipedia's standards for inclusion for persons. I'm just exploring what I think the crux of the issue is: some editors want to use a standard for inclusion that requires sufficient source material to make a reasonable non-stub article. I apologize for not making it clear that my suggestion was just a starting point for discussion. I toyed with language to say that articles must cover the portion of the persons' lives during which they were engaging in the activities that make them notable, but I settled on what I wrote above as a compromise. The key problem of course is how to determine what makes a reasonable non-stub. (I know it will be hard to reach any agreement as Wikipedia's version of consensus has structural issues, including an inability to scale up, but nothing ventured, nothing gained.) isaacl (talk) 00:49, 21 April 2017 (UTC)[reply]
Disagree - Our notability criteria are here to make sure that everything on Wikipedia is able to be verifiable and neutral. With NSPORTS and such, we are setting the bar for what we can verify and write in a neutral manner. Without this criteria, we are excluding certain people without any noticeable benefit. Thus, getting rid of this criteria will be a detriment to the encyclopedia. RileyBugz^{Yell at me} | _Edits 20:09, 23 April 2017 (UTC)[reply]
Disagree - We should seek to establish a high quality of articles on all athletes, but not having a great abundance of material on an athlete should not preclude us from including that person. Every major competition, MLB team &c. should be able to blue-link to players involved, since even so-called "non-notable" athletes often "get around." Editosaurus (talk) 03:57, 24 April 2017 (UTC)[reply]
Agree, I've long believed that NSPORTS allows us to keep articles on living persons that we would not otherwise have due to a paucity of reliable sources. This has led to a situation where there are a lot of low-quality directory style entries that are not adequately maintained and create an enormous risk of serious BLP violation. I'd be in favour of wholesale repealing it and moving back to the standard that substantial non-routine coverage in reliable sources is required for sportsperson biographies. Lankiveil ^{(speak to me)} 04:56, 24 April 2017 (UTC).[reply]
- As discussed in the Wikipedia:Notability (sports)/FAQ, that is the standard. The sports-specific guidelines provide rules of thumb when this standard is expected to be met, but they do not replace the general notability guideline. Any AfD closers ignoring this portion of the sports-specific guidelines are substituting their own opinion for the consensus agreement that established the guidelines. isaacl (talk) 05:08, 24 April 2017 (UTC)[reply]
  - I'm fine with that being the intent, but that's not what is happening on the ground at AFD. My observation from closing and participating in many, many, many discussions is that "Passes WP:NSPORTS" or similar is being used as a substitute for actually locating sources and presenting them for discussion, on the grounds that "well, they should be there". Potentially one could tweak the guideline to emphasise that it should not be used as a substitute for finding good references. I'd also suggest that for many more obscure sports, merely competing at the top level is probably not a good indicator of notability in the same way that it is for major sports. Lankiveil ^{(speak to me)} 01:34, 26 April 2017 (UTC).[reply]
    - I have argued for improved wording (in particular the second sentence and the nutshell summary), but to be honest, if closers are not reading the instructions that are right there in the first sentence and the third paragraph of the lead section, plus the FAQ which puts the context of the guidelines front and centre, tweaking the wording isn't going to help. It's an odd position where there is really agreement between the editors who have crafted the guidelines (*) and those who are concerned that the guidelines are not being applied in accordance with the general notability guideline, but closers are not following suit.
    - (*) For the sports that have been discussed on the talk page in the past few years; others can benefit from further review. isaacl (talk) 01:54, 26 April 2017 (UTC)[reply]
Disagree - Someone who competed at the highest level of a major sport is notable. These events receive significant coverage, both for them and the participants, and, as others have pointed out, it's useful to bluelink the participants. Smartyllama (talk) 18:16, 24 April 2017 (UTC)[reply]
- The general standard for a subject having an article in English Wikipedia is not based on attaining achievements, but by receiving significant, independent, non-routine, non-promotional secondary coverage from reliable sources. Accordingly there is no consensus that those who compete at the highest level of a sport are inherently notable. isaacl (talk) 19:23, 24 April 2017 (UTC)[reply]
  - And therein lies the problem - a perfectly sensible real world definition of notability dismissed based on a guideline (WP:GNG) that is (generally) reasonable when applied to argue that a subject is notable, but absymal when used to argue that they are not. --Michig (talk) 19:32, 24 April 2017 (UTC)[reply]
    - The issue is that consensus doesn't scale upwards. As long as the English Wikipedia editing community tries to follow its version of consensus (which is really a straw poll followed by people trying to convince each other to change their stated view), and doesn't / can't take into account expert opinion (*), it cannot agree upon what achievements should be used to determine if a subject should have an article. Instead, it tries to evaluate the question one step removed: are there adequate sources to indicate if a subject should have an article?
    - (*) There are pros and cons to this; a pro is that English Wikipedia can more easily expand into areas not covered by other encyclopedias. A con is that editor opinions end up weighted roughly evenly, regardless of their level of familiarity with the topic in question. This particular con plus the fact that (real-world) consensus only works when the interested parties are strongly aligned in their goals are huge barriers in trying to establish an achievement-based standard for having an article on a subject. isaacl (talk) 20:43, 24 April 2017 (UTC)[reply]
    - @Michig: It is a very good point that we need to use common sense, and I totally support the idea that top sportspeople - ones who represent the nation at top events - are notable. But the problem is that there is no agreement on what is a top event: for someone like me, it is just the Olympics, but for some fans of cycling, apparently it is 10-20 events per year, and that seems excessively inclusive. Hence the default to GNG makes sense. If there is coverage, we are good. But if there is nothing but a stat site saying that someone competed in event X, didn't win any prize, just competed... I don't see what makes them notable. Again: I am fine agreeing to a GNG/BIO exception for Olympic participants and winners of several other major competitions, but run-of-the-mill participants of events who did not win any medals or such... why should they get special treatment? What makes them more notable than let's say a random average academic, who is also listed on an academic stat(citation) site? --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 07:03, 26 April 2017 (UTC)[reply]
      - In most sports, there are far more major events than the Olympics, and taking part in those events represents a competitor getting way beyond 'run-of-the-mill', e.g. in cycling there are world championships and other international track events, and on the roads major tours - your average club competitor won't get anywhere near these events. For some sports the guidelines are very inclusive (e.g. someone who played one first class cricket match), others quite the opposite. What a lot of people don't appreciate is that having a guideline for notability doesn't mean every subject that satisfies it has to have an article - if we have 100 minimal articles about people who played in one match for Glamorgan they can all be covered in one list article - if this isn't happening that's a fault of editors, not the guideline. This whole discussion is a bit bit pointless since there is no proposal for anything specific to be changed. If someone feels any part of the guideline is pitched at the wrong level of inclusivity they should make a specific proposal of how to change the guidelines. To answer your other point, as someone who used to be a random average academic, they don't tend to get covered because nobody is generally interested in them; Any that achieve something out of the ordinary and become top-level in their field will likely get articles, just like sportspeople who do that get articles. --Michig (talk) 07:26, 26 April 2017 (UTC)[reply]
        Wikipedia:Notability (sports) and the general notability guideline do in fact specify standards for having an article on a subject. So in the context of this discussion, the issue is what standards should be required for a sports figure to have a standalone article. I think everyone agrees a lower standard is required to be mentioned in another article. isaacl (talk) 15:46, 26 April 2017 (UTC)[reply]
        No, I disagree. There is no proposal here. If you read WP:N, it is quite clear, despite what a lot of editors seem to believe, that whether or not a subject is considered notable, and whether or not they should have a standalone article are separate considerations. After spelling out the notability criteria it states "This is not a guarantee that a topic will necessarily be handled as a separate, stand-alone page. Editors may use their discretion to merge or group two or more related topics into a single article." Given the editor who started this discussion is trying to get stub-level articles on world-class athletes deleted to try to prove a point, it appears that some people think we should either have a detailed article on sportspeople or should have no mention of them at all, which is ridiculous. If someone were to come up with a proposal on how to incorporate minimal articles on SNG-satisfying subjects in wider articles that would at least be constructive. The SNG tells us whether someone is likely to be notable, not necessarily whether they should have a standalone article. Maybe if more people understood that they would see that NSPORT isn't really the problem. --Michig (talk) 16:18, 26 April 2017 (UTC)[reply]
        I didn't say anything about there being a proposal. I referred specifically to the general notability guideline, and not the notability page in general. However, the first sentence in Wikipedia:Notability is On Wikipedia, notability is a test used by editors to decide whether a given topic warrants its own article. The sentence saying that a topic will not necessarily have a standalone article doesn't negate the previous paragraph, which starts A topic is presumed to merit an article. It just means even if a topic meets the standards for having a standalone article, there doesn't have to be a standalone article if editorially the content would better fit into another article. Also note the discussions that resulted the sports-specific notability guidelines specifically decided to create criteria for the creation of standalone articles. In my comments, I usually refer explicitly to "standards for having an article", or a similar phrase, to avoid any confusion regarding what aspect is being discussed. isaacl (talk) 17:44, 26 April 2017 (UTC)[reply]
Agree, god, yes. I have been slogging through the orphan backlog at Category:Orphaned articles from February 2009 (over 17,000 articles) and you see so many awful sub-stubs for random sportspeople who have played one single game of cricket or football or whatever other sport at a top-tier level and that is the sum total of their "notable" lifetime achievement, and yet because of WP:NSPORT, people insist that we maintain an article on them.

In almost all of the cases I've seen of these people, there is simply no WP:RS coverage of them that rises above the brutally routine sports stat collections: "Dude McDude of Some Second-Tier Team 1 game 0 points 0 whatevers 0 other important metric". No in-depth commentary, no discussion of their function in the game, not even trivial personal history. Nothing. Hell, I'd love to see an article that mentioned why the guy only played one game, but you don't even get that. We are not an indiscriminate collection of stats, and yet in many cases NSPORT forces us to appear that way.

As Masem noted above, "people that are doing their profession in a routine manner, while while may be very visible to the world via sources, is just not necessarily an encyclopedic topic." In many cases these people aren't even visible to sources except stat-farms. If there is something non-routine (ie, notable) about these people, it should be in the articles, but otherwise, there is no point to maintaining thousands of articles about people who did a sport once in a manner so dull as to be recorded as "1 game 0 points 0 whatevers". ♠PMC♠ (talk) 21:39, 24 April 2017 (UTC)[reply]

Comment A bunch of commenters seem to be missing the point that NSPORT and simiar guidelines are all meant to be subordinate to GNG. If there aren't enough sources to write a decent article, then the article shouldn't exist. The fact that a bunch of AFD commenters and a lot of article-creators also miss this point isn't a failure of the policy; it's a failure of the individual editors, and editors who repeatedly fail in this manner should be politely notified of the policy first, cautioned second, and (if both of those fail) sanctioned. Hijiri 88 (聖やや) 09:37, 25 April 2017 (UTC)[reply]

Incorrect. As WP:N spells out: "A topic is presumed to merit an article if it meets either the general notability guideline, or the criteria outlined in a subject-specific guideline". WP:N, GNG, etc. are guidelines (rules of thumb), not policies - big difference. You can often have sufficient sourced information without finding anything constituting significant coverage, which is where GNG falls down. --Michig (talk) 09:58, 25 April 2017 (UTC)[reply]

Okay, I was using "policy" in the figurative sense; if we were nitpicking about what are "policies" and what are "guidelines" then this whole thread, which is about guidelines, would be off-topic per the name of this page (which I also take as figurative). Also, you appear to be right to quote WP:N that way, but that wording is clearly in error, as it incorrectly assumes that all the subject-specific guidelines include explicit "significant coverage" requirements. Wikipedia doesn't contain standalone articles on topics for which enough sources don't exist to write a decent standalone article. This fact is so intuitive that it shouldn't need to be explicitly codified in any PAG, but it is in the "significant coverage" clause of GNG. All the other notability guidelines are subordinate to that in that if a topic has not received significant coverage, and the only way we could write a decent article is by engaging in OR, the topic doesn't merit a standalone article. I really think we should be discussing either (a) the removal the "either/or" part of WP:N, if it is being interpreted as allowing articles on non-notable topics, or (b) the addition a "significant coverage" clause to NSPORT and whatever other guidelines currently lack it. Hijiri 88 (聖やや) 01:47, 27 April 2017 (UTC)[reply]

No, right at the top of the page it states this page is for discussion of policies and guidelines, but there's a big difference between the two. The wording on WP:N is not in error, and the subject-specific guidelines do not all insist on "significant coverage". You can have enough sources to write a well-enough sourced short article without having anything that strictly constitutes "significant coverage" - short articles just like you would find in any other encyclopedia. --Michig (talk) 13:11, 27 April 2017 (UTC)[reply]

The key word of WP:N is "presumed". We allow the standalone from a subject-specific guideline, presuming the criteria in that guideline are chosen appropriately, that in the future and with time and effort, a detailed article with significant coverage (which we do consider notable now, no presumption necessary) can be obtained. The presumption that allows for a standalone article can be challenged by demonstrating that a proper search of sources, as outlined at WP:BEFORE, brings forth no new sources or a lack of information that can bring the article up to meet significant coverage. That's why the subject-specific criteria are selected in a manner that assures few false positives (articles that can't be expands). --MASEM (t) 13:23, 27 April 2017 (UTC)[reply]

You keep stating that the SNGs are just indicators that GNG can be met, but it isn't true. We had SNGs before anyone came up with the GNG. And no, satisfying the GNG is not a guarantee of notability. We really need to get away from this doctrinaire GNG-obsession because it simply doesn't make sense. --Michig (talk) 13:59, 27 April 2017 (UTC)[reply]

WP:NSPORT was created on 29 August 2007. WP:N was created at least as early as 19 May 2005, and while it has undergone several moves and renames and changes of wording, the earliest expression of what we now call the GNG is found on 20 November 2006 when it was merged into WP:N by User:Uncle G who had been working it up in his userspace before that. Called "Primary Notability Criterion" it is essentially (with some minor wording changes) nearly identical to what we are using now. So you seem to be mistaken that NSPORT predates the GNG. The GNG had existed for 10 months before NSPORT did. Also, 11 years of continuous acceptance as a guiding principle on the minimum standards for a stand-alone article is a pretty good sign of consensus. --Jayron 32 14:19, 27 April 2017 (UTC)[reply]

I didn't say NSPORT predated the GNG, and WP:N has for as long as I can remember stated that subjects are presumed notable if they satisfy either the GNG or one of the SNGs. WP:N is the guiding principle, not GNG. --Michig (talk) 14:27, 27 April 2017 (UTC)[reply]

Please see Wikipedia talk:Notability/RFC:compromise, an RFC in 2008 that established the relationship between the GNG and SNG. --MASEM (t) 14:31, 27 April 2017 (UTC)[reply]

And multiple discussions since with differing outcomes, as this old chestnut seems to get discussed about every 6 months. --Michig (talk) 14:37, 27 April 2017 (UTC)[reply]

Except it still remains valid, there's a reason we use the word "presumed" in both GNG and the SNGs, and link that to Rebuttal presumption. I am pretty confident that while I don't see eye-to-eye with many of the drafters of the NSPORTS often, that we have been in agreement that if you can satisfactorily prove there are not enough sources or significant to expand an article beyond a stub (that otherwise meets NSPORTS) to meet WP:N in general, then deletion is an acceptable route; it's just that with sports, satisfactorily proving that requires a great deal of effort for those seeking deletion (print source exhaustion for example). --MASEM (t) 15:47, 27 April 2017 (UTC)[reply]

Right, but the problem is that WP:NSPORTS/FAQ is not consistent with the general notability guideline: "The topic-specific notability guidelines described on this page do not replace the general notability guideline. They are intended only to stop an article from being quickly deleted when there is very strong reason to believe that significant, independent, non-routine, non-promotional secondary coverage from reliable sources are available, given sufficient time to locate them." However, it seems evident that in practice this is not actually observed: see for example this deletion discussion where there does not seem to be any effort to determine whether the subject passes the GNG, and instead going down the WP:NCYC checklist. Furthermore, to make matters worse, WP:NSPORT is internally inconsistent, since it also says "The article must provide reliable sources showing that the subject meets the general notability guideline or the sport specific criteria set forth below." This is consistent with the guidance put forth at the GNG, and it seems like it captures the way NSPORT is applied in practice, but defenders of the permissiveness of NSPORT seem fond of pointing out whenever asked that meeting NSPORT implies that the subject also meets GNG, which is false for the majority of articles about participants in minor sporting events. Sławomir Biały (talk) 10:42, 25 April 2017 (UTC)[reply]

I suspect the topic-specific guidelines were all written with the assumption that Wikipedians understood that topics need to have had significant coverage to merit standalone articles. In other words, they were written with the same assumption I made when I wrote my comment above, but WP:N incorrectly assumes that they were written with the assumption that they could act as a substitute for GNG. Hijiri 88 (聖やや) 01:47, 27 April 2017 (UTC)[reply]

Michig, that is only partially true. There's actually three pieces of guidelines involved: the core concept of WP:N that we can flesh out a topic in detail with secondary sourcing, the GNG that established a presumption of notability based on at least some secondary sources existing, and the subject-specific notability guidelines that provide an alternate means of presumption (from the GNG) that WP:N can eventually be met, using criteria that generally assure secondary sources can be found for a topic, which in turn is generally based on reaching some level of achievement of merit. So to this end, the GNG and the SNGs are both subsidary to the core WP:N principle (that is, the SNG is an equal but alternate path of presumed notability), but in the long run, we do expect topics much meet WP:N with demonstration of the secondary sources. In all fairness, with something like NSPORT, this means that editors wishing to delete do need to do the footwork of exhausting to a reasonably good degree the available pool of resources (including print versions!) to show no sourcing exists, making it a difficult barrier to delete and avoids petty deletions. And in my past interactions with most editors behind NSPORT, they fully understand and accept this in drafting the NSPORT guidelines. What I think is the newer challenge that is being recognized is that most of those sources that NSPORT anticipates people will find when they actually search run afoul of being routine coverage for sports; for example, one criteria is for anyone that has played in (that is, actually on the field at some point, not just bench warming) a professional league game is presumed notable, the argument having been that one does not simply get into the pro leagues for a sport without having some career in the past (college, amateur, whatever), and those merits can be documented, at minimum. But it seems to me more and more that this type of reporting of team sports often is routine and very often focused only from local sources. Technically that all meets the WP:N/GNG, but... taking NORG and the recent discussion of SCHOOLOUTCOMES as examples, there's a general trend that in-depth coverage from only local sources is not sufficient for a WP article. Eg, its not clear how to resolve this but it is one of the factors I see what NSPORT is seen as too inclusive. --MASEM (t) 13:51, 25 April 2017 (UTC)[reply]

I mean, our rules are suppose to codify practice. If they don't, change 'em. If you can't change 'em -- and it's hard to change anything here -- ignore 'em. There are exceptions for core constitutional provision -- NPOV, V, BLP, and a (very) few others, where even if there's a local consensus to ignore, this is overridden by the functionaries. Whether or not we have an article about Joe Shlabotnik is not a core constitutional question.

So if you've got a 12-2 "Keep" vote based on WP:NFOOTY wielded in the a manner that makes it clear that the sentiment is actually "But we like us some football player articles", what are you gonna do? Sputter "Bu-bu-but our Rules, which were handed down from the Holy Mount"?

Since the days of Nupedia, it has been assumed that if

There are sufficient good sources to have an article, and
Some reasonable number of people would be interested in the article, and
Somebody can be bothered to actually create the article,

then the default is that we should have the article. WP:GNG came later. It is reasonable and certainly something to consider, but is not Holy Writ and we can override whenever we want to, since this is a wiki. Herostratus (talk) 14:10, 25 April 2017 (UTC)[reply]

Well, GNG and your point 1 sorta go together. Cas Liber (talk · contribs) 14:34, 25 April 2017 (UTC)[reply]

I think ignoring the explicit statements within a given guideline, which includes context on how to apply the guideline, is supervoting. Enacting the guideline was a result of long discussions held amongst a relatively (for English Wikipedia) large group of interested parties who were looking at the overall best interest of Wikipedia. If consensus agreements are readily discarded, there isn't much motivation to work towards a common understanding. isaacl (talk) 15:30, 25 April 2017 (UTC)[reply]

I can affirm there was a site-wide RFC run several years ago for NSPORT, no question. But things change. Again, looking at the recent SCHOOLOUTCOMES discussion, while not reaching any actionable steps, shows there's a larger concern about the use of local sources as primary sourcing for a topic, which at the time NSPORT's RFC was run, wasn't as large an issue. --MASEM (t) 15:33, 25 April 2017 (UTC)[reply]

I agree things can change, but they should change as a result of a new general agreement. As the context behind the original approval of the guideline has been reaffirmed several times since then, I don't see any reason why closers of AfDs should take a statement "per NSPORTS" to mean they should ignore the portion of NSPORTS that defers to the general notability guideline as the final word. Before this can be done, a new consensus should be established, and this may mean rewriting the sports-specific criteria, which were not developed with the aim of replacing the general notability guideline. (Note local sources were indeed discussed in the enacting RfC, and the basic criteria has qualifications on how local sources should be used. For example, with hockey players, local sources are regularly discounted as indicative of meeting the standards for inclusion as defined by the general notability guideline.) isaacl (talk) 16:10, 25 April 2017 (UTC)[reply]

While the 2010 RFC did include aspects related to local sources, and passed with that, keep in mind that the broader communicate consensus towards local and more routine sources has changed; NORG has specifically added language towards this and the recent SCHOOLOUTCOMES discussion showed there's less tolerance to articles that rest too much on routine or local sourcing (they're fine to augment significant coverage from broad sources, however). This is a relatively new consideration that would absolutely require some thought towards planning an RFC, including principles for grandfathering and avoiding AFD flooding via fait accompli, and can't be changed just on the basis of this current discussion. But this is a new wrinkle that didn't exist when NSPORTS was passed in 2010. --MASEM (t) 16:22, 25 April 2017 (UTC)[reply]

There has been discussion on the sports-specific notability talk page regarding local coverage, most recently in this thread. There are some enthusiastic editors who feel differently, but from the discussion over the past few years I think most editors are in agreement with what I wrote in that thread regarding sports journalism and local sources. Thus I believe the most interested parties in the sports-specific notability guidelines are in agreement with ensuring adequate sources are present for a sports figure to have an article, and that the quality of sources must be examined carefully to determine if they are suitable for establishing if the standard for inclusion has been met. The problem is English Wikipedia's straw poll approach to evaluating consensus isn't conducive to this type of examination. Closers will mostly look at an initial straw count, and so those pushing for inclusion with shaky sources are at an advantage, since their opinion will get counted with a lot less effort than those who take the time to look into the sources. Plus often what happens is someone dumps the first set of hits from a search engine, a bunch of people pile on say per person X, and person Y goes through the links and explains how they are routine coverage. By strength of argument considerations, the first set of opinions should be given no or little weight, but often the closer just goes by the raw count. isaacl (talk) 19:30, 25 April 2017 (UTC)[reply]

This actually might out there's two separate issues at play: 1) the changing perception of how much reliance the notability of a topic (not just sports-related) can have from local or routine sources, which affects all of WP, and subsequently 2) how this change, if agreed by consensus, impacts the various criteria of NSPORTS. Or another way, to jump at NSPORTS before establishing and documenting a new consensus about local/routine sources is the wrong order. Otherwise, I can certainly understand editors at sports-related projects balking at this undocumented change that would be forced on them. --MASEM (t) 23:55, 25 April 2017 (UTC)[reply]

In addition to the promotional/local interest nature of local coverage, and the entertainment aspect alluded to above, there is also a promotional aspect of sports journalism that is specific to sports. I think most editors understand that not every highschooler covered in a local paper should have an article, and that the same underlying issues with local coverage extend to higher levels of competition. Drawing the line is, as always, challenging. And if closers continue to ignore general consensus in favour of a local head count, any guideline changes will be fruitless. isaacl (talk) 02:13, 26 April 2017 (UTC)[reply]

FWIW, I disagree the guidelines are too inclusive. Plenty of sports have notable players who didn't represent their country nationally. And article quality. There will be loads of stubs of higher notability as well, its about folks actually writing them. One of WPs drawing cards is its inclusiveness so to tamper with that is risky. Cas Liber (talk · contribs) 14:34, 25 April 2017 (UTC)[reply]

Disagree - If a project area with many active participants creates a reasonable notability criterion I don't think we should second guess them without a strong showing of harm. I haven't seen anything close to such a showing here.--agr (talk) 15:00, 25 April 2017 (UTC)[reply]

This applies to several of the more recent comments, but notability guidelines are not meant to be developed in isolation of community expectations. I point to the problem with Mixed Martial Arts (MMA) problem from about 3-4 years ago, where a small group of editors decided on their own notability guidelines, and required a lot of cleanup from that. This is not to say that NSPORTS was developed in isolation, there's enough different projects' eyes on it, but they also add and modify criteria without what should be put to wiki-wide review (I know that there was an RFC years ago for this, but more recently there hasn't). The subject-specific notability guidelines should have roughly equal acceptance across the project, and what this discussion seems to suggest is that there may need to be a better review of these criteria by the community to make sure they follow with other subject-specific guidelines. --MASEM (t) 15:18, 25 April 2017 (UTC)[reply]

Arbitrary break

Looking over all the above, my personal impression -- this is strictly my own personal opinion -- is that there are three classes of arguments for tightning NSPORTS:

Arguments which, while reasonable and I understand where the the person is coming from, sort of come down to personal opinion, don't demonstrate any harm, and so I don't think ultimately will gain much traction, given the number of sports completionist enthusiasts here. These are arguments of the nature "these are too short", "we are too inclusive", "these will never advance beyond stubs", "This is not what the Wikipedia is for" and so forth.
Arguments which make a claim of harm, but a weak one. "It is clutter, and clutter has bad effects in subtle but pervasive ways". "It encourages the wrong sort of editor". "Each article must be watchlisted, and that takes resources", and that sort of thing ( don't see much of this above, but you do hear this.) I consider these to be unproven mostly, or can be reasonably refuted (yes you need more watchlisters, but then then the articles bring in watchers), or even if true are not that big a deal.
One very good argument, which is this: "A lot of these are referenced to a single source or maybe two at most, and you can't add more refs since they don't exist, and a lot of these sources are unreliable or at any rate we can't be confident of their reliability.

(Actually the point "No article should have just a single source, since no source is reliable enough to entirely support an entire article -- it is too dangerous" is also made. Fair point also.)

Speaking for myself, this third point and only this third point makes me sit up and take notice. We do not want to present false data and if NSPORTS is leading us down the path where we are doing so to an unacceptable degree, then that is a problem. Is it true? Yes, it is true in some cases if you take the "we can't be confident of their reliability" part.

I am mostly familiar with baseball, and we use sources like Baseball-Reference.com. Believe me, baseball sources are very reliable. Are they always right? Of course not. The question for any source is "are they wrong seldom enough that, generally, we can consider them reliable to our standard, such that we can be confident that their statements of fact are true, sufficient to put our name and reputation behind such a statement" and Baseball-Reference.com meets that standard.

For other sports, not so much. At least, editors above have vouchsafed that.

I wonder if the solution is not to mess with WP:NSPORTS but to seek out and blacklist the bad sources? Yes blacklisting is severe, but if the source is mainly used for statistics, and we can't trust their statistics, what use are they? Blacklisting some sports sources will cause many sports articles to fall as unsourced of course. What we'd need to know though is what sources are bad, with some indication of that beyond just assertion. This sounds hard, and I wonder if its possible.

Another solution would be rule against single-source articles generally (if a second source can't be found). This has nothing to do with NSPORTs, and would be tricky and easy to game (e. g., add the text "is from Barcelona, a city in Spain" and then a ref that Barcelona is indeed in Spain). So dunno about that. Herostratus (talk) 16:04, 25 April 2017 (UTC)[reply]

Agree But it's no crisis. And just fixing the SNG wouldn't balance it out. Because Sports "coverage" is itself entertainment, sports tends to be "coverage"-heavy in proportion to actual encyclopedic notability as compared to other topics. Someday we can re-write wp:GNG to calibrate itself to such things and then get rid of the SNGs. There's a little project for someone. :-) North8000 (talk) 19:53, 25 April 2017 (UTC)[reply]

Agree Why should the bar for inclusion be so much lower for sportspeople than for other professions? Taking part in athletics does not confer notability to anyone, even if it means they will be forever listed in almanacs of sports statistics. Even competing professionally does not mean there will be coverage of an individual, especially because there are many levels of professionalism and lengths of participation. The requirements for actual coverage of people in multiple reliable sources, not just listings on rosters and event reports in sports-related media, should still be met for sportspeople. This oversaturation of articles about individuals who have done nothing of note is an embarrassment to Wikipedia. Reywas92^Talk 06:38, 26 April 2017 (UTC)[reply]

Agree I do not agree that verifiability and the completionist desire by sports project enthusiasts justify stand-alone articles for every participant in say a dozen competitions a year in a sport which gets little coverage. I do not mind a "one-at-bat" rule for baseball, because the few such unfortunate players probably have some coverage in books and articles about the odd cases of major sports.See a one-game pitcher, for instance. There are unlikely to be reliable sources analyzing the one-time cycling competitors who competed in one of 12 events in a year and finished 99 out of 100. Sport-specific guidelines should not trump GNG. It is inappropriate to close an AFD as "Keep" when the only argument presented is "Satisfies NSPORT." It is more encyclopedic to include the participant in a list than to have a one-line permastub article which only conveys the same information that inclusion in a list would convey. Edison (talk) 18:35, 26 April 2017 (UTC)[reply]

Agree-- NSPORTS is intended to support, not supplant, GNG. Focus on the quality and reliability of the sources. That said, I think that at times NSPORTS is a bit on the permissive side, but I also think that NPROF is often too restrictive. So really, just keep working on GNG to clarify standards for the pre-internet age where sources may be more difficult to locate, be careful not to let cultural bias sneak in (as in excluding an eastern European but not a similarly-situated American), and onward through the fog Montanabw^(talk) 17:37, 27 April 2017 (UTC)[reply]

Disagree-- There's biographical dictionaries and encyclopedias of sports that are dedicated to documenting every single major league player released nearly every year (Where are the dictionaries of professors?), why should we deprive our readers of that knowledge? It will create more problems (Who meets the criteria etc) then it's worth. Also the fact that this will bring more bias to American sports as they are easier to find sources for due to dominant American media coverage. This policy change will result in more Americanization of the project. Having thousands of articles deleted is not in the projects best interest. GuzzyG (talk) 03:04, 28 April 2017 (UTC)[reply]

Comment. In a perfect world, we'd have articles on just about everyone and everything, with access for everyone to reliable sources (even if they're behind a paywall or in a print-only medium) and an endless supply of volunteers to look after them. The reality of it is far different. Anyone who has patrolled recent changes will be aware that a large percentage of edits are to articles on obscure athletes and sports teams. Mousing over the diffs, I see updates to standings and other statistics that I cannot verify easily (if at all), news of trades and injuries and on-field dramas that might be verifiable if I happened to be fluent in the relevant language. Rarely do I see an edit summary. Checking the page histories, I see innumerable unexplained changes from innumerable IPs, occasional evidence of a low-profile edit war between supporters of rival teams, unreverted vandalism by someone who was upset with someone for playing a less-than-stellar game, and so on. And I have to wonder: how can anyone be confident this content is even a little bit sound? There are way more articles about way more athletes (and teams) than we have the resources to curate. I don't have any solution in mind here, but I do think there's a problem. Rivertorch^FIRE_WATER 16:41, 28 April 2017 (UTC)[reply]
Agree Oh yes, this has long irritated me. That someone who once kicked a ball in a single game is presumed notable with very little else being needed, as it often appears, is, to me, ridiculous. LadyofShalott 20:28, 28 April 2017 (UTC)[reply]

I have to agree with that comment. The proposed blacklisting of these unreliable sources(and the subsequent deletion of otherwise unsupported articles via BLPprod/AFD) is, however, a step in the right direction. In a direction where articles not meeting WP:V are no longer kept.
To the other points: We have three places where we have something that should go against this sort of articles that aren't curatable in total. First is the essay WP:Run-of-the-mill, which sums up that sportspeople get more coverage, especially in local newspapers; second is the WP:GNG, specifically the sidenote that routine coverage may not be sufficient to establish notability, something that both these statistics sites and newspaper sources often are, routine. Third, we have the policy WP:What Wikipedia is not; especially WP:NOTNEWSPAPER and WP:INDISCRIMINATE, and the articles that are solely based on newspaper coverage and broad statistics databases should NOT pass WP:NOT.
The question immediately forming itself is:Why do we have so many articles that are being kept when they don't meet these expectations?

There are two possibilities: The first one is that these policies and guidelines are ignored because they are outdated, and no longer apply; at least not to sports articles; they have no consensus anymore. The second one is that we have a WP:Local consensus that keeps working to undermine a global consensus that these pages should actually not be kept. A WP:RFC might help here, however, it might have to be one that starts here, on this noticeboard, with a central notice, and especially without every sports fan going to inform every sports project that their precious articles are in danger, which would be disruptive to the consensus-building as canvassing.Burning Pillar (talk) 20:42, 28 April 2017 (UTC)[reply]

As discussed above, the participants at Wikipedia talk:Notability (sports) largely agree that stats sites and descriptions of games are routine coverage and not indicative of meeting English Wikipedia's standards for having an article. The need to eventually meet the general notability guideline has been reaffirmed multiple times in conversations involving those who aren't sports enthusiasts as well as active members of sports WikiProjects. ("Eventually" is typically interpreted very leniently, though, given that there is no deadline to complete articles.) Thus a local consensus seems to be occurring at AfD where the consensus achieved in broader discussions is getting ignored. isaacl (talk) 20:54, 28 April 2017 (UTC)[reply]

Disagree with changing the current standards. If it ain't broke, etc. Carrite (talk) 04:03, 30 April 2017 (UTC)[reply]
Disagree - the problem with just using WP:GNG is that we'll end up with a lot more work debating whether (at least in the Football world) this player meets GNG and that player meets GNG. It's bad enough with just English sources, with various debates about what is a valid source, and what isn't - it becomes a nightmare using many other languages. Therefore, to simplify it, we need some kind of Bright-line. Our various notability criteria, such as WP:NFOOTBALL is that bright-line, and saves us a lot of work - requiring only focus on those who don't meet it, but still meet WP:GNG, and those rare individuals who do meet it, but have an extreme lack of verifiable sources. This discussion should be whether we abandon the Bright-line, but instead whether we shift or tweak it. Is it in the right place for Football? We certainly have enough players who don't meet it, and yet still meet WP:GNG - so one might argue that it's too high. Or perhaps, it needs to be modified by increasing the minimum number of appearances or something. But that's another debate - and it's a different one for different sports. Meanwhile, what's the damage? There's little indication that material here is false. There's a lot of verifiable information for many athletes who come know where meeting any criteria. There's probably less information that's wrong in a short stub, than a long article on a well-known player. Nfitz (talk) 20:03, 30 April 2017 (UTC)[reply]

Agree -- I've seen a number of sports related AfDs where the discussion went along the way of "Meets NSPORT-X, speedy close! And let's trout the nominator while we are at it." All the while the article contains one sentence, no sources have been presented and it's not at all clear whether they exist in the first place. I think things are at a point where the sports SNG are not in step with GNG; that is: they are not predictive of sources on these subjects actually existing to write an NPOV bio article. K.e.coffman (talk) 05:41, 9 May 2017 (UTC)[reply]
Agree – I created Tom Babson because the guy appeared in Cheers, not because he is an ice hockey coach. I saw a few months ago that it was nominated for deletion last year, indicating how NSPORTS is slowly becoming challenged. And I saw a few years back that "Sam Scarber", whose subject appeared also on Cheers, was created. On the general issue, if the guideline is trimmed or something, the notabilities of sports personalities will be challenged, leading to more and more AfD nominations. Nevertheless, I think WP:notability and WP:V#Notability will be adequate, especially if NSPORTS is deemed redundant. Also, notabilities of those personalities will become determined case-by-case, making NSPORTS less effective anymore. George Ho (talk) 19:10, 9 May 2017 (UTC)[reply]

Comment -- The people who are the subject of these problematic sports stubs just don't randomly appear, but by necessity are pulled into the sport for some reason; for example, another player was sidelined due to injury. This suggests that the stub should be merged into another article, say the one about the missing player. Or the team. Or the year of that sport. Doing so would force an improvement in the article the sub-stub was merged into: the article would have more detail about the fortunes of the team or sport. But if we are faced with a sub-stub about Some Guy who played a single game of Professional Tiddly-winks on 1946 for Lower Elbonia, & there is no article on Professional Tiddly-winks in 1946, let alone one on Professional Tiddly-winks in Lower Elbonia, then there is a larger problem than WP:GNG & WP:NSPORTS conflicting. -- llywrch (talk) 17:08, 15 May 2017 (UTC)[reply]

Agree -- I believe the problem arises from league/competition/championship classifications. E.g. tier 3 soccer (football) league (which is basically one step above amateur league ) is classified as "fully-professional", that's absolutely wrong. Some leagues on this list of fully professional leagues kept by WikiProject Football are not even close to professional level (sponsorship, stadium capacity etc.). Also I noticed that at least 3 people mentioned WP:NFOOTY. I absolutely agree that section 2 of this policy is way too lax: Players who have played, and managers who have managed in a competitive game between two teams from fully-professional leagues, will generally be regarded as notable. Basically, this sentence is implying that you can ignore WP:GNG & WP:NSPORTS as long as the person meets WP:NFOOTY, Jone Rohne Nester (talk) 16:29, 24 May 2017 (UTC)[reply]

Comment That very question was asked today at WP:TEAHOUSE and the answer given was yes, WP:GNG, WP:BASIC, and WP:NSPORTS are all overridden by WP:NFOOTY. MB 21:42, 24 May 2017 (UTC)[reply]
Disagree with a general statement like this. If you have concrete proposals, make them. Otherwise, I can only say that in my experience athletes that meet NSPORTS would be proven notable if editors bothered to access offline sources. ~ Rob₁₃^Talk 17:59, 25 May 2017 (UTC) Moved from the "How to best close this discussion?" section. --George Ho (talk) 17:42, 26 May 2017 (UTC)[reply]

Hi again, BU Rob13. Wasn't your comment supposed to go to the "Arbitrary break" section? George Ho (talk) 08:15, 26 May 2017 (UTC)[reply]

Probably. I'm contributing on mobile so it's difficult to find that location - I just put my comment at the end. Feel free to move it. ~ Rob₁₃^Talk 14:54, 26 May 2017 (UTC)[reply]

Moved as permitted by BU Rob13, for whom I thank him. --George Ho (talk) 17:42, 26 May 2017 (UTC)[reply]

Agree – Look at the list of notable alumni for any US high school or college and it's almost all sports figures who are there for no reason except that they played in some game that qualifies them for NSPORTS. You'd think US schools turn out nothing but athletes. Kendall-K1 (talk) 02:42, 1 June 2017 (UTC)[reply]

Questions:Deletion

RFC: Conflicts between WP:GNG and WP:NSPORTS

No satifactory discussion and consensus has arose w.r.t to imposition of any arbitrary time limit.Broadly, the community is currently opposed to any such timeline. Winged Blades^Godric 14:03, 20 May 2017 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

What should be the maximum time limit for sufficient sources according to WP:GNG to be found before an article that meets WP:NSPORTS is to be deleted at WP:AFD?Burning Pillar (talk) 21:50, 28 April 2017 (UTC)[reply]

Survey

I think one problem is that people keep insisting on WP:There is not deadline when it doesn't work: If (as indicated by a patroller) the number of articles with poor sourcing we have cannot be curated well, then we need to cut down that number. A possibility is to set a time when this article should absolutely have something, because the current approach seems to end with "can be improved" in articles that are 7 years old. I'd set the cut at 3 months after the first AfD nominationor 1 year after page creation, whichever happens first.(The article would still have a last chance time period at WP:AFD, unless speedy deletion criteria apply)Burning Pillar (talk) 21:50, 28 April 2017 (UTC)[reply]

Comment -- this is a rather non-specific question. For all other AfDs, it's usually "put up or shut up" -- i.e. sources must be presented at the AfD for the article to be kept. I don't see why NSPORT would be an exception. K.e.coffman (talk) 05:28, 9 May 2017 (UTC)[reply]

Threaded discussion

It is normally agreed that the issue is that those seeking deletion of a NSPORTS-meeting article that they believe fails WP:N must do the burden of searching for sources to convince others that no such sourcing likely exists (you can't prove a negative completely, obviously). And so it is whatever time it takes for a reasonable search to be completed and the results presented. If you give me an athlete that played for a few seasons since 2010 in a US professional sport (as to meet NSPORTS) but took a life-threating injury and retired to a normal job, I would expect that someone can search for sources within a week, if not less, since most of that will be online. If you give me a similar athlete from 1960, I expect that to take a lot more work since that absolutely requires print source exhaustion. Moreso if we are now talking Indian or Chinese athletes from the same timeframe. --MASEM (t) 22:45, 4 May 2017 (UTC)[reply]
I am opposed to any such arbitrary timeline. There is no reason for such a thing to exist and as others have said it would lead to lots of unnecessary deletion activity. From the prior discussion there is absolutely no consensus for such an approach either. Spanneraol (talk) 22:42, 4 May 2017 (UTC)[reply]
The question is not limited to the timeline I proposed(and the answer could be infinite). However, a WP:BEFORE search only asks you to make a reasonable search. It does NOT require you to look everywhere where sources may exist if it would take undue time and effort. Ultimatively, the sources should be in an article, and if you have only sources not meeting our standards, then the article isn't reliable. And an article meeting WP:NSPORTS with the possibility of sources can get some period of time to allow for longer searches, but if the current approach is followed(articles about old subjects are kept infinitely without meeting WP:GNG, then we allow unreliable articles to persist. And if you compare the number of articles needing improvement with the number of articles getting improval, then you will see a gap. A gap that, by the way, is not only present in sports- rather in nearly every field, and not limited to verifiability, reliability but also neutrality.Burning Pillar (talk) 03:28, 5 May 2017 (UTC)[reply]
How do you know the article isn't reliable? An article sourced only to a dependable stats site is actually reliable and accurate so if your only complaint is reliability most of these articles pass that bar. Spanneraol (talk) 15:15, 5 May 2017 (UTC)[reply]
We should simply adopt NSPORTS, and deprecate the use of the GNG in this field except as an exception (or where there is notability in another field). This is exactly what we do in NPROF, and it avoids most of the arguments. Obviously we still want WP:V. The simpler the better. (I point out that we only keep from having WP flooded by millions of school and locally known amateur athletes by deciding in almost all cases that the sources aren't substantial or discriminating enough. It's amazing the number of undistinguished people in all fields that get covered by local newspapers. DGG ( talk ) 19:11, 5 May 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Speedy deletion criterion X3

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

This is being agreed on as a bad ideaBurning Pillar (talk) 18:41, 2 May 2017 (UTC)(non-admin closure)[reply]

Should the following speedy deletion criterion be created as exceptional criterion; for a limited time?: Articles related to sports that are either unsourced or only sourced via broad statistics databases and/or routine newspaper coverage. This must apply to all versions of the article, and the article must have been created before the 1st May of 2015."Burning Pillar (talk) 21:50, 28 April 2017 (UTC)[reply]

My answer is yes, to remove all these problematic articles that seem to be possibly too numerous to let them all go through AFD without disrupting that process. It might generate a higher number of false positives than normally expected when using a speedy deletion criterion, however. This criterion should stop to be applied after 1 year.Burning Pillar (talk) 21:50, 28 April 2017 (UTC)[reply]

This is a gross misuse of Prod - using speedy to get rid of articles which are sourced would ensure that many salvageable articles are deleted just because some people don't like them.Nigel Ish (talk) 22:32, 28 April 2017 (UTC)[reply]

I'm sorry, user:Nigel Ish, but I don't understand what you mean with "This is a gross misuse of Prod". I also don't know why it is a problem-articles that are sourced may still be speedily deleted under some critera, for example a garage band that is sourced via facebook et al. These sources aren't establishing any notability, so why is that a problem?Burning Pillar (talk) 22:41, 28 April 2017 (UTC)[reply]

I mean that the proposal would be a gross misuse of speedy - to delete articles many of which could be salvageable. These are not articles that do not assert notability. (if so they would qualify for A7) - these are articles that you don't think are notable. Articles that are sourced from facebook etc should NOT be speedied unless they meet one of the very specific criteria, and not meeting GNG isn't one of them.Nigel Ish (talk) 22:51, 28 April 2017 (UTC)[reply]

Strong oppose WP:IDONTLIKEIT is not a reason for deletion, yet alone speedy. If it has an assertion of notability, then it shouldn't qualify for speedy. This seems to make it seem like you're just being lazy and/or know these articles would survive AfD and are trying to go around it. Mass speedy deletion is never good, especially for articles which could be salvageable. Smartyllama (talk) 17:41, 29 April 2017 (UTC)[reply]
Oppose - First off, the RfC/survey above has not even been concluded—oppose votes are roughly equal to support votes. Second off, speedy deletion is to be used for unsalvageable articles. Third off, this would be... very alarming to some experience editors working in these areas, possibly discouraging contribution. Overall, a terrible idea. RileyBugz^会話_投稿記録 17:50, 29 April 2017 (UTC)[reply]
Oppose The only problematic articles are those that fail WP:V. Even those people who don't like the sort of stubs we are discussing, if they care about the project at all, should be looking at merging them into list articles. --Michig (talk) 18:59, 29 April 2017 (UTC)[reply]

Magdalena Zamolska has been deleted. The reason -- which I'm not necessarily saying is wrong -- was more or less "Well, it was a lively discussion, and the votes were about equal, but in terms of strength of argument, nobody made a credible assertion that she meets WP:GNG, therefore deleted on policy grounds".

OK, this is interesting. By my count the vote was 10-6 in favor of Keep (the closer characterized that as "roughly equal", which, maybe). It is extremely usual (although certainly not universal) for a result like that to be closed as "no consensus" if not outright "keep". But as the closer pointed out WP:CONSENSUS says "Consensus is ascertained by the quality of the arguments given on the various sides of an issue, as viewed through the lens of Wikipedia policy" (which BTW shows that "consensus" as used here is quite different from the normal dictionary definition of the word). And "does not meet GNG" was certainly true -- there's no question of that, since there was not "significant coverage". There was not one sentence on where she went to school, or what her class background was, or her path to becoming an athlete, or if she is married, or what her opinion on state of sports in Poland is, or anything. Just raw statistics, essentially.

Bot on the other hand... we are a community. 10-6. That's quite a poke in the eye to a segment of our community. And GNG is not technically a policy -- it isn't labeled as such, and hasn't gone through the process by which guidelines are elevated to policy. But the quote by the closer was "...viewed through the lens of Wikipedia policy".

Anyway, this suggests that as practical matter GNG has attained the status of NPOV, BLP, V, and other core principles as policy de facto. In my personal opinion this is a new thing and an unwelcome sea change.

But it is what it is. If that's what it is, then adding a Speedy Deletion criteria "Incontrovertibly does not meet GNG, by any reasonable interpretation of 'significant coverage', and is judged unlikely to be able to do so with a reasonable level of effort" or something. We could certainly clear out a lot articles quickly that way.

Wouldn't this be logical? If an article is AfD'd, and the vote is 17-2 to keep, but it clearly doesn't meet GNG and can't, and so is going to be deleted anyway -- what's the point of a long discussion?

That's a separate discussion, to be had over at Wikipedia:Criteria for speedy deletion. And probably at Wikipedia:Notability should be proposed to be formally promoted to policy, or perhaps WP:GNG split off and just that promoted.

At any rate we can possibly cap this discussion -- if NFOOTY and the rest are not going to be considered, what are we even talking about here? Herostratus (talk) 18:34, 29 April 2017 (UTC)[reply]

A guideline is only ever a rule of thumb (and should never be more than that), and consensus within an AfD discussion, as long as people are not putting forward arguments that contradict policy, should never be overridden because of a line in a guideline. I think that's a very poor close. --Michig (talk) 18:59, 29 April 2017 (UTC)[reply]

I have mixed feelings (I voted to delete BTW). Closing is hard (its amazing we get smart people to do this stuff for free). Going the other way could have been criticized too. Anyway what matters is not what you or I think but what closers, as a class, are going to to generally. If this is the trend... Herostratus (talk) 21:16, 29 April 2017 (UTC)[reply]

The purpose of a discussion is to verify that indeed the article doesn't meet the GNG. This is why I oppose the notion of a temporary speedy deletion criterion. In my book the AFD for Magdalena Zamolska has gone exactly as it should have; when you take away the bolded words in that discussion and don't treat it as a vote, it's clear that nobody was able to demonstrate that the article met our criteria. Lankiveil ^{(speak to me)} 04:28, 30 April 2017 (UTC).[reply]

Besides, if anyone feels the AfD was closed incorrectly, they are free to use WP:CLOSECHALLENGE. — Gestrid (talk) 04:31, 30 April 2017 (UTC)[reply]

I have left a message on the closer's talk page. We'll see what happens. Smartyllama (talk) 15:10, 1 May 2017 (UTC)[reply]

That was an excellent close from Scottywong, who (correctly) judged consensus based on policy-based arguments and not a vote count. The keep voters, while numerous, completely failed to make a case that Zamolska meets GNG; many of them even seemed to regard meeting GNG as optional, which is the exact opposite of what NSPORT says and therefore undermined their appeals to NSPORT. Sideways713 (talk) 23:13, 1 May 2017 (UTC)[reply]

Opposed - Disimprovement of the encyclopedia. GNG works, leave well enough alone. Carrite (talk) 04:06, 30 April 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Tightening up NCYC

Trying one small step at a time, I proposed tightening of NCYC with detailed rationale why I think that particular subguideline needs rewriting, at Wikipedia_talk:WikiProject_Cycling#Tightening_up_of_WP:NCYC. Sadly, it has been 48h and nobody commented there - I guess the project is sem-active. Perhaps the discussion needs more light, so I am posting it here (and feel encouraged to repost the link to any other relevant forum). --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:38, 4 May 2017 (UTC)[reply]

Changes to WP:NSPORTS

I changed two parts of WP:NSPORTS, one after a discussion of one day, one without discussion(that one seemed obvious to me). I was quickly reverted, and the comments were very negative. It was pointed out that any substantial change should be discussed througly before to avoid changes that are not welcomed by the community as a whole. I have to agree with that. Now, I know that there are definitely problems with WP:NSPORT, so I quickly checked if someone else had done such things in the past. The substantial changes are:

Resulting from a WT:NSPORTS discussion:[2];[3](both)
Resulting from a discussion of a WP:WikiProject:[4];[5];(both)
Changes without discussion on WT:NSPORTS without attribution to a WikiProject:[6];[7];[8];[9];[10]

These are the changes of one year. Quickly reverted changes are not included.Two (significant) changes were discussed at WT:NSPORTS only, two were discussed at a Wikiproject only, one was discussed at both, and five changes seem to be undiscussed. Burning Pillar (talk) 20:57, 13 May 2017 (UTC)[reply]

Discussion is still ongoing - you need to wait for some sort of consensus to form in the discussion before trying to change the guidelines if you want your changes to be accepted by the community.Nigel Ish (talk) 21:17, 13 May 2017 (UTC)[reply]

Those changes I proposed are under discussion, and I agree with you. I just wanted to share this because there were concerns about undiscussed changes, and there are some that are indeed undiscussed changes(and those that were not discussed at the right talk page).. And this could be related to the discussion above.Burning Pillar (talk) 22:09, 13 May 2017 (UTC)[reply]

Thank you for the analysis. I find edits like [11] quite problematic, as they are likely edits made by fans of a certain sport, pushing for yet more inclusiveness, seemingly without discussion. And attempts to revert such changes later are criticized as 'no consensus', even when there often is 'no consensus' to broaden the criteria in the first place. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:08, 17 May 2017 (UTC)[reply]

@Piotrus: The problem you uncover is another - a part of a guideline or policy should have to get consensus to be added, but also to be retained, after all, guidelines and policy should have strong community support, so the burden to get consensus for the part of the guideline/policy should be on those who want the text to be in the guideline or policy. Oh, and you know that there is something wrong with a special notability guideline that states "Motorsport figures are presumed notable if they[...]2.Predate the sharp distinction between professional and amateur (prior to World War II)."(this has been in the guideline for more than a YEAR) And no, this sentence is even being defended on the talk page, even if it should be obvious that this criterion cannot have been developed with WP:GNG in mind. In addition to shifting the burden on those who want something in the guideline, we probably have to make sure that proposed changes are scrutinized well enough, which just does not seem to happen(e.g your example above, and that sentence I provided). There are probably some methods to assure that. One of them would be full protection of WP:Notability (sports), another would be requiring consensus on the talk page, in a RFC or here, before the change is to be made, for any substantial change and for any change that someone assumes to be substantial. I hope there are more, less problematic but still effective methods.Burning Pillar (talk) 00:25, 18 May 2017 (UTC)[reply]

How to best close this discussion?

Before asking a request for closure, I just want to be sure how complex or simple the issue is before requesting it in one venue, either WP:ANRFC or WP:AN. Seems that the issue is notability of sports personalities and the inclusivity criteria of NSPORTS. I listed this discussion in the "Centralized discussion" template. Numerically, the consensus appears divided. However, if the issue is more complex and requires more than one closer, I would go to WP:AN to request teamwork closure. Thoughts? --George Ho (talk) 03:29, 17 May 2017 (UTC)[reply]

It is too soon to close this discussion. It is still active, and a new section was started that generated recent comments.Burning Pillar (talk) 00:01, 18 May 2017 (UTC)[reply]

No, I think it should be closed... The conversation has too many threads and its clear that nothing will get resolved from this as there just is not consensus for change.. or any real concrete proposals. Spanneraol (talk) 18:00, 19 May 2017 (UTC)[reply]

Okay, Spanneraol. Is just one closer okay, or do you think more closers are needed? --George Ho (talk) 19:16, 19 May 2017 (UTC)[reply]

I don't see why we would need multiple ones. Spanneraol (talk) 19:49, 19 May 2017 (UTC)[reply]

Two closers might be needed, just to check for consensus, as whether there is a lack of consensus or not might be a problem. RileyBugz^会話_投稿記録 21:43, 21 May 2017 (UTC)[reply]

I requested closure at WP:ANRFC and then WP:AN three days later, just in case. --George Ho (talk) 18:49, 22 May 2017 (UTC)[reply]

Update: Tazerdadog says that he'll be one of the closers. If necessary, at least one more may be needed. --George Ho (talk) 02:52, 6 June 2017 (UTC)[reply]

I don't see any need for two closers. isaacl (talk) 03:16, 6 June 2017 (UTC)[reply]

I don't know. 1) The one who started the thread was very undecided on how to request a closure. 2) The votes are split, and the arguments of both sides may be logical. 3) It was listed in the "centralized discussion" template. 4) The discussion got very lengthy (if not heated), and two subsequent proposals failed. 5) It's been nearly two months since the discussion started (including nearly three weeks after first 30 days), indicating that it's not an easy close yet. 6) Amending the NSPORTS is still on the works. 7) Two above editors disagree on how many closers are needed, i.e. one or two closers. 8) A closing rationale may affect multiple biographical articles of sports personalities. --George Ho (talk) 04:17, 6 June 2017 (UTC); amended. 04:24, 6 June 2017 (UTC)[reply]

Forgot #9: One exemplified article shown in the OP was deleted per AFD; the deletion review ended with "no consensus" defaulting to "deleted". George Ho (talk) 04:29, 6 June 2017 (UTC)[reply]

Whether it demands two editors or not, I've started working with Tazerdadog on a possible close. Eggishorn (talk) (contrib) 04:31, 6 June 2017 (UTC)[reply]

Cool. Two should suffice if third closer is unnecessary. Okay. I'm awaiting the closure then. George Ho (talk) 05:04, 6 June 2017 (UTC)[reply]

These are common characteristics of many discussions with dissenting views that have been closed by a single closer (because nearly all discussions have been closed by a single closer). One person saying, hey, maybe two closers would be useful, doesn't mean that two closers are needed. Two closers don't necessarily help discussions get closed faster (the two co-ordinating can in fact slow things down); discussions often remain unclosed because they are lengthy with a lot of views to dig through. isaacl (talk) 12:44, 6 June 2017 (UTC)[reply]

Sure, they may be common for most discussions. However, the topic isn't about one article or another. It's about a bunch of them and their notabilities. Also, the consistency between Notability guideline (longstanding one) and NSPORTS has been questioned here. I don't know whether you believe the matter is simple to close and resolve. Well, no point on arguing further whether a second closer is necessary; we already have two closers working on the closure. Moreover, I'm not seeking a rushed closure but an accurate one because, as you also said, arguments have to be thorough checked, not missed. BTW, the person said earlier, "Two closers might be needed." George Ho (talk) 14:55, 6 June 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Labeling illegal (or undocumented) immigrants correctly

Consensus is not to have a policy or guideline requiring the use of "illegal alien" or "illegal immigrant".

The terminology should be decided on a case by case basis, and the argument should be about what term reliable sources use to refer to the persons in question and not based on the terms used in law.

And there is a weak auxiliary consensus against using the word "alien".

Bryan Henderson (giraffedata) (talk) 02:25, 5 June 2017 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

This has been a term that has been debated heavily on Wikipedia forever. Whether or not to call them "illegal immigrants" or "undocumented immigrants". I propose that we settle this debate by creating a policy or a guideline that requires all pages to either use the term "illegal alien" or "illegal immigrant" given that this is the term used by the US government. as seen in Arizona v. United States. Federal district court Judge Andrew Hanen of the Southern District of Texas was quoted as saying: The Court uses the phrases ‘illegal immigrant’ and ‘illegal alien’ interchangeably…The Court also understands that there is a certain segment of the population that finds the phrase ‘illegal alien’ offensive. The Court uses this term because it is the term used by the Supreme Court in its latest pronouncement pertaining to this area of the law. See Arizona v. United States, 132 S. Ct. 2492 (2012). I am open to amending the details of my proposal and encourage anyone to think before opposing it and suggest how to make it acceptable. THE DIAZ ^{talk • contribs} 12:52, 5 May 2017 (UTC)[reply]

See Antonio Vargas history and this now-removed comment on my Talk page. On the merits, this proposal is fatuous as what term to use depends on the article and the context. It would be impossible to have even a guideline creating a presumption as to what term should be used.--Bbb23 (talk) 13:03, 5 May 2017 (UTC)[reply]

@The Diaz: Snarky but not snarky question: You are aware that Wikipedia exists outside of Texas or indeed the United States, right? --NeilN ^{talk to me} 13:42, 5 May 2017 (UTC)[reply]

Illegal immigration to Australia cause some debates in Australia about what to call the people. And Wikipedia has a whole Category:Illegal immigration by country on the topic of other countries. Mitch Ames (talk) 13:58, 5 May 2017 (UTC)[reply]

The BBC has actually gone into this (language RE migrants) in some detail in the past, both covering others and how they use it themselves. You will often find disclaimers about why they use a particular phrasing on their online articles in the area. One of the related articles is here. FYI - almost no one uses the word Alien anymore except the US. Only in death does duty end (talk) 14:08, 5 May 2017 (UTC)[reply]
The pronouncement of one court in one case should have little if any bearing on Wikipedia, which should be based on the usage demonstrated through a preponderance of reliable secondary sources. What term is used by newspapers, journals, magazines, reliable books, etc. on the subject? If there is a single term which the preponderance of them use, we should go with that. If there appears to be interchangable terms, then we default to WP:ENGVAR guidance. --Jayron 32 14:13, 5 May 2017 (UTC)[reply]

The BBC does it depending on their (confirmed) status. They are otherwise 'migrants'. (eg, someone jumping off the boat from Cuba may be an economic migrant, a refugee/asylum seeker, an illegal immigrant etc, if there is no confirmed status it just reverts to 'migrant' as a neutral word to denote someone crossing borders - migrating from country to country. It does mean they tend to get blasted from both ends of the political spectrum, the right have a go at them for not labelling them the filthy illegals they are, whilst the left complain they are not labelling obvious refugees as such. Only in death does duty end (talk) 14:25, 5 May 2017 (UTC)[reply]

Whether to use "illegal immigrant" , "undocumented immigrant", "migrant", etc, is as Jayron points out something that should be determined topic by topic based on the common usage in reporting sources. However, I would agree we should avoid "illegal alien" and use "illegal immigrant" instead. "Alien" is not as wrong but has the same negative connotation as something like "negro", as well as for non-native English speakers, potentially confusing, and "immigrant" works as an equivalent replacement without loss of meaning. --MASEM (t) 14:33, 5 May 2017 (UTC)[reply]

I think the word "alien" applies to the legal documents/status. - Knowledgekid87 (talk) 14:38, 5 May 2017 (UTC)[reply]

I do know it is a term defined by law in some areas (US for example it is anyone not a citizen or national), and not meant derogatorily there, but I've seen it used in media opinions in a more derogatory way. More often, though, when we are talking what the media says about these cases, it is a very uncommonly used word over "immigrant". --MASEM (t) 15:06, 5 May 2017 (UTC)[reply]

I am in agreement with Jayron32 on going by what the reliable source calls them. For comparison there is also Native Americans, American Indians, or just Indians. - Knowledgekid87 (talk) 14:36, 5 May 2017 (UTC)[reply]

The first post in the thread says in part "Whether or not to call them "illegal immigrants" or "undocumented immigrants". [Emphasis added]. The correct term all depends on who they are. A person who goes illegally to a country with the intent of staying indefinitely is a illegal immigrant. The term "illegal alien" would cover illegal immigrants, and also those who only intend to stay for a brief visit, and those who's intent can't be determined. An undocumented immigrant could be an illegal immigrant, but cold also be a foreigner who is entitled to be in the country, but who's documents are lost, stolen, in transit, etc.

WP:NOTFORUM

As a notary, I've dealt with a substantial number of undocumented persons. These persons give the appearance of being not only citizens, but members of families who have been living in my state for many generations. But they couldn't get regular driver licenses and had to settle for driving privilege cards instead. I'm sure nearly all of them could cure their undocumented status if they put their mind to it; if they're citizens, they are under no legal obligation to possess evidence of citizenship (although such a lack is inconvenient from time to time). Jc3s5h (talk) 14:54, 5 May 2017 (UTC)[reply]

Best to have no rulebook on this whether it's the Wikipedia or Political Correctness rulebook. E.G. calling drug dealers "undocumented pharmacists" :-) Since the PC rulebook rears its head by default, a positive "use the most accurate, descriptive term" statement would be good. Then the most accurate relevant term for the context will be used which is "illegal". And usually the actual context for the second word would dictate "immigrant" and only occasionally "alien".....the distinctions on that second terms are minor enough to give flexibility there. For all of the other wilder stuff we don't censor, we should certainly not PC ness to informally censor the most accurate, informative terms. North8000 (talk) 15:10, 5 May 2017 (UTC)[reply]
Odd. Who are you arguing with? No one here mentioned political correctness as a rationale before you brought it up. The most accurate, informative term is the one used in a preponderance of reliable, independent, secondary sources. --Jayron 32 16:56, 5 May 2017 (UTC)[reply]
It's also a straman since no one has ever suggested using terms like undocumented pharmacists to describe drug dealers.--64.229.167.158 (talk) 19:56, 7 May 2017 (UTC)[reply]

North8000, come on man. Let's not have this nonsense about supposed PC speak here; that tune is getting old. Drmies (talk) 16:23, 8 May 2017 (UTC)[reply]

My "undocumented pharmacists" was a sidebar chide on the choice of words. On the main thread, I think that nearly all of the time the "undocumented" term is used, it is in a sentence is about illegal presence. Generally something that is illegal is called illegal (especially if that status is germane to the statement of the sentence) even it were possible to make it legal by the right permit or paperwork. My point is what, what in this case would dictate many to avoid the more direct, informative term and substitute a vague euphemism?... and I think that the answer is the PC rulebook. North8000 (talk) 18:21, 8 May 2017 (UTC)[reply]

How about just making it a policy or guideline for any topic relating to illegal immigration in the United States? THE DIAZ ^{talk • contribs} 15:44, 5 May 2017 (UTC)[reply]

Instruction creep... not needed. Blueboar (talk) 18:19, 5 May 2017 (UTC)[reply]

@Blueboar: What is an instruction creep? THE DIAZ ^{talk • contribs} 21:10, 22 May 2017 (UTC)[reply]

See: WP:Avoid instruction creep. Blueboar (talk) 21:30, 22 May 2017 (UTC)[reply]

Actually, a/c general BLP principles, we can't use "illegal" unless there is a RS that they have been found to be illegal by a court of comp tent jurisdiction. That they are expelled from a country (much less not admitted) does not necessarily mean that a court has formally ruled. "undocumented" avoids some of the possible difficulties here. DGG ( talk ) 19:16, 5 May 2017 (UTC)[reply]

@DGG: Actually, the term "illegal immigrant" or "illegal alien" are the terms for an alien who enters the U.S. illegally. It's a fact that the Supreme Court uses that term regardless of any immigration court ruling, as their purview is only to declare an immigrant or alien "removable". THE DIAZ ^{talk • contribs} 21:10, 22 May 2017 (UTC)[reply]

No, let's not codify our contributions to the marginalization of The Other. When we're directly quoting a source, we obviously go with whatever terminology the source uses. In Wikipedia's voice, "migrant," "immigrant", and "emigrant" are all reasonable choices, depending on context, but "migrant" is safest because it works in either instance, coming or going. "Alien" is either legal jargon or pejorative, so that's definitely never optimal (except in a direct quote). In many cases, I suspect that preceding it with an adjective, whether "undocumented" or "illegal", is quite unnecessary. Both are primarily used in American English and may be confusing to readers abroad, but if one or the other must be used in Wikipedia's voice, "undocumented" should be preferred because it lacks the negative connotations of "illegal" (which, when used by itself as a noun, is derogatory). Rivertorch^FIRE_WATER 04:45, 6 May 2017 (UTC)[reply]
"Illegal alien" just sounds weird, That aside we should use whatever the sources state which would be the easiest way of doing it (Personally I prefer illegal immigrant" however it's not about preferring it's about what sources etc use). –Davey2010^Talk 16:51, 26 May 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

SCHOOLOUTCOMES again

A couple months ago we had an RfC about WP:SCHOOLOUTCOMES. About 100 editors participated and it was closed by four editors in good standing. In the closing statement is a summary that includes this text: "Secondary schools are not presumed to be notable simply because they exist."

Today, I was disappointed to come across Wikipedia:Articles for deletion/Korea Kent Foreign School, which looks like it's heading towards being kept based almost entirely on arguments along the lines of ~"there's a long-held consensus that we keep articles about secondary schools if they verifiably exist".

Putting aside the specifics of that AfD, if school AfDs are primarily attended by those who see the February RfC as illegitimate or otherwise don't agree with or don't like the close, what purpose did the RfC have? (That's a real question, not simply rhetorical). — Rhododendrites ^talk \\ 20:15, 6 May 2017 (UTC)[reply]

The outcome of that RfC was no consensus to the actual question asked as at least one of the closers (Primefac) has stated when asked about it. The rest of the close was an observation of the implications of no consensus. The close was difficult, and I think muddied the waters, to be honest (not intended as a critique of the closers). SCHOOLOUTCOMES now shouldn't be cited, but you have admins saying that referencing school outcomes and then explaining the reasons behind it is a fine argument since that makes it cease to be circular. From a practical standpoint, this leaves us exactly the same place we were before the RfC, which is what a finding of no consensus is in the end. TonyBallioni (talk) 20:37, 6 May 2017 (UTC)[reply]

Economical with the verité or what? Primefac also said that the arguments being used by admins like Necrothesp - which are identical to the ones Rhondodendrites mentions - are a de facto invocation of SCHOOLOUTCOMES and are not valid. This end-running by admins in particular needs to stop. I am particularly appalled that DGG is attempting the trick, even though he does generally have a reputation as an inclusionist (eg: extremely broad application of SYSTEMIC). - Sitush (talk) 20:53, 6 May 2017 (UTC)[reply]

Sitush I wasn't trying to give the impression that Primefac was endorsing the SCHOOLOUTCOMES arguments (its why I stated that it shouldn't be cited, which I am in agreement with you on.) I was simply noting that the part of the RfC that is very rarely mentioned is that the close to the actual question asked was no consensus. This basically leaves us where we were before: fighting over the notability of schools in South Asia and Africa on a case by case basis.. TonyBallioni (talk) 21:01, 6 May 2017 (UTC)[reply]

Fair enough, sorry. But I'm still pissed that people who should know better are in fact citing it in all but name. It's about time we scrapped all these notions of presumed notability and enforced WP:BURDEN, ie "use it or lose it". - Sitush (talk) 21:17, 6 May 2017 (UTC)[reply]

I've said it before, and I'll say it again now... instead of a "presumption" of notability our SNGs should talk about the "likeliness" of notability. Blueboar (talk) 21:27, 6 May 2017 (UTC)[reply]

Which will just enable even more crap that no-one has time to maintain and which in many cases is fundamentally just promotion. - Sitush (talk) 21:30, 6 May 2017 (UTC)[reply]

SCHOOLOUTCOMES isn't even a SNG, the RfC tried to effectively tried to elevate it to one, but that failed. It is a description of the typical outcome of AfDs. I've said this before, but I think there is going to be uncertainty on what to do with secondary schools for a while, and then there will be a de facto compromise reached at AfD, which will then likely cause OUTCOMES to be updated. DGG's argument for that is strong, in my opinion, but the compromise will probably be different than what we had seen pre-RfC. TonyBallioni (talk) 21:38, 6 May 2017 (UTC)[reply]

DGG's argument is usually that something might turn up. It's not strong at all. Take that road and we may as well not delete anything on notability grounds. - Sitush (talk) 21:44, 6 May 2017 (UTC)[reply]

We need closers of such AFDs to recognize that one clear outcome of that SCHOOLSOUTCOME is that we can't have the Catch 22 of using SCHOOLOUTCOMES as the only !vote reason (WP:ATA was updated to include this). It doesn't meant secondary schools can't be kept, but the strength of argument has to be towards sourcing and notability (but it doesn't have to be fully GNG-meeting) and not just regurgitating SCHOOLOUTCOMES. I wonder if we need a template message for such deletion debates to remind editors that !voting only on a SCHOOLOUTCOMES-based argument should be discounted by the closer. --MASEM (t) 14:27, 9 May 2017 (UTC)[reply]

I feel like Wikipedia needs *some* policy regarding SCHOOLOUTCOMES. I am opposed to the status-quo of "ignore this policy entirely". For example, would evidence of notable alumni be sufficient for notability? Proof of enrollment? There are very few online newspapers from Nigeria; a requirement for that type of coverage may cause needless deletions. Power~enwiki (talk) 01:46, 23 May 2017 (UTC)[reply]

@Power-enwiki: SCHOOLOUTCOMES is not a policy, nor a guideline, so there is no policy to ignore entirely. Please do not mis-state it as such. --Izno (talk) 03:07, 23 May 2017 (UTC)[reply]

@Izno: then what is it? If it's nothing, it should be deleted entirely, or it should be replaced with an actual policy or guideline. Power~enwiki (talk) 03:56, 23 May 2017 (UTC)[reply]

@Power~enwiki: The page itself says it is a supplement, but neither policy nor guideline. Why do you think it should be deleted entirely if it is not replaced by a PAG? --Izno (talk) 12:36, 23 May 2017 (UTC)[reply]

Delete SCHOOLOUTCOMES and replace with an actual guideline - If the school doesn't have reliable and indepth sources of any kind (regardless of the country) then it should be deleted, We're strict with everything else on this site so we should be strict with Schools too. –Davey2010^Talk 16:46, 26 May 2017 (UTC)[reply]
- We already have a guideline for schools and other educational organizations; it's at WP:NSCHOOL. WhatamIdoing (talk) 00:00, 12 June 2017 (UTC)[reply]
  - That's not a guideline for schools, it's a mere pointer to general guidelines. It is also circular, as it includes a link to WP:ORG, the guideline of which it is a section (albeit a single sentence). As far as it applies to schools, WP:ORG does no more than restate WP:GNG. Ultimately, NSCHOOL is a mere informational statement (just like SCHOOLOUTCOMES). We would lose nothing if it was removed. Jack N. Stock (talk) 06:22, 12 June 2017 (UTC)[reply]

pre-RFC on user talk page archival

Okay, so I'm not going to put this up as an official RFC because I might get thrown bodily out of the room, but I wanted to gauge people's opinions on user talk page archiving. While I know it is entirely up to individual editors when and if to archive their pages, but I have seen multiple times on multiple talk pages recently where there are hundreds of sections and some old discussions going back to 2010. Maybe it's just me, but when you get user talk pages reaching 400k (or even more ridiculous at over 900k) it gets hard, not only for the processor to load, but to read through. I'm on a fairly new computer, and it still throws a hissy fit any time I visit these pages.

Now obviously, there are some people whose talk pages are huge because they're heavily involved in areas where they simply get a lot of commentary on their talk pages, and that's not really what I'm concerned about. I'm concerned about the people who are apparently unconcerned with ridiculously long pages.

So here's my general question: should we force users to archive their talk pages if and when they reach a certain length? For the sake of this argument, let's say 200k, since that seems to be about as long as I can find of people who do have archiving. Primefac (talk) 17:18, 22 May 2017 (UTC)[reply]

Attaching "must do's" to userspace is something we should be very careful with. There are obviously necessary limits on userspace for the safety and security of the project that I think nearly everyone accepts: Don't have hate speech on a user page, don't have copyright violations, don't have libelous material, etc. We need those restrictions but there's a philosophical difference between "you can't do these things" and "you must do these things." I think the closest we have to the latter we have currently is leaving declined unblock requests and a few other notices (Arbcom, some deletion ones) on the user talk pages, but even that is a restriction that actually says: "A number of important matters may not be removed by the user..." It goes back to preventing a user from taking an action rather than imposing an action that the user must perform. This proposal would be the first example of the latter category I'm aware of and that means it breaks a lot of new ground. Eggishorn (talk) (contrib) 17:44, 22 May 2017 (UTC)[reply]

I think a lot of users probably don't know how to archive pages or don't think it's worth the time. Forcing them to do so would just make them mad.. Perhaps a friendly note about archiving with some helpful instructions would be better? Spanneraol (talk) 17:49, 22 May 2017 (UTC)[reply]

I agree, so long as it's bot driven and not involving any effort on the user's part. We often create policy in order to make Wikipedia accessible to all, I don't see why this should be an exception as it cripples browsers (mine included). In general, I feel it would be a courtesy to the user as archiving either manually or bot-assisted can be a bit of a hassle to some. Drewmutt (^ᴥ^) talk 17:53, 22 May 2017 (UTC)[reply]

Perhaps an essay describing the benefits of archiving and how to set it up would be the way to go. Then, if a talk page is getting overly long, we would have something to point to. Encourage rather than force. Blueboar (talk) 17:57, 22 May 2017 (UTC)[reply]
I'm genuinely curious how your browser struggles with something like this. Stock Chrome reports the page requires 2.2 MiB of data and a mere 83 MiB of memory. By comparison, scrolling past a dozen posts in my Facebook newsfeed fetches 4 MiB of data and requires 118 MiB of RAM; CNN's home page necessitates a whopping 5.4 MiB data transfer and fills 245 MiB of RAM for the tab. My tentative feeling is that a user whose talk page is already substantially leaner than many contemporary websites should not be forced to archive it; 900 KiB is not an unreasonable amount of data in 2017. Rebb ing 17:58, 22 May 2017 (UTC)[reply]
- PCs, sure, but iPads? Tablets? Phones? People don't just browse on PCs these days. There are semi-weekly complaints that WP:ANI is getting too long, and some of these talk pages are longer than ANI. Primefac (talk) 18:37, 22 May 2017 (UTC)[reply]
- User:Rebbing, that is true, and probably the reason the Facebook crashes my browser every time I visit it, whereas, with a few recent exceptions, Wikipedia behaves well. All the best: Rich Farmbrough, 23:16, 10 June 2017 (UTC).[reply]
I'd quite strongly oppose enforced archiving. While I agree it's an issue (and have recently commented at length on the talkpage I assume prompted this proposal to that effect), for third parties to try to second-guess which threads the editor will no longer keep is a recipe for endless revert-wars and frayed tempers. Unless it's set so ridiculously high as to be useless, a bot-enforced maximum talkpage size will be completely impractical; as a straightforward example my talkpage is currently 170,000kb in length and generates a server load more than twice as heavy as EEng's, but that's because there's one very long thread currently there which contains multiple large images, and an enforced bot archiving to bring it below the recommended 75kb would have the perverse effect of removing the one thread currently on my talkpage in which people are still currently commenting. Some situations just don't lend themselves to firm rules, and this is one of them; the way to force people to keep talkpages to manageable sizes is through social pressure and the hope that if enough people say "this causes problems for some people and keeping it doesn't have any obvious benefits to outweigh those problems", the editors in question will at some point decide that they don't want to be known as "that jerk with the disruptive talk page". ‑ Iridescent 19:08, 22 May 2017 (UTC)[reply]
- I must be missing something obvious here, but yours is currently 154k (I guess the bot came by since you posted) and EEng's over 900k according to the page information. That translates to 277k and 1.6M HTML, respectively. Opabinia regalis (talk) 19:35, 23 May 2017 (UTC)[reply]
  - @Opabinia regalis In terms of bandwidth rather than just "nuisance to scroll through" it's not the text that causes issues on big pages, it's images and templates. (These are the actual bandwidth loads broken down by component for EEng's talk page and my talk page. Note that for my talk, virtually all the load comes from downloading the two big screen-captures.) You probably don't see it because you've visited my talkpage repeatedly between the images being added, so they're in your cache and not being downloaded each time. The easiest way to see how long a page will take to load for someone coming to it fresh is to open the page in the Wikitext editor (the real one, not the love-child-of-VE experimental new one) and select "preview"; because of how MediaWiki operates that forces your browser to re-load all templates and images from scratch rather than using the cache. (With images, bear in mind that except for a few exceptions like animated gifs it's not the size of the file that causes issues, it's the size of the displayed image as the actual image processing is done server-side; a tiny file displayed at 300x200px will create more load than an 18Mb high-definition scan displayed at 200x150px. If you really want to fry your browser, go to a high-def architectural blueprint like File:Noel Park plans.jpg and compare the time it takes to load the thumbnail with the time it takes to load the full-size image.) ‑ Iridescent 20:45, 23 May 2017 (UTC)[reply]
    - You're right, I forgot about the images. I also forgot about your randomized top image, till I just now got the cat video. Unrelatedly, I kinda like the love child of VE.... Opabinia regalis (talk) 22:52, 23 May 2017 (UTC)[reply]
I may support it provided a provision that this excludes threads which have been active in the past week - that is, if the threads which were active within the last 168 hours add up to a very long total, you don't need to have it archived. עוד מישהו Od Mishehu 07:35, 23 May 2017 (UTC)[reply]
I'd probably oppose a 'mandated archiving' solution, but just for the record, holy shit is this annoying. I catch up on wiki stuff on my phone a lot and I really have no interest in using my mobile data to download someone's talk page threads from 2010. Rebbing isn't wrong that these are not unreasonable amounts of data, but they are unreasonable amounts of unwanted and irrelevant data. I already have a mental list of "those jerks with the disruptive talk pages" and really, I want those neurons back. Opabinia regalis (talk) 19:35, 23 May 2017 (UTC)[reply]
- For whatever it's worth, I much prefer talk pages (user and article) that are kept tidy and somewhat relevant. I have the bot set to keep mine under fifteen threads; even with pictures (I probably should archive those), it weighs in at only 229 KiB. Rebb ing 20:03, 23 May 2017 (UTC)[reply]
  - OR, not wanting to commit lèse-majesté but I feel I ought to point out that by far the worst offender for "volume of irrelevant crap on a user page" is one of your esteemed fellow arbs, whose talkpage is currently hosting considerably more threads than does EEng's, and doesn't even have EEng's virtue of occasionally being interesting or entertaining. ‑ Iridescent 20:52, 23 May 2017 (UTC)[reply]
    - Who reads arbs' talk pages? There are arbs there! Well, except Drmies; everybody reads his talk page. Opabinia regalis (talk) 22:52, 23 May 2017 (UTC)[reply]
      - Drmies isn't actually top among the current arbcom. Why anyone would want to read DGG's talkpage when there's perfectly good paint to watch dry, I leave as an exercise for the reader. ‑ Iridescent 22:59, 23 May 2017 (UTC)[reply]
        Hmm I thought I was more popular than DGG. Far be it from me to criticize my fellow arbs, but I agree that DGG's talk page is a heavy load to handle. As for the matter at hand, I also don't want to enforce certain limits--but EEng's talk page (he knows I feel this way) is torture. His I try to avoid; DGG's I visit rather frequently but only when I have to: those two pages are difficult if not impossible to load navigate on a mobile device, and even on my desktop. Drmies (talk) 04:24, 24 May 2017 (UTC)[reply]
        DGG may be the most watched, but you're the highest traffic by a long shot: [12]. In other news, I'm somehow not the most boring arb, and I beat NYB by six whole views in the last three months. Neener neener! Opabinia regalis (talk) 04:27, 24 May 2017 (UTC)[reply]
Like other editors who have commented, I would oppose anything that is mandatory. But I think that it could be a reasonable advisory to have on a guideline page. --Tryptofish (talk) 20:35, 23 May 2017 (UTC)[reply]

I'm sure glad Iridescent brought that up because I had the bytes tabulated and names of admins lined up...but the sign says "Please do not feed, touch, poke, prod, tease, pester, annoy, torment, worry, irk, harass, disturb, bother, or ruffle. While we tease EEng about archiving, he is not an admin, so discourse is optional unless there is something that requires an elevated amount of attention. There's always the option of pinging him from one's own TP. He does archive, and like others who commented, I would oppose anything that requires mandatory archiving. ^{Atsme📞📧} 21:25, 23 May 2017 (UTC)[reply]

*coughs quietly* Flow would fix this problem if I'm not mistaken. Eman 235/talk 21:47, 23 May 2017 (UTC)[reply]

You are mistaken (except in the sense that Flow's handling of images and templates was so fubar that it would have rendered the issue moot). Even text-only pages using Flow sucked up bandwidth like a hoover owing to all the kludgy javascript (example); once you started scrolling through them, the lazy-loading setup was spectacularly inefficient. (Try scrolling through a Facebook or Twitter feed—or anything else using lazy-loading—on a metered connection and watch how quickly your data goes. You'll be shocked.) ‑ Iridescent 21:58, 23 May 2017 (UTC)[reply]

Comment -- Users' Talk page archiving looks like a non-starter. Although, automatic archiving of article Talk pages would be quite welcome. I.e. if one could add a template that would automatically transclude to set it up, it would be great. K.e.coffman (talk) 04:28, 24 May 2017 (UTC)[reply]
Thank you everyone for the input. While I can't stop people commenting, I'm going to echo coffman and consider this discussion Closed. Primefac (talk) 14:04, 26 May 2017 (UTC)[reply]
I have received as many comments that people are glad I have a system of manual category archives, as I have received complaints that the main page is too long. I also have a system for follow-up that I must manage manually--I typically have a considerable number of in-progres follow-ups dating back to initial inquires a few months back. I know I have fallen behind on both, but I can & will fix that. I absolutely do not want it archived automatically; it will make things even more complicated for me. DGG ( talk ) 16:02, 26 May 2017 (UTC)[reply]
FWIW I would probably oppose the RFC but instead would support a detailed guideline, Another user that springs to mind is Hullaballoo Wolfowitz who received quite alot of shit for the size of his talkpage, Anyway I'd happily support a guideline. –Davey2010^Talk 16:33, 26 May 2017 (UTC)[reply]
Mandatory user talk page archival (and talk page archival in general) is a non-starter until the archiving bots work reliably. They currently don't, and it's not clear if anyone is maintaining them. Ivanvector (^Talk/_Edits) 16:33, 26 May 2017 (UTC)[reply]
This sort of policy will be akin to a legislation saying that no person in the country will weigh more than 80kg⮘FORCE RADICAL⮚ 10:46, 30 May 2017 (UTC)[reply]
I think I would support something like this. There's plenty of archive bots, and I could add Femto Bot fairly easily.
- At some point, not archiving becomes disruptive.
Perhaps there should be a gentle approach though.

All the best: Rich Farmbrough, 23:23, 10 June 2017 (UTC).[reply]

Appel de décision administrative de blocage

Hello,

I'm really sorry for my middle-level english language, it is enough to read and good enough to write basic sentence but for this question I need a better accuracy than I'm able to be in english language, so I will write in french language, with the hope and expectancy that some people will help translating my words.

Le problème est le suivant : je pense avoir été la victime d'une sorte de complot. Non réellement un complot, plutôt la coalition provisoire de contributeurs qui, par esprit de groupe et par solidarité, ont réalisé une série d'actions qui font que je ne suis plus en moyen de me défendre, étant complètement bloqué, au point qu'un robot automatique à fonctionnement permanent a été activé pour réagir instantanément si un contributeur se connecte à partir d'une certaine adresse IP. La question n'est pas de savoir qui a raison ou qui a tort, cela peut se régler après, si j'ai la possibilité de mettre le problème sur la place publique.

Voici mon interrogation: la présomption d'innocence ne fait-elle pas partie des principes même de Wikipédia, et en ce cas mettre en place un mécanisme automatique de blocage préalable d'un IP dont la communauté n'a pas décidé qu'elle devait être bloquée, n'est-ce pas une violation d'un principe intangible? Et si c'est le cas, est-ce que quelque personne lisant cela pourrait, en quelque sorte, se faire mon avocat non pour le fond mais pour la forme, la demande du respect des formes pour décider du blocage d'une IP. Quoi qu'il en soit, merci d'avance pour les réactions à ce message. Olivier Hammam (talk) 15:36, 29 May 2017 (UTC)[reply]

Anyway, I will try to translate myself (it's easier than writing directly in a non-native language :-) If somemone can improve this translation, je remercie d'avance cette personne.

(translation)
Here is the case: I think I have been a victim (more or less in the meaning of patsy) of a kind of plot (more or less, a conspiracy). Not really a plot, something like a temporary trust where, with an "esprit de groupe" (I think you know the meaning) and in a spirit of solidarity (more or less, interlock), they chained several actions that leaded to a complete lock and now I really can't do nothing to try to defend myself. The point is not to know who is wrong or good, this will be decided later, but to obtain the ability to get explanations about this.

Here is the thing: is the "presumption of innocence", not in a legal but in a general meaning, a part of Wikipedia fundamentals or no, and if it is, does the activation of a robot (a web "bot") to prevent any connection through another user account than this I use here currently with the same IP address a lack of respect of Wikipedia fundamentals?
(end of translation)

Kind regards. — Preceding unsigned comment added by Olivier Hammam (talk • contribs) 2017-05-29T15:36:26 (UTC)

@Olivier Hammam: I think you are asking about two related topics, the use of alternate accounts - and the use of these as bot accounts. You can read the English Wikipedia policies about these on the following two pages: Wikipedia:Sock puppetry, Wikipedia:Bot policy. — xaosflux ^Talk 15:40, 29 May 2017 (UTC)[reply]

Also if you have a specific alt-account that is blocked, you can use the Template:unblock to request a review. — xaosflux ^Talk 15:41, 29 May 2017 (UTC)[reply]

I really can't understand nothing in your purposes, User:Xaosflux, sorry, and it's not because of my, so said, «mauvais english», I'm fluent enough in this language, as you can see, but I really can't understand your writings. Olivier Hammam (talk) 17:38, 29 May 2017 (UTC)[reply]

User:Xaosflux directed you to three English Wikipedia pages that may help you understand how to bring your case to a good end. Please follow these links and read the suggestions.
To be utterly blunt, if your English is too weak to discuss in that language you should seriously reconsider whether you can deliver relevant input into the English Wikipedia version. After all, following another Wikipedia page explains that sufficient control of the English language is required to be a valued editor (see WP:Competence under language issues). Perhaps it would be a wiser choice to become an editor on French Wikipedia rather than having the ambition to be one here. Arnoutf (talk) 17:47, 29 May 2017 (UTC)[reply]

@Arnoutf: FYI this user is now blocked for personal attacks after warning; it appears they are also blocked for personal attacks on frwiki - it may possibly be a compromised account. — xaosflux ^Talk 17:50, 29 May 2017 (UTC)[reply]

<edit conflict> Actually, this is about them being blocked at the fr-wikipedia, because of nothere-contributions to the bistro over there (about the same as the village pump here), and now being blocked here because of nothere-contributions to our village pump (Conspiracy!!!!). Kind of a rant actually, interspersed with rather heavy personal attacks (in French) added to their talk-page after the block. When I look at older contributions, this looks like a compromised account. I actually would up the block to a week at least, for them to calm really down. And remove talk-page access to boot. Lectonar (talk) 17:56, 29 May 2017 (UTC)[reply]

And yes, they are blocked over there for repeated personal attacks; some edits had to be oversighted in fr-wikipedia. Lectonar (talk) 17:58, 29 May 2017 (UTC)[reply]

Je ne comprends rien, mais alors strictement rien aux bavardages stériles des personnes (ou supposées personnes) qui bavassent ci-dessus à propos de choses dont elles ne savent rien et croient tout savoir, et cette incompréhension n'a rien à voir avec mon supposé niveau insuffisant en anglais, beaucoup à voir avec la vacuité de leurs supposés échanges, même s'il me semble n'entendre qu'une voix, celle de l'écho. Je regrette fort que mon intervention ici n'ait, à ce qu'il semble, attiré que des mouches du coche. Triste chose... Je me demande si ces propos anodins et factuels vont encore provoquer un blocage, et me demande encore en quoi le précédent avait la moindre validité. Olivier Hammam (talk) 06:28, 5 June 2017 (UTC)[reply]

@Olivier Hammam: Notez que sur Wikipedia Anglais les reglements ne sont pas necessairement les memes que sur sur Wikipedia Francais. La pluspart des utilisateurs ici (incluant moi-meme), nous ne sommes pas necessairement familiers avec les reglements du Wikipedia Francais. Il semblerait d'apres Lectonar que vous ayez ete bloque pour raisons d'edition disruptive ou "POV"/point-de-vue non supporte par les sources fiables ainsi que pour motif d'incivilite (incluant l'utilisation de la page personelle de facon abusive apres avoir ete blocke). Xaosflux evoque meme le fait qu'il serait possible que votre compte aie ete pirate. Nous avons ici WP:APPEAL, WP:OFFER, WP:ABK (autoblock), etc. Il est possible que le Wikipedia Francais aie aussi l'equivalent. Les administrateurs etant egalement differentes personnes sur les differents projets, je doute qu'ici soit la meilleure place pour votre requete. Desole pour le manque d'accents dans mon texte, et bonsoir, —Paleo Neonate - 07:00, 5 June 2017 (UTC)[reply]

Merci pour votre contribution, Paleo. J'ai tout-à-fait conscience de ce que vous me dites quant aux différences dans les usages des wikipédias anglophone et francophone, mon problème est l'incapacité où je me trouve de faire valoir mon point de vue, qui est possiblement mal fondé, sut WP:FR, du fait de mon blocage complet. Je sollicite simplement une bonne volonté de WP:EN, non pas pour défendre mon point de vue mais pour l'exposer, tel que présenté ici, sur ma page de discussion francophone et aussi, si elle le souhaite, sur la gage où a été décidé ce blocage. Si donc quelque personne de bonne volonté peut relayer la part de mon premier message qui commence par «Voici mon interrogation» jusqu'à ma signature avec un peu de contexte (lien vers la source du message, si la personne le juge opportun motifs qui l'ont conduite à relayer ma demande...). Merci encore à vous, et merci par avance à qui voudrait accéder à ma requête.

Amicalement. Olivier Hammam (talk) 09:19, 6 June 2017 (UTC)[reply]

Pre-RfC: Political Memes

Withdrawn; no interest. Power~enwiki (talk) 22:29, 12 June 2017 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

It seems that every time Donald Trump sneezes, somebody creates an article for it. We have Covfefe (AfD) and Trump orb (AfD) created and sent to AfD in the past week.

I propose that there be a new Speedy Deletion criteria for any pages related to new political memes of the past two weeks. It's somewhat difficult to define what should be included, though. If Congress creates and passes a new health care bill, we don't need to wait two weeks to create an article. For things which are fundamentally noteworthy because people talk about them a lot on Twitter (and a significant portion of the press will cover whatever political gossip people talk about on Twitter), the articles should be deleted (or redirected) as a matter of policy.

If any of these memes or news stories are encyclopedic, they will still have coverage and still get an article in two weeks. If they're not, we discourage dozens of editors from making unproductive contributions. Power~enwiki (talk) 23:28, 31 May 2017 (UTC)[reply]

No, you can't create arcane policy pages at Wikipedia that will prevent good-intentioned people from creating articles you don't like. There's no impending harm from these articles going through the normal AFD process. Wikipedia is in no rush, and it gets it right eventually. This would be an abuse of the speedy deletion process, which is not for deleting articles like the two cited, and never was. --Jayron32 23:37, 31 May 2017 (UTC)[reply]
- I am suggesting a change to the policy, of course it doesn't fit in the existing speedy deletion process. Perhaps "redirect and protect" is a better suggestion? Power~enwiki (talk) 23:40, 31 May 2017 (UTC)[reply]
  - It's not merely that it doesn't fit existing criteria, it completely and utterly violates a core principle of CSD. There's a distinction to be made here, CSD does not exist merely to allow us to delete stuff we can't be bothered to wait around to delete. CSD is entirely and totally about extreme examples of WP:SNOW-level deletions, things like an article which is nothing but a copyvio of another source, or articles which are gibberish, or written about your cat, or things like that. There's no need to hold an AFD discussion deleting an article which consists of someone who blindly smashed their keyboard and hit "save". However, potentially encyclopedic content is never suitable CSD material, even if we find it really annoying. Such content should always be discussed. That the discussion ends up deleting the content does not mean the discussion wasn't necessary in the first place. Indeed, the discussion should happen, because deciding which recent events have met the threshold of suitability for an article is something that requires more than one person's feeling. That was my point, your proposal was not to create a new definition of gibberish, it completely violates a core principle of CSD. --Jayron32 03:42, 1 June 2017 (UTC)[reply]
    - With multiple keep votes cast by users with non-automatic user rights, and at least one keep vote by an admin for each, this is clearly not speedy material. עוד מישהו Od Mishehu 05:01, 1 June 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Cease using Rotten Tomatoes in film articles

I know next to nothing about Wikipedia policies and how they work, and, from what I understand, the closed-shop attitude of many users therefore means that my proposal may well be dismissed out of hand, which is fine. This problem is likely to be exacerbated by the fact that I do not have any awareness of what policies currently support the current tendency to which I am objecting. Nevertheless, I believe it should at least be stated: it is detrimental to the encyclopedia, to cinema, and to culture in general that film articles on Wikipedia almost universally refer to the "Rotten Tomatoes" score for a film. This score, composed of a sum of thumbs up and thumbs down, represents the absolute nadir of critique. It indicates nothing of significance or importance about a movie, and only contributes to the general lowering of critical capacities, and probably to film producers being even further encouraged to worry about nothing except the lowest common denominator opinion. Use of this score represents, to me, nothing but the laziest approach to composing encyclopedic articles on cinema. In my view, whatever policies or guidelines support the constant reference to this website should be changed to prevent this use, and articles that currently refer to this "score" should be edited to no longer do so. Perhaps this question has already been debated and decided upon (I would have no idea where that debate may have taken place), but if so, it is time to revisit it: Rotten Tomatoes really is something rotten in the heart of Wikipedia's film pages. — Preceding unsigned comment added by 49.194.22.210 (talk) 04:37, 1 June 2017 (UTC)[reply]

Thank you Kirk Cameron, for your input. It is nice to see a filmmaker of your caliber editing Wikipedia. --Guy Macon (talk) 05:13, 1 June 2017 (UTC)[reply]

Hmm... Although I happen to disagree with you, I think that it is best to provide context to things, and that there shouldn't just be a section saying "It was rated highly by critics" or something like that. I do think that stuff like that should be removed, but I think that if there was any context, like "It was rated highly by critics, especially for its score [music, in this case]" then it should probably be good. I do try and do stuff like this when writing the articles I write (in my case, I write about Japanese politicians, and I try to give the reader an idea of some of their main policies). But, it would almost certainly be discouraged for one to remove stuff like this wholesale—the best method would to try and fix it and context to the issues. I think that some information which could be helpful is better than none. RileyBugz^会話_投稿記録 14:48, 1 June 2017 (UTC)[reply]

We have a noticeboard to discuss and determine if sources can be used, the Reliable Sources Noticeboard (WP:RSN). I don't personally have an opinion about using Rotten Tomatoes or not. Thank you for your comment, —Paleo Neonate· 17:53, 1 June 2017 (UTC)[reply]

This isn't a debate on how reliable the source is, it is a debate on how much information is actually passed along by the score. RileyBugz^会話_投稿記録 18:23, 1 June 2017 (UTC)[reply]

In my experience with film articles, review aggregators like RT and Metacritic are used for a quick summary of critical reception before the article launches into a round-up of specific reviews and critiques. Well-written articles certainly don't rely on the aggregate score alone. Since this discussion relates to film articles, WT:FILM is probably the best place for it. For more info on how film articles are structured, see MOS:FILM (especially the bit on critical response). For an essay on review aggregators, see WP:ROTTEN. clpo13_(talk) 18:05, 1 June 2017 (UTC)[reply]

The problem is that you are bringing your own biases to Wikipedia. Whether you think it's impacting film criticism or box office performance should not be a reason to avoid using the site. Wikipedia is, ideally at least, about undeniable facts. The fact of the matter is that the site, and Metacritic, are viewed as viable metrics by the vast majority of the general public at large. That is why we use them, not necessarily because we agree or disagree with it on a personal level. --Deathawk (talk) 04:15, 11 June 2017 (UTC)[reply]

Maybe we could aim a little higher than "majority of the general public at large". In most areas, the goal is to reflect what academic or expert sources say, no matter what the general public at large believes. WhatamIdoing (talk) 00:15, 12 June 2017 (UTC)[reply]

The Rotten Tomatoes scores are by movie critics who would be considered experts in their field. Secondly there is no good reason to not have this score other than someone stating that "I don't like this". It provides data that is not easily replicated elsewhere. --Deathawk (talk) 06:19, 14 June 2017 (UTC)[reply]

This very website is a product of the general public at large. Dismissing scores from public barometer sites such as Rotten Tomatoes, MetaCritic, and Alexa is to dismiss Wikipedia itself. Therefore, this conversation is self-destructive. Robert The Rebuilder (talk) 12:01, 16 June 2017 (UTC)[reply]

RfC: Wikimedia referrer policy

Please consider joining the feedback request service.

An editor has requested comments from other editors for this discussion. This page has been added to the following list:

Wikipedia policies and guidelines

When discussion has ended, remove this tag and it will be removed from the list. If this page is on additional lists, they will be noted below.

What referrer information should Wikipedia send to an external website when a reader clicks on a link? --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

Limitations: The Wikimedia foundation is not bound by Wikipedia RfCs, so this RfC is purely advisory. If there is a strong consensus for a particular referrer policy, a request will be made to the Wikimedia foundation to implement that policy. The request will contains the words "if possible" and "as far as is practical" to make it clear that we are saying what we want, not how to accomplish it.

Background:

Extended content

Previous discussions:

Overview:

When someone who is reading Wikipedia clicks on an external link their web browser may be given referrer information to be passed on to the external website. Depending on various factors, this information can range from telling the external site exactly what Wikipedia page the reader was on when they clicked on the link, telling them that the link was from Wikipedia but not telling them where on Wikipedia, to telling them nothing at all about the site that the link was on.

For example, consider the case of someone reading the Wikipedia page at Bomb-making instructions on the internet#References who clicks on the link to The Low Cost Cruise Missile: A looming threat?. (example ignores our recent switch to HTTPS for clarity).

Depending on how we configure Wikipedia, the aardvark.co.nz website (and anyone monitoring the connection between the user's browser and aardvark.co.nz) could be told that site/page the link was on (the referrer) was:

en.wikipedia.org/wiki/Bomb-making_instructions_on_the_internet#References (full URL)
en.wikipedia.org/wiki/Bomb-making_instructions_on_the_internet (page)
en.wikipedia.org (full domain)
wikipedia.org (partial domain)
No referrer information (a silent referrer)

Furthermore, Wikipedia might send additional information that some sites may (or may not) choose to honor in various ways. These are discussed in the technical comments section below.

Many web sites have a legitimate desire to know who links to them. Alas, spammers also desire to know who links to them so that they can refine their spamming. Most users, including those who fear surveillance by governments, corporations or criminals, desire to send the minimum amount of information that is consistent with them being able to use the website.

History: (needs to be double checked; please post corrections in the comments section)

Before 2011, Wikipedia was an HTTP site with full URLs sent in the referrer.

In 2011, Wikipedia added support for Hyper Text Transfer Protocol Secure (HTTPS). Users who accessed Wikipedia with HTTP still sent full URLs in the referrer. Users who accessed Wikipedia through HTTPS and clicked on an HTTPS external link also sent full URLs in the referrer. Users who accessed Wikipedia through HTTPS and clicked on an HTTP external link sent no referrer information.

In 2015, Wikipedia stopped offering HTTP and only offered access to the site with HTTPS.

In February of 2016 Wikipedia added the following to the HTML code on every Wikipedia page:

<meta name="referrer" content="origin-when-cross-origin"/>

Various websites talk about this leaking information to an eavesdropper when a user clicks on an HTTP link, and some recommend...

<meta name="referrer" content="strict-origin-when-cross-origin"/>

For example, [13] says "Warning: Navigating from HTTPS to HTTP will disclose the secure URL or origin in the HTTP request." and "Likewise if you're thinking of using origin or origin-when-cross-origin then I'd recommend looking at strict-origin and strict-origin-when-cross-origin instead. This will at least plug the little hole of leaking referrer data over an insecure connection."

What this RfC is not

This RfC is not binding on anyone. It is purely advisory.
This RfC is not a discussion of the technical details regarding what is and is not possible using current technology, which may change. It is an RfC about policy, not implementation. There is a section on the bottom for technical discussions such as what HTTPS does and does not allow, etc. All comments that focus on "how we should do it" rather than "what should we do" may be moved to the technical discussion section by any user.
This RfC is not about links to other Wikipedia pages or to other projects that are under the control of the Wikimedia foundation. It is assumed that, as far as possible/practical, other WMF sites will receive as much information as they want to receive.

Questions

Question 1: As far as possible/practical, should referrer information contain full URLs? (en.wikipedia.org/page#section)

(This was the status quo before we switched to HTTPS in 2015)

Support And nobody was hurt, right? It is useful in cases such as an outreach museum/library/GLAM project finding out that its contributions are generating links back. However, for the security conscious, we could consider adding some opt-out, or we could add a mechanism for making some pages have more silent referrers. Both seem like a waste of our developers time, IMHO. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 05:41, 18 June 2017 (UTC)[reply]

Question 2: As far as possible/practical, should referrer information contain full domains and pages? (en.wikipedia.org/page)

Support example support comment.

Question 3: As far as possible/practical, should referrer information contain full domains? (en.wikipedia.org)

(This is the status quo as of February of 2016, default HTTPS behavior was the Status quo before that)

Support since this is already the same level of information disclosed by usage HTTPS/SNI (slightly reworded after initial post). I think referrer information is a useful information about site-2-site relations, and i think it doesn't disclose so much reader behavior that we should be too concerned about it. Considering how many people were railing against the introduction of HTTPS as a default a couple of years ago, some of the comments in this RFC leave me absolutely stunned. It seems like everybody has jumped towards the other extreme now.. —TheDJ (talk • contribs) 12:20, 11 June 2017 (UTC)[reply]

I am pretty sure that this is not the default amount of information disclosed by HTTPS (but please double check and post any corrections; this is exactly the type of thing that it is easy to get wrong). If there is no meta referrer tag, the default is "no-referrer-when-downgrade". This means that if a link from an HTPPS website such as Wikipedia goes to an HTTP site, the site gets no referrer information at all, but if the link goes to another HTTPS site, the link gets the full URL including the page (not just the full domain) in the header.

From [ https://www.w3.org/TR/referrer-policy/ ]:

"3.2. 'no-referrer-when-downgrade'

The 'no-referrer-when-downgrade' policy sends a full URL along with requests from a TLS-protected environment settings object to a potentially trustworthy URL, and requests from clients which are not TLS-protected to any origin.

Requests from TLS-protected clients to non- potentially trustworthy URLs, on the other hand, will contain no referrer information. A Referer HTTP header will not be sent.

If a document at https://example.com/page.html sets a policy of 'no-referrer-when-downgrade', then navigations to https://not.example.com/ would send a Referer HTTP header with a value of https://example.com/page.html, as neither resource’s origin is a non-potentially trustworthy URL.

Navigations from that same page to http://not.example.com/ would send no Referer header.

This is a user agent’s default behavior, if no policy is otherwise specified.

(Emphasis added). --Guy Macon (talk) 16:43, 11 June 2017 (UTC)[reply]

That's the referrer policy, which has nothing to do with HTTPS (other than that it is carried in side of it). I've already discussed below how SNI discloses similarly information below. —TheDJ (talk • contribs) 22:32, 12 June 2017 (UTC)[reply]

Simple answer. You are wrong. Read the passage from w3.org above. Besides explicitly mentioning HTTPS and comparing it to HTTP in multiple places, "TLS-protected" means HTTPS. Referrer policy is an inherent part of HTTP and HTTPS, and is carried in the header. The meta referrer tag in the body simply allows finer control instead of accepting the default HTTP/HTTPS referrer behavior. --Guy Macon (talk) 02:53, 13 June 2017 (UTC)[reply]

Support per my comment below about the benefits to the Wikimedia community of sharing this information, which I believe to be a net positive. Sam Walton (talk) 12:08, 2 June 2017 (UTC)[reply]

Support—This is valuable information that enhances the visibility of Wikipedia to other sites, while providing little additional information about the involved user.--Carwil (talk) 18:59, 15 June 2017 (UTC)[reply]

Question 4: As far as possible/practical, should referrer information contain partial domains? (wikipedia.org)

Second choice. I'd prefer the symbolism of option 5—"Wikipedia doesn't give away any of your information" is an important tool in building trust—but I could live with this if the WMF really feel it's important that other sites see how much of their traffic is coming from Wikipedia. ‑ Iridescent 22:21, 10 June 2017 (UTC)[reply]
Support it provides a service to webmasters to inform them that Wikipedia is the source of traffic to their website. Providing additional information on a per-user level is unnecessary and has risks to privacy. If it is technically easier to implement, I might support option 3. Power~enwiki (talk) 22:26, 12 June 2017 (UTC)[reply]
comment: It should be noted, that this is not technically feasible. Not including the subdomain is not one of the options at: https://www.w3.org/TR/referrer-policy/#referrer-policies . BWolff (WMF) (talk) 23:35, 13 June 2017 (UTC)[reply]

Question 5: As far as possible/practical, should referrer information contain no information? (silent referrer)

Support Here is the reason why, whenever possible, we should reveal as little as possible about what pages our users read or what links they follow:

Locals Questioned by Suffolk County Police Department after Googling "Backpack and "Pressure Cooker"

Google Pressure Cookers and Backpacks, Get a Visit from the Feds

Update: Now We Know Why Googling 'Pressure Cookers' Gets a Visit from Cops --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

Support The HTTP-Referer header is one of the basic security bugs of the web. While there are ways to suggest to browsers how to act, that is also a type of bogus opt-out feature software may or may not properly observe. It may be useful for users to use specialized software to forge those (pretending to originate from the destination site itself). However, I support doing what is possible to limit this issue for users with common software and default configurations. Another reason to support this is that link spammers are interested to know where clicks originate from. We already attempt to limit unnecessary tracking information from URLs themselves, and we use nofollow for search engine indexing (and noindex where appropriate), this is another step forward. Thank you for posting this RfC. —Paleo Neonate - 06:36, 3 June 2017 (UTC)[reply]
Support Protecting our users' privacy is of paramount importance. We should share as little as possible with external entities, both as a matter of principle, as well as because of the meatspace consequences described by Guy above. James (^talk/_contribs) 16:45, 9 June 2017 (UTC)[reply]
Support revealing as little as practical in most cases. I lack technical expertise, but I see this as being within the concept of Wikipedia as a free resource, and just as we protect the private information of editors, we owe it to our readers worldwide to feel that they can use us as a resource without undue fear of repercussions. I'd make some exceptions for complying with law enforcement in the countries where servers are located. --Tryptofish (talk) 23:55, 9 June 2017 (UTC)[reply]
Support. We don't need referers. In fact, none of the Wikimedia sites need referers. They may be fancy HTML5 Web 2.0 stuff, but not all fancy HTML5 Web 2.0 stuff is useful. KMF (talk) 00:40, 10 June 2017 (UTC)[reply]

@KATMAKROFAN: it does not have much to do with HTML5 or Web 2.0, it is part of the HTTP protocol since version 1.0 (1996) ^[14]. To disable it, users needed to use special proxies, modify their web clients or configure them to not send the HTTP-Referer header along with queries. More recently, it begins to be possible for sites to suggest referrer behavior to browsers which support this. This RfC is about making those suggestions to browsers to limit the security and privacy implications of referrers for users using common browser software with common (generally too permissive) default configurations. —Paleo Neonate - 03:58, 11 June 2017 (UTC)[reply]

KATMAKROFAN, It sounds as if you are supporting Question 5: ("As far as possible/practical, should referrer information contain no information? (silent referrer)") In other words, no referer. Should your comment be moved to that section? --Guy Macon (talk) 20:34, 10 June 2017 (UTC)[reply]

I went ahead and WP:BOLDLY moved KATMAKROFAN's comment and all responses to the proper section. --Guy Macon (talk) 14:17, 12 June 2017 (UTC)[reply]

Support – Protecting the privacy of those visiting our encyclopedia is much more important than reaping any benefit that comes from sharing referrer information. — Godsy^(TALK_CONT) 08:05, 10 June 2017 (UTC)[reply]
Support - anyone can add a reference - which could be a front for surveillance, by not providing referrer information we help protect the privacy of our readers. — xaosflux ^Talk 21:12, 10 June 2017 (UTC)[reply]
Support While I sympathise with the benefits of partners seeing the incoming traffic, and this encouraging them to contribute content, there are other ways of achieving this. And even if we made exceptions, the principle of being a silent referrer is important as outlined above. All the best: Rich Farmbrough, 21:37, 10 June 2017 (UTC).[reply]
Support: especially considering latest developments towards surveillance and censorship in countries like, as non-exhaustive examples, Turkey or the UK, it would seem outright irresponsible for Wikipedia to provide information that could help identify users and what they have connected to, and in a very non-obvious way at that, since users typically expect privacy-conscious HTTPS sites not to leak details like the page or subdomain being visited. LjL (talk) 22:00, 10 June 2017 (UTC)[reply]
Support From what I can make out with the technical details, the Turkish government could potentially see anyone coming in from *.wikipedia.org to a Turkish website and then block that IP's access to Wikipedia (if they somehow see that it is, in fact, a Turkish user). The WMF's goal was to have Turkey unblock access to Wikipedia, not reinforce their ability to do so. (Whether we as a community will support the WMF's statement remains to be seen, but I personally do support it regardless of whatever political problems we may face because of it.) — Gestrid (talk) 22:08, 10 June 2017 (UTC)[reply]

I thought of another reason, which actually has to do with our current English Wikipedia policies. Webmaster X sees, hey, I'm getting a lot of traffic from Article A (or some variant of the web domain, as illustrated in the other options). That webmaster then tries to spam Wikipedia articles with links to his article because more clicks means more money/ customers/ etc.. Now imagine a bunch of webmasters doing that. Maybe even some companies. We already have some problems with undeclared WP:PAID violations by companies. (Specifically, I recall one SPI maybe the middle of or late last year where I believe it was brought to ANI, and I actually ended up emailing Wikipedia Legal about it to report the problem because it was a TOS violation. I got permission from them to post the email from them, if anyone wants to look it up, but I don't believe it's pertinent.). We don't need more violations of it because companies now know how much traffic is coming from us. — Gestrid (talk) 01:29, 11 June 2017 (UTC)[reply]

Went ahead and found the ANI archive myself. Thanks to the WMF for upgrading Special:Contributions so we can search specific date ranges, by the way. The link to the ANI archive is Wikipedia:Administrators' noticeboard/IncidentArchive929#Undisclosed Paid Editing Farm. I was still pretty new to Wikipedia, so I apologize in advance for my older contribs. — Gestrid (talk) 01:53, 11 June 2017 (UTC)[reply]

Support. I've never had much sympathy for the "privacy is our top priority" view—I opposed the switch to https, for instance—but I find the arguments in favour of referrer information to be utterly unconvincing. If the best argument that can be made is "some commercial organisations might not want to donate in kind if we're not giving them this information", that really doesn't add up to much. No disrespect to The Wikipedia Library, which has the best of intentions, but the number of people who would even notice if it disappeared could probably be counted on two hands at most; besides, I really don't buy the argument that other organisations won't support Wikipedia unless they're allowed to track incoming traffic, let alone that their potential loss outweighs the potential problems this could cause (and the potential negative press if the WMF lines up alongside Facebook, Google et al in the "our readers are the product" camp). ‑ Iridescent 22:15, 10 June 2017 (UTC)[reply]
Support Jclemens (talk) 23:39, 10 June 2017 (UTC)[reply]
Support. Wikipedia should protect its users -- and the consequences of a negative news story in which this feature is used against readers could be severe. That said, I think it makes sense to have computer experts look over the possibility of other ways of notifying sites that they are linked to. We have overall statistics for how many times articles are accessed. It would be possible to take all the articles that are accessed >1000 times monthly, look at all the references in them, and notify all the webmasters of all those sites that they are linked by Wikipedia but referrer (or "referer"...) information is not being provided, giving a report of all the pages that link to them and how much traffic each gets. That is, if there's a polite way of doing it that is not considered spamming, which is why I want computer experts to think of something. If need be, I ~~could put up~~ (umm, wait, no -- see below) with a general citation to Wikipedia in the referrer (i.e. the previous option), which then could link directly to the report. Wnt (talk) 00:17, 11 June 2017 (UTC)[reply]

I believe that this information is already available, but of course we could make it more convenient to access it. For example, imagine that you own the American Institute of Physics website. Our linksearch tool tells you that exactly one Wikipedia page (Age of the universe) links to your The Expanding Universe: Theories of a Static Universe page, and our Pageviews Analysis tool tells you that Age of the universe got 145,317 pageviews in the last six months. Putting a partial domain (wikipedia.org) in the referrer would also allow you, the imaginary owner of the American Institute of Physics website, to figure out the IP address of every Wikipedia user who clicked on the link to your page and to know exactly what Wikipedia page they were reading when they clicked on that link. If the AIP website happened to be one of those pages where users log on and cookies are stored about that logon, this would allow the AIP to link the real identity of certain Wikipedia users with which Wikipedia page they were reading when they clicked on that link. Being a silent referrer would prevent this. --Guy Macon (talk) 11:41, 11 June 2017 (UTC)[reply]

OK, you've convinced me -- let's not even have the domain there. You're right that it would only take a moment for a unique site to search itself on Wikipedia -- and a "sting" site, like the Russian government (or Dianne Feinstein!) looking for people reading about drugs, would obviously arrange to get itself added in one unique place for this specific purpose. Wnt (talk) 20:49, 11 June 2017 (UTC)[reply]

Support Because referrer information only makes it easier for third party sites to correlate IP numbers with Wikipedia readers and editors; and because it potentially creates temptation to spammers by linking Wikipedia with page traffic. Neither benefits the project. Geogene (talk) 00:58, 11 June 2017 (UTC)[reply]
Support. MER-C 02:47, 11 June 2017 (UTC)[reply]
Support TonyBallioni (talk) 03:17, 11 June 2017 (UTC)[reply]
Support This is an obvious security vulnerability. I can't believe there needs to be an RfC for something like this. A Quest For Knowledge (talk) 05:45, 11 June 2017 (UTC)[reply]
Support any benefit, real or imagined, enjoyed by the WMF or the projects themselves, is outweighed by the chilling effect of knowing that encyclopedia browsing habits are being exposed to unknown actors. –xeno^talk 10:41, 11 June 2017 (UTC)[reply]
Support The entire referrer model is, at bottom, a technical option that for most of its existence was not actually optional, thus leading most users to accept it as a fait accompli. Now that it is optional, we should recognize that its two main uses are commercial and monitoring-related - neither of which Wikipedia should want any truck with. Potential negatives, outlined by various editors above, are very likely to outweigh any slight reciprocal/funding benefits. --Elmidae (talk · contribs) 11:44, 11 June 2017 (UTC)[reply]
Support - we stop spamming of enforced referral links as much as possible (i.e., links with '&referrer=' type of parameters). There are enough of those examples where spammers add those parameters just to know the efficacy of their spamming. This is basically a hidden version of that same information. To protect the privacy of Wikipedia editors, to protect readers of Wikipedia, and to hide the efficacy of spamming attempts, it does seem best to hide where people are coming from. --Dirk Beetstra ^{T C} 12:39, 11 June 2017 (UTC)[reply]
Support. In the current day and age of global espionage this is a glaring security hole. I'm astounded that this hasn't been fixed already. Daß Wölf 03:03, 12 June 2017 (UTC)[reply]
Support The whole concept of privacy (which WMF claims to support, though they keep trying to weasel against it) is to withhold info from third parties unless users themselves decide otherwise on a case by case, individual basis. That Wikipedia should avoid revealing any referer info on privacy grounds to the extent possible should be a no-brainer. Getting referers also is very useful for SEO's, another source of disruption and bias in Wikipedia that we should avoid trying to help. (I think we shouldn't even disclose article view counts, partly for that same reason).
That said, on a technical level, I didn't even know there was a way to suppress referers from the server side, so it surprised me considerably to hear that we had previously been doing it. I remember trying pretty hard to figure out how to let people click links on some of my personal pages, without revealing the referring page.
I see the whole concept of referers as invasive, a leftover from the more communal and mutually trusting days of the academic WWW built at CERN. HTTP headers in those days also contained the user's email address: as you can guess, when the web went commercial, they had to get rid of that pretty fast. Referers should have gone at the same time. If I read an online article about how to fix my shoes using duct tape, and then go to the store to buy some duct tape for that purpose, the store is not entitled to know what article I read or what I'm going to do with the tape. That's precisely what referers do.
FYI note: in Firefox, you can turn off referers on the browser side by viewing about:config and setting network.http.sendRefererHeader to 0. It stops a very few pages from working properly but usually causes no trouble, so I do this and recommend it. 173.228.123.121 (talk) 06:58, 12 June 2017 (UTC)[reply]
Support. Well, it is not really a Wikipedia thing: the referer field was an abomination from the start and should be killed with fire. Maybe a rogue browser will send the referer against our wishes, or tech-savy users have already deactivated it, but I see zero positive side to encouraging it to be sent. The only argument to send referers is the "Wikipedia Library" argument (see below), but sorry, that is just a hidden form of monetizing users' history, which I oppose on principle (but again, not a Wikipedia thing). Tigraan^{Click here to contact me} 16:22, 12 June 2017 (UTC)[reply]
Support Yup. Only in death does duty end (talk) 16:27, 12 June 2017 (UTC)[reply]
Support I initially leaned towards Q4 above, as I don't believe there are privacy issues there. But after considering the potential for increased link-spamming (as already discussed above), it would be best not to pass any information at all. -- Tom N talk/contrib 01:37, 13 June 2017 (UTC)[reply]
Support. In the absence of any persuasive argument to the contrary, this seems smarter and safer. Rivertorch^FIRE_WATER 05:23, 13 June 2017 (UTC)[reply]
Support Tell them nothing! The privacy of our users and editors is not negotiable - only a legally enforcible warrant or court order should allow the release of any user information. It is common knowledge that people in some countries are persecuted (and prosecuted) for accessing "prohibited" information, facilitating such persecution is morally reprehensible and arguably a human rights violation. We must also WP:DENY linkspammers and SEOers the "benefit" of their bad faith acts. Roger (Dodger67) (talk) 21:47, 13 June 2017 (UTC)[reply]
Support Anonymity on all fronts. →Σ σ ς. (Sigma) 06:54, 16 June 2017 (UTC)[reply]
Support - Doing otherwise would just make it so that the corporations can refine their algorithms and make more money. That would also open Wikipedia to link spam attacks. So, I oppose anything else. RileyBugz^会話_投稿記録 18:29, 16 June 2017 (UTC)[reply]
Support Privacy is paramount. Doug Weller talk 19:37, 17 June 2017 (UTC)[reply]
Support Cas Liber (talk · contribs) 01:37, 19 June 2017 (UTC)[reply]

Question 6: As far as possible/practical, should referrer information contain something else not listed above (please be specific about what you think it should contain)

There are a couple of possible referrer policies that we have had in the past that were not mentioned in the above list.

From 2001 to 2011, we had the following referrer policy:

Access Wikipedia through HTTP, click on an HTTP link: Full URL sent in referrer.
Access Wikipedia through HTTPS, click on an HTTP link: Not possible. Wikipedia did not support HTTPS.
Access Wikipedia through HTTP, click on an HTTPS link: Full URL sent in referrer.
Access Wikipedia through HTTPS, click on an HTTP link: Not possible. Wikipedia did not support HTTPS.

This is the same as Question #1.

In 2011, Wikipedia added optional HTTPS support, so we then had the following referrer policy:

Access Wikipedia through HTTP, click on an HTTP link: Full URL sent in referrer.
Access Wikipedia through HTTPS, click on an HTTP link: No Referrer information sent.
Access Wikipedia through HTTP, click on an HTTPS link: Full URL sent in referrer.
Access Wikipedia through HTTPS, click on an HTTPS link: Full URL sent in referrer.

This possible referrer policy is not mentioned in the above list.

In 2015, Wikipedia stopped supporting HTTP, so we then had the following referrer policy:

Access Wikipedia through HTTP, click on an HTTP link: Not possible. Wikipedia does not support HTTP.
Access Wikipedia through HTTPS, click on an HTTP link: No Referrer information sent.
Access Wikipedia through HTTP, click on an HTTPS link: Not possible. Wikipedia does not support HTTP.
Access Wikipedia through HTTPS, click on an HTTPS link: Full URL sent in referrer.

This possible referrer policy is not mentioned in the above list.

In February of 2016 Wikipedia started using origin-when-cross-origin in the meta referrer, so we then now have the following referrer policy:

Access Wikipedia through HTTP, click on an HTTP link: Not possible. Wikipedia does not support HTTP.
Access Wikipedia through HTTPS, click on an HTTP link: Full domain sent in referrer.
Access Wikipedia through HTTP, click on an HTTPS link: Not possible. Wikipedia does not support HTTP.
Access Wikipedia through HTTPS, click on an HTTPS link: Full domain sent in referrer.

This is the same as Question #3.

--Guy Macon (talk) 01:38, 13 June 2017 (UTC)[reply]

Discussion

General policy comments

Please comment about what we should do, not how we should do it. Misplaced comments may be may be moved to the proper section by any user.

User:Guy Macon... can you please state what problem you are trying to solve in this RFC, before trying to ask a wide variety of very technical questions that most people will have trouble answering with full understanding ? Why not ask less technical questions like "Do you think it is important to indicate to other websites and partner organisation where there traffic is from" "Do you think this should be limited to partners that make use of of https" etc.. —TheDJ (talk • contribs) 15:40, 1 June 2017 (UTC)[reply]

mess up my ping User:Guy Macon.. —TheDJ (talk • contribs) 15:40, 1 June 2017 (UTC)[reply]

No need to ping me. When I make a comment I check for replies.

I thought that the following was clear:

"Overview: When someone who is reading Wikipedia clicks on an external link their web browser may be given referrer information to be passed on to the external website. Depending on various factors, this information can range from telling the external site exactly what Wikipedia page the reader was on when they clicked on the link, telling them that the link was from Wikipedia but not telling them where on Wikipedia, to telling them nothing at all about the site that the link was on."

The problem I am trying to solve in this RFC is that the Wikipedia community was never consulted before the Wikimedia foundation made decisions about what information about what pages/websites our users visit is given to external sites that the WMF does not control.

As for "partners". I believe that they are covered by the following:

"Many web sites have a legitimate desire to know who links to them. Alas, spammers also desire to know who links to them so that they can refine their spamming. Most users, including those who fear surveillance by governments, corporations or criminals, desire to send the minimum amount of information that is consistent with them being able to use the website."

A "partner" would seem to be an example of "a web site that has a legitimate desire to know who links to them".

What would be helpful would be a list of these "partner" websites, and why they might want this information about what websites and/or pages a user visited before visiting their site. --Guy Macon (talk) 16:01, 1 June 2017 (UTC)[reply]

It could have helped to state that you wanted to validate the current position of the foundation, this is unclear from the current consultation and now it feels like a random Q/A without effects. The simpler RFC seems to be (in pseudo): "This is the current status quo, and this is what it means. For the following scenarios please indicate which information (none, hostname, url etc) you would be comfortable with disclosing" maybe a sibling question of "Do you think the Foundation's current referral disclosure should change from what it is now ? "

Also i didn't really ask about partner's, it was an example. But I see that you have a keen interest in that particular area. —TheDJ (talk • contribs) 16:33, 1 June 2017 (UTC)[reply]

I don't want to validate the current position of the foundation. I want the users of Wikipedia to have a say about whether the sites that they visit when they click on an external link are sent information that allows the site to figure out what Wikipedia pages the users read. --Guy Macon (talk) 20:37, 1 June 2017 (UTC)[reply]

For what is worth, as a lay Wikipedia editor, I found the question posed in this RfC clear enough to answer it. Still from the point of view of a lay Wikipedia editor, I don't really care if some of the options technically end up validating the Foundation's choices and others invalidating it; the important thing is that it be made clear the results are not binding, and it was. Knowing these facts, I will subsequently be able to assess for myself whether the Foundation heeds the community's wishes, or ignores them. LjL (talk) 23:42, 10 June 2017 (UTC)[reply]

The problem is assuming the Foundation still believe their decision is the correct one, they will probably prefer to ignore this RfC. But they'd probably also prefer to be able to do so without saying "sorry but we don't care what you think". It seems to me this RfC unfortunately is likely to give them that. For starters, they're likely to argue that it wasn't sufficiently neutral since there seem to be a major focus in the opening on the evils of referers with only a brief mention of how "Many web sites have a legitimate desire to know who links to them" which is then immediately followed with talk of spammers and and something which if it were in an article would be called up for both WP:Weasel and lack of citation about "Most users, including those who fear surveillance by governments, corporations or criminals" (not even especially). Yes there is this mention here of one advantage of referrers, but realistically most participants aren't going to read it.

In fact, it seems clearly a number of people either didn't read or didn't understand the opening part. At least one seemed to think a referer was something to do with HTML5 rather than, as me and someone else pointed out, something that had existed since 1996 or whenever. Others seem to think the info is being sent by wikimedia rather than the pages served by wikimedia telling the browsers, "this is what we suggest you do" and the browsers following it.

The fact it isn't clear how many people even understand the history of referrers, is something I'd imagine they could really use to their advantage. While nominally there's nothing wrong with people deciding to this is what we should do without understanding the history, it's also easy to say "people would have made a different decision if they did understand the history". Some comments indicate a clear understanding of the history of referrers but for the others, it's easy to ask "do you understand that this is what was happening when people were using HTTP, and the default even with HTTPS is the info will be sent provided the other page is secure it's only with non secure environments it isn't sent to avoid leaking of excessive info".

One more thing that concerns me is there seems to be a strong focus on how this will allow people to figure out what pages are being read even if only the domain is being sent. Yet there's only very limited (yes it was mentioned below where the issue of unique URLs came up) mention of the fact that it's unclear how much this is is genuinely a concern since anyone likely to have nefarious aims for this info is probably just going to find some way to ensure their views for that URL only come from wikipedia or mirrors. (It seems unlikely anyone with nefarious aims is going to be particularly concerned whether the info is being read on wikipedia itself, or on some other page which copied from wikipedia.)

In reality I'm not sure how many people would have !voted differently even if they knew all this, but the fact we can't rule it out give the WMF a good excuse for not following this. Although it's also worth noting that people may not !vote simply because it seems clear where this RFC is headed. When I first saw this I considered !vote probably for option 3 or 4, but it seemed clear at the time, and even more clear now, that there was no point. It would be lost in a sea of option 5s. While as said, I suspect that things wouldn't have been any different if the RfC was different it's also impossible to say so there's all those who were put off by !voting either by the RfC or by the current results too. Further even if we were to organise another RFC when they reject this one for these reasons and more, they could easily say the whole situation is tainted now because of any stink so they feel they will only consider an RFC in a year or two when it's died down.

In case it's not clear, I have a strong suspicion this is precisely what's going to happen, the WMF will reject this because of the way the RFC was done and say the whole thing is too messy now. A bunch of people will get angry and talk about how the WMF is once again ignoring the community. We may or may not have a second RFC, and if we do people will probably get angrier. The WMF will make some noises about reaching out to the community etc and nothing much will happen. Maybe if we really get angry enough they will relent, but even if that does happen it could probably have happened with a lot less fuss. If we want to properly engage with the WMF and get our voices heard, we need to IMO seriously consider how we approach stuff like this. That means for example that any RFC telling the WMF something needs to be very carefully drafted so that they can't pick holes 10 kilometres long in it.

Note that even if we put aside the WMF, I'll be blunt if this was solely a community decision, I'd likely strongly oppose its implementation based on this RFC.

Nil Einne (talk) 16:46, 12 June 2017 (UTC)[reply]

P.S. Of course some may say my comment has tainted the RFC although I think the chance many will read it is slim. But anyway as may be obvious, I've felt this way since I first saw the RFC when I posted below, but waited to post for various reasons including to give more time to confirm my suspicions which unfortunately have been and also to reduce any possible effect. I wish this RFC well since it's obvious some people feel strongly about this and I don't really care that much myself. But I just don't think it's actually going to achieve anything and if it does, it will only be with a lot of unnecessary angst. I couldn't hold off for longer since I've been spending way too much time on wikipedia recently and need to cut it out so didn't want to leave it hanging but also felt this had to be said. There just seems to be too much unnecessary drama between the WMF and the community and while yes, they've definitely been a major contributing factor I feel so have we and people don't seem to appreciate that part.

Nil Einne (talk) 16:58, 12 June 2017 (UTC)[reply]

Interestingly, I had the same thought while writing the RfC. The problem is, I literally could not think of any benefit to Wikipedia from sending referrer information. I could think of a benefit to the WMF; they are likely to think (rightly or wrongly) that sending referrer information somehow increases large donations. But I didn't want to list that because I thought that a lot of editors would see it as a reason not to send referrer information.

If Wikipedia-WMF relations weren't broken, and RfC like this would lead to a productive discussion with someone at the WMF where they explained in detail why they are rejecting this proposal. Alas, Wikipedia-WMF relations are broken, and the all-too-predictable response from the WMF will be stonewalling; Not saying yes, not saying no, and not saying why. --Guy Macon (talk) 21:12, 12 June 2017 (UTC)[reply]

As I have mentioned on my talk page to Guy: I am going to have a bit more of a response tomorrow or the next day, but I want to be clear: this has nothing to do with money or a relationship with commercial interests for commercial interest sake. This has more to do with the ability of the Wikimedia movement and individual Wikimedia communities to signal their impact to aligned communities working within a large system of reliable sources and information on the internet. Guy asked me "what exactly, will Wikipedia be getting out of sending referrer information, and of course the same question regarding the WMF as opposed to Wikipedia"? I wanted to point out two main components that led to myself (longtime Wikimedia editor and outreach volunteer) and Dario Taraborelli (long time Wikimedia researcher and advocate) advocating for the current referral policy:

First, we at WMF are assuming that Wikimedia is central to a much larger knowledge ecosystem, rather than only serving Wikipedia community/reader needs, per an understanding of impact and mission closer to what is being discussed around themes #4 and #5: in the movement strategy process. To participate in that ecosystem, and benefit aligned knowledge organizations: we have to acknowledge that their motivations are different than ours. WMF and Wikimedia might be largely self-funding and open/free/privacy purists, but those institutions that we work with are not: and need some way to justify either funders or sponsors (rarely commercial) the impact of their work, and most orgs operate with two metrics: citations and referrals/pageviews on their own websites. As good citizens in that ecosystem, we need to provide at least some signal that helps folks evaluate their impact through us -- if we signal total silence (Dark traffic), then they assume the impact is from Google or some commercial or not-aligned-with-Wikimedia's-values source and invest money and time in SEO rather than free-knowledge infrastructure or they never get found by the public because they don't have financial resources/savvy-- either way, this is bad for the free and open internet's reliable content.
Second, I personally, in my volunteer capacity, have experience working with a GLAM, that only wanted to support my Wikimedia contributions, because they increased the visibility of their resources (not my main goal by far, but it opened the door to working with them), see: WP:Blake. I hear that story from a lot of long-time Wikimedians who are working with GLAMs, and we also meet a lot of librarians who choose to join the community because of realizing referral stats, but end up staying as valuable allies and contributors in the community well beyond that initial metric (a few published case studies: here, or for example, or example, or example). Our theory of change around this in my team (Community Programs) and Dario's team (Wikimedia Research), is that without first realizing the value of Wikimedia projects through the internal metrics, they never have a motivation to join us, or at least recognize we are allies. Without that first layer of discovering our impact, the rest of our arguments for being allies die (including ones dependent on the values of free, open, secure and private).

When WMF Security and Legal describe full silence as a referrer as having marginal impact on end-user security or privacy compared to our current situation: to me it doesn't make sense to sacrifice the above, high-impact and scale opportunities (movement orgs allying with aligned digital-knowledge-sectors, and recruiting expert communities to participate and invest in Wikimedia projects). I will have a bit more tomorrow, hopefully, but if there are any questions or clarification, let me know. Astinson (WMF) (talk) 22:25, 12 June 2017 (UTC)[reply]

I would encourage everyone to give serious thought about the view expressed above. Yes, it disagrees with my view in many ways, but this is a good thing, and I am glad to see the the RfC is becoming less one-sided. We all want what is best for Wikipedia, and to my way of thinking a vigorous debate about what, exactly is best for Wikipedia is much more helpful than a bunch of editors who agree with me. Too many times in my life I have had that happen and later realized that groupthink kept us from making the best decision. So let's avoid knee jerk reactions and give Astinson (WMF)'s arguments careful consideration.

Minor housekeeping detail: Astinson (WMF), you wrote "...WMF Security and Legal describe full silence as a referrer as having marginal impact on end-user security or privacy compared to our current situation...". Do you have a link to the discussion where they said that? --Guy Macon (talk) 00:59, 13 June 2017 (UTC)[reply]

Found it. See Wikipedia:Village pump (policy)/Archive 126#Review of the change in terms of privacy concerns. --Guy Macon (talk) 12:25, 13 June 2017 (UTC)[reply]

Yep, that is the one I am referring to. We are re-reviewing this right now, but the initial indications are much the same conclusion. We may have some other additional thoughts as well, Astinson (WMF) (talk) 15:00, 13 June 2017 (UTC)[reply]

Also, wanted to ping a few folks that asked for a better arguement or might have thoughts based on how they are commenting above:@WNT, Godsy, Iridescent, Elmidae, Tcncv, and Rivertorch:. I will add a bit more later too, and will ping other folks, Astinson (WMF) (talk) 15:00, 13 June 2017 (UTC)[reply]

In a purely personal capacity, I didn't sign up to be and don't recognise myself as being central to a much larger knowledge ecosystem, nor do I want to be part of a project that amends its policies to make it easier for Facebook et al just because the WMF have arbitrarily decided they are "strategic partners". Of the organisations on the list you link above, at a rough guess I'd say I'm actively hostile to working with at least 25% of them, and at best neutral about the remainder. ‑ Iridescent 15:16, 13 June 2017 (UTC)[reply]

@Iridescent: As regards the graphic linked above. They aren't all partners or allies in that map, but rather folks who influence our movement in one way or another (or act on us whether or not we work with them-- for example Facebook and Google). There are some groups, that we are much more closely aligned with: and those groups are the ones that we are most interested in supporting.Astinson (WMF) (talk) 17:44, 13 June 2017 (UTC)[reply]

At this risk of making it seem like Wikipedia Library is particularly impacted by or invested in this decision, I really want to echo and emphasize the points Astinson is making above. Wikipedia has become a cultural phenomenon, a part of news cycles, university curriculum, scholarly research, institutional engagement, community outreach, diversity campaigns... we have become visible to the point of ubiquity.

As individual editors it may be tempting to think that our service to readers ends with what we write and what they read, but that is putting blinders on our increasingly central role in the entire knowledge ecosystem. Our massive amounts of traffic to our site tell us that we are doing a good job, but it's the traffic that we send to other sites that tells the rest of the world Wikipedia matters--that it is important, authoritative, un-ignorable--and this leads to a myriad of individuals and organizations seeking to engage with us.

Everything from university libraries to museums to newspapers to human rights groups to scholars... they do care more about Wikipedia because it is visible to them in the traffic we send. I'm terrifically opposed to the privacy-shredding infrastructure of the modern web and take ample steps myself to avoid being tracked, targeted, and advertised to. If we were sending information about the specific pages that users were coming from it would be an egregious risk to them, a loss of privacy, a violation of our ethos as a research service, and a contributor to the beast of internet identity tracking. That, however, in all but a very small number of marginal vulnerabilities (that we still need to look at), is not what happens now.

Sending only the *domain* of a website like ours tells external sites essentially nothing about the individual readers of Wikipedia. All it tells them is that Wikipedia is big, it is critical, it is significant, and it matters. We don't need to erase that last piece of data and go silent, dark, because we're already dark at the individual level. At the site level, this minimal amount of information shines a spotlight only on our projects as a whole, and we as a movement benefit greatly form that in many ways, some of which are also invisible to individual users. Keep users safe in the dark, put Wikipedia in the spotlight--that's what benefits us most. Ocaasi (WMF) (talk) 18:19, 13 June 2017 (UTC)[reply]

@Astinson (WMF): Some of the technical specifics of this issue are way over my head, but since you asked, I'll say this much more. Wikipedia is the only top website that is totally noncommercial and doesn’t attempt to monetize its visitors. That’s nearly miraculous, considering what the World Wide Web has become, but it won’t stay that way if it chooses to forge relationships with other entities whose motives are less pure or whose financial position is more precarious. Wikipedia’s users and visitors have an expectation of…if not benevolence exactly, then at least detachment, when coming and going from the site.

In your reply to Iridescent above, you say that [t]here are some groups, that we are much more closely aligned with: and those groups are the ones that we are most interested in supporting. But how can we be sure who those groups are aligned with now, or who they’ll be aligned with in the future, and how can we be sure what data might ultimately be passed along? I realize that slippery-slope arguments are sometimes alarmist and illogical, but when it comes to Internet user data, I think these are questions worth asking.

On a more practical level, referrals are a concern for a at least two reasons that I won’t go into here, except to say that the more important of the two conceivably could affect the physical safety of vulnerable people in certain places. I can even imagine instances in which the Wikipedia domain alone could pose a problem. It’s not likely but it could happen, and if we accept that, then we should accept that it is incumbent upon the Foundation to take that into account when deciding how much information its servers pass along to other sites. Eliminating risk is impossible, of course, but minimizing it should be the goal—and that means providing no information whatsoever, whenever that is possible. Rivertorch^FIRE_WATER 19:32, 13 June 2017 (UTC)[reply]

In the comment above, Ocaasi (WMF) claims that "Sending only the *domain* of a website like ours tells external sites essentially nothing about the individual readers of Wikipedia." That isn't true. Consider the case of someone reading the Wikipedia page at Bomb-making instructions on the internet#References who clicks on the link to The Low Cost Cruise Missile: A looming threat?. A quick check with the link search tool we helpfully provide will tell the owner of aardvark.co.nz (and anyone who hacks that site or who has a court order) that the only two articles on Wikipedia with that link on them are are Bomb-making instructions on the internet#References and Bruce Simpson (blogger)#DIY Cruise Missile. Thus it is not true that the owner of aardvark.co.nz knows "essentially nothing". Instead he knows that the Wikipedia user was reading one of two Wikipedia pages. In cases where there is only one page with the link, he not only knows exactly which page the Wikipedia user was reading when they clicked the link, he knows exactly which sentence they were reading. And we already know of at least one person who got a visit from the police after Googling pressure cookers and backpacks. (See Locals Questioned by Suffolk County Police Department after Googling "Backpack and "Pressure Cooker", and Google Pressure Cookers and Backpacks, Get a Visit from the Feds.) There are places where you can be arrested and tortured for reading the Wikipedia page Bomb-making instructions on the internet. We should not take revealing who reads that page lightly. --Guy Macon (talk) 20:57, 13 June 2017 (UTC)[reply]

Guy, I'm out of my depth here technically and so would rely on WMF privacy, security, and legal advice for risk specifics. In other words, take this with a grain of salt. I think that you're not wrong about risk, but not right enough to make a meaningful distinction in the risk-benefit analysis. The scenario you presented, of a clicked-on link to a site about low-cost missiles traced back to the Wikipedia article on bombmaking is a great example. It's a scary scenario, but misses the bigger point and actual privacy risk readers face. If the reader is visiting the site on low-cost homemade missiles and their ip address or browser information is exposed, that alone and in itself is problematic--and no moreso than adding "and I came from Wikipedia's page on the subject to get here". If a user needs protection from that level of ip surveillance off of Wikipedia, then they would need to be blocking/hiding/masking their ip to the websites regardless, and Wikipedia's referrer info wouldn't make a significant difference. It even gives a false sense of security that having a silent referrer protects you from tracking once you leave Wikipedia; indeed, far more cautious and rigorous (but available) methods are needed for that. User privacy once someone leaves Wikipedia is very important, important enough that we need to inform users to equip themselves with tools sufficient to manage their actual risks. The referrer policy being minimal vs. absolutely silent tips the scale only slightly, but the larger issue of exposed/unmasked ip addresses is the real concern: that is the weight which tips the whole scale over on its head--and it happens outside of our control. So would we choose to lose massive network benefit and visibility for a negligible sliver of anonymity that isn't sufficient to protect privacy in the first place? Ocaasi (WMF) (talk) 22:20, 13 June 2017 (UTC)[reply]

I think there are two adversaries here that are kind of being conflated in this discussion:

First adversary - A passive eavesdropper in the middle of the network path. (e.g. Someone in a coffee shop monitoring the wifi, some governments, etc)

Changing our referrer policy has little affect on this adversary as the adversary can sniff the SNI TLS extension (which contains the domain name), the IP destination address (Its easy to find out who owns the destination IP address given the IP). Depending on where in the network the adversary is located, they may also be able to sniff DNS requests. Thus the referrer policy has basically no affect on this adversary.

Second adversary (The target of the link in question)

The entire point of the current policy is to allow the targets of links to be able to generate bulk statistics about where there traffic is coming from without revealing what specific page. To that end, it does leak enough information for such third parties to be able to say x% of our requests come from en.wikipedia.org. However, the referrer is limited to the origin (origin is fancy webspeak for the part of URL before the "path" portion. so https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(policy) for example, has the origin https://en.wikipedia.org). I believe most people consider it acceptable that third parties can make bulk statistics about how many of their visitors come from Wikipedia vs other places on the internet (Am I correct in this)?

From what I understand - the argument here is that if there is a unique-ish url (perhaps with a unqiue identifier in a query parameter) that's only used a few places on wikipedia, the adversary can then use Special:linkSearch to narrow down which specific page on Wikipedia the user is coming from. My counter to this would be:

Is the full url someone is coming from actually "sensitive"? Ideally that wouldn't be exposed as it kind of sounds creepy, but is it actually harmful to expose it? Prior to the introduction of our current referrer policy, this information was regularly exposed (The default referrer policy in browsers exposes this information in many circumstances) with no complaints afaik. Are there realistic scenarios where users would be hurt by exposing this information, and what are they?
Even with the most restrictive no-referrer policy [15], an adversary could still figure out what page a user is coming from by putting the url only on Wikipedia. This type of attack relies on the url being relatively unique, thus it makes sense that someone doing this attack would only use the link in one place, meaning that they don't even need the referrer. It should be noted that google webmaster tools allows site owners to see all the people who link to them (Like Special:Linksearch but for the entire internet).

Thus, I think the privacy concerns for the current referrer policy are overstated, and even if they are an issue, changing the referrer header will not "fix" the problem well it would seriously inconvenience people who want to run bulk traffic statistics on their website. BWolff (WMF) (talk) 23:28, 13 June 2017 (UTC)[reply]

@BWolff (WMF): I said the same thing above. However, I don't think that is the last word. The thing is, if you simply put a "sting" link on the Internet, you don't really know how a user came to it. Unless you're told. It's still at least plausible that there is some other way your link ended up sitting clickable somewhere. And the other thing is, well ... I'm not really feeling accommodating toward Internet commerce. I mean, there are so many companies and governments trying to figure out everything you do on the internet and have one God-AI Who Simply Knows Everything About You, and if Wikipedia were to not go along with them there would be a certain non serviam satisfaction to it all. Let them write articles and reports and estimates how much Wikipedia traffic there is. What does it hurt? Even if they misestimate how much there is, and therefore underestimate the value of advertising their company on Wikipedia ... is that a bad thing? Really? Wnt (talk) 23:57, 14 June 2017 (UTC)[reply]

@Wnt: [Switching to my personal account to emphasize this is just my personal/volunteer opinion]. The flip side of not really knowing the sting link came from Wikipedia even if it is very likely it did, is that even with referrers, there's no guarantee. People can hack their web browser to make the referrer be anything, including making it be Wikipedia even when it shouldn't be (Or use command line tools like curl or wget). However that's pretty unlikely.

From what I understand, one of usecases for referrers is GLAMs who freely license photographs (or other Media) so that the media can then be used on commons and then Wikipedia and want to track the "impact" of their media "donation". GLAMs who open up their media often do get a traffic bump due to people clicking on the source link, or perhaps because upon seeing the image in question users become interested in the object in question, and want to do follow-up research around the internet. Having the referrer data helps these groups quantify the effect that free licensing their collection has on their web properties. This in turn helps their staff make the case to their bosses internally that investing effort to re-license media is worth it. GLAMs are after all like any other organization - staff have to prove that their pet projects further the organization's mission. The result of all this is we are able to get access to multimedia we wouldn't otherwise have access to. Now if I believed the referrer data actually compromised the privacy of our users, I would say that the trade off is not worth it. But I do not believe that it does, and as a result of GLAM collaboration we have a lot of cool photos that we would be unable to otherwise have (Can't exactly send a volunteer back in time to photograph an event that happened 50 years ago). Additionally, I don't consider this type of third party to fall into the evil-world-dominating-corproation category. Bawolff (talk) 07:25, 15 June 2017 (UTC)[reply]

A solution that satisfies privacy and GLAM requirements?

@Bawolff: I went back to the original specification and found something interesting: it is actually possible to set referrer policy on an individual link using <a referrerpolicy=... >. That means that if GLAM volunteers feel a dire need to get this information out for some specific links, WMF could accommodate them. (There are also supposedly some funny Javascript methods to set all the referrers everywhere to a deliberate fake address e.g. www.wikimedia.org ([16]) though, as often the case with Javascript, it's a crapshoot what happens.) But in general, privacy should be the policy, and we should look for very compelling reasons before tolerating even very limited exceptions. When push comes to shove we don't have to get this stuff today; but we do have to resist a flawed culture. Treading water won't seem so bad once the ship sinks. Wnt (talk) 14:33, 15 June 2017 (UTC)[reply]

Good catch! This really looks like a good solution. Right now we are heading down the road to a situation where the consensus of the Wikipedia community is to be a silent referrer, with the inevitable result that those who are more concerned about GLAM will lobby the WMF to ignore the consensus of the Wikipedia community. This is a recipe for conflict. If instead the GLAM people get on board with being a silent referrer and take the initiative to send referrer information to those GLAM partners that need it, then we will have reached a compromise that makes everyone happy. In addition, we can send full URLs to the GLAM partners not just the domain, thus giving the GLAM partners more information than they get now. --Guy Macon (talk) 15:12, 15 June 2017 (UTC)[reply]

This has a lot less to do with specific partnerships, and more to do with all of the different, diverse, and often-not-yet-contacted organizations that ought to be partnering with our community. There are thousands of Libraries with published content cited on Wikimedia projects, similarly there are dozens of new academic journals being added to our citations each month, expecting Wikimedia communities in many languages to conform to a rather complex format for the citations: even if we automated this in one way or another, you still are actively damaging the impact of these knowledge organizations, who share the same the same fundamental end-goal and mission as us. Astinson (WMF) (talk) 16:28, 15 June 2017 (UTC)[reply]

Astinson, I appreciate that everyone at the WMF pushing this is acting in good faith, but when you have every non-WMF-affiliated person to express an opinion disagreeing with you, this is the recipe for a repeat of Flow or Superprotect. While this is my personal opinion, I'm fairly confident that I'm expressing the general will of the community when I say that if a particular self-appointed "partner" is making demands that are totally at odds with Wikipedia's basic ethical and cultural standards, that's an argument for breaking ties with that organisation, not for redesigning Wikipedia to accommodate the whims of whichever publisher the WMF is currently trying to schmooze. If a particular institution refuses to cooperate with Wikipedia because they don't like our policies, that's their loss not ours. It's to the WMF's credit that they don't generally try to throw their weight around, but that doesn't mean we should lose sight that we are the cultural phenomenon leading the revolution in how information is disseminated, not the William Blake Institute, the Metropolitan Museum of Art or even Elsevier, Gale and Google, and we shouldn't be afraid to dictate to other organisations the terms under which we are willing to allow them to work with us, rather than go cap-in-hand asking them what compromises they'd like us to make in order for them to give us the crumbs from their table. Put bluntly, there is no partner whose loss would cause significant damage to Wikipedia; it seems important to the handful of editors who make use of it, but if the whole of WP:TWL shut down tomorrow the impact would not even be noticed by 99% of Wikipedia's active editors. (I've managed to get by thus far without once having interacted with either TWL or a GLAM partnership, and I'm writing primarily on the visual arts where one would expect that kind of thing to be more common. The whole TWL/GLAM/WIR alphabet soup is really not as important as those who are connected to it think it is.) ‑ Iridescent 17:00, 15 June 2017 (UTC)[reply]

Astinson (WMF), the problem is that you are asking the WMF to ignore the wishes of the Wikipedia community, which quite frankly values user privacy over GLAM. Are you "sure" that you want to fight that fight? It is likely to end up with you losing the fight, but only after a huge shitstorm. --Guy Macon (talk) 17:30, 15 June 2017 (UTC)[reply]

(edit conflict) @Iridescent: I would agree with you: if this were about satisfying a handful of organizations or momentary partnerships with the change, it wouldn't be worth it, but we are talking about a web-wide utilized standard for metrics among informational websites. Our hard-line on other concerns throughout the movement (including free and open copyright, community based editorial decisions, etc), actually are leading in a practice shift as you mention.

Yet these external communities are a significant component of our impact. If you haven't worked with these other knowledge organizations, or haven't found them useful yet, that doesn't mean that your experience is one shared by all facets of the movement; there is ample indication that partnership and collaboration with organizations and communities outside of our projects is a top-level priority for many parts of the Wikimedia movement: see the various themes emerging in the strategy process. Moreover, in this conversation, there are several indications that folks who might have opinions can't meaningfully engage here (take User:Mike Christie's comment). I, and other WMF staff, am engaging in good faith, in part because we recognize that the privacy concerns expressed here effect a certain subsection of the community, but the project-wide way of addressing those concerns which the RFC proposes doesn't fully account for the various other stakeholders in the conversation. As I mentioned a couple days ago, we are consulting internally to see how we can support the privacy concern, while not discounting the impact of this whole other part of the movement. Astinson (WMF) (talk) 17:58, 15 June 2017 (UTC)[reply]

@Astinson (WMF): I have to wonder how far you're planning to go with that. I mean, if Facebook tells you they'll pay all your bandwidth but in exchange you have to put their tracking scripts and like buttons and web bugs on every page, what do you say? It seems like it would be better to stick to a hard line now when the bribes are small and resisting them is easy, rather than going down the garden path and having to say no when the money is huge but the consequences are a loss of fundamental values. Wnt (talk) 20:12, 15 June 2017 (UTC)[reply]

Again, FUD. what are you even talking about here. As alex says at the end of his comment, if anything the Wikimedia movement has been leading the privacy debate.. You ask the security staff for information, and you blame them for engaging ? You ask people who have a good use for the current referrer policy to explain this, and you are surprised they answer ?? —TheDJ (talk • contribs) 08:01, 16 June 2017 (UTC)[reply]

I did not mind that he answered - but I do want to take that opportunity to point out to him what direction such an answer could eventually lead toward. Wnt (talk) 10:50, 16 June 2017 (UTC)[reply]

I have been reading this thread quietly, and want to respond here, as Astinson has articulated a key thing. The editing community is what creates and maintains the key asset here - content. Without content generated and maintained by the community, WMF would have nothing to do. It is content that people come to find.

It is absolutely true that WMF enables content creation and maintenance by providing the servers and software, and that is a great thing. And I appreciate WMF's efforts to grow the movement via things like alliances. But what folks in WMF seem to forget all the time, is that this peripheral stuff - like creating alliances - is not content creation and maintenance. It might enable content creation and maintenance, but it is not itself content creation.

So - don't put the cart before the horse. The WMF does not lead -- it serves. When the WMF (and WMF employees) forget that what they are doing is supportive or peripheral to the core asset-generating miracle here (volunteer content creation and maintenance) and act as though WMF and its goals are central, it has lost its way.

What is the the generator of the miracle? People have spent lots of time studying that, but at the bottom it is probably values. We do this for free, because it is made available freely. We do this anonymously, and we fiercely protect our privacy and want to protect the privacy of readers.

What you are failing to hear (quite completely) is that what you are advocating runs hard against these two core values of no-money and privacy. You want to "monetize" the link-trail. And why? To further the WMF's goals. You should be very wary of violating core community values to pursue WMF goals. That is ass-backwards. Jytdog (talk) 04:43, 17 June 2017 (UTC)[reply]

Just to make sure that we are all talking about the same thing, if Wikipedia puts the following in the head of the HTML...

<meta name="referrer" content="same-origin">

...we will send no referrer information when a user clicks on a link to a non-Wikipedia page and full referrer information when a user clicks on a link to another Wikipedia page.

If we then add the following to selected links...

<a href="http://example.com" referrerpolicy="always">

...this will override the meta tag in the head for that particular link.

Thus we can use the meta tag to make Wikipedia a silent referrer for all outgoing links and then override that policy with referrerpolicy on the link for any sites that are of particular interest to GLAM or The Wikipedia Library. This can all be done with a bot; all we humans would need to do is to make a list of what sites we want to send referrer information to.

Several sources say that all major browsers support setting a referrer policy for the page and then overriding it with a referrer policy for the link (which is what the standard at [ https://www.w3.org/TR/referrer-policy/#referrer-policy-delivery ] says they should do) and an extensive web search has turned up zero evidence of any major browser not supporting both. A few seldom-used browsers (Internet Explorer under Windows 95, Opera Mini (not to be confused with Opera) don't support either. --Guy Macon (talk) 07:05, 18 June 2017 (UTC)[reply]

Arbitrary break 02

Unsourced claim that has been proven to be false

The following discussion has been closed. Please do not modify it.

Only Chrome supports this attribute so far, and none of the other vendors have indicated they are going to be adding this any time soon. —TheDJ (talk • contribs) 15:32, 15 June 2017 (UTC)[reply]

@TheDJ: If Chrome supports the attribute, and ... unfortunately ... I fear Wikipedia knows exactly how many visitors use Chrome, then it should be deducible almost exactly how many total visits were from Wikipedia. Now, you can say that site admins are uninterested in figuring out such details, except... the premise is that someone at those museums *is* deciding whether to share with Wikipedia, which means they should have a moment to think about it. Also, if people at WMF really care, remember that Firefox is a public project sort of like Wikipedia, and perhaps you should be coordinating with them more anyway. As for the Chinese company that has the TOS that says "Opera" has access to six months of users' browsing history, well, should they be dictating what you do? Wnt (talk) 20:02, 15 June 2017 (UTC)[reply]

Where is the evidence that only chrome supports use of <a href="http://example.com" referrerpolicy="always"> (which is what we need to override a "same-origin" (silent referrer) referrer meta tag on selected links)?

According to [ https://caniuse.com/#search=Referrer%20Policy ] Referrer Policy is supported by:

Microsoft Edge
Firefox
Chrome
Safari
Opera
iOS Safari
Android Browser
Chrome for Android

It is not supported by Internet Explorer, but because Internet Explorer also doesn't support the referrer meta tag or referrerpolicy on the link, IE will always have the default HTTP behavior no matter what we do with our meta tags and links.

According to [ http://searchengineland.com/need-know-referrer-policy-276185 ], There are many ways you can deliver the referrer policy:

Via the Referrer-Policy HTTP header
Via a meta element with a name of referrer
Via a referrerpolicy content attribute on an a, area, img, iframe, or link element
Via the noreferrer link relation (rel=) on an a, area, or link element
Implicitly, via inheritance

--Guy Macon (talk) 03:44, 16 June 2017 (UTC)[reply]

No answer, so I asked at Wikipedia:Village pump (technical)#Help me, Obi-Wan Kenobi! You're my only hope!! (Referrers again)!. --Guy Macon (talk) 13:36, 17 June 2017 (UTC)[reply]

Also at Wikipedia:Reference desk/Computing#Help me, Obi-Wan Kenobi! You're my only hope!! (Referrer policy). --Guy Macon (talk) 13:47, 17 June 2017 (UTC)[reply]

Bottom Line: The user who claimed "only Chrome supports this attribute so far" has failed to provide any evidence to support that claim, there are sources that say otherwise, and the users has made other claims in this discussion[17] that appear to be things he just made up. Until he provides evidence, I suggest ignoring / disregarding his claim about what browsers support setting a referrer policy for an individual link that overrides the referrer policy for the page. --Guy Macon (talk) 07:05, 18 June 2017 (UTC)[reply]

...aaaand, after I wrote the above, the refdesk came through[18] and proved that the claim ("Only Chrome supports this attribute so far, and none of the other vendors have indicated they are going to be adding this any time soon") that started this section is utter bullshit. I am archiving this section as the total waste of time that it is. --Guy Macon (talk) 15:48, 18 June 2017 (UTC)[reply]

Examples where sharing referrer information is a benefit to Wikipedia

It's easy to point to examples where sharing this kind of information is bad, but I've seen sharing some information be a real benefit to Wikipedia, and wanted to make sure we talk about that. I have an obvious conflict of interest for this particular example, but it's the one I'm most qualified to talk about; The Wikipedia Library. As much as we would love for all the publishers and databases who give free access to Wikipedians to be doing so purely out of the good of their heart, some (not all) are doing so at least partly from a business perspective. More Wikipedians citing their content on Wikipedia means more people clicking through to their content - that's unavoidable. As such, one of the metrics by which some organisations judge whether they want to continue giving Wikipedians free access to resources is traffic from Wikipedia, which they can only track if we're at least sending them the current amount of information (wikipedia.org). Removing that information would have a demonstrable impact on TWL's ability to expand, as well as to continue existing access donations. My main point being, the ability to monitor traffic levels from Wikipedia is a really important way to convince organisations to get iolved with the Wikimedia community; if they can see obvious benefits to doing so they're much more likely to work with us to make the Wikimedia projects a better place. I know there are other examples where sharing this data is ultimately beneficial to Wikipedians, and would encourage anyone who has a similar case study to share it. Sam Walton (talk) 12:06, 2 June 2017 (UTC)[reply]

I am looking at Wikipedia:WikiProject Resource Exchange/Resource Request, and I am not seeing any examples where a publisher or databases gave free access to the world, which they would have to do in order to get traffic from Wikipedia or anyone else. Of course me not being able to find it does not equal it not existing. Do you have any specific examples that lead to links on Wikipedia? --Guy Macon (talk) 22:27, 3 June 2017 (UTC)[reply]

@Guy Macon: My example wasn't about the resource exchange, but rather the access donations. See File:Wikipedia Incoming Traffic Graph.png for some example traffic data, showing the increase for one TWL partner, which led them to be enthusiastic about continuing to give free access to Wikipedians. That plot is also interesting to see the HTTPS dip, before the referrer tag was put in place :) Sam Walton (talk) 23:11, 5 June 2017 (UTC)[reply]

The HTTP referer article gives an in depth explanation of referrers for anyone else who was largely unfamiliar with the concept before this like me. — Godsy^(TALK_CONT) 07:50, 10 June 2017 (UTC)[reply]

((EC) Yes perhaps one significant point in light of the comment above it it's something that's existed and been used since 1996 and has nothing to do with HTML5. As I understand it, the particular issue of concern here is that HTML5 has a "meta referer" tag that can be used to overide default browser referer policy. Default policy for most browsers is to not send any referer tag when HTTPS was used to connect to the originating website. The WMF did not used to implement that tag meaning once wikimedia sites were switched to HTTPS only, referers would not normally be sent for any user. (Prior to that, any user using HTTPS would generally not have sent a referer but anyone using HTTP would generally have done so.) They've now added the tag meaning that a referer will be sent again, similar to was done before with HTTP but now for everyone using HTTPS (unless they get their browser to overide default behaviour) except they specific origin meaning only the hostname will be sent unlike before where it would have been the full URL. Meta:Research:Wikimedia referrer policy So while this specific question has to do with HTML5, referrers are not some sort of fancy HTML5 web stuff. Nil Einne (talk) 04:11, 11 June 2017 (UTC)[reply]

That chart raises a question in my mind, and perhaps the suspicion that there is something I don't understand about referrers in the HTML handshake vs referrers in the meta tag. The file description says "Trend line does not include this dip or the recovery period shortly afterwards". Why would there be a recovery period? The trend line instantly dropped to zero when we stopped sending referrer information. Why didn't it instantly recover when we resumed sending referrer information? --Guy Macon (talk) 11:57, 11 June 2017 (UTC)[reply]

Yeah, good question, I wondered that too, and don't know enough about the technical details to say. Sam Walton (talk) 07:16, 12 June 2017 (UTC)[reply]

@Samwalton9: To begin with, if there's a quid pro quo, it's not a donation. And if there's a quid pro quo, what is that? If all we're paying is eyeballs, then any method to estimate those eyeballs should be sufficient. Now that they know how many people were referred, we don't need to prove it again. But if what we're paying is user privacy, then this devil's deal needs to be stamped out posthaste. If the companies are that eager to see referer data that they would give real value and then take it away if it's not given, that's exactly the reason not to give it! But I think the real reason is just that the access donations don't really cost them anything at all; they know that they lose only a handful of sales, and get them straight back again in free advertising, no matter how you count the eyeballs. Wnt (talk) 23:19, 11 June 2017 (UTC)[reply]

@Wnt: Problem is, there is no real way for the "sponsor" (not "donator", I agree on that point) to quantify the "eyeballs" without the referer info. In the example given above, Wikipedia-originating pageviews are a few hundreds per day. If the website receives 1000+ hits per day with a pattern of increase (as is usually the case when you are taking SEO measures), it would be extremely hard to determine the impact of Wikipedia exposure if you just have to guess. This particular sponsor may now know how many WP-originating pageviews there are, but for another new sponsor, or that sponsor after a change in the IT department, without the referer there is no easy way to quantify the page views.

If the only way to count the eyeballs is to allow the sponsor to know the browser history of whoever's brain is at the other end of the optic nerve, I would support not counting eyeballs at all ("devil's deal", as you say). I would expect the WMF to take a similar position (unless/until finances get tight), but it is not a noncontroversial position, as much as I would like it to be. Tigraan^{Click here to contact me} 16:34, 12 June 2017 (UTC)[reply]

@Tigraan: So? They guess. You make it sound as if otherwise it were an exact science -- like they have any real idea how much the added traffic really translates to sales, given that someone who finds out they exist via Wikipedia will probably come back and look at them on some other occasion before putting down any money. Yeah, they might not like not being in control and they might offer us fewer shiny baubles in exchange ... but how much difference do those really make anyway? Wikipedia is swimming in cash, could buy all the subscriptions it wanted if it felt like it. The fact is, old fashioned advertising didn't come with links - you put your ad in the paper and maybe you got a chance to ask a few people what brought them to your shop. We should strive to be a precedent for turning the entire web back to old fashioned advertising - no more tracking, no more scripts, no more spies, and, sadly for many of us, but happily for the companies, no more ad blockers. Because you can't adblock a straight image or a link on a page. That's the leadership we should be showing. The corporates who invaded our educational internet a decade and a half ago have made a trash heap of it, only good for getting your computer ransomed, and we ought to stand up for a better way. Wnt (talk) 11:48, 13 June 2017 (UTC)[reply]

How much views translate to a sale is not an exact science. But how much views originate from which referer is. Maybe the sponsors are idiots who don't realize this, and they are wrong in wanting referer data; maybe they are evil megacorporations or useful idiots of those and they are morally bankrupt for wanting the data. But the fact remains that they want the data, and will stop sponsoring if we don't give it to them. Should the WMF become 100% convinced of your point, which I doubt, it would still be unlikely to have the inclination to argue with every and each sponsor to convince it. (I agree with you 100% about the adverspying in today's internet, but SEO and referers were already a part of the previous model, so I don't see how it is relevant here.)

As for buying all possible subscriptions, well, see Elsevier#Pricing. The WMF is probably not thrilled at the idea of coughing north of $1m per year and giving it to one of the worst offenders in terms of restricting access to knowledge. Tigraan^{Click here to contact me} 14:02, 13 June 2017 (UTC)[reply]

Surely there is a middle way to be found, e.g. enabling origin-when-cross-origin only for .edu sites. I would expect an organisation with $80 million in revenue to do a better job negotiating instaed of giving away all of this data just to solicit more donations. This is technically speaking unidentifiable/anonymised data, but in practice some corporations out there can pretty much reconstruct large parts of a person's browsing history based on such data, and I'm quite annoyed that many people refuse to acknowledge this. Daß Wölf 00:26, 14 June 2017 (UTC)[reply]

@Tigraan: If they don't want to cough up a million in subscriptions, they should be leery of coughing up a million in privacy. Do they really need all those subscriptions? Are they really worth a million? Elsevier knows full well that many of the Wikipedians do not work at a well-heeled private scientific company that can afford to pay per article or by subscription, and therefore, they lose no money or value by letting them subscribe. And they also know that many of us prefer to use Sci-Hub even above applying for free Wikipedia-related access. I mean, it costs them nothing to give this access, it's worth very little, we don't need it, but it ends up leading to them getting valuable exposure even if they can't quantitate it reliably. And if all that fails ... we can go back to the old status quo, no subscriptions, we get the articles whichever way we can, whether it's Sci-Hub or Twitter tags or just plain emailing the author for an eprint like in antediluvian times, or even, God forbid, going down to the library for a formal interlibrary loan. I mean, the thing about bait is you don't actually have to eat it, not if you're not clever enough to figure out what it's sitting on. Wnt (talk) 23:44, 14 June 2017 (UTC)[reply]

Actually seems I misunderstood something I read. I now believe the default with HTTPS is as outlined in the comment above namely referer will be sent only if the other connection is also secure. See e.g. [19] [20] BTW the comment above I referred to in my comment was KMF's comment no under option 5, which was here before it was moved. Nil Einne (talk) 15:45, 12 June 2017 (UTC)[reply]

I posted a list of the various behaviors and when Wikipedia exhibited them under question #6, in case someone wants to !vote for a previous behavior that I didn't list in my list of questions. It may also help everyone to understand what the default HTTP and HTTPS behaviors are. --Guy Macon (talk) 01:45, 13 June 2017 (UTC)[reply]

I think the points in favor of wikipedia.org or en.wikipedia.org referrer information are compelling (@Ocaasi (WMF):), and I would certainly act differently as a content provider analyzing bulk statistics if I saw many such referrals vs. zero, thereby perhaps contributing to the Wikimedia movement. Further, I would say that such benevolent users don't need referring page data, since they are perfectly capable of searching Wikipedia for their links and running pageviews tools on the pages that link to them.

Conversely, I think the hostile actor problem is a real one, if a site is seeking to either track individual users invasively or censor Wikipedia (Turkey). However (contrary to opinions expressed above), I think the solution is to blacklist hostile webhosts and act as a silent referrer to them, rather than whitelist sites we want to offer clear statistics to. In the hostile nation-state case, this could mean becoming a silent referrer to an entire top-level domain. I leave it to others with more technical skills to do that work, but I don't think silencing all our external referrals is a smart first step to address that problem.--Carwil (talk) 19:11, 15 June 2017 (UTC)[reply]

Comment - How would one of options receiving unanimous support affect external websites covering Wikipedia, like Wikipediocracy, especially if a user clicks a link to that website from the article about it? --George Ho (talk) 02:01, 13 June 2017 (UTC)[reply]

It would have zero effect unless Wikipediocracy is secretly trying to gather information on which pages Wikipedia readers have read or on how effective some spam campaign is, in which case being a silent referrer would Foil Their Evil Plans. :)

We know that being a silent referrer will not hurt Wikipediocracy, because we were a silent referrer to HTTP sites such as wikipediocracy.com for five years, from 2011 to February of 2016. Wikipediocracy is probably the #1 site most likely to complain if Wikipedia does something to hurt them. --Guy Macon (talk) 16:09, 16 June 2017 (UTC)[reply]

Thanks to @Guy Macon: for pointing out a use case where this referrer information could be materially harmful to someone. There are clearly other cases where having this referrer data (typically in aggregate form) would be beneficial, notably:

Partner institutions and others who have chosen to make their content freely available discover that doing so is bringing traffic to them.
External institutions discover that Wikipedia is a major driver of readership to their sites and choose to reorient their content in way that is more freely available.
Paywalled sites discover that large numbers of readers come from a free-content site, shifting conversations about the expectations of their readers to access content.
Data like that presented here shows academic content providers that the free-content Wikipedia community is a central audience for them, prompting shifts towards open access at various levels.

I come to this conversation as both a Wikipedia and an academic aware of numerous ongoing conversations about the future of academic research and whether or not it will be open access. Individual scholars, academic associations (like the American Anthropological Association), funding agencies (the National Institutes of Health and the National Science Foundation) are all discussing whether and how research might be made more publicly accessible. One compelling argument in that process is the fact that open access content is read more, and read more widely; and this referrer data makes that argument more compelling. We're really talking about whether other arenas of content, including academic knowledge production, orient themselves around free and open sharing sites like Wikipedia or around various fee-for-service models.

So, can we address the negative cases, while continuing to reap the aggregate benefits? Some mechanism might be

Silence referrals on particular categories of Wikipedia pages.
Create a silent referrals user preference.
Blacklist surveillance-oriented destination sites to receive only silent referrals.

Fundamentally, there are real tradeoffs here, and I believe the Wikipedia community should recognize that.--Carwil (talk) 17:57, 17 June 2017 (UTC)[reply]

I agree that there are real tradeoffs here. I simply disagree that any of the cases where sending referrer data could be beneficial are even 1/1000 as important as protecting our users from governments that will imprison, torture, or kill based upon what Wikipedia pages they read. I don't think they are 1/100 as important as protecting our readers from marketers who would love to add "favorite Wikipedia pages" to the information they gather and sell about us. I don't think they are 1/10 as important as making the job of Wikipedia spammers more difficult. And, it appears, the vast majority of those who have responded to this RfC agree.

That being said, even though the cases where sending referrer data could be beneficial are not as important as user privacy or inconveniencing spammers, they are important, and we should provide that information to them if at all possible without violating user privacy.

So, does your suggested solution above meet both needs? No. It does not.

We know that only a tiny percentage of Wikipedia users will opt out of sending referrer information, and besides, every major browser already has that capability.

We know that we lack the manpower to maintain a blacklist of all surveillance-oriented destination sites, and we know that the vast majority of high-traffic sites are surveillance-oriented. Facebook doesn't spend millions of dollars giving you a website for free. No, they sell information about you. You aren't the customer, you are the product being sold.

A blacklist does nothing to address the one scenario that gets people imprisoned, tortured, or killed, which is a government accessing the logs of a website through a court order.

The first time some drug dealer ends up in a US federal court and it is revealed that part of the prosecution's case against them includes the fact that they accessed our pages on Clandestine chemistry and Rolling meth lab -- information that the police obtained only because we send referrer information -- the shit will hit the fan, and this RfC will be extensively quoted in the press. All the more so if the Government is China and the Wikipedia page is Falun Gong.

No, blacklisting is not the answer. Whitelisting is the answer. Unlike spammers and governments looking to find dissidents, the use cases you list above are few in number, easy to identify, and mostly eager to cooperate with us. And, if it turns out that we cannot give them the information they want without compromising our user's privacy, then too bad. They will simply have to do without. --Guy Macon (talk) 19:16, 17 June 2017 (UTC)[reply]

If our goal is to protect Wikipedia users from state monitoring and surveillance at the external links they click on, then silencing referrals for all outbound links is a very blunt tool to achieve that goal. We're talking about two presently hypothetical cases where either (1) "thoughtcrime"-level content on Wikipedia (e.g., Falun Gong page for an aggressive PRC prosecutor) links to innocuous external content, but someone gets traced to reading it; or (2) Wikipedia searches are used to provide evidence of criminal intent (Bomb-making, clandestine chemistry, rolling meth lab). Now, it's far more likely that an interested user in any of these cases clicks through to also objectionable content, and far more damning for them criminally. If we're worried about thought-crime and criminal-intent-through-reading for the user who is on these pages, we should be worried about the same thing when they click on non-innocuous content. They would be even more endangered for those clicks.

Also, even when we silently refer people to pages, those servers get to have the IP address, location, and system fingerprint of the user that connects to them. (Unless browser or Tor-like countermeasures are taken). So the protection silent referral offers is limited at best. If there's a state actor capable of demanding server logs, they could easily request both the destination website's server logs and the corresponding IP address's ISP logs. Again, silent referral isn't so protective.

This suggests two possible response. One, which I mentioned before: silence referrals on pages associated with repressible activities, including the pages we've been discussing. Create a by-page parameter to silence those referrals. Two, and this would take rather more work, use Wikipedia to advertise the privacy risks associated with external links. Every "external links" subhead could include a [your privacy on external links] link in the banner, offering users the chance to understand how Wikipedia treats browser and user data differently than other websites, with click-through links to countermeasures that users can take (including a user preference to silence all referrals, but more importantly on how to anonymize their traffic). That might actually keep people from being jailed or tortured in a way that simply setting a default silent referral never could.--Carwil (talk) 21:42, 18 June 2017 (UTC)[reply]

So, since we can't fully protect our users, we should give up? I agree that a by-page parameter, especially a cookie-based (and sufficiently anonymised!) one for unregistered users would raise awareness of ad surveillance methods, but it's a lot of extra work, and it requires users to take a poorly informed stance on a fairly open-and-shut question (how could forwarding referrer to non-sponsor sites be of any use to the reader?). I can't imagine this could ever be as effective as even an unadvertised low-key switch to silent referrer. As for disclaimers, any smoker will tell you how much they're worth. Daß Wölf 23:39, 18 June 2017 (UTC)[reply]

Comment - I have two more questions. How will this RfC affect search engines, like Bing and Google? Also, how will it affect traffic statistics of Wikipedia articles, like Pageviews Analysis tools? --George Ho (talk) 03:50, 18 June 2017 (UTC)[reply]

Response to "Review of the change in terms of privacy concerns"

This is a response to Wikipedia:Village pump (policy)/Archive 126#Review of the change in terms of privacy concerns.

Expanding on the responses to User:Astinson (WMF) by User:Rich Farmbrough and myself on that page, I would like to add the following;

(I will sign each point to make it easy to respond to it inline.) --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

The claim was made "We did a review of the concerns you raise with both the WMF Security and Legal teams in light of the implemented change". Why have we not heard from WMF Security or WMF Legal directly? Instead we have a second-hand paraphrase by The Wikipedia Library Projects Manager. Astinson has a bias in the direction of putting the needs of the Wikipedia Library above the privacy of Wikipedia users. In particular, it is extremely implausible that WMF legal said anything other that "sending referrer information or being a silent referrer are both perfectly legal". If one of the options we are discussing has legal issues, we would have heard from WMF legal directly by now. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

The claim was made "This change does not change HTTPS in any way for users while within our sites". This is true, but not relevant. At the top of this RfC is the clear statement "What this RfC is not: This RfC is not about links to other Wikipedia pages or to other projects that are under the control of the Wikimedia foundation. It is assumed that, as far as possible/practical, other WMF sites will receive as much information as they want to receive". --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

The claim was made "We're not in any way changing the nofollow policy, so there is no net benefit for search engine optimization or other traffic generation strategies" the first part is mostly true, but the second part is demonstrably false. Nofollow only helps with traffic generation strategies that depend on search engines. Inserting a spam link on a Wikipedia page is itself a traffic generation strategy. The only issue is whether we want to give the spammer feedback on how his Wikipedia spam strategy is working. This is extremely valuable information to the spammer. Is it better to spam a high-traffic Wikipedia page and risk your spam link being quickly removed, or is it better to spam multiple a low-traffic Wikipedia pages in the hope that nobody will notice? Being a silent referrer denies the spammer this information. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

The claim was made "From a privacy evaluation perspective, organizations or individuals with enough sophistication to maliciously track referral data from URLs on Wikipedia, could also track the SSL interactions for the same exchange".

This confuses several distinctly different eavesdropping scenarios.

Consider the case where I, a Wikipedia user, read our page on Bomb-making instructions on the internet and then click on the link to Feinstein Amendment SP419 at Cornell university.

Scenario 1: An eavesdropper who I will call "Eve" is monitoring my internet usage (this would require a court order to my ISP, but is certainly doable). Eve sees that I accessed Wikipedia and then accessed www.law.cornell.edu, but does not know that I accessed t SP419 page on cornell.edu and does not know that I accessed our bombmaking page. Wikipedia has 8,665 links to www.law.cornell.edu. All of this remains true no matter what we set our referrer policy to. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

Scenario 2: "Eve" is monitoring Cornell university's internet usage. Eve only sees that I accessed www.law.cornell.edu, but does not know what page I accessed and knows nothing about me accessing Wikipedia. All of this remains true no matter what we set our referrer policy to. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

Scenario 3: "Eve" gets a court order giving them access to Cornell university's server logs or simply hacks the server to get the logs. Note that this has the advantage for Eve of revealing my previous visits to cornell.edu, not just the ones that happened after Eve started listening in. Eve know knows what page on cornell.edu I accessed (www.law.cornell.edu/uscode/text/18/842)

Now it gets interesting. If we are a silent referrer, the logs at cornell.edu will not contain any information about me accessing Wikipedia. If we send domain-only referrer information, the logs at cornell.edu will say that I came to the www.law.cornell.edu/uscode/text/18/842 from wikipedia.org, and it turns out that the only link to www.law.cornell.edu/uscode/text/18/842 on Wikipedia is from our Bomb-making instructions on the internet page. It is, however, linked to by many other websites. So, by sending referrer information, we just turned Eve knowing that I accessed the text of the Feinstein Amendment SP419 -- a perfectly innocent act in itself - to eve knowing that I accessed it while reading the Wikipedia bomb-making instructions on the internet page. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

The claim was made "Malicious link spammers could discover which pages editors are coming from anyway by creating unique URLs that signal their source." This is a classic case of the Nirvana fallacy. Burglars could get into my house and steal my possessions by picking a lock or smashing a window, but that doesn't imply that I should store my valuables on my front lawn with a big "please steal me" sign on them. After we stop giving the spammers the referrer information they need to get better at spamming us, I fully intend to spearhead a multi-faceted effort to get rid of any such unique URLs on Wikipedia. Plus, it is a demonstrable fact that most spammers are stupid, and if we only interfere with the 99% of spammers who lack the sophistication to create a unique URL, we will have done a good thing. --Guy Macon (talk) 07:58, 17 June 2017 (UTC)[reply]

General technical comments

Please comment on the technical aspects of this proposal, not on what our policy should be. Misplaced comments may be may be moved to the proper section by any user.

Comment: Allowing the domain in the referrer (en.wikipedia.org) does not protect the Wikipedia reader from the target site figuring out what Wikipedia page they were on when they clicked the link. Most sites only have a few links, and we provide a linksearch tool that tells anyone who asks exactly where those links are. Only being a silent referrer protects against this. --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

You convinced me with that. Indeed, even just the wikipedia.org is too much -- because if a malicious bureaucratic entity wants to punish people for reading a Wikipedia page, they'll put their honeypot on that one page only. Indeed, they may go a step further and have no other link to it on the web, in which case omitting referer is no protection -- except, they don't absolutely know that nobody took the text and copied it out of context to a forum, or copied just the link, and I would hope some quick post facto muddying of the waters by supporters of the political prisoners (i.e. by putting those links where they would appeal to government supporters, say, in a context where it is hard to get a firm date for when they first appeared) could leave the prosecutors very uncertain whether they have nabbed the people they want. Wnt (talk) 23:42, 11 June 2017 (UTC)[reply]

Excellent analysis. If a silent referrer policy is implemented, we should work together to create a Wikipedia help page explaining this and giving everyone detailed instructions on how to obfuscate any such honeypot efforts as efficiently as possible. And of course we at Wikipedia should do whatever we can to protect our readers, even if we know that our countermeasures are not perfect. Being a silent referrer is part of that, but we should be thinking hard about what more we can do. I just put that on my calendar to revisit after this RfC closes. The comment by PaleoNeonate about removing UTMs that they posted at he bottom of the "query strings and similar tactics" section is a good example of something else we can do. --Guy Macon (talk) 00:08, 12 June 2017 (UTC)[reply]

Comment: Twitter has figured out a way to hack the default HTTPS behavior, See Hacking HTTPS -> HTTP referrers --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

Comment Given that it's the browser that decides to send the referrer information, I think it would be more fruitful to lobby your favourite browser vendor to include a mechanism to disable sending this info. For example, Firefox has the network.http.sendRefererHeader configuration setting and various other settings starting with network.http.referer that allow you to control if and how the referrer information is sent (see [21] for more information). Firefox and Chrome both seem to have addins that allow you to control when the referrer data is sent. Perhaps the browser you use has equivalent functionality available. isaacl (talk) 03:48, 2 June 2017 (UTC)[reply]

Already done. As far as I can tell, every modern browser obeys the Meta Referrer tag, so Wikipedia can decide not to send referrer information for all users without them having to make a special configuration change. The key point here is that the Wikimedia foundation decided to send information to every website that we link to containing information that allows the site to figure out what Wikipedia page the user was reading, and they did it without asking us if that is what we want done. Shouldn't that be our decision as a community to make?

My post was regarding the need for a policy, so not sure why it was moved to this section. I think the user should control the decision on what the browser should send, as this can be applied generally to all web sites. Thus I'm not convinced of a need in a change in policy. Of course, any editor or group of editors are free to put forth a proposal for changing the meta tags sent, as is being done here. isaacl (talk) 14:02, 2 June 2017 (UTC)[reply]

I think it is obvious that not every user (alas!) is going to download Firefox, nor change these settings. Not all users are ever going to know this. Our question is what are our responsibilities. I mean, it's like making nitroglycerin. Every kid in an intro college chem class should know not to make nitroglycerin during the lab exercise. But if one of the eager students happens to do it anyway? Then people go to the college and say why didn't you stop this? It's our responsibility to try to prevent dangers to our readers -- though this is subordinate to our defining purpose to collect and disseminate knowledge, even when it is knowledge proscribed by some regime. Wnt (talk) 23:49, 11 June 2017 (UTC)[reply]

Let me put it another way: I don't see a need for the level of consensus required for a policy to be agreed upon, since users can ultimately decide for themselves what gets sent, so a change to the server configuration does not inhibit their ability to send the information they prefer. But given that the primary issue for the WMF appears to be the desire to provide referrer data for partners, then certainly the more support collected to support a change to the meta headers, the better. isaacl (talk) 01:18, 12 June 2017 (UTC)[reply]

Maybe there is no need for a policy-level RfC, but "users can change it" is certainly not an argument to that effect. If the WMF decided that all logged-in editors who did not tick a box in the preferences get some javascript that makes their computer mine Bitcoin for Jimbo's wallet, I would certainly expect a lot of opposition even if it is opt-out. Tigraan^{Click here to contact me} 16:15, 12 June 2017 (UTC)[reply]

To put it a third way: it's unusual for a website to enact a policy to enforce what the user's browser sends, since it can't control what happens on the user's side. But the distinction doesn't really matter for the purposes of this discussion. isaacl (talk) 00:36, 13 June 2017 (UTC)[reply]

It's not unusual (see Twitter example above) and control is possible in this case via <meta> tags and other tricks. Of course, the user's browser can choose to ignore all that stuff, but in most cases it doesn't, and most users are too ignorant to change it or even care about it. We can't expect everyone to be an expert on everything, and especially so in an area dominated by advertising empires (who are responsible for some of the top browsers of today: Chrome, Edge, IE...) Daß Wölf 01:26, 13 June 2017 (UTC)[reply]

What Wölf said. We don't want to "control" the browser behavior (and we have no way to do that), but we can encourage it to behave in a certain way. If the meta tag is going to be ignored by most browsers, or if it should in general not be used (for a reason yet to be specified), one wonders why it was introduced, really. Tigraan^{Click here to contact me} 09:35, 13 June 2017 (UTC)[reply]

I'm not saying it is unusual for a website to send meta headers or to forward links through a redirection mechanism; I'm saying it is unusual to establish a policy (in the Wikipedia sense of the policies that can be enacted by community consensus) for its users which cannot be enforced. We aren't going to sanction users, for example, for what their browsers send. However I am in agreement with all of the discussion above that the community can certainly develop a view on what the web servers should be sending, regardless of what this view is labelled. isaacl (talk) 12:47, 13 June 2017 (UTC)[reply]

Comment: Is a phabricator task needed to suppress an attempt to send information to an external website? Seems that option 5 is having an unanimous support right now. --George Ho (talk) 03:27, 12 June 2017 (UTC)[reply]

A phabricator task would be a much later step. One can't just post a phabricator task about a major policy change and expect it to be implemented. If this RfC ends up the way it is going so far, we are going to have to get the WMF to agree to a no-referrer policy. After that the phabricator task is just paperwork.

Please don't jump the gun and attempt to make this happen. That will just muddy the water. After the RfC closes I will create a page where we can coordinate our efforts, choose who to make the request and where to make it, carefully edit exactly what we will be asking for, etc. --Guy Macon (talk) 06:07, 12 June 2017 (UTC)[reply]

Okay. I'll await the upcoming project page then. :) --George Ho (talk) 06:17, 12 June 2017 (UTC)[reply]

HTTPS

Comment: The fact that Wikipedia uses HTTPS and whenever possible uses HTTPS links makes some of the listed options impossible using current technology (the technology may change and this restriction may change in the future). In particular, when you click on a link from Wikipedia (Which is HTTPS) to an HTTP site, the default "no referrer when downgrading from HTTPS to HTTP" policy protects you against someone monitoring your connection (someone intercepting your WiFi, your ISP, the police...).

The default HTTPS behavior means that the eavesdropper only knows that you visited en.wikipedia.org, not what pages you went to. Without the default HTTPS behavior, when you click on a HTTP link and leave Wikipedia, the Wikipedia page you were on would be sent in an unencrypted HTTP packet to the new site, and the eavesdropper would see it. (This appear to be incorrect per [ https://www.w3.org/TR/referrer-policy/ ]. See section 3.2 for details)

The default HTTPS behavior means that an eavesdropper monitoring you when you click on an HTTP link sees nothing in the header that shows what page you were on when you clicked the link. If you click on an HTTPS link the eavesdropper sees nothing - it is all encrypted -- but the website owner sees the full URL of the Wikipedia page you were on when you clicked the link. --Guy Macon (talk) 15:00, 1 June 2017 (UTC), Modified 18:22, 11 June 2017 (UTC)[reply]

But also remember.. It's pretty easy to draw out which IP addresses you connect to begin with. So even when you have https with all the referrer protections you could think of, computer models can be built to follow your path of connections you make (and how often). Unless many websites share IP addresses between eachother or you are using TOR, you can be tracked / data mined by ISPs or the government. (actually even with TOR, this tracking was applied in the past by governments using a bug in TOR I seem to remember.. It was speculated that this type of profile building was even used to identify TOR users by matching two similar profiles). I'm mentioning this, because while yes, some of the referrer information can be used to identify you, but at the same time very similar OTHER information will remain even if you hide some more of this referral information. And that is an important element to this consideration. —TheDJ (talk • contribs) 16:01, 1 June 2017 (UTC)[reply]

We are talking about different threats. Consider the example I gave in the overview: someone reading the Wikipedia page at Bomb-making instructions on the internet#References clicks on the link to The Low Cost Cruise Missile: A looming threat?, but this time assuming that both sites are HTTPS and wikipedia does its best to be a silent referrer. What does an eavesdropper know? For convenience, I will call the user "Alice", the person who controls the external site "Bob", and the eavesdropper "Eve".

If Eve is monitoring Alice specifically, you are correct in that Eve knows which IP addresses Alice connects with, and thus knows that she accessed Wikipedia.
If Eve is monitoring Bob, She knows that Alice visited Bob's site, but she knows nothing about Alice visiting Wikipedia, and neither does Bob.

Now assume the same scenario, still using HTTPS, but with Wikipedia sending domain-only referrer information. Now Bob (but not Eve) knows that Alice came to his site from the English Wikipedia, and a quick check with the link search tool we provide will show that the only two articles on Wikipedia with that link on them are are Bomb-making instructions on the internet#References and Bruce Simpson (blogger)#DIY Cruise Missile If if Bob gets a court order or his servers get hacked, other people know what Wikipedia page Alice visited.

It is my contention that the users of Wikipedia should have a say about whether some site in New Zealand is sent information about what Wikipedia pages they read. --Guy Macon (talk) 19:14, 1 June 2017 (UTC)[reply]

So your concern is the leaking of referral information from, for instance the hacked server's (Bob) access logs and then reverse engineering that information back towards the behavior of Alice... I guess that's a possibility... Although again, i would argue that domain information is already leaked and that as you point out, page level information can be reverse engineered with the link search tool from that. I would like to point people to these extensions btw: Referrer Control for Chrome or Firefox. — Preceding unsigned comment added by TheDJ (talk • contribs) 20:56, 1 June 2017 (UTC)[reply]

Meta Referrer Tag

Comment: Including a Meta Referrer Tag in the <head> section of a Wikipedia page allows us to specify what information is sent when people follow links to any other site, whether it uses HTTPS or HTTP. This means that we can specify referrer behavior that HTTPS normally would not allow. --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

Rel=nofollow on links

Comment: Adding rel="nofollow" to external links instructs search engines to ignore the link when ranking of the link's target in the search engine's index. It makes adding spam links to Wikipedia less effective, but otherwise does not have any effect on the referrer information. --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

No: using "nofollow" also conflicts with other standards, for example rel="me". And it punishes non-spam sites as much as spammy ones. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:40, 2 June 2017 (UTC)[reply]

Rel=nofollow doesn't "punish" anyone, it just refrains from assisting them. We don't owe such assistance to anybody and we should not provide it. There were two RFC's back in the day that both concluded in favor of nofollow, reportedly because the discussions were manipulated by SEO's. Eventually nofollow got activated when there was some kind of SEO contest that worsened our linkspam problem to ridiculous levels (even before that contest, it was much worse than it is now). Activating nofollow was of tremendous benefit and while I acknowledge being avant-garde about this topic, I think we should activate noindex as well (that says: block the entire encyclopedia from external search engines forever) and enjoy the screams of SEO's as they tear their own hair out all over the world. 173.228.123.121 (talk) 07:16, 12 June 2017 (UTC)[reply]

"we should not provide it" that's a bold statement - or should I say "opinion". If we find a source useful enough to cite it, or add it to an 'External links' section, we should allow it the crumb of benefit that accrues from us doing so; that's simple courtesy. As for past RfCs, we're not beyond seeing the error of our ways, and WP:CONSENSUSCANCHANGE applies. The "tremendous benefit" you claim sadly threw that baby out with the spam bathwater. To suggest that we "block the entire encyclopedia from external search engines forever" beggars belief. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:12, 12 June 2017 (UTC)[reply]

Query strings and similar tactics

Comment: It is possible for a spammer who inserts a link to create a unique URL. this would let them identify the exact page that the incoming link was on no matter what we do with the referrer. There are multiple ways that this can be done, including

www.example.org?ID=myevilpointer
myevilpointer.example.org/
www.example.org:3717/
www.example.org/reasonable_page/my_evil_pointer/

Some of these could be fixed by a robot looking for such links. A simple robot could mass-remove all query strings in external links, but some links do not function without them. A more sophisticated robot could remove them only in cases where the page is identical with and without the query string, but so far Wikipedia has not shown itself to be technically advanced enough to create such a robot. --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

This does not seem simple at all to me really... The false positives would likely be gigantic. —TheDJ (talk • contribs) 16:09, 1 June 2017 (UTC)[reply]

Sorry for being unclear. What I was trying to convey is that the problem is complex and that a simple robot most likely would not be able to solve it. It would be a fair amount of work, but I believe that I could create a program that solves the problem with a minimum of false positives (the key to doing that is doing a byte-by-byte comparison and only removing the suspected evil pointer in cases where the page is byte-for-byte identical after the change to the link), but I wouldn't want to try using anything other than C/C++. Some of the languages listed at #Programming languages and libraries seem to be poorly suited for such a task, but to be fair some of the others have a good reputation -- I am simply not familiar enough with them to know whether they could handle such a task. --Guy Macon (talk) 20:30, 1 June 2017 (UTC)[reply]

Even if a robot could solve it, do you really trust that the pages in question wouldn't be able to detect it, and serve the robot different text? —Cryptic 21:40, 1 June 2017 (UTC)[reply]

Anything having to do with spammers is an arms race. Every measure generates a countermeasure, witch in turn generates a counter-countermeasure. Along the way you lose the dumber and less skilled spammers, but nobody has found the magic bullet that stops all spam forever. --Guy Macon (talk) 01:23, 2 June 2017 (UTC)[reply]

I'm not sure if there's already some robots checking for this, but by routine among citation cleanup tasks if I see unnecessary tracking information in urls I remove it (other than what is needed to reference to a page or relevant query of course). I remember reading that it was discouraged, but I now fail to find it quickly via Wikipedia:Citing_sources#Citation_style. It may be a good idea to make this more obvious if that is considered good practice to avoid. —Paleo Neonate - 06:50, 3 June 2017 (UTC)[reply]

I made a note on my calendar to get back to this after this RfC closes. We should definitely have an easy-to-understand help page explaining how spammers create unique URLs and how to tell them from legitimate uses of things like query strings. Thanks for suggesting it! --Guy Macon (talk) 23:33, 10 June 2017 (UTC)[reply]

Also a side note: we do remove UTMs which are related to the above, Primefac has also successfully obtained permission for a PrimeBot task to remove those, for instance (and thanks for working on this, Primefac). —Paleo Neonate - 02:04, 11 June 2017 (UTC)[reply]

I clean such links up by hand when I come across them. I'd support some narrowly targeted and vetted bot operations to do it on a wider basis. 173.228.123.121 (talk) 07:18, 12 June 2017 (UTC)[reply]

Generally this is a productive area to look at. I don't expect a silver bullet but I do expect a few shots to connect. I encountered the UTM parameters thing in action, which I'd never heard of previously, and was pleased to see that someone had taken a potshot back at the dismaler science. Though this is definitely not the most difficult instance.

Where unique domain names are concerned, I think it might be worth doing some screening for "hapax legomenon" links. Most Wikipedia references come in multiples, so if we have something go through and find links that occur from only one page, odds are they are typos and they don't work anyway and a bot should check to see if they're 404 and tell the person who posted them. And if they do work, they may be spam or unreliable sources, and maybe they could go in a queue for volunteers to look through, thought obviously that last part is easier said than done. And among all those links, any "sting" or spam-testing links might be spotted by editors looking through. (Possibly, having editors look through on a volunteer basis also provides an alternate theory to why someone looked at the link other than that they were reading a proscribed page, but I doubt that would help them, since censorship and presumption of innocence rarely turn up in the same proceeding) Wnt (talk) 00:48, 19 June 2017 (UTC)[reply]

Server Name Indication

Comment: Server Name Indication may be able to leak some limited information about a Wikipedia session to an eavesdropper who is monitoring the SSL handshakes. I believe that an effective countermeasure to this would be becoming our own certifying authority. Because wildcarding is allowed *.wikipedia.org will work with all Wikipedias, *.wiktionary.org" will work with Wiktionaries, etc. --Guy Macon (talk) 15:00, 1 June 2017 (UTC)[reply]

This seems to have little to do with the referrer discussion, but also, I think you are mistaken. As far as I remember, SNI is a client-side driven part of the negotiation and unless you modify your browser and it's TLS libraries, you cannot plug this leak of information. If this has you concerned, you should probably be using Tor. —TheDJ (talk • contribs) 15:48, 1 June 2017 (UTC)[reply]

"SSL handshakes (part of the HTTPS encryption) *already* include our domain and the IP address of our servers (through Server Name Indication), so any passive observer of the HTTPS traffic could uncover this data with minimal effort. Source: Astinson (WMF),[22] Posted 23:09, 4 April 2016 (UTC) at Wikipedia:Village pump (policy), archived at Wikipedia:Village pump (policy)/Archive 126[reply]

"SNI could be simply disabled by becoming our own certifying authority, and re-issuing certs as a standard part of rolling out new projects. This would be an improvement in security, but is not part of the issue I raised. (Note: Since wild-carding is allowed "*.wikipedia.org" covers all present and future wikipedias, "*.wiktionary.org" all Wiktionaries and so forth, so we would only need a new cert once in a blue moon anyway.)" Source: Rich Farmbrough [23] Posted 00:33, 5 April 2016 (UTC) at Wikipedia:Village pump (policy), archived at Wikipedia:Village pump (policy)/Archive 126

BTW, I already use TOR. See my user rights list[24] and you will see that I am IP block exempt, allowing me to edit through TOR. --Guy Macon (talk) 16:21, 1 June 2017 (UTC)[reply]

I think Rich misinterpreted that. While yes, that strategy avoids using SNI on the server side, a browser supporting SNI, will send you the SNI extension with the hostname information regardless. At least that is my understanding from the protocol. —TheDJ (talk • contribs) 20:35, 1 June 2017 (UTC)[reply]

Yes SSL/TLS is designed to be proxy-able (with proxy-generated alternate certificate signed by an authority trusted by the browser, often added when installing antivirus or other software). Because of this some information is visible in the handshake, this still is a different issue than passing information to the destination site when following a link from another (about the origin of that link). —Paleo Neonate - 06:58, 3 June 2017 (UTC)[reply]

So, bottom line, does becoming our own certifying authority help user privacy or not? Is it desirable/undesirable for other reasons? --Guy Macon (talk) 00:15, 12 June 2017 (UTC)[reply]

I'd imagine it's not desirable for a pretty big reason: our root certificate wouldn't be in any of the browsers, meaning that nobody could access our sites without the big glaring "THIS SITE IS INSECURE" warnings. Yes yes, people could import our certificates but that's too much of a burden for our readers (nor can they all do this--think people at work, embedded browsers like smart TVs, etc...). FACE WITH TEARS OF JOY ^[u+1F602] 04:10, 12 June 2017 (UTC)[reply]

Basically you don't just "become a CA" - you can just "self-sign" - but then you have all the problems 1F602 mentioned. — xaosflux ^Talk 04:34, 12 June 2017 (UTC)[reply]

No, we are not talking about self-signing. We are talking about becoming a certificate authority.

Becoming a X.509 Certificate Authority

Normally when you configure a server to use TLS or SSL you have two choices; Either you pay someone like Verisign or Thawte to sign a certificate or you generate a self-signed certificate. However there is an alternative, which is to generate your own certificate authority or CA. ... Which route you choose depends on your circumstances and why you need a certificate. For a large public service like an e-commerce website, you’ll want a certificate signed by an established trusted root CA, who, like Verisign, have their root keys bundled with web browsers and operating systems. This allows anyone to trust your server is the server it claims to be and traffic is encrypted, without having to install any additional certificate. The downside to this is the cost of getting a certificate. At the time of writing, Verisign were charging $2,480.00 USD for a 3 year 128bit certificate. Source:[25]

We can cover *.wikipedia.org with one CA, and *.wikimedia.org with another. We may choose to not do this for wiktionary.org, wikisource.org, wikidata.org, etc. --Guy Macon (talk) 05:55, 12 June 2017 (UTC)[reply]

Yeah real handy. Or if you get it signed, then you are just another intermediate CA, just with less expenses on issuing new certificates. And again, it doesn't change anything regarding the leak of the hostname information, as certificates are exchanged and checked AFTER the client has initiated the SSL/TLS handshake which contains the SNI host information.[26] —TheDJ (talk • contribs) 22:49, 12 June 2017 (UTC)[reply]

Motion to close this subsection Is there any reason to discuss this here? AFAICT this has zero relevance to the referer policy. Beware of WP:TRAINWRECK. Tigraan^{Click here to contact me} 09:42, 13 June 2017 (UTC)[reply]

Are you sure that Rich Farmbrough is wrong? See Wikipedia:Village pump (policy)/Archive 126#Review of the change in terms of privacy concerns. --Guy Macon (talk) 12:24, 13 June 2017 (UTC)[reply]

I am inclined to agree that this is a little off the track of silent referer. As far as the technology is concerned, it is true that, if we provide a solution that renders SNI un-needed, that does not mean the client software won't use it. The solutions we could provide include subjectAltName if we were self-certifying (or prepared to spend enough) - either by chain-of-trust, which I haven't kept up on, or by becoming a certifying authority, or by other means - or by having a mapping of domains to IP numbers. In any event this would allow readers to choose a stack that deprecates SNI, and regain some privacy (depending on the solution) while still accessing our projects: doing nothing denies them this possibility. All the best: Rich Farmbrough, 13:01, 13 June 2017 (UTC).[reply]

Just found this description which seems to be rather accessible explanation. —TheDJ (talk • contribs) 14:35, 13 June 2017 (UTC)[reply]

This section has several misunderstandings about how TLS/web-pki works. Keeping in mind that this has really very little to do with referrer policy, I would like to point out we already use wildcard certs. If you click the little lock icon in your browser, and go to advanced information about the certificate, you will see under "Certificate Subject Alt Name"

Not Critical
DNS Name: *.wikipedia.org
DNS Name: *.m.mediawiki.org
DNS Name: *.m.wikibooks.org
DNS Name: *.m.wikidata.org
DNS Name: *.m.wikimedia.org
DNS Name: *.m.wikimediafoundation.org
DNS Name: *.m.wikinews.org
DNS Name: *.m.wikipedia.org
DNS Name: *.m.wikiquote.org
DNS Name: *.m.wikisource.org
DNS Name: *.m.wikiversity.org
DNS Name: *.m.wikivoyage.org
DNS Name: *.m.wiktionary.org
DNS Name: *.mediawiki.org
DNS Name: *.planet.wikimedia.org
DNS Name: *.wikibooks.org
DNS Name: *.wikidata.org
DNS Name: *.wikimedia.org
DNS Name: *.wikimediafoundation.org
DNS Name: *.wikinews.org
DNS Name: *.wikiquote.org
DNS Name: *.wikisource.org
DNS Name: *.wikiversity.org
DNS Name: *.wikivoyage.org
DNS Name: *.wiktionary.org
DNS Name: *.wmfusercontent.org
DNS Name: *.zero.wikipedia.org
DNS Name: mediawiki.org
DNS Name: w.wiki
DNS Name: wikibooks.org
DNS Name: wikidata.org
DNS Name: wikimedia.org
DNS Name: wikimediafoundation.org
DNS Name: wikinews.org
DNS Name: wikiquote.org
DNS Name: wikisource.org
DNS Name: wikiversity.org
DNS Name: wikivoyage.org
DNS Name: wiktionary.org
DNS Name: wmfusercontent.org
DNS Name: wikipedia.org

Thus we already use wildcard certs, and SNI is not technically needed for most Wikimedia sites (atm. This could of course change in the future). It should be noted that some subdomains do not use that main cert (e.g. lists.wikimedia.org is on a different server and uses a different cert). However, this is a mostly moot point since you cannot disable SNI in your browser. Also disabling SNI really does not gain you any privacy because DNS queries go in the clear, and the destination IP address go in the clear. If you're sending a request to one of 208.80.154.224, 91.198.174.192, 198.35.26.96, 208.80.153.224, 2620:0:861:ed1a::1, 2620:0:862:ed1a::1, 2620:0:863:ed1a::1, 2620:0:860:ed1a::1 then its obvious your are visiting a Wikimedia website. BWolff (WMF) (talk) 22:16, 13 June 2017 (UTC)[reply]

That is good news. Does *.wiktionary.org not cover *.m.wiktionary.org though? All the best: Rich Farmbrough, 11:07, 17 June 2017 (UTC).[reply]

The Alert Reader will have noticed that at the very top of this RfC I wrote "This RfC is not a discussion of the technical details regarding what is and is not possible using current technology, which may change. It is an RfC about policy, not implementation". and that every RfC question contains the words " As far as possible/practical".

Server Name Indication is an good example of why I wrote those qualifiers. The above discussion is all about how SNI behaves now, but how SNI behaves is likely to change in the near future. See [ https://www.ietf.org/proceedings/interim/2014/07/20/tls/slides/slides-interim-2014-tls-2-6.pdf ].

As I write this, the overwhelming consensus of the Wikipedia community is to be a silent referrer with exceptions made for certain selected sites. The most common counter-argument has been "but they can still get information about our users using another method". This is a classic case of the Nirvana fallacy; burglars could get into my house and steal my possessions by picking a lock or smashing a window, but that doesn't imply that I should store my valuables on my front lawn with a big "please steal me" sign on them.

As we can see from the above IETF document, there are some very smart people working very hard to make it so that these other methods of finding out what pages our readers access no longer work. We just need to do our part, controlling what we can. In other words, once we make the policy decision " As far as possible/practical, referrer information should contain no information (silent referrer)", we should expect the Wikipedia developers to keep up with things like changes in how SNI behaves and respond appropriately to implement the policy we have chosen. --Guy Macon (talk) 07:37, 18 June 2017 (UTC)[reply]

Other

Comment: I think that Wikimedia has to be on guard against the syndrome of falling into line with other computer-intensive sites. (For example, comparing itself to Google and Facebook) Remember: they all exist to spy on their customers and make money by underhanded means. The WMF projects are something noble, something far removed from the use cases that characterize the rest of the web. For example, the WMF often cannot use their software, of course, because of licensing/copyright issues. It should also follow that the WMF sometimes has to transgress their spoken or unspoken cultural norms -- such as that everyone is joined together in a project to spy on and exploit the user. Because they are not always up front about their purpose, this may require some vigilance. Wnt (talk) 13:38, 11 June 2017 (UTC)[reply]

Comment This whole discussion just makes me utterly sad. So much FUD, such a isolationist view.. When did we get this self obsessed ? —TheDJ (talk • contribs) 22:35, 12 June 2017 (UTC)[reply]

It makes me sad to see that you are sad about our good-faith effort to balance user privacy against giving external websites information that they want to see. Might I humbly suggest not reading things that make you sad? If, by some chance, you are strapped to a chair with your eyelids tied open in front of a monitor showing a Wikipedia:Village pump (policy) feed with The Wikipedia Song blasting in the background, then let me address this message to your captors: First of all, keep up the good work. Secondly, please take away his keyboard. :) --Guy Macon (talk) 01:56, 13 June 2017 (UTC)[reply]
The cure to Fear, Uncertainty and Doubt is to not have things to worry about, have a simple certainty to your life, and know that there is agreement on this. In other words, don't give worrisome pseudo-private information, don't wonder how it is (not) being used, and have an RFC to say so. Wnt (talk) 00:00, 15 June 2017 (UTC)[reply]

Comment. I've made a good faith effort to read the discussion, but it's too long to go through in detail, so I won't be !voting either way as I can't assert in good faith that I am doing so in full understanding of the issues. I am concerned by the comments about the value to Wikipedia Library relationships of the referrals, and suspect I could be persuaded to oppose if it those concerns are real. As it stands I'm afraid it feels like this RfC was drafted by someone with an hoped-for outcome in mind -- sorry, Guy, I'm sure you wrote it in good faith and didn't intend to have your thumb on the scales, but it does feel that way. I would prefer to see an RfC in which the benefits (if any) of the current scheme are laid out by someone who is a proponent, and the RfC puts more time into making the case for both sides. Mike Christie (talk - contribs - library) 22:32, 13 June 2017 (UTC)[reply]

Comment: If some people have concerns about the outcome of the RfC, ~~why not "Question 7: Do nothing"?~~ --George Ho (talk) 00:03, 14 June 2017 (UTC); partially struck (see below). 11:10, 16 June 2017 (UTC)[reply]

Define "do nothing". If you mean "do what we are doing now", that is the same as Question #3. Or do you mean "go back to the default HTTPS behavior? --Guy Macon (talk) 03:56, 16 June 2017 (UTC)[reply]

Oh.... I overlooked the "status quo" notes mentioned in Questions #1 and #3. Oh well, I guess I should strike out the Question #7 suggestion then. --George Ho (talk) 11:10, 16 June 2017 (UTC)[reply]

Comment: This has previously been discussed on Meta at [27]. English Wikipedia users @Nemo bis: @Pundit: @Piotrus: @Denny: and @Halfak (WMF): may want to comment here.--Carwil (talk) 19:29, 16 June 2017 (UTC)[reply]

"Ethnicity" on templates.

Last year there was a discussion that removed "religion", and "ethnicity" from infoboxes all being centered on the U.S. American politician Bernie Sanders, and the people "reached consensus" (the "strongest possible oppose" "votes"were all literally just different people repeating the same arguments that either ethnicity was always irrelevant, or that "it's hard to define"), despite there being several counter-arguments that explained that in many people's infoboxes it was needed because simply placing one's citizenship wouldn't be enough (it's like saying that all notable Jews that died 💀 during the shoah were just "Germans"), but the same repeated arguments keptpersisting. Over that year I've seen numerousarticles about E.G. Austro-Hungarians or Yugoslavians where the individual's ethnicity mattered that look now like a complete mess.

The rules should've been amended that it should be only mentioned if relevant to the individual's circumstances (like many have pointed out, for example an Iraqi Kurd being "a Kurdish national" makes no sense, as It's now replaced by "nationality"), my other annoyance comes from replacing it with "Nationality", in general English the word "nationality" means citizenship, and "a nation" is often a geographical and political entity, I've seen several ethnic Italians or various other people that have lived centuries before Italy was even a thing as "Nationality = Italian", that makes no sense, for many historical people their ethnicity could play a lot into who they were, and oversimplifying it by saying "it's hard to define, therefore don't use it", or "it's never relevant" only take into account modern perspectives, and yes, modern taboos. This violates the WP:NPOV, if we write about historical figures we shouldn't treat them like "living people".

42.112.158.179 (talk) 05:34, 2 June 2017 (UTC)[reply]

I think you've misrepresented the consensus in your argument above. The primary issue is that discussions of ethnicity require some context, and a single-word mention in an infobox does not provide the necessary context. As an aside, let such matters go. Infoboxes are a minor issue anyways. Focus on the text of the article, where ethnicity can be explained in detail. --Jayron 32 05:49, 2 June 2017 (UTC)[reply]

42.112.158.179 misstates the result of the RfC(s). We didn't "remove 'religion' from infoboxes".

There were actually three RfCs, and they all dealt with listing non-religions such as "none", "atheist", "secular" and variants thereof in the "religion=" field of the infobox. All three RfCs specifically allowed such information to be added to the body of the article, subject to our usual sourcing rules, and all three RfCs specifically allowed listing actual religions in infoboxes.

15 June 2015 RfC: Template talk:Infobox person/Archive 28#RfC: Religion infobox entries for individuals that have no religion

This RfC had a clear consensus for removing the religion parameter from the infobox for individuals (living, deceased, and fictional), groups, schools, institutions, and political parties that have no religion, but that RfC was determined by the closing administrator to not apply to nations.

17 June 2015 RfC:: Template talk:Infobox country/Archive 11#RfC: Religion in infoboxes of nations

This RfC had a clear consensus for removing the religion parameter for countries, nations, states, regions, etc., all of which were determined to not have religions.

31 December 2015 RfC: Template talk:Infobox/Archive 11#RfC: Religion in infoboxes.

This RfC was a response to certain individuals insisting that the previous RfCs did not apply to their favorite pages (schools, political parties, sports teams, computer operating systems, organized crime gangs...) and had a clear consensus that in all all infoboxes in all Wikipedia articles, without exception, nonreligions should not be listed in the "Religion=" parameter of the infobox.

It took a lot of effort, but we finally had a community decision that "religion=atheist" and "religion=agnostic" would no longer be allowed. This should not have been an issue, because...

"Atheist" is not a religion. Atheism is the lack of any religion. "Agnostic" is not a religion. Agnosticism is the lack of any religion. Bald is not a hair color. Bald is the lack of any hair color. Off is not a TV channel. Off is the lack of any TV channel. Barefoot is not a shoe. Barefoot is the lack of any shoe. Silence is not a sound. Silence is the lack of any sound. Never is not a date. Never is the lack of a date. Clear is not a color. Clear is the lack of a color. Not collecting stamps is not a hobby. Not collecting stamps is the lack of a hobby. --Guy Macon (talk) 23:41, 2 June 2017 (UTC)[reply]

This, a million times this. Calling atheism a religion shows a complete lack of understanding of what it even is, and is offensive to atheists. Beeblebrox (talk) 00:42, 3 June 2017 (UTC)[reply]

With regard to religion, I believe the OP is referring to a fourth RfC, Wikipedia:Village pump (policy)/Archive 126#RfC: Religion in biographical infoboxes, closed on 11 April 2016 with consensus to remove the religion parameter from Infobox person. --Worldbruce (talk) 01:24, 3 June 2017 (UTC)[reply]

Ah. I had missed that one. Companion RfC: Wikipedia:Village pump (policy)/Archive 127#RfC: Ethnicity in infoboxes.

So we have an overwhelming consensus to remove the religion= and denomination= parameters from Infobox person and against adding an ethnicity parameter to infoboxes (that RfC doesn't specify what kind of infoboxes).

Based upon my experience with with the three RfCs it took to finally stop certain editors from calling atheism a religion in the infobox, in my opinion the two RfCs I just listed will not prevent Wikilawyering ("The RfC didn't say that I can't add religion to the infobox at Rwanda and Pikachu. It says "person"). ("The RfC says 'no ethnicity parameter in infoboxes', so I am removing the ethnic_groups = parameter from Japan"). --Guy Macon (talk) 17:29, 3 June 2017 (UTC)[reply]

Atheism is a religion. Atheism implies a lack of belief in gods, while "religion" refers broadly to beliefs about the supernatural that include animism, which deals with spirits rather than gods, Daoism, which in its purest essence e.g. Laozi concerns the "way" in which things work, rather than deities, and Scientology, which makes statements about souls ("thetas") but doesn't postulate anything higher-level than a really old war between crazy bureaucrats. Atheism implies thoughts about how things came into existence which supposes that the force(s) that created the Mandelbrot set and cockroaches are non-conscious; this is a specific religious philosophy.

Now as to all these parameters... how about lumping them all into one term, "Demographics:"? That way, (a) we don't have to dance on the head of a pin over what atheism is, (b) we don't have so much compulsion to include every aspect when it's stuff we may be unsure about or, more likely, is just too hard to put in a tiny box on a form, and (c) we can be really precise about the Pope being Christian (Roman Catholic) while we can be a bit vaguer about whether Muhammad was a Sunni or a Shi'ite. Wnt (talk) 16:58, 11 June 2017 (UTC)[reply]

Atheism is a religion like not collecting stamps is a hobby. Atheism is a religion like silence is a language. Atheism is a religion like barefoot is a shoe. Atheism is a religion like off is a TV channel. Atheism is a religion like never is a date. Atheism is a religion like transparent is a color. Atheism is a religion like bald is a hair color. Atheism is a religion like total vacuum is matter. Pretending that atheists might not reject animism as if they only reject theism goes against the common (if not technically accurate) use of the term, which is "rejecting all religions and other forms of the supernatural" --Guy Macon (talk) 03:28, 13 June 2017 (UTC)[reply]

"Religion" may by an abbreviation for "religious position". All the best: Rich Farmbrough, 13:05, 13 June 2017 (UTC).[reply]

"Atheism is just another religion! You need to have faith to not believe in God!!" is an extremely popular argument among fundamentalist Christians, and is vigorously denied by multiple annoyed atheists.[28][29][30] We don't call people names created by their enemies that they deny. We don't call abortion opponents "anti-choice". We don't call those who oppose them "anti-life". And we don't call atheism a religion. This was settled by the consensus of multiple RfCs. --Guy Macon (talk) 07:13, 14 June 2017 (UTC)[reply]

@Guy Macon: Atheism is a religion like not believing in phlogiston is a scientific position. Atheism is a religion like Independent is a political affiliation. Atheism is a religion like asexual is a sexual orientation. Atheism is a religion like Single is a marital status. Wnt (talk) 00:50, 15 June 2017 (UTC)[reply]

It seems clear to me that by consensus it is preferable to avoid contentious infobox fields for this, which are invitations to fill them for article writers. Where religious affiliation or ethnicity is a particularily important topic in relation to a BLP, there is usually enough important material to have an article subsection about it, or at least a mention in the prose. In other cases, it's something that is often contentious and is better left out. I don't understand why this thread is still open... —Paleo Neonate - 08:26, 14 June 2017 (UTC)[reply]

"Religion: atheism" says nothing about whether atheism is a religion or not. There are a bunch of concepts that fall under the general rubric of religion, atheism being one of them. If we were discussing a sentence in an article reading "Atheism is a religion", I would say that sentence was incorrect. But we are not discussing a sentence. We are discussing a word pairing, which is not a sentence, in which an entirely appropriate concept is placed after the word "Religion". Bus stop (talk) 01:54, 15 June 2017 (UTC)[reply]

The infobox could also say "Religion: none" if the subject is known to follow no religion. Jack N. Stock (talk) 03:34, 15 June 2017 (UTC)[reply]

Yes, it could. These are editorial decisions that should take place on a case-by-case basis and by the limited group of editors trying to develop a given article. Yes, "Religion: none" is perfectly acceptable, in my opinion, but we should not be deciding this at the level of project-wide policy applicable to all articles. Bus stop (talk) 14:27, 15 June 2017 (UTC)[reply]

I agree. Jack N. Stock (talk) 04:03, 16 June 2017 (UTC)[reply]

The three of you are allowed to disagree with the result of an RfC, and you are allowed to post another RfC to see if the consensus has changed or whether you have an argument that will change minds, but you are required to abide by the result of the RfC whether you agree with it or not. --Guy Macon (talk) 07:48, 16 June 2017 (UTC)[reply]

In short, some people consider the use of the word in a religion description appropriate, some not. Many but not all people who very strongly & publicly consider themselves atheists would not be willing to state it formally as a religion, if only to discourage people from asking about religion at all, or because they think the entire concept area does not apply to them. The only way to avoid disputes between camps is not to use it at all in this manner. DGG ( talk ) 14:43, 16 June 2017 (UTC)[reply]

Do redirects encourage poor spelling?

discussion started by a banned user in violation of ban --Jayron 32 04:16, 5 June 2017 (UTC)[reply]

The following discussion has been closed. Please do not modify it.

Our article Street hierarchy is a redirect from street heirarchy, which means that someone typing in the incorrect spelling may never realise that they are spelling the word wrong. Another example - the word "calendar" is often mis - spelled "calender". If you type "Julian calender" you will get the article you want and may be none the wiser. If you type in "religious calender" the search result asks "Did you mean: religious calendar". Should we adopt the second approach to improve readers' spelling? 86.148.116.248 (talk) 09:25, 3 June 2017 (UTC)[reply]

I am not sure that it is a Wikipedia's mission to improve reader's spelling. Ruslik_Zero 10:34, 3 June 2017 (UTC)[reply]

I agree about not improving reader's spelling, but there is another huge advantage to asking "did you mean...", which is that you can answer "no". Maybe I really did want to search for "heirarchy"[31] instead of "hierarchy". --Guy Macon (talk) 17:51, 3 June 2017 (UTC)[reply]

We are supposed to be assisting our readers find the info about the topic, more so than helping them spell. Perhaps redirect from misspelling can cause the title of the target article to show in flashing red with a pop up saying they had a spelling error. But it is just annoying. Graeme Bartlett (talk) 22:03, 3 June 2017 (UTC)[reply]

Google search for "street heirarchy" says "Showing results for street hierarchy" and Street hierarchy is the first hit. Wikipedia search for "street heirarchy" says "There is a page named 'Street heirarchy' on Wikipedia" and Street hierarchy is the first hit. I think readers can find what they want easily enough without deliberately misspelled redirects, and I also question whether benefit exceeds cost here. It is not Wikipedia's mission to teach people how to spell, but it is our mission to spell correctly in articles if we wish to call ourselves remotely professional and seek public respect. If an editor misspells an English word in a link, the red color would be a great way to tell them that they should consult a dictionary (or the target article's title). That said, I know that the chances of a change here are infinitesimal so this discussion is academic. ―Mandruss ☎ 22:35, 3 June 2017 (UTC)[reply]

I think that this is fallacious—if the misspelling redirects to the correct title, then most readers should be able to notice that they were redirected, and that the title of the article differs in spelling from the redirect. One can also relate this problem to what some philosophers thought of books in ancient Greece—they claimed that books might endanger our ability to remember things. But, since having memory is useful even with books, the capacity to remember things did not decrease. Same thing here—Wikipedia isn't the only time when people have to spell things, and it isn't the only chance that people have to be reminded of their spelling. Look at Google. They use the system you suggested. So, with that combined with the fact that readers will probably/might notice the fact that the redirect title is different than the article title, there is no benefit in this proposal. But, there are drawbacks. First, the fact that we aren't really in the business of being a spell-checker, we are in the business of giving people general information. Second, it will, although not much, slow down the process of getting to an article by a bit. Third, how would we implement such a system? Since redirects aren't just for misspelling, we wouldn't want to get rid of all the redirects. So, how would we get rid of all the redirects related to misspellings? It would likely require humans to check it, and even if you automatically get rid of the redirects that are categorized as misspellings, there are a lot of redirects that are not categorized. Thus, due to the minimal (or no) benefit and the significant detriment, this is not a good idea. RileyBugz^会話_投稿記録 23:40, 3 June 2017 (UTC)[reply]

Look at Google. They use the system you suggested. I do look at Google, and Google does not go out of their way to facilitate incorrect spelling on their own websites. That is a very flawed comparison. there is no benefit in this proposal except for that silly professionalism thing which, contrary to widespread opinion, is not actually something to be avoided in an encyclopedia. we are in the business of giving people general information. As I suggested, I'm not aware of evidence that deliberately misspelled redirects are needed to help people find the general information; in fact, the only actual evidence presented in this particular discussion belies that notion. As far as I can tell, then, that is an unproven popular myth that probably originated before search engines were as smart as they are today. I'm willing to be corrected with actual evidence. ―Mandruss ☎ 00:16, 4 June 2017 (UTC)[reply]

For Google (I'm talking about search here), if one types in "Heirarchy" it says "Showing results for hierarchy". But, for some search results that are further off, it says "Did you mean: term". So, I'm not exactly wrong there. I don't really understand the argument behind about the professionalism thing (I mean that I don't understand where it came up and why it relates to this). Next, whether they are needed or not. The redirects aren't really needed, so it might do good to delete them. But, the argument of them not being needed also brings up whether this is needed. It would be much easier to keep the current system, which might not have any benefit, than implementing another system, which might not have any benefits. I think that whether we should delete these redirects is something for another day, so I think that we should focus on the current debate. RileyBugz^会話_投稿記録 01:31, 4 June 2017 (UTC)[reply]

Note: I just realized something, so ignore most of my previous comments. Basically, I had thought that this proposal was on the implementation of a feature, which is incorrect, so I withdraw my comments. RileyBugz^会話_投稿記録 01:36, 4 June 2017 (UTC)[reply]

Any likely typo, if it doesn't have a different meaning, should exist as a redirect, to make it easier for our readers, Their spelling isn't our problem; them finding the page they're looking for is. עוד מישהו Od Mishehu 18:03, 4 June 2017 (UTC)[reply]

Jayron32 is spitting into the wind here. The person who typed "street heirarchy" into the search box and was redirected to "street hierarchy" was ... Jayron32. 217.34.36.106 (talk) 11:34, 7 June 2017 (UTC)[reply]

I would disagree that most readers notice when they have been redirected, at least where the letters in the misspelling are the same. In point of fact, people don't read by picking out each individual letter in a word - they scan the whole word and rely on context. In the case in point [32] Jayron would have read the article "street hierarchy" before linking to it, and the fact that he maintained the misspelling in his post indicates that he didn't pick up on the error. Both method "A" and method "B" lead the reader where she wants to go, but method "B" educates her along the way. It might be useful to mention this discussion at the Teahouse and see whether editors joining for the first time have a preference. Anyway, spell checkers are obviously in demand because all word processors have them. 217.34.36.106 (talk) 12:15, 8 June 2017 (UTC)[reply]

All becomes clear. See this statement of intent User:Jayron32/Jayron32's Principles of Vandal Fighting, particularly Principles 3 and 4. 217.34.36.106 (talk) 10:43, 9 June 2017 (UTC)[reply]

Petition to re-add the categories "ethnicity", and "religion" to "infobox person".

It's generally considered bad form to resurrect an issue a year after it was settled, especially one that had so much participation and debate. Even if you did, it would need an RfC. The community consensus is that the downside of these parameters exceeds their upside. I'll stick my neck out and close this per WP:SNOW and WP:BOLD. ―Mandruss ☎ 17:58, 4 June 2017 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Last year the categories “religion”, and “ethnicity” were removed from the infobox template “person”, and in response to this sudden removal a new infobox “religious person” was created that could only be applied to a person whose entire notability and coverage is based on their status within a religious hierarchy. On the part of the editors who removed these two categories this move seems rather unfounded, and worst of all deletionist, the repeated arguments were that what someone is, or what they believe in aren’t a part of their notability so these categories should be removed but if we'd go by that logic we could scrap not just the entire infobox, but remove anything from the article ‘’other’’ than what made them famous, basically making Wikipedia into a redundant add-on to the Wiktionary.

Let me dissect and debunk every major argument that lead to this “consensus” (it was more of “a popular vote” since those in favour to keeping these categories were often ignored, and if responded to only got addressed as to why these categories were irrelevant to the U.S. American politician Bernie Sanders).

The idea that “religion/ethnicity” are “deeply personal things that can’t be summed up in an infobox”, what this argument misses is that for various people their religion, ethnicity, Etc. are important factors in their political or public life. And I mostly have to agree with the proposals that these categories should be kept but only included according to a strict guideline but that argument was “shot down” multiple times by being responded by with “this is not the place for such suggestions” or “talk about it elsewhere” basically making it into a “yes or no” “discussion” (more like a vote), for example some politicians may want to add aspects of their religion into their legislature I don’t see how this makes their religion “irrelevant”, another thing could be how rarely “ethnicity” was used.

The same argument above requires another kind of debunk for the “ethnicity” argument, yes It’s impossible to simply say “that person is that ethnicity”, and for Bernie Sanders the argument was used “according to American laws he would be White, not of Hispanic origin”, even if the argument was limited to Bernie Sanders it would still be wrong, “White, not of Hispanic origin” is a ‘’racial’’ and not an ethnic grouping, him being a Non-Hispanic Jewish American would fill the category, in America you have races (who can, or can’t be Hispanic, you can combine any race with “Hispanic” thus having a Black, Asian, and/or White politician all being “ethnically Hispanic”, though it’s more of both a linguistic and geographical thing, but that’s another discussion), I must agree that “ethnicity” could be hard to define as not all ethnic groups are genetic, you can have ethno-linguistic groups such as “Hispanics”, and “Arabs”, ethno-religious groups such as Croats, Muslims (Southern-Slavs), Serbs, and Muslims (genetically Han Chinese), and yes it could be hard to ask whether an “Ethnic Muslim” from Bosnia and Herzegovina who is an Atheist should be seen aa “a Muslim” but again, that’s a while different debate, and if they would still ethnically define themselves as “Muslim” despite not believing in God (yes, I’m calling “Allah” God as we write in English), and (s)he lives in a predominantly Serbian area where discrimination is rampant and news 📰 stories constantly repeat that the person is “brave fighting Serb supremacy” than removing “ethnicity” from that person’s infobox does the reader a disservice. The same argument could be said for Jews in the shoah, their “citizenship” would still say ”German”, and placing “Jewish” at “Nationality” would confuse some readers with little idea of history that the person was an Israeli citizen.

Which brings me to “Nationality” simply doesn't replace “ethnicity”, the main argument I read for the cases where ethnicity was relevant was that we should “simply use Nationality”, not only does the term “nationality” only describe “ethnicity” in America, Canada, and Bolivia (thus making the term confusing for readers who aren’t from there), even in those places “nationality” could still mean which country’s passport you hold. To give a good comparison in the Republic of China you have 3 types of “Chinese/Taiwanese” people, you have “Chinese/Taiwanese citizens” (people who are registered in a R.O.C. family book 📚), “Chinese/Taiwanese nationals” (people who hold a R.O.C. passport, but don’t necessarily live in the free area), and “ethnic Chinese/Taiwanese” (people of Han, Mongolian, Tibetan, Manchu, Muslim, Etc. Heritage in the world, thus making it well over a milliard “Taiwanese” people worldwide), in this context it could be easily understood that “citizenship” is residential, “nationality” is which passport they hold (thus “national”), and “ethnicity” is genetic.

Another argument I would like to propose is that in some occupations ethnicity is important such as acting, for the sake 🍶 of keeping up with the WP:NPOV policy I never clarify who, or what I am, and what I believe in, but I could tell that some of the arguments proposed came from “politically left” people that felt that ethnicity, and religion should NEVER be disclosed, and that It’s ALWAYS irrelevant. Well, I’m far-left myself, and I think that in real life someone’s ethnicity, race, gender, sexuality, Etc. Should be irrelevant, but if I would put that into Wikipedia I would actively violate the neutral point of view where they are possible, look no further than acting. To me personally I can enjoy a good visual story regardless of if the protagonist or antagonist is male, female, homosexual, heterosexual, non-Hispanic White, non-Hispanic Asian, non-Hispanic African/non-Hispanic Black, Hispanic and/or Latino, if they speak Wolof, Hausa, Cebuano, Quechua, Etc. A good story is a good story, but in reality people (and by extend reliable secondary, and tertiary sources) do care about these things, if Bruce Lee is known for being a Non-Hispanic Chinese-American in Hollywood who broke various barriers for other Orientals than removing his Han ethnicity would remove crucial information from the infobox, the infobox should be “an article within an article”, a condensed “article at a glance” look and withholding information for any reason is simply un-encyclopedic, it’s not something that should be done, let alone celebrated. Yeah, by my own above argument I could watch “yellowface”, and yeah, I could if the story is good enough, and so could people during the 1920’s, if there’s a yellowface actor removing their ethnicity would be equally of a disservice as the fact that they weren’t Oriental makes it relevant to their profession, and yes even to their notability. (if a White (wo)man is famous for playing Asian that shouldn't be omitted).

Further, and this is beyond Wikipedia’s scope, but infoboxes help make search engines like Microsoft Bing, Google, Etc. In categorising data at a glance, though I do think that there should be strict guidelines when or when not “religion”, and “ethnicity” can be used (and I’d even propose an “irreligion” category for non-theists like Atheists, Deists, Free Thinkers, Agnostics, Etc. But I digress), removing them simply doesn't do anyone a arrive.

Which brings me to my final argument, why such a weak attempt? The categories at the bottom of the pages for religions, races, and ethnicities still exist, clearly it's worth categorising people as such when it’s relevant, and to claim that “it’s private”, Wikipedia talks about people’s sex lives if that’s been covered by the news 📰/historical sources so “WikiGossip” is already a thing, and if they openly disclose these things to the public, why should we hide it? Wikipedia is a free encyclopedia about notable, and relevant information, removing information from hundreds of thousands of articles because 1 American politician’s page was subject to edit warring is not only childish, drastic, and unnecessary, it’s unproductice and removes relevant information from countless of other articles.

Sent from my Microsoft Lumia 950 XL with Microsoft Windows 10 Mobile 📱.

I hope that someone who read the above is willing to debate this. 1.54.210.203 (talk) 11:49, 3 June 2017 (UTC)[reply]

Oppose - In my opinion, both religion and ethnicity are irrelevant details for the majority of our bio subjects... and certainly not worth highlighting in the infobox. In most cases, it does not matter if someone is/was of French or Scottish or Serbian heritage... it is what they did that makes them notable, not where their ancestors came from. In most cases, it does not matter if the person is/was Jewish or Christian or Muslim... What matters is what the subject did, not what their religion is/was. OK... sure... there are some subject who are notable because they are of a given religio-ethnic background. And in such cases we can create a custom info box to note this. But the template should be for the majority of cases where this is not the case.

Secondly, all too often, adding a subject's religion or ethnicity to an info-box ends up fomenting argument... as editors with an agenda try to claim the subject as "one of us" (or perhaps more importantly, not "one of them"). Such arguments can be avoided by simply omitting the data from the info-box in the first place.

Finally... you mention the religious and ethnic categorizations... I am of the opinion that adding religious or ethnic categories to bio articles is all too often a case of Wikipedia:Overcategorization. I think we should enforce WP:NONDEF with more vigor, and remove such categorizations in cases where the subject's religion or ethnicity is not really central to what makes them notable. Blueboar (talk) 15:40, 3 June 2017 (UTC)[reply]

There were too many arguments before about this. Having a parameter in an infobox tempts people to fill in the value, even when really unknown or poorly sourced. So it is better to leave it for the text. The facts and any nuances can be expressed in a sentence with a reference. Graeme Bartlett (talk) 22:00, 3 June 2017 (UTC)[reply]
Oppose. The issue was thoroughly considered at the RFC, and the reasons for the outcome have not changed. I think this discussion can be boxed up early. The significance and details of religion and ethnicity are better addressed in the article text. (With an allowance for boxes such as clergy, where religion is intrinsically significant.) Alsee (talk) 02:22, 4 June 2017 (UTC)[reply]
Oppose the results of the previous discussions were sound. Two points for the OP 1) please take note of WP:TLDR and 2) nobody cares what device you made your post from. MarnetteD|Talk 02:41, 4 June 2017 (UTC)[reply]
Oppose per Blueboars arguments. Gråbergs Gråa Sång (talk) 08:57, 4 June 2017 (UTC)[reply]
Oppose" per the overwhelmingly clear consensus at Wikipedia:Village pump (policy)/Archive 126#RfC: Religion in biographical infoboxes and Wikipedia:Village pump (policy)/Archive 127#RfC: Ethnicity in infoboxes.

Sent by whistling into my telephone into a 56K modem line and running data compression, TCP/IP Stack, and HTTPS encryption in my head. --Guy Macon (talk) 11:33, 4 June 2017 (UTC)[reply]

Oppose Put simply, if a field is there, some numbskulls will think it has to be filled in. But in the overwhelming majority of cases, someones ethnicity and/or religion is irrelevant. —Farix (t | c) 12:06, 4 June 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Comment on the close, which appears to have happened about 29 hours after the start:

Mandruss, I think that waiting a whole year before re-opening a contentious discussion is very good form. I believe that we normally hope that people will wait only half that long.

Sure, it would have been good to add an RFC template, but that's an easily solved problem.

On the merits of the question: it probably also would have been good for participants to first understand that ethnicity and religion are absolutely enormous factors in parts of the (non-Western) world, and only second form their opinions. I imagine, for example, that Tutsi people in Rwanda have a different view about the relevance of ethnicity than white people in the US. If your country had an ethnic war recently, then the ethnicity of people in your country is likely critically important information from your POV, even if someone from somewhere else thinks that you "shouldn't" care. WhatamIdoing (talk) 00:36, 12 June 2017 (UTC)[reply]

I'll reply since I was pinged. 1. I feel that my close represented the consensus, as closes are supposed to do. 2. If you read WP:RFC you'll see that the absence of an {{Rfc}} template is far from the only thing separating this discussion from a properly-formed RfC. You can't take a discussion and just slap the template on it and call it an RfC.
If you still feel this shouldn't have been closed, re-open it. ―Mandruss ☎ 01:21, 12 June 2017 (UTC)[reply]

Mandruss, I've written a good deal of WP:RFC, so I don't think I need to read it again to say that I think this was salvageable beginning. If it were me, I'd have added a handful of examples, with sources saying that ethnicity or religion was a decisive factor in the local culture, and I'd have added a brief summary question before the long explanation, but my preferred approach isn't the only one. (Actually, I'd probably not have opened it, because it's not a question that I care about personally, and our track record for understanding cultures where ethnicity or religion is a rigid, involuntary, and important fact of life is poor.)

I believe that it is very difficult to know whether the comments that appeared during 29 hours over a weekend, without contacting any of the groups that know something about the subject area, actually represents consensus, especially since a large fraction of the commenters never got much further than making an anti-wiki and anti-policy "consensus can't change" statement. I think you should have left it open, or at least have refrained from telling people that they shouldn't discuss decisions that were made last year. WhatamIdoing (talk) 02:00, 12 June 2017 (UTC)[reply]

The RfC that established the consensus lasted a bit longer than 29 hours, so you may be missing some of the point. In my opinion, the chances of a change here closely approximate zero, so I'm not going to re-open it; that's the thrust of WP:SNOW as I understand it. As I said, if you think it should be open, re-open it, but I'm not going to do so. I cited WP:BOLD in the close statement for a reason. ―Mandruss ☎ 03:17, 12 June 2017 (UTC)[reply]

It would have been good for Whatamidoing to first understand that the participants know perfectly well religion and ethnicity are very important for some people, and even a brief perusal of the previous RFC's will show that lots of arguments were made that it was too nuanced an issue in many cases to be left to an infobox which cannot go into detail. Only in death does duty end (talk) 08:20, 14 June 2017 (UTC)[reply]
A 29-hour closure was absurd in this case. And SNOW seems to have been applied on the basis of a handful of aligned signatures, not because there are overwhelming number of reasons presented here. On the merits, religion and ethnicity are in many contexts both objective and relevant (take the Northern Island conflict and Apartheid South Africa as places where they represented hard-edged groups with near constant political and social relevance). The stance of the community is, quite frankly, anomalous with respect to the human community about whom the encyclopedia should be written, and the original decision privileges convenience of editing ("it causes arguments") over encyclopedic utility (what social parameters ought to characterize people in a systematic fashion). We ought to have this conversation at some point in the future, and two days on Village Pump isn't enough time to establish we're not ready to have it now.--Carwil (talk) 03:00, 15 June 2017 (UTC)[reply]

This means that in such contexts, the religion or ethnic aspect is sufficient to have significant coverage and for the article to discuss the issue. If it's important enough, the lead will also mention it. Why is the infobox necessary then? —Paleo Neonate - 03:39, 15 June 2017 (UTC)[reply]

PaleoNeonate—why wouldn't the Infobox be used to convey to the reader an attribute of identity that reliable sources support as being applicable to an individual, be that attribute ethnic or religious? The real question is, what do sources say? Obviously in the absence of substantial support in sources we are not going to convey to the reader that such an attribute applies to an individual. But given sufficient support in sources I don't think the Infobox is an improper format in which to convey such information to a reader. Bus stop (talk) 14:10, 15 June 2017 (UTC)[reply]

PaleoNeonate — There are at least two reasons to have this in the infobox:

The infobox is designed for quick reference of relevant biographical data, and in many cases, this information is highly relevant to place someone, while not necessarily belonging in the lead describing them. Consider Rwandan politician Victoire Ingabire Umuhoza, an ethnic Hutu; and German-born philosopher Hannah Arendt, an ethnic Jew (but eventually not a religious Jew, hence Jewish would go under ethnicity) born in pre-war Germany and exiled from Nazi Germany. For both, ethnicity would become "an indispensable datum of my life" (Arendt's words), defining their situation in times of turmoil, while not characterizing their professional life more broadly. That is, it should be easily reference-able, but doesn't merit being in the lead.
The infobox services the systematization of knowledge in a machine-translatable form via the Wikidata project, and as an element easily translated onto other language Wikipedias.

In both these capacities, ethnicity and religion stand alongside other parameters of the Infobox person template and the person properties on Wikidata, like height, movement, and organization, that should not be filled out for many humans, but are relevant to others.--Carwil (talk) 14:30, 15 June 2017 (UTC)[reply]

Thank you for outlining the precise reasons why labels such as ethnicity and religion do not belong in infoboxes. Readers wanting a quick summary and machines scraping factoids would often be totally misled by a short label which cannot capture a complex story. Johnuniq (talk) 22:19, 15 June 2017 (UTC)[reply]

Agreeing with Johnuniq here; :If it were possible to always state it unambiguously and without dispute, it might have been thought acceptable, but it is not always possible, and there is no place to make a clear distinction obvious or not. There has been a large number of contentious disputes over these parameters, and thee disputes have had a tendency to turn into extended and unresolvable arguments between some members of the different groups involved. The shorter the statement, the more difficult it is to write a universally acceptable description, and in practice not having these disputes had been very beneficial to the encyclopedia . The only way to not have them is to never use the parameters. DGG ( talk ) 14:36, 16 June 2017 (UTC)[reply]

Overlinking

MOS:DUPLINK says, "Generally, a link should appear only once in an article, but if helpful for readers, a link may be repeated in infoboxes..." I have seen many editors remove linking on this basis, on the grounds that the first instance should be linked, but not the subsequent ones. For example, in an article on "Widgets", "John Smith" is linked to the article on "John Smith" the first time "John Smith" is mentioned. However, later in the article, John Smith is not linked to. I can understand the logic in this, but it seems to imply that people read Wikipedia articles from start to finish. I don't think this is true. People read what they are interested in. For example, someone might read the section entitled "Widgets in Spain" and see a mention of "John Smith". Who is this "John Smith"? they ask. There is no link. There is a link to "John Smith", but it is buried halfway through the article under "Widgets: Technical Aspects". Is there any research which says that people read Wikipedia articles from top to bottom?--Jack Upland (talk) 02:17, 6 June 2017 (UTC)[reply]

I don't think there's any research, but instead it's part of our MOS. Things like linking, abbreviations/initialisms, last names over full names, etc. are all based on the assumption that there are two ways we anticipate readers use articles: They read the lede only, and they read the lede followed by the rest of the body. We know people will jump to a section, but it makes it very clumsy for us to account for this mode and still make the other modes cleaner to use. We're still writing that each article should be usable in a standalone mode. I do agree that people shouldn't strip away links without any thought, but their inclusion does require careful thought. --MASEM (t) 02:28, 6 June 2017 (UTC)[reply]

Just slightly above in WP:OVERLINKING: "An overlinked article contains an excessive number of links, making it difficult to identify links likely to aid the reader's understanding significantly". I think that this concern is still valid in relation to duplicates. Also, because links appear in another color or style, even if not reading the whole article, it is rather easy to visually locate the important links (this too would be more difficult if the article contained too many links)... —Paleo Neonate - 18:31, 7 June 2017 (UTC)[reply]

However, if I was looking at a particular section of a long article, I would have no way of knowing there was a link to John Smith elsewhere. I could search the page for the name, but I might assume there was no article on John Smith because of the lack of link. Removing the link to John Smith in that section has made navigation difficult with no real gain.--Jack Upland (talk) 20:43, 7 June 2017 (UTC)[reply]

My point is that we've made a decision in the MOS and several other places that we work on the assumption the reader will read from top to bottom, which has requirements that are mutually exclusive from those that presume the reader can jump in and start in any other place in the article (eg we establish what an initialization stands for once in the article, at its first use in the body, rather than the first time in each major section). We know this is not ideal for the reader that jumps to a section, but we had to pick one way or the other. Fortunately, linking is not as mutually exclusive compared to other MOS parts. We do want editors to use good judgement in such links not to spam one reuse in each section but fair reuse is reasonable. --MASEM (t) 01:36, 12 June 2017 (UTC)[reply]

@Jack Upland: I don't really see a problem to use common sense and link more than once in very large articles, since in this case the readability issues I mentioned don't apply. Here was an obvious overlinking example (and I think that some work had already started to reduce them). What you propose would not necessarily result in this

. Duplicates was not the only issue, but general unnecessary overlinking... —Paleo Neonate - 08:47, 14 June 2017 (UTC)[reply]

Agree If I had to make a call, it would be to allow for duplicate linking between large sections in noticeably large or complex articles. People might not read the whole article and jump between sections, so when they come across a term – if it wasn't linked in the immediate section above, I see no harm linking it again. That's essentially how I see a lot of articles having been edited. There's always duplicate linking inevitably, and can be helpful to the reader as long as it isn't cluttered. DA1 (talk) 08:06, 10 June 2017 (UTC)[reply]

Just thought I'd mention that links should actually appear twice in an article, once in the lead and once in the body. I'm personally more concerned with everyday terms being overlinked than with a legitimate link appearing an extra time. Primergrey (talk) 13:01, 14 June 2017 (UTC)[reply]

I agree. The overlinking of everyday terms is far more of a problem. Not only is it unnecessary, it is also potentially misleading. A reader can be lead to believe the link will lead them to something specific and relevant to the article whereas they are just led to a general article about an ordinary term that most people understand.--Jack Upland (talk) 19:00, 15 June 2017 (UTC)[reply]

I agree about the over-linking of everyday terms (such as ISBN). But try to avoid "easter-eggs", which should help minimise the second issue you raise. All the best: Rich Farmbrough, 11:11, 17 June 2017 (UTC).[reply]

Agree I also agree that in long articles, repeating links in different sections would facilitate navigation between articles. I also agree that common sense needs to be applied to ensure that there is not overlinking that hinders readability. As long as the section isn't already full of links and it's not a trivial term, adding a link would be helpful and could prevent needing to search the page (and losing your place) to find a link. Adding this to the MOS would reduce the chance that someone removes helpful links because "it's policy". Ry's the Guy ^{(talk|contribs)} 14:01, 14 June 2017 (UTC)[reply]

Agree I agree as per Rystheguy above. MB 15:35, 14 June 2017 (UTC)[reply]
agree also, though we would need someguidelines on how to use it. Major sections of new articles is one good use. though not for every term. Another is the olacewhere the concept is being primarily discussed. DGG ( talk ) 14:12, 16 June 2017 (UTC)[reply]
Agree with repeating links in long articles, especially when you get to a section well below the lead which is the main one to deal with the link topic. I must say I thought this was what the policy said, rather more clearly than it does now (ans the language doesn't appear to have changed for a few years). Also I usually link names & work titles etc. in captions. I note FAC seems fine with both of these, and requests by reviewers for 2nd links in text or caption are not unusual. Johnbod (talk) 14:53, 16 June 2017 (UTC)[reply]

The MoS is a guideline - this particular element says "Generally, a link should..." and goes on to list some ten or so explicit situations where the guideline might not apply. Where there is a good reason to re-link, go ahead and do it.

As far as research, I am pretty certain there is research that shows people generally read the lead and sometimes a little more. The injunction against over-linking is to avoid a "sea of blue".

Perhaps we should consider functionality to highlight a section of text and provide a search option - there are plugins that do that.

All the best: Rich Farmbrough, 11:22, 17 June 2017 (UTC).[reply]

Is there a policy about writing articles about the same subject at 2 different technical levels?

Resolved

Hello again. I've seen several articles with the "too technical" banner produced by {{technical|date=}} at the top. Is there a wp policy about how to deal with this, other than what is stated in the banner "Please help improve it to make it understandable to non-experts, without removing the technical details"? What I want to suggest is to have TWO articles about a topic. For example, there could be an article Endoplasmic reticulum (which currently has a too technical banner) that was written so that a freshman high school student could understand it, and a second article (red link for now) Endoplasmic reticulum College level written with all the details expected by senior university students. If somebody searched wp for endoplasmic reticu... both articles would appear as suggestions. The reader could decide which to access. Readers of the basic article would be alerted to the more detailed article by hatnote at the top and "see also" entries. Is there a precedent for this? Is there a policy against it? Thanks, JeanOhm (talk) 16:34, 6 June 2017 (UTC)[reply]

@JeanOhm: There is precedent (see Introduction to evolution and Evolution and also Category:Introduction articles) but it is rarely done and usually only for "popular" subjects. --NeilN ^{talk to me} 16:39, 6 June 2017 (UTC)[reply]

If we were to do that then we might as well have a whole separate encyclopedia for advanced learning. It would also cause issues as one person's opinion on what is "college level" differs from the other. - Knowledgekid87 (talk) 16:39, 6 June 2017 (UTC)[reply]

Yeah, this is best avoided. We're an encyclopedia, which means we present information in WP:Summary style. That often excludes highly-technical details. You can always make use of further reading sections to suggest more advanced texts and liberally use wikilinks to provide in-line explanations for technical words/concepts. ~ Rob₁₃^Talk 16:42, 6 June 2017 (UTC)[reply]

Not a policy here, but you can always fork an article to simple:Simple English Wikipedia. — xaosflux ^Talk 17:26, 6 June 2017 (UTC)[reply]

See also the featured article Introduction to viruses. An other such article is Introduction to M-theory. The others can be found here from [[Introduction to cooperative learning to Introduction to systolic geometry. עוד מישהו Od Mishehu 12:07, 7 June 2017 (UTC)[reply]

In this specific case, I'd recommend improving the lede, but leaving the rest of the article as written. Cell is the page that should discuss the Endoplasmic reticulum at a level suitable for a high-school freshman. Power~enwiki (talk) 20:49, 7 June 2017 (UTC)[reply]

Thank you all for the replies. JeanOhm (talk) 01:05, 8 June 2017 (UTC)[reply]

Maybe one last note: to avoid are POV forks (WP:POVFORK), however, of course. These are not about simplification but about POV pushing, a different matter. —Paleo Neonate - 03:27, 8 June 2017 (UTC)[reply]
In most cases, the key is to make the lead, and perhaps a section or two after it, as simple as possible, then, if very technical detail is unavoidable, put that below the lead. WP:MEDRS stresses this for medicine, & it is accepted (if often not followed) for articles on advanced mathmatics. Johnbod (talk) 14:57, 16 June 2017 (UTC)[reply]

WP:Citation overkill RfC

Opinions are needed on the following matter: Wikipedia talk:Citation overkill#Should this essay be changed to encourage more citations?. A WP:Permalink for it is here.

Although WP:Citation overkill is not a policy or guideline, it is often treated as one, and the RfC is framed under a policy RfC because it is the only RfC template that fit it. Flyer22 Reborn (talk) 01:57, 9 June 2017 (UTC)[reply]

Inclusion of non-encyclopedic pages in mainspace as a technical workaround

There is an AfD here hinging on the question of whether non-encyclopedic pages can be retained in mainspace as a technical workaround. It's not clear yet how the community feels about this, but it is clear that more attention is needed to the matter. —swpb^T 12:41, 9 June 2017 (UTC)[reply]

Page LGBT conservatism Takeover

Over the last several months the page LGBT conservatism has been modified, a few edits at a time, to characterize LGBT conservatism as alt-right. Although some edits seem constructive, the first edit was to change all words "conservative" to "right-wing" and to try to move the page (change its name to LGBT Right-Wing) which was refused. Links and headings to "alt-right" and "fascism" were added; some users deleted some of the more egregious changes. I was thinking of doing a global change of "right-wing" back to "conservatism" but found that this is a huge task at this point, so, on the advice of another user, before I proceeded I thought I should come here with my concerns. I have three problems: I'm a very junior editor, I have no time or interest in an editing war, and I'm not particularly knowledgeable about the subject matter. The site owner does not seem to be in evidence. The policy issue is editing that replaces normal labels like "conservative" with pejorative labels like "right-wing". -motorfingers- 08:37, 10 June 2017 (UTC)

The best place to discuss this is at Talk:LGBT conservatism. I see you did start a couple of threads there. (I'll reply to one of them presently.) I'd suggest dropping a neutrally-worded note at a couple of relevant WikiProjects, Wikipedia talk:WikiProject LGBT studies and Wikipedia talk:WikiProject Conservatism, if you want to get more eyes on this. Rivertorch^FIRE_WATER 16:44, 10 June 2017 (UTC)[reply]

I did post neutrally toned remarks three times on Talk:LGBT conservatism with no result, apparently. I do notice that the most egregious changes, e.g. references to "alt-right" and "fascism" are gone. I don't recall page watch e-mails for those changes and haven't looked in awhile; perhaps I watch-flagged just the talk page.

I haven't seen LGBT studies. Looking there, I don't see a reference to LGBT conservatism at all. A global search reveals no other mentions of the exact phrase "LGBT conservatism" on any other Wikipedia page. I don't see a good place for a link to LGBT conservatism on the main project page, unless LGBT conservatism has been designated a "good" article or other category there. Perhaps there is another page in the project that should link to LGBT conservatism. -- motorfingers : Talk 06:43, 11 June 2017 (UTC)[reply]

If you believe the article is not neutral, you could tag it with {{POV}} or any of the other Wikipedia:Neutrality templates that would be appropriate. —Farix (t | c) 11:23, 11 June 2017 (UTC)[reply]

@motorfingers: Not LGBT studies, Wikipedia talk:WikiProject LGBT studies. And Wikipedia talk:WikiProject Conservatism. Follow the links to the talk pages. WikiProjects can be good resources when you want to draw attention to issues at a relevant article. Rivertorch^FIRE_WATER 18:47, 11 June 2017 (UTC)[reply]

Should global renamers/stewards be authorized to process requests at WP:CHUU?

Please see: Wikipedia talk:Changing username#Should global renamers/stewards be authorized to process requests at WP:CHUU? –xeno^talk 16:47, 10 June 2017 (UTC)[reply]

WP:OWN and essays

I have been thinking about WP:OWN and how it relates to WP:Essays. Essays often contain the opinions of a specific group of Wikipedia editors. But let's say that another group of editors comes along and significantly edits the essay... to the point where the essay no longer reflects the opinions of the original group of editors (perhaps to the point where it now states the complete opposite). If the original editors try to object (perhaps saying "Look, you can go write your own essay... please leave our essay alone"), the new group can just point to WP:OWN (and say "Nope... you don't own the page"). This bothers me. If one of the reasons for essays is to give editors a place to state opinions, shouldn't we provide some mechanism to protect those opinions? I am thinking that we should allow some degree of ownership when it comes to essay pages. Please discuss. Blueboar (talk) 19:56, 11 June 2017 (UTC)[reply]

I believe that an essay may ber edited by othger users only to the degree that it still expresses the same opinion. Of course, any user may create an attributed copy, and edit this copy to say anything they want. עוד מישהו Od Mishehu 20:09, 11 June 2017 (UTC)[reply]

No firm opinion here, just some thoughts:

Is there a distinction to be made between user space essays and Wikipedia-space essays? WP:POLICIES already says that any essay that the author does not want changed should be in user space.
A policy which says that a Wikipedia-space essay cannot be changed in a way that it no longer expresses the same opinion could be read to say that any change whatsoever causes it to vary from that opinion, since the Devil is often in the details.
If we have a policy which says that essays can be OWNed, is that going to lead to a proliferation of essays with just tiny differences? For example one essay says that infoboxes ought to have colored backgrounds (and the color should be blue) and another says color, yes, but light blue and another says color, yes, but royal blue, etc. And if so, is that a problem?
What about an amendment to OWN which says something like, "Since essays are the opinion of the author, a much higher degree of ownership-like conduct is to be tolerated in order to prevent the essay from moving away from the editor's original position. If an essay is located in an editor's user space, the author has the prerogative to modify or revert any changes made to it by any other user. In either case, however, the right to engage in ownership-like conduct does not allow the essay's author to delete or modify other editors' edits to the essay's talk page. Nothing here, however, is intended to allow edit warring over the content of any essay, wherever located."
Who should be entitled to exercise OWNership of an essay? Only the original author? If the original author fails to defend an essay from changes, should that simply make it fair game for changes or should it trigger deletion or userfication of the essay?

Best regards, TransporterMan (TALK) 21:57, 11 June 2017 (UTC)[reply]

It should be noted that most (I said most) essays in the Wikipedia namespace are there because they reflect a large portion of the community's views, though not large enough to elevate the essay to guideline status. Some of these essays are even used (successfully) as reasons for deletion at AfD. They may also be used to explain a portion of a policy page. If any of these essays were changed, this would likely be immediately noticed and reverted. The editor who made the initial change would then have to show why community consensus has changed to the point of having to rewrite the essay (essentially WP:BURDEN, but for the Wikipedia namespace). — Gestrid (talk) 00:07, 12 June 2017 (UTC)[reply]

For anyone curious about the context, see the RFC at Citation overkill, where there is some discussion about whether the essay should be changed to (somewhat) encourage citation overkill, instead of to oppose it. WhatamIdoing (talk) 02:10, 12 June 2017 (UTC)[reply]

There is no discussion about whether any essay should be changed to encourage citation overkill. The RFC is malformed and misinterprets others intentions. QuackGuru (talk) 02:38, 12 June 2017 (UTC)[reply]

While that RFC certainly inspired my question, my question here isn't really about that RFC. I would like this discussion to be more general in scope... it can apply to many of our essays (especially those essays that express a minority viewpoint... cases where a group of editors disagree with broader consensus, and have written an essay to express their opinion) Blueboar (talk) 14:08, 12 June 2017 (UTC)[reply]

IMO, people should be able to OWN their essays, as that's what an essay is, it's an expression of one person's thoughts and opinions. If you disagree with those opinions, write your own essay. If an essay becomes important enough to be commonly cited, it should probably be promoted to a guideline, as happened with MOS:ICON (which was originally Wikipedia:Don't overuse flags). Kaldari (talk) 23:33, 12 June 2017 (UTC)[reply]

You have accurately described essays in user space. Those in the Wikipedia namespace represent the views of far more than one editor, and some of those gain enough support to be called "widely accepted" essays. Notably, WP:BRD is widely considered to be widely accepted, and it carries more weight than the average essay, but repeated attempts to promote it to guideline have failed. You are suggesting that it should therefore be moved to user space, and I don't think that's going to happen. ―Mandruss ☎ 23:45, 12 June 2017 (UTC)[reply]

Self-correction: When did BRD get promoted from essay to "explanatory supplement"? Regardless, it was essay for many years, even years after it was widely accepted, and that was seen as a normal situation. WP:OSE is another example that still exists. ―Mandruss ☎ 23:49, 12 June 2017 (UTC)[reply]

BRD was promoted from essay to "explanatory supplement" in May 2012. Hawkeye7 (talk) 02:44, 13 June 2017 (UTC)[reply]

At the suggestion of another editor, an essay of mine (WP:RECEPTION) was recently moved from my user space to project space. As far as I'm concerned I lost ownership when that happened. If it evolves in a direction I don't like, I'm free to recapture my original ideas in another user essay. Mike Christie (talk - contribs - library) 02:52, 13 June 2017 (UTC)[reply]

I agree with Mike. If the community disagrees with the content so much that a consensus emerges that changes it beyond what you originally intended, you can always write another essay expressing your own unfiltered views. That said, it's perfectly legitimate for the community to 'protect' namespace essays even if they don't fully agree with them to preserve the original opinion, as is happening in the case that prompted this discussion. So I don't see a need to introduce a concept of ownership in namespace. Scribolt (talk) 06:52, 13 June 2017 (UTC)[reply]

Why should those who spent time and effort to write an essay to express their viewpoint have to write another essay to express their viewpoint? After all, they have already gone through the time and effort to write one. Shouldn't the onus to write another essay be on those who don't share the original view of the one already written? And if the original (minority view) essay writers do write a second essay to express their unfiltered views, do they "own" that? what's to stop the majority from changing the second essay as well? When does it stop? Blueboar (talk) 10:57, 13 June 2017 (UTC)[reply]

Because as Mike said, the second essay would be a user essay. Scribolt (talk) 12:23, 13 June 2017 (UTC)[reply]

It shouldn't be a user esssy if there are multiple editors who support it. Say the original version (now the second essay) reflects the views 20 editors... but 200 disagree with that view. The 200 can effectively silence the view of the 20 by constantly editing that view away. Blueboar (talk) 21:14, 13 June 2017 (UTC)[reply]

Well, it's not effectively silenced because anyone who wants to can still read it. It's not like it disappears. In your hypothetical example, I think that if 200 people felt strongly enough that something was inappropriate and needed to be removed or amended, then it should be userfied or changed. I disagree with your premise that just because multiple (2? 3?) editors think agree on something that that means it automatically deserves protected status in namespace. Namespace should mean that enough people tolerate it (even if they don't agree with it) for it not to be significantly changed in tone or content (as is happening on citation overkill, where editors are explicitly defending the essay's right to exist and not be compromised). If not enough people care enough about what an essay says to stop others from disrupting, the chances are that it doesn't say anything that other people find useful or relevant and if the author really wants to preserve it, then they can have it as a user essay and the multiple people that agree can still refer to it. Scribolt (talk) 06:52, 14 June 2017 (UTC)[reply]

hmm... a lot to think about here. thanks for all the replies. Blueboar (talk) 15:02, 17 June 2017 (UTC)[reply]

It can be tricky to distinguish between following the original author's topic for the essay versus the author's personal opinion on the topic. Take Wikipedia:Administrators' best practices as an example. Leaving aside the question of whether or not the topic would already fit into another page, are future contributors required to preserve the same best practices as written by the original author, or can they introduce new best practices? Can the original best practices be refined further or modified, or would this be considered a dissenting view that should be covered in a separate essay? In that case, does it make sense for one of these essays to be called "best practices"?

Another challenging aspect is how much of the original author's writing style should be considered part of the original message to be preserved. For this example, I thought the original tone was overly conversational in a way that personalized the advice from a particular non-neutral point of view. I felt this was detrimental to communicating the recommended best practices, as administrators could dismiss them as supporting that point of view. However I appreciate that others may think this is part of the original author's message and so is untouchable. Again some of the problem lies with the essay's title, which implies the content has gained general consensus. If the title more clearly identified the essay as one person's thoughts, there would be less need for the essay to be written from a neutral standpoint. isaacl (talk) 00:03, 18 June 2017 (UTC)[reply]

Widely held (but minority) views

A question related to the one above. WP:POLICIES states that "Essays that the author does not want others to edit, or that are found to contradict widespread consensus, belong in the user namespace." My question is... how "widespread" does a "widespread consensus" have to be? Can an essay be considered to have "widespread consensus" if it is supported by a reasonably large and diverse group, or must the essay be supported by the majority of the community? In other words... how large/small a consensus should an essay have before it is moved from main policy space to userspace (or conversely from userspace to main policy space)? Blueboar (talk) 15:02, 17 June 2017 (UTC)[reply]

Proposal for new response parameter to COI edit requests

Hi all. As a COI editor who sometimes makes requests on behalf of clients, I've been struggling with the length of the COI edit request backlog. It pretty consistently hovers around 170 requests, with many that are four months old or more.

I understand that the community is busy, and these requests can't take special precedence. However, an uncertain response time that spans months or more is a tough sell when I'm advocating to potential clients to work with me through Wikipedia's proper COI disclosure process.

As a partial solution, I propose creating a new response parameter to the {{request edit}} template, along the lines of "Revision needed." Editors could use this response to ask the original requester to update their request with additional detail or sources as needed before resubmitting. This would reduce the number of poorly formatted or ill-thought-out requests in the queue, so that editors can focus on reviewing the requests that are ready for thoughtful consideration and inexperienced COI editors can receive useful guidance on how to collaborate with the community. I also propose that COI editors be permitted to provide this response to certain COI edit requests, where it's applicable. To be clear: I am not suggesting that COI editors be empowered to accept or reject others' COI edit requests wholesale. This would simply be a way for COI editors to "give back" to the community by providing guidance to their peers and helping make the backlog less overwhelming for the NPOV volunteers who assist us.

Looking forward to everyone's thoughts and feedback. Thanks for your time! Mary Gaulke (talk) 21:21, 11 June 2017 (UTC)[reply]

Sounds too much like the tail is wanting wag the dog. It is COI's who are causing the backlog. Aspro (talk) 21:38, 11 June 2017 (UTC)[reply]

This proposal doesn’t reduce our workload but just creates a temporal displacement. It take more of an editors time to review (as other edits may have been made before an article can be reviewed and so the history will have to be searched as well to make sense of it all) so it will increase our workload. Suggestion to remove your frustration -Stop using WP to promote your clients business interests and encourage them into paying for an entry in trade directories etc. Aspro (talk) 22:01, 11 June 2017 (UTC)[reply]

Note that WP:EPH exists that you can point users to. --Izno (talk) 23:36, 11 June 2017 (UTC)[reply]

This discussion is from Wikipedia_talk:Conflict_of_interest/Noticeboard#Potential_process_changes. Read there for more of the history. Most edit requests from COI editors aren't good edits; I mentioned a half dozen bad ones back there. Still, a bit of help at reducing the backlog at User:AnomieBOT/EDITREQTable would be nice. I've dealt with about 10 of those, but can't do it alone. Thanks. John Nagle (talk) 05:16, 12 June 2017 (UTC)[reply]

Think the OP is playing with semantics here. There is no difference between Request edit & proposed "Revision needed." Has anyone noticed the second proposal : ”I also propose that COI editors be permitted to provide this response to certain COI edit requests, where it's applicable.” Translate: Paid COI's can then advise other COI's to use Request edit. There, you will find, unpaid volunteer editors that will be eager guide you through how to promote your company etc. with out you having to learn how to edit WP competently. After all, you're not interested in WP mission to spread free knowledge to all but just promote your own self interests. So let some other jerk do the work. Ref: ibid and in OP's own words: “Editors could use this response to ask the original requester to update their request with additional detail or sources as needed before resubmitting”. Notice she did not say We editors. To the OP: what did your last slave die of? Aspro (talk) 15:24, 12 June 2017 (UTC)[reply]

@Aspro: I hope you'll lighten your tone. Whether you like the idea or not, the OP is offering a suggestion to reduce the workload for volunteer editors, by proposing a recognized method for knowledgable COI contributors to offer advice to less knowledgable ones. I think it's a good idea, and I'd support it. As a longtime COI contributor (and longer-time volunteer) myself, I'd be willing to help offer feedback to make requests easier for volunteers to follow up. Moreover, there already is a request edit queue on Wikipedia for this purpose, and Jimbo Wales has asked volunteers to help, so paid editors feel less compelled to edit directly. And speaking from experience, a lot of this work is not about promoting clients, but simply correcting outdated information. WWB Too (Talk · COI) 15:59, 12 June 2017 (UTC)[reply]

Just as quick follow-up note, I think that Aspro might have missed that Mary isn't suggesting pointing COI editors to {{request edit}}, what she's saying is that "revisions needed" be added as a response parameter to close existing open edit requests that aren't appropriate in their current form. COI editors would help adding this response where it's needed, to close open requests that are problematic for editors to respond to, and give the requestors feedback on what they need to do to make their request appropriate to Wikipedia. Same as WWB Too, I'd be more than happy to jump in and help out with such an effort. 16912 Rhiannon (Talk · COI) 16:10, 12 June 2017 (UTC)[reply]

@Aspro: I think you may misunderstand me. As 16912_Rhiannon noted, I'm proposing a new response parameter that would allow COI editors to help triage the incoming queue of COI edit requests, so that NPOV editors can focus on requests that are of high quality, and COI editors (including myself) can help their peers learn how to format and source their requests properly in order to create the least possible friction. Per WP:COI, of course, COI editors are not supposed to edit directly articles on which they have a COI; if they believe changes should be made (including basic factual corrections and updates), they must either use this edit request process or violate Wikipedia guidelines. And to be clear, regarding your "We editors" comment: I fully count myself as one of the COI editors who would take time to use this new response parameter to help reduce the load on other editors. I'm a Wikipedian too who regularly works to improve articles on which I'm NPOV. I'm sincerely trying to help the project here. Mary Gaulke (talk) 17:21, 12 June 2017 (UTC)[reply]

Don't misunderstand me either. Think you have a brilliant idea to set the fox to guard the henhouse Aspro (talk) 17:51, 12 June 2017 (UTC)[reply]

To be completely clear: I don't suggest empowering COI editors to approve or implement others' requests, just to provide guidance for improving the quality of poorly sourced or ill-conceived requests. Mary Gaulke (talk) 17:59, 12 June 2017 (UTC)[reply]

Well, You have these rights already! As a WP editor you already have the right to offer guidance to other less experienced editors! Also, if you have been grated reviewers rights, you can also improve the quality of poorly sourced or ill-conceived requests on flagged protected pages! Your argument appear to revolve around paid COI's being given a special privilege to have their edits short-cut our policies where we endeavour to treat all editors as equals. Reminds me of George Orwell's Animal Farm. Quote: "All animals are equal but some animals are more equal than others”. Aspro (talk) 23:33, 12 June 2017 (UTC)[reply]

True we can do that now, although the new parameter could help to flag it in the queue temporarily (if in fact that's what MaryGaulke is suggesting) and allow volunteers to focus on requests that are more "ready for prime time". It would also probably help somewhat for the less experienced COI contributors to get some helpful response, even if it's not a final decision. WWB Too (Talk · COI) 19:03, 13 June 2017 (UTC)[reply]

Correct—that is indeed exactly what I am suggesting. Mary Gaulke (talk) 00:24, 14 June 2017 (UTC)[reply]

In general, there is such a big fixation on the COI aspects that there is not enough awareness of how they serve it up to the reviewing editor.....whether their method makes it a big or small job for the editor that might put it in. My experience is that even very intelligent editors lack this empathy/understanding, and propose it in a way that would make it a very big job to put it in, thus impairing response to such requests. The best remedy/ forcing statement that I can give briefly is to say to give simple, explicit instructions for the requested edit. North8000 (talk) 01:11, 14 June 2017 (UTC)[reply]

COI does not necessarily refer just to paid editors--it applies also to the person being discussed in the article, or those with a close relationship. Their help can be very useful in building up articles, and they should be encourage to contribute--in the proper manner. One good way of encouraging them is to respond to their queries quickly, and anything which will help people sort out which to reply to is useful. DGG ( talk ) 14:18, 16 June 2017 (UTC) .[reply]

Sports-specific criteria and GNG discussed at WT:NSPORTS

The relationship between sport-specific criteria of WP:NSPORTS and WP:GNG is discussed at Wikipedia talk:Notability (sports)#Proposal re GNG/SSC relationship. --George Ho (talk) 02:06, 12 June 2017 (UTC)[reply]

As this was just discussed, I think it is far too premature to re-open this topic. isaacl (talk) 03:33, 12 June 2017 (UTC)[reply]

Many of us did not even see the discussion here. There are too many forums on WP, we can't possibly watch them all. Such discussions will go on, in multiple forums, as long as WP exists, and you can't stifle them just by saying "we already talked about that." Jack N. Stock (talk) 05:19, 12 June 2017 (UTC)[reply]

There was a notification placed on the talk page for the sports-specific notability guideline. As you can see from the discussion thread above, there were many participants in the discussion. Should anyone wish to discuss the matter again, they can of course proceed. Just be forewarned that there may well be a lack of receptiveness to continuing a discussion that already took over a month and a half. isaacl (talk) 05:35, 12 June 2017 (UTC)[reply]

I guess I didn't visit WP that day, or maybe we had just finished the same discussion elsewhere and I considered this one redundant. Jack N. Stock (talk) 05:58, 12 June 2017 (UTC)[reply]

First, I ask everyone to read (or re-read) the previous discussion. The next question is... does anyone have anything new to add to that discussion? If we are simply going to rehash the debate we recently had, then there is no point in discussing it again now... but if someone can raise a new point or concern (something that was not already discussed to death) then it is worth reopening the discussion. Blueboar (talk) 22:38, 12 June 2017 (UTC)[reply]

Related proposal to demote WP:NSPORT

Discussion closed against proposal per WP:SNOW. Nothing more to do here. SkyWarrior 18:28, 12 June 2017 (UTC)

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

For those who watch this page and not WT:NSPORT, there is a related discussion to demote WP:NSPORT from guideline to information page or essay. Please discuss at Wikipedia talk:Notability (sports)#Proposal for demotion of NSPORT. Jack N. Stock (talk) 05:58, 12 June 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Proposing to bring all wikipedia policies in accordance with NPOV

IP blocked for block evasion.--Bbb23 (talk) 19:51, 13 June 2017 (UTC)[reply]

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Making wikipedia policies as neutral as it's encyclopedic content. Do not forbid people with particulate believes or medical conditions to participate in the community for example. 178.187.10.7 (talk) 17:51, 13 June 2017 (UTC)[reply]

What leads you to believe that Wikipedia policy or guideline cares about the particulars of a person's beliefs or medical conditions? --Izno (talk) 18:00, 13 June 2017 (UTC)[reply]

Some of it's policies. But my question is: Why we can't make this rule: Wikipedia be neutral by itself, not just it's content. Policies and everything be neutral. — Preceding unsigned comment added by 178.187.10.7 (talk) 18:23, 13 June 2017 (UTC)[reply]

It will be difficult to convince other editors that your ideas are valuable if you do not specify, in some length, the problems with certain, specific ,policies and guidelines. Which policies? --Izno (talk) 18:28, 13 June 2017 (UTC)[reply]

Problem is not with policies. The problem is that Wikipedia is not neutral by itself. It could stand for certain political, ideological or moral believes, and there is no rules forbidding such stands. — Preceding unsigned comment added by 178.187.10.7 (talk) 18:39, 13 June 2017 (UTC)[reply]

I don't understand the request or the problem. Our policy on use of copyrighted material is stricter than required by (US) copyright law. Would that be considered an idiological stance? — Arthur Rubin (talk) 19:05, 13 June 2017 (UTC)[reply]

Was it implemented from ideological or from practical reasons? Did it help it's encyclopedic purpose or harmed it? I don't know details.

Oppose there's no proposal here, and I would oppose any of the proposals I suspect the IP-editor has in mind. Power~enwiki (talk) 19:07, 13 June 2017 (UTC)[reply]

Oppose First, WikiP is already is as neutral as possible. Next, please see the items at Wikipedia:What Wikipedia is not#Community and WP:CIR for reasons that it shouldn't be 100% neutral. The IPs suggestion would leaves things open to all sorts of WP:FRINGE articles which are best avoided. MarnetteD|Talk 19:16, 13 June 2017 (UTC)[reply]

My proposal was very clear: Make Wikipedia's neutrality obligatory. It will only help it's purpose. — Preceding unsigned comment added by 178.187.10.7 (talk) 19:18, 13 June 2017 (UTC)[reply]

We already have WP:NPOV. If you don't propose specific changes to the wording of policies/guidelines then all you're doing is wasting everyone's time. --NeilN ^{talk to me} 19:21, 13 June 2017 (UTC)[reply]

Propose to change "All encyclopedic content on Wikipedia must be written from a neutral point of view (NPOV)" on "All encyclopedic content on Wikipedia and it's policies must be written from a neutral point of view (NPOV)" — Preceding unsigned comment added by 178.187.10.7 (talk) 19:24, 13 June 2017 (UTC)[reply]

Thank you and oppose. Wikipedia is "biased" towards summarizing mainstream viewpoints found in high-quality reliable sources and our policies and guidelines reflect that. I'm comfortable with this. --NeilN ^{talk to me} 19:31, 13 June 2017 (UTC)[reply]

Did you read WP:NPOV? Your post is basically the first sentence of that policy. MarnetteD|Talk 19:30, 13 June 2017 (UTC)[reply]

I will give it a cautious welcome. Yes, it would be positive for Wikipedia. No, it will certainly lead to wiki-war. For instance, it will cut down the way WP:MEDRS is used to strangle subjects that only touch to health. For instance on "organic food" where MEDRS claims more importance than sources provided by agricultural universities and the like. It would also lead to equal treatment of schools and colleges across the board, demanding the same notability from American schools and colleges as required for non-American schools and colleges. Is it necessary to do that? Yes, it is. Is it realistic to do it? No, as the vested interest will fight it to the bone and beyond. The Banner talk 19:48, 13 June 2017 (UTC)[reply]

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Do we need different standards for fictional and 'actual' content?

If I write that George Bush was born in 1904, that can easily be checked, likely from the references on the page. If I say the NYT reported that JFK was assassinated, that can be checked as well. There are citations, and information without citations is flagged or taken out. However, it is different when it comes to fiction, or tv shows, or things of that sort.

Take an example of these articles: Fictional universe of Harry Potter. Or Produce 101 Season 2. How can the facts in these articles be verified? Well, you need to have read all the books, or have seen every episode. Lots of vandalism occurred to Produce 101 today, but frankly it was hard to tell what was real and what wasn't. How do I know who was taken off the show and who wasn't?

And Harry Potter: "Some wizards are the products of unions between humans and magical creatures of more-or-less human intelligence, such as Fleur Delacour and her sister Gabrielle (both quarter veela), Professor Flitwick (a quarter goblin), Madame Maxime and Hagrid (both half giant). Prejudiced wizards (such as Umbridge) often use the insulting term half-breed to refer to mixed-species wizards and werewolves, or other beings such as house elves, merpeople and centaurs (who are separate species). The centaurs within the series prefer to exist amongst themselves, with little interaction with humans." Is that entirely made up? I have no idea. Maybe you do. But even if you read the books, you'd have to look back to check. And there are thousands of pages of Harry Potter.

You probably get the picture. A lot of my time is spent in articles re current events, especially terrorist attacks. The things that are argued over are pretty minute, and they involve dozens of editors. Was the Manchester attacker a Muslim? There was quite a bit of argument over that. But is Harry Potter a union between a quarter veela and a Umbridge? I don't know. Someone could put that in and frankly unless someone just read the books or knows them by heart, how can any of these things be confirmed? It's clear that there is just a discrepancy between the two categories.

So, what's the solution? I don't know. You could add disclaimers: This plot summary may not be correct. Or you could make it so that extensive plot summaries need to go to Wikia or some other outlet. I don't know. This is just something I have been thinking about. WP is not a paper encyclopedia, but I don't know if it has any business re-stating the plot of anything. It seems to go against what it's good at-- cited factual information. ‡ Єl Cid, Єl Caɱ̩peador ^ᐁT₳LKᐃ 06:11, 17 June 2017 (UTC)[reply]

This is already being discussed in an AfD. As the page title starts with "Fictional", I don't particularly care as long as it's not egregiously bad, e.g. a copyright violation. Power~enwiki (talk) 06:54, 17 June 2017 (UTC)[reply]
If I write that George Bush was born in 1904, that can easily be checked, likely from the references on the page. If I say the NYT reported that JFK was assassinated, that can be checked as well. There are citations, and information without citations is flagged or taken out. However, it is different when it comes to fiction, or tv shows, or things of that sort. Firstly, I don't actually agree that the ease of verifying something is that strongly related to whether or not it is fictional. It would be much easier to verify that "in JK Rowling's Harry Potter novels, Harry was the son of James and Lily Potter" (which I would bet good money can be cited to a dozen places other than the Harry Potter books) than it would be to verify "the British artist Henry Moore had a pet dog called Fawkes, named after Guy Fawkes" (really true, but unless you happen to know where that fact comes from, not easy to verify). The difference is not whether or not a fact is relevant to the real world or to a piece of fiction, the difference lies in how significant a fact is to the subject. In-Universe, Harry's parentage has plot significance, while Fleur Delacour's doesn't.
Secondly, current events (especially terrorist attacks) are often held to a much higher standard of sourcing than any other article for two reasons: firstly, the BLP implications, and secondly, they often contain information which is "challenged or likely to be challenged": that is, meeting two of the three circumstances according to WP:V and WP:BLP where inline citations are actually necessary rather than required. By contrast, if you read articles on ancient history, lots of them, especially in less-popular topic areas, have been less fully cited than I would prefer, and/or cited mainly to primary sources, for years.
Personally, I would like to see more citations in plot summaries, but the fact that they mostly don't have them does not appear to be holding plot summaries to a different standard of citation than other information. Caeciliusinhorto (talk) 08:11, 17 June 2017 (UTC)[reply]

I also would like to see more citations in plot summaries. I often see detailed plot summaries which are completely unsourced, in my view this indeed would be a de-facto lower standard. It probably doesn't matter until someone begins contesting. It seems that since fictional information is non-critical information (i.e. versus medical claims, BLP claims, fringe theories, subjects with political implications (some may include myths considered as fiction by many, but is really of higher gravity), etc), primary sources can also be used and would be better than no source. This may become required when a content dispute occurs on things which are easily verifiable using a primary source. However, if the interpretation of those primary sources becomes difficult to the point of content disputes (primary sources not being enough), then resorting to secondary sources may be necessary. If the subject is notable and difficult to interpret, notable secondary sources which already did such interpretation, research or synthesis, are likely to exist. If such secondary sources are reliable (if they even exist) is probably another matter... —Paleo Neonate - 08:48, 17 June 2017 (UTC)[reply]

if the interpretation of those primary sources becomes difficult — there should be no interpretation at all. State the plot in the simplest terms possible, and leave any interpretation to reliable sources. Bright☀ 12:33, 17 June 2017 (UTC)[reply]

Plot summaries do not get citations because the work itself is supposed to be the citation. Stuff that isn't in the original work needs a citation but doesn't belong into the plot summary in the first place. Secondary sources are not good for this, they tend to have accuracy problems. Jo-Jo Eumerus (talk, contributions) 10:22, 17 June 2017 (UTC)[reply]

It may help to differentiate between a broad plot summary and a narrow plot detail. Broad summary is verifiable by reading/viewing the primary source (the book or performance itself). Narrow plot details are also verifiable by reading/viewing the primary source ... but narrow details are much more likely to be challenged than a broad summary, and so may need to be cited to specific locations within the primary source. Certainly, a detail needs to be cited to the specific location if it actually is challenged (per WP:BURDEN). Then we have plot analysis... analysis always needs to be cited to a secondary source (Per WP:NOR). Blueboar (talk) 10:50, 17 June 2017 (UTC)[reply]

The big issues with plot summaries are that they either get too detailed, or that they are "hooks" like jacket blurbs or magazine listings. Some are also awful English. All the best: Rich Farmbrough, 11:25, 17 June 2017 (UTC).[reply]

This is covered under WP:SYNTH. I agree, plot summaries are often coat-racks for original research and for that reason they should be kept to the bare minimum, and preferably cited to non-primary sources. Bright☀ 12:26, 17 June 2017 (UTC)[reply]

Actually, unless you are explicitly writing analytic, evaluative, interpretive, or synthetic plot summaries you should refrain from using secondary sources as they are often wrong. Jo-Jo Eumerus (talk, contributions) 12:32, 17 June 2017 (UTC)[reply]

You should never write analytic, evaluative, interpretive, or synthetic plot summaries. Bright☀ 12:35, 17 June 2017 (UTC)[reply]

Thanks for all the opinions guys, really. I continue to stand by what I said, that the content in those sort of articles are of a lower quality level and with lower reliability. But it probably is just that it will stay how it is, I just had to share my opinions, and thanks for considering them. ‡ Єl Cid, Єl Caɱ̩peador ^ᐁT₳LKᐃ 15:53, 17 June 2017 (UTC)[reply]

We have chosen to have higher standards of sourcing in two areas, Biographies of living people and medical matters. Both are subsets of "actual information". The implication of that is that we already have a higher standard of sourcing for the average non fiction article as opposed to the average article on a fictional subject. There are good reasons why we want to have stricter standards in those areas. The risk of people acting on false information in an article on a fictional subject is hopefully less than if the subject is medical or a living person. I'm not currently aware of a similar need in another area, but if there is I'd suggest we start the discussion with an explanation as to why the consequences are more significant of an error in the area where you are proposing stricter standards. Ϣere SpielChequers 18:36, 17 June 2017 (UTC)[reply]

Discussion of interest

You may be interested in this discussion, which relates to an interpretation of the verifabilty policy and the citation guidelines. Wikipedia:Miscellany for deletion/Wikipedia:Citation underkill. Scribolt (talk) 19:01, 17 June 2017 (UTC)[reply]

Relisted RM at Template talk:2016 US Election AE

The ongoing RM discussion at Template talk:2016 US Election AE is relisted, so feel free to comment there until closure. --George Ho (talk) 01:12, 18 June 2017 (UTC)[reply]

Neutrality and Morality

One of the larger problems that I've been seeing develop over the past 4 or 5 years but is growing to a head is the issue of neutrality with respect to morality in discussing subjective opinions about persons, groups, and other organizations.

Where we don't have problems is when someone or some group is seen as a subjective term but that is near-universally morally acceptable. For articles on people/groups that are regarded as geniuses, as the best creative person, as a strong businessman, etc., we are acutely aware of the appropriate language to use in these articles to stress that these are widely-held opinions and not so much fact. We rarely use these subjective, if not universally-applied labels, in the lede sentences, but introduce this later in the lede with statements that clearly make it attributable opinion rather than fact. This all is properly in line with WP:NPOV and other policies.

Where we do run into problems is when the subjective term is something that is seen to a larger group as immoral. It is human nature to call someone out when their morals vastly differ from ours, so it is natural to rush to push this immoral stance front and center. This is part of the problem.

Now, it is important we're not trying to play down someone or some group that has been convicted in court or a similar body of a immoral crime that directly harms persons or personal property. These types of actions are near universally accepted as immoral, and we should treat those as such. We shouldn't coddle subjective language around convicted murderers and terrorist groups, for example, though one must stress that if there's only an accusation pending trial, we presume innocent until proven guilty, which is where this morality question plays out.

When we turn to a subjective label that are particularly about someone's viewpoint absent of any actual harmful actions, such as "white nationalist", for example, this is clearly a view that to most would consider immoral, but it doesn't have the universal agreement; there are parts of this world where that is considered morally acceptable. This sets up the problem that I've seen: since by the nature of what we have deemed to be reliable sources for such commentary, they are generally going to share a specific moral center, which would agree with the larger moral center shared by Western, English-speaking regions. There is no question that this larger view should be included, but far too frequently, these views are presented front and center in the lede sentence of the appropriate article and without clarifying them as attributed opinion. Editors when challenged with this often argue that since the majority, sometimes near unanimity of the press, says this, it must be fact and must be presented as soon as possible per WP:UNDUE.

However, I would argue that WP needs to have more an amoral center, one that aligns with how we treated "positive moral" terms. Unless we're talking about convicted crimes, WP should not adopt the same moral center that the bulk of the RSes does as to reflect a more proper world view and to stay much more neutral on these types of topics. This means that we should not be so focused on introducing the subjective labels as early as the lede sentence and sacrifice factual content, nor presume the "factual" nature of these subjective labels. Again, we don't do this at all for "positive moral" terms, we shouldn't at all be doing the same for "negative moral" terms.

This also avoids a potential slippery slope, even if the press fall along that. Having views that are generally seen as "negative moral" like bleeding-heart liberal, fascist, anarchists, racist, misogynistic, white nationalist, alt right, far left, or far right is just having that view. Unless their actions actually lead to universally immoral crimes, they haven't done a single thing wrong from a legal standpoint, they just share a vastly differ viewpoint from what the moral average does. While the press may want to condemn them for this type of thinking, that seems extremely out of line with our job to be objective and neutral. Again, this is not saying that the majority stance shouldn't be included, but it should be tempered as to be attributed opinion and not as early as some wish to write. It also avoids the slope where less problematic labels or "thought crimes" deemed by the press are treated the same by us, such as "climate change denier", "conspiracy theorist", and so forth.

Most of the time, this is easy to edit to fix problematic articles, adapted a more amoral tone that does not seek to condemn the person or group. Take Jared Taylor which has come up before (specifically this version for discussion [33]) where the lede calls out his white nationalist and supremacist labels before actually introducing him properly. When it has been discussed before on BLP/N [34] many editors think that because the majority of sources treat him as such that these should be the facts and presented first. If WP was more amoral, we'd start by describing him as an author, and founder and editor of the magazine, followed by a sentence that says that he is generally considered a white nationalist and supremacist. Tidying up that lede to be more amoral and not treating "white nationalist" or the like as a crime to mark such persons with would go a long way towards meeting NPOV for the article overall. Here's another case happening just now [35] where just because the majority of sources call Breitbart "far right", it should not be forced as a factual statement in the lede sentence, though still clearly must be emphasized in the lede. But because other media sources label has such, editors insist it must be front and center. I stress: the goal here is not to eliminate such information from articles, but to temper it appropriately with attribution and outside of WP's voice while maintaining the proper balance.

The other problem is that without treating things with a more amoral approach, we then allow the moral center of WP to be determined by the press. As noted, this reflects only a portion of the world view, but more importantly, this view shifts over time, which affects our articles over time. For example, with Trump's election, the moral center has significantly shifted to the left, which if we held to that moral center, we'd have to update a lot of articles to reflect how people and groups are now treated this way. Since WP aimed for long-term stability, it would be much better that we work from a more amoral center, so that we're not forced to make changes like this even as the media's moral center changes.

Technically, NPOV has the language towards all this under "Morally-offensive views", WP:IMPARTIAL and at WP:OUTRAGE, and WP:LABEL also has some, but these are routinely ignored putting WP:UNDUE first and foremost.

The TL;DR of all this is that I propose we develop a guideline to support NPOV that has a stronger establishment that WP should take a more amoral stance towards morally-offensive views (just as we do for morally-accepted ones), and guidance for how to write such articles where the morally-offensive view is something you can't avoid writing about. This guideline would make it clear that there are places we accept are morally wrong (convicted crimes against others), but in most other cases, we should be much more neutral in tone and writing to not treat it as a condemnation even if every major press source wants to treat it that way. There needs to be a better balance of UNDUE and IMPARTIAL in these areas and how to be more amoral, otherwise we are going to continue to reflect what the press deems is the moral center.

Moreso, there needs to be advice to editors to put aside their own opinions when writing such articles; just as we have acute awareness when an article seems to factually present a morally-acceptable label to rework it as attributed option, we need that same acute awareness for offensive ones, which is often clouded by one's personal opinions.

(Cavaet: I in no way personally endorse, approve, or support any of the example views/labels, I'm just pointing out where there are problems.) --MASEM (t) 15:25, 18 June 2017 (UTC)[reply]

Please, let Gamergate go. Johnuniq (talk) 23:26, 18 June 2017 (UTC)[reply]

Why would any special policy be necessary? We should go with the consensus of reliable sources. When do we say a person is Jewish? When most of them say he's Jewish. When do we say he's white supremacist? " " " " " " " " " But if the sources say he's white separatist, then by golly, we say he's white separatist! We're here to shuffle the cards and deal them out, not to toss any under the table nor pull any out of our sleeves. Wnt (talk) 00:56, 19 June 2017 (UTC)[reply]

Oppose for the specific Jared Taylor article, there's no claim he's notable apart from his actions as a white supremacist, and he publicly attests to those views [36]. The only conceivable option apart from mentioning them in the lede is to delete the article entirely. The details of language policing ("white supremacist" v. "white nationalist" v. "alt-right" v. "race realist") aren't relevant; we should use terms understandable by most readers and used by most secondary sources. Power~enwiki (talk) 01:04, 19 June 2017 (UTC)[reply]

Notability for electronic devices

Moved from User talk:Mz7 § Notability for mobile phones

Hi @Kudpung, Oshwah, and Primefac: and Mz7.
Recently, there are way too many articles being created for mobile phones. User talk:Usernamekiran#Notability. A perfect example for ~~WP:NEE~~ WP:ENN. I can come up with a draft for an essay for "Notability guidelines for electronic devices". But I dont know what happens to that draft later lol. I mean, I know there would be a discussion, many contributions from other editors, and a consensus before making it an actual essay/guideline. I request suggestion/guidance from you guys. —usernamekiran(talk) 23:16, 12 June 2017 (UTC)[reply]

There are also way too many articles about footballers (soccer players) based on a single mention in a team's website. For some reason Wikipedia considers them more important than scientists and university professors who have to jump through a whole row of hoops before they are allowed an article. I don't think there's much you can do about phones as long as they meet WP:GNG and are not blatantly promotional (generally phone manufacturers are so big they don't need extra advertising in Wikipedia). Kudpung กุดผึ้ง (talk) 00:01, 13 June 2017 (UTC)[reply]

@Kudpung: yes. I also noticed the same thing about football clubs. I have no idea about football (never played, never watched, dont understand), so i usually skip these articles. And WP:NACADEMICS always surprised me. I first came across this when I saw an article of an actress in AfD. She has an article in encyclopaedia only because she has minor appearances in few "notable films". And a parson who has studied, and worked hard; has done reasearch, has taught students cant have an article.

These days, there are websites that have pages for a mobile as soon as it is announced/launched. It doesnt matter how trivial that mobile is going to be. There are websites for reviews, and for "unboxing gadgets". So these gadgets get mentioned a lot no matter how "non notable" they are. Currently, for these new articles everybody just looks for current mentions. Nobody thinks about "notability is permanent" or WP:DEGRADE. If this goes on like the way it is, soon wikipedia will become brochure for electronic gadgets.

— Preceding unsigned comment added by Usernamekiran (talk • contribs) 06:55, 13 June 2017 (UTC)[reply]

To return to your original question - you are welcome to create an essay regarding notability of electronics. I think once it's done I'd get input from WP:TELECOM and maybe a few other projects in the Science and Technology space. They'll be able to give other thoughts, input, advice, and if there's enough support it can be converted into a formal guideline (though I'll be honest, I'm not sure if an RFC would be required for that). I support this venture. Good luck! Primefac (talk) 12:43, 13 June 2017 (UTC)[reply]

@Usernamekiran: I'm afraid I have to agree with Kudpung. In most cases a subject has to either pass a subject-specific notability guideline or the general notability guideline. If all we have currently for electronic devices is the GNG, and too many articles about phones are being created, I'm not sure creating an SNG is the proper way to go, since editors can argue that the subjects still pass GNG. Consider instead starting a centralized discussion about what level of coverage should justify a phone article, given the guidelines we already have, similar to discussions like Wikipedia:Village pump (policy)/Archive 133#RfC on secondary school notability or the discussion that ended the Wikipedia:Pokémon test. What kind of criteria were you thinking a new Wikipedia:Notability (electronic devices) should have? Mz7 (talk) 13:19, 13 June 2017 (UTC)[reply]

To put it in precise words, "devices that have created an impact on market/society, are mentioned in the sources which are not related to technology itself (including websites of reviews, and technical specifications). They should be mentioned in other sources (news?). The devices should have something different/revolutionary or something notable." We can also add a guideline/case for number of consumers. That is, even if the mobile doesnt have anything different/revolutionary feature, if it made an impact on market/society like Nokia 1100 did.

Most of the recntly created articles have issues with WP:MILL, WP:ENN, and WP:SIGCOV. I am still thinking about the wording though. This comment uses words very loosely. —usernamekiran(talk) 13:40, 13 June 2017 (UTC)[reply]

Hi @Kudpung, Oshwah, and Primefac: and Mz7.

I created a very preliminary draft for the essay at User:Usernamekiran/sandbox2. Kindly let me know what you think of it. I haven't added a declaration stating its just an essay and not a guideline. Also, there are some inline notes/comments. Your feedback is much appreciated. Thanks. —usernamekiran(talk) 21:04, 15 June 2017 (UTC)[reply]

Hi, and sorry for the late response here. My first thought here is that putting the idea out there that having notability requirements for electronics wouldn't be a bad idea. Do keep in mind that the general notability guidelines exist and are guidelines that all articles must meet, but if you feel that further guidelines need to be written specifically for this area (such as "Wikipedia:Notability (electronics)"), I'm open to seeing what the reason behind it is, as well as what proposals are made. ~Oshwah~^{(talk) (contribs)} 04:29, 18 June 2017 (UTC)[reply]

Hi everybody. Mz7, and Primefac; I made some changes to the draft essay.
@Oshwah: I am not stuck on making it "solid criteria" (rule?), but I think there should be an essay for sure. I mean, a loose guideline. An essay will be able to discuss/explain the already established WP:GNG in context to the devices. Its just that, without this explaining and coverage on internet, there are lots of articles being created for every mobile being introduced recently. This is also spreading to other electronic devices gradually. If this isnt taken care of soon, then we would have an article each for every model of 5.1 home theatre system manufactured by a famous company like Sony or Yamaha. —usernamekiran(talk) 19:45, 18 June 2017 (UTC)[reply]

As mentioned in the discussion above, the essay is still under construction. I am not sure if it would be made into a wikipedia policy; but the situation is calling it for sure. I would like opinions from other editors. Thanks a lot for the consideration. Best, —usernamekiran(talk) 02:15, 19 June 2017 (UTC)[reply]