Wikimedia meetings/May 12, 2010

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search

Raw minutes for the May 12 meeting. These need revision and condensing.

I hope this will be the first of a series of regular open discussions, since a number of people feel the community as a whole needs places to meet and discuss.

gmaxwell: notes that IRC isn't ideal for large groups -- perhaps the next discussion won't be on IRC :) thank for those who took time out for the chat today. and to jyothis for moderating. the bulk of discussion was kept to ~90 minutes, with some freeform ideas at the end. SJ · talk | translate


In attendance[edit]

jyothis, huib, mindspillage, sj, stuwest, thedj, patrol110, sir48, gmaxwell

qcoder00, killiondude, guillom, philippe, la2, gigs, huib, esby, ottava

dragonfly6-7, bishakha, stu, michaelsnow,



prepared topics[edit]

commons audience[edit]

  • about commons: who is its audience?
    • - like any other WMF project, the audience of commons is the whole world.
    • it is its own curated library. also a 'client' library for local projects


  • if commons has its own audience, what determines when media is apprporiate? (for the media that is /not/ used on projects)
    • What happens when the consensus in the community on contentious topics is questioned?
      • questioned by others in the community, or by people outside of it?
    • it has to be able to illustrate something.
    • the role of Commons as repository that accepts comprehensive donations of media goes against any considerations regarding redundancy and quality



  • what is the value of 100 images of a ford pinto?
    • big value - redundancy is ok for Repository-Commons
    • it's not about the number of pictures; unless it correlates with diversity, which it often does. angles, color, model, time, location & background, notable owner, etc.
    • Libraries are choosey in what material they have and remove material that is 1. of low quality or 2. would cause clutter in their holdings. but wiki is not paper.
      • "not paper" - how about if you can sort them by quality ratings?



inclusion standards[edit]

  • quality: commons inclusion standards
    • we regularly delete promotional band images from obsecur brazilian bands.
    • for reasonable quality, is it enough for material to have been curated by some other institution? (one 'yes')
    • these seem to reflect a liberal US viewpoint. Other Wikimedian communities have different cultrual traditions compared with the US (and different legal codes)
    • we don't need all NASA shuttle foto's, NASA shoots the same 100 shots for every launch, but we do have some from each.
    • I want only high quality work because I try to only produce high quality work. I would rather have 5 high quality than 50 low
      • I would rather have 55 images, the 5 high quality ones, AND the 50 low
    • we already have a practice of putting selections of high quality into galleries, while pushig "random shots" into the categories.


  • impact
    • Wikiversity has trouble recruiting college professors to work on our site because of the reputation of other projects.
    • usage is not a good indicator of value


  • context
  • Some would argue a low wuality image is worse than none, it depends the domain and the usage
    • but it is not enough reason to store houndreds of dicks or another pics ;p
  • quality has different factors: how it relates to the text and adds to the text (context), and how it is visually/aesthetically appealing.
  • does it help to have context and good descriptions provided for each image?
    • yes... I think commons does a fairly decent job with descriptions, but they don't show on cats/galleries


  • ideally, images (and other media files) should be ordered by quality when viewing search results or categories
  • I like the idea of separating (in some way) in-use items from the vast "repository" of useful (but currently unused) images. Perhaps with something as simple as sorting in-use items at the top of categories, a "more" button, or other technological tweaks at Commons.
    • in category browsing, search results
    • we should have a repository-Commons one click away from the Commons with Project-media
      • and when something is put into use it becomes more visible


  • In an ideal world Commons would have everything and everyone would be able to find exactly that subset of everything which they want with little effort. I realize this is not possible. :-)
    • a real library of babel? with any offensive image one can imagine as well? ...yes.


contentious topics[edit]

  • policy formation on contentious topics
    • suggestion: don't concentrae on commons; take it as an example of what can happen in any project
    • I'm here really only to get an idea of what policy will become - as it affects the image tagging I'm doing on enwiki
    • What happens when the community is not able to reach a consensus? (examples wanted - no two issues are the same...)
      • where things were not claear, there is no consensus, but the system was calm before recent deletions


  • sexual content
    • sexual content should be stored at the sites dedicated to this aim. 1000-5000 potential files while commons now have 6000000 of them
      • that would means creating a commons2 which would contain only sexual content, that would only amplify the issue
      • that seems to be an idiotic idea :P Porno Commons or what erhm?



Do you think that our policies should be applied with the same rigor in al topics? Perhaps in controversial topics we should apply the policies more strictly.


commons specifics[edit]

  • concentrate on Commons also.
    • it has a maintenance problem: some old deletion discussions are not closed. (some consensus).
      • But enwp also has pages voted for deletion if not merged... back in 2003.. which still exist. some debates span years.
      • Is there any technical reason why Commons doesn't have a timeout for Deletion Requests? not a simple keep/delete decision.


Can we all agree that Commons needs a lot of technical improvements, and that those alone could solve a lot of Commons' problems?

  • yes... mediawiki is not designed to do what commons does. It's amasing that it works as well as it does.
  • (partly due to the awesome JS work that commons contributors have done over the years)




communication, different perspectives[edit]

  • we should respect wikimedians from other projects who are not able to speak English. These discussions are (usually) in English. How to resolve?


  • descriptions should be multilingual, multicultural
    • a limiting factor given how desciptions are shown now. (but better technical solutions are possible there)
    • like auto-translate? yes, and more translators
      • it's ok to have impossible goals. NPOV can be regarded as an impossible goal. But helpful.
    • Commons nak-mem suunosigconsatosi



on hosting content and implications[edit]

should we host all art? all images?

  • host all media donated from institutions - for other uploads: host useful, legal and free media but remove bad quality and excessive redundancy
  • there is rarely direct harm (or benefit) from the fact of hosting images. It's the way they are used that matters... which is tied to how hosted media are displayed.
    • there is rarely direct harm from having stub pages on non notable people from the 12th century, but we would still probably delete them per policy. :)


if there is no harm to simply hosting images, and the negative reactions are to how they are shown/discovered in context perhaps we shouldn't be focusing on notabiliyt/inclusion principles but on principles of presentation

  • I agree (no harm in hosting, concern from presentation)


what legally accepted and free media should not be hosted on Commons and why?

  • can someone speak to the potential harm of hosting / widely publishing controversial images?
    • - legal risk
      • I think we should keep that to the legal counsel.
    • - financial hit from uncomfortable donors - Crazies with guns
      • mostly irrelevant, (financial hit) WMF have specific policies which avoid taking enormous sums of money from single donors. We turn down large sums of money fairly frequently.
    • - boycot by countries (china, iran)
      • i doubt banning sexual content is the problem for country like iran & china



do we need to propose any policy formations atm?


I wish: to get community consensus about neutrally labeling items using something like ICRA ( http://en.wikipedia.org/wiki/Internet_Content_Rating_Association ) so third parties could filter if desired... while making it possible for those concerned to to address problematic and controversial content I understand it can be seen as a "step toward censorship" but to me it seems much more logical to be neutral and descriptive (as we are in all things) in a way that enables others to take action..

  • You support the view that 'self-content-rating' is better then Externally opposed direct censorship? it wouldn't protect Commons etc from actually ILLEGAL material
    • it allows concerned bodies such as schools, donors, and the god-forsaken FoxNews team to see that there is some "hook" that can be used to handle controversial content. But we don't have to censor ourselves, or provide a censored website.



is there any general harm in hosting images?[edit]

one of the strong criticisms of making controversial images available: seeing images considered taboo in a reader's culture is often seen as harmful to them emotionally or psychologically. before seeing such images it's hard to know if it will be harmful or not; and seeing them has an irreversible effect; cannot be unseen.

  • likeimages of the 'prophet'?
  • like showing images of a country's royal family?
  • that is a wide group of objections, impossible to manage with content scope criteria
    • this particular group fuels many ardent concerns, fears, criticisms of commons -- that hosting everything - and doing a great job at finding everything, while not serving different audiences differently, (and while being very popular) leads to harm to unwitting audiences.
    • all we can do here is offer tagging/metadata

Note: ja:wp bans use of criminals' names on their pages. this causes conflict with people using interwiki links to articles. would WMF need policies on such issues?

To give another example, someone asked for deletion of this image because he was 'shocked' [1] ...

Blocking / filtering / boycotting[edit]

Is the 'sexual content' issue a real problem? How big is it? - Cimon.

  • from Sweden, having talked a lot to schools and libraries about Wikipedia, the sexual content discussion has never been mentioned. is this a real problem for Wikipedia, and if so, how big?
  • Data needed. We don't even know the order of magnitude of the problem.
  • Where this is a problem: not just the US; at least parts of latin america, africa, the mideast. UK schools, libraries filter for 'adult' content (anecdotal). in the US schools *have* to filter to receive federal money. (but do any of them currently filter Commons?)
    • Not sure. we should find out before acting.
  • If commons were the problem, technically only it would be blocked, not WP.


differentiate b/t blocking (technically), filtering, and boycotting (socially)

blocking might be unrelated to specific content; just avoiding information reaching people; or avoiding 'social' websites. ...to actually work with schools we need to have more information on what they need

Differentiate 'rating' from 'filtering'. (technically china is filtering) I have not heard a persuasive argument that flagging is a significantly greater evil than being blocked wholesale etc.


people have been proposing

  • evaluating category effectiveness
  • tools for a customized experience (avoiding surprise).
  • actively removing material from the projects

custom experience:

  • we need either to be able to rate content or to use the category to filter out what the user does not want

filter design:

  • http://commons.wikimedia.org/wiki/Commons:Village_pump#Designing_filtering_software
  • someone should talk to the filtering-software makers. maybe they have technical ideas. Find out what their threshholds are.
    • we shouldn't worry about other people's threshholds. We can't control that, we shouldn't try, it will be different from country to country and org to org anyway. All we do is provide a tag/category/etc that says "female nudity-medical" and let those who fuss decide (rationally or not) whether Breast cancer should include that image for their kid/school

i would like to warn for the danger in deleting parts of donated material without relaying this information to the donating institution.



unwanted surprise[edit]

unwitting: people being surprised by media they didn't want to see.

  • we have images of the prophet in our categories on food... or something equally surprising.
  • Some christians also find certain types of Catholic art objectionable...
  • fortunately, surprising people is not a crime
    • surprising people is much more relevant that whether it's a crime or not!
    • is the categorization of the content the problem then?
      • not primarily
  • we don't want to censor what is useful to some, we don't want to cause offense to those who will be distressed.



  • some wikipedians find articles about primary schools objectionable. (however, often this means they get to delete them)


Why don't we just get a dutch and a german and other specific commons?


so then the question is how active one needs to be to reach the content? It takes a while to find a proper category and images one actually wants to use (e.g. for WIkipedia), is it so easy to fall into Category:Vulva?

only by visiting the article "Vulva" as far as i know, or by searching.

is it a good goal to avoid shocking people? there's the question of context for images. can this be decided on a project by project basis?

  • To establish context one idea I'd heard for controversial images was to extend the usage rationale system to otherwise 'free' images
  • I'm not sure it can. I would normally want it decided on a case by case basis for the use

[12:15:56] <gmaxwell> Lack of context is a big issue in my view. Perhaps we could improve it by making description text visible in cats / gallery pages?

other topics[edit]

why people browse commons[edit]

I don't think many people browse commons for fun


self-governance[edit]

  • self-government. foundation role: Guidance only like with BLP


how do we deal with admin under the age of 18 opining on images that are of pornographic material when it would not be legal for them to look at such material?


communities are free to establish more restrictive rules than is required by law, of course. notability is not a requirement on commons, is it?


how images are used - that's more about articles than categories, although categories play a role. Wikipedia articles primarily, but again, Commons itself and its categories are relevant

  • also somehow passing through the categories of a page's images to the page itself...
  • context can have legal implications: The lawfulness of some of the works we include in the US stem from their literary, artistic, political, or scientific value which isn't clear without some context in some cases.


Ideally, people should be able to know (not just guess, or expect) what they will see before they see it.. The difference between putting images of Muhammad in Muhammad as opposed to putting them in Depictions of Muhammad? Yes, a category/article clarifying you will see pictures of him, not just calligraphy or maps of his journeys

  • I think the images should be available to those who want them, but not to those who won't expect them.
  • I think people will often look up a subject to get information about it, and need that information before they can decide for themselves whether they want to see the picture of it.


so where do we draw the line for a reasonable expectation?


most of the heatd debate has been about deletion criteria, which is about hosting. (perhaps inspiring people to make up non-notable articles to get around a firm deletion rule)


Imagine someone wants the article and a description first, and then perhaps the option to look, but doesn't want it forced on him

  • this argument applies just as well to the text ... some 'porn' can be literature...
  • this is similar to spoilers in articles about films and books
  • does singling out images really make sense?


people are more shocked or offended by visuals than by text. We don't get complaints about profanity in Wikipedia articles, by and large

  • I think we get compariable amounts of complaint to explicit images as explicit text. Do we have any real data on it?
  • I have no idea, but that's where NPOV and our very academic/clinical approach comes in.
    • I got interesting complaints txt-wise, women who complained their birth dates were mentioned


providing service to children[edit]

A priority was to provide a free encyclopedia to be used by African school children.

  • Empty citation (help)  ?
    • valid point: do we exist to serve children as an important audience / does outreach to children make sense, given a goal of not being censored?

how to meaningfully reach children: a question of preentation rather than inclusion

  • "the sum of human knoweldge", isn't always great for kids.
    • Would you rather a child found about sex from i) their teachers ii) Wikipedia iii) the rapist? was a comment I heard someone make once...


  • does anyone know how many comes to Commons directly and not from projects?


  • the answer there is definitely yes; they're a large part of the world


Over categorization: there's no point in . Or "articles that have very graphic descriptions of disgusting things" (although that might be popular with schoolchildren...)"


providing excellent information about human sexuality is an important service we provide (I grew up in a part of the world where sex education was poor, and that has its own unfortunate consequences)

  • I too think that providing accurate quality info about sexuality is critical, specially in countries like India where we can't otherwise get it the Indian govt just banned sex ed in schools, so lots more young people are turning to the Net.


project-level decisions about presentation and appropriateness[edit]

individual Projects should decide on presentation guidelines... appropriate for their audiences

  • yes.... sometimes people need a certain level of information to help them decide whether to seek more

How do the projects communicate with one another about this? share best practices and differences?

  • do they?
  • Commons is very "international" but bad with communicating with its "customers"
    • it needs to be even more international
  • agreed. there's no easy way for someone on a small wiki to easily know where to go to complain about an image (or the disappearance of an image), or to learn how larger wikis from similar cultures have grappled with the same issues.
    • Non-english speaker are unable to participate in decisions - media are deleted without contacting projects etc.


Is part of the communication about what is available? how to participate? or simply translation limitations?

General thoughts[edit]

[12:16:19] <LA2> Wikipedia and all its sister projects have limitations. Maybe its a good way to create an encyclopedia with 100 - 100M articles, just like a steam engine locomotive is good for 1 - 100 mph, but a steam engine can't propel a rocket into space. Same with "international" for Commons, it works great for some countries, but maybe not for all countries.


was the 'spoiler' idea discussed for filtering content, like putting an intermediate 'must click' warning for displaying some content ?


  • MichaelSnow: hm. It probably should have a video, if the thumbnail is innocuous, and the caption warns you that it will be very graphic! That is the kind of barrier I am thinking of--you know it's there, but you also know you might not want to see it.
  • mindspillage: Yes, exactly


Wikimedia community: past, present, future[edit]

I worry about the community in the future because it seems that most people want to remove any restrictions in their way and push forward without any set standards.


bishakha raises a very interesting problem - that compliance with some government/cultural(?) policy would actually be against our mission

  • we cannot control all the world governements for sure.



it's strange to me that commons discussion and policy are all in english

clearly a failing; collections of media are at least as expansive and interesting and difficult as collections of encyclopedic knowledge


the project itself is based n en and there is no individual lang versions of it. is there systematic lanaguge bias?


there is no cross-project policy body

for the future, I'd like to see some body that could discuss project-level policy solutions (and policies that cross multiple projects). they could consider questions of scope

with representatives from different community, so that there is no confusion about foundation roles there.

the tough thing about cross-project policy is precisely that it's hard to coordinate discussions of that magnitude or length on a wiki page


On a related point about language issues, How many board/Foundation members speak a foriegn (ie non English language)

  • Arne (german), Ting (german, chinese), Jan-Bart (dutch), Bishakha, SJ (german, hebrew), others?




future planning, future of the foundation[edit]

I am a little afraid of a huge foundation, requiring tons of money and being scared of losing donations

responsibility should be distributed the community could do many more jobs voluntarily

a bit of hosted in Russia, a bit in Sweden, we'll manage

  • the community could do many more jobs voluntarily. no travel costs needed probably


create working groups on wikies

and keep away from editorial decisions

take action in case of legal trouble and otherwise do guidance

  • local projects have tried to accomplish a lot of that. creating clear space for growth is important


The main "customers" of the foundation are the chapters - not the projects :)

  • O_O
  • that depends on how you look at it... or who you ask
    • at any rate, finding ways to let the volunteer community do more is extremely important

encouraging distributed work and creativity is one of the current foundation priorities. there's a need for better cross project communication. language will always be a hindrance for that

  • but that's ... different from the size of the foundation. you could do that with both a small or a large foundation.
    • are we actually finding ways to let the community do less? if we define additional guidelines/ways of prioritizing/categorizing/filtering content. or pre-screening content as some suggest... we will see a drop in productivity I suppose.


we don't hear opinions from Africa here, do we

  • Wikimedia needs translators?
  • mmm, regular communication in 200 languages. nice vision.


and natural growth of organizations often expands to fill available space.

  • yes, hence my warning.

we should continue the discussion about how to help people do more on the wikis (later).

"that's the real discussion."

Wikipedia, Luhteranism, and distributed work[edit]

whether religous or not, I think many wikipedians come from a Lutheran background, when it comes to things like we don't believe in having "a pope" or a large hierarchical organization.. the idea that everybody is empowered and nobody is your boss, this is a Lutheran idea, resonating very well in North America and northern Europe


the idea might have roots that are far older than Martin Luther himself, it resonates well with the Icelandic sagas, while the Roman Catholic church resembles the Roman empire

the idea to distribute work (everybody is entitled, empowered) and to spread our ways of working to other countries is the classic Lutheran missionary we don't hear much from Africa, but we go there and tell them


age of admins[edit]

one more point brought up earlier: about young admins dealing with explicit material

Should there be an age limit of admins?

  • that seems to me a social decision that each project needs to make. guidelines for such things differ by community.
  • in the one decision the Japanese wiki blocked Korean wiki users who never went on jp and then demanded global blocks on meta


ideas not pursued this time[edit]

  • Is there a scope for a technology project to detect what is a user's language selection in their wiki project and add their email id to a mailing list for that specific language? Foundation could use that to send out regular communication in that language


a discussion about WMF's future in 3rd world countries would be interesting, but perhaps some other time


splitting content by legal jurisdiction[edit]

hosting commons in Germany would solve the sexual content issue, but then we'd have to remove all swastikas, right?

  • wait - You want content splits on jurisdictional boundaries?


Luckily, national laws have little to do with what we do and don't host. (it's more our own policies and guidelines that govern this) . If there were a significant legal problem with hosting material in the US, that might be a reason to move the server farm or even to incorporate somewhere else.


I think a separate commons type area for de and nl in the Netherlands would ease many concerns. not all of it moved, just have a sub commons type section

  • no, a Commons mirror!


The question of what is potentially illegal in differnet jurisdictions is hard to solve. but as a community of reusers, we need to help one another solve it.=

each of us has to solve this question when we make subcollectoins or extract categories for use in books, or or or.

  • tough, since many grey areas in the law too
    • yes. but how better to get at those tricky nuances than by encouraging thousands of people to iterate on a shared body of knowledge?