Wikimedia Apps/Share a fact

From mediawiki.org

Share a fact is a feature on Wikimedia apps that allows sharing text extract from Wikipedia articles. It is available on both Android and iOS, where users can highlight text from within articles, and share it with a background of the lead image. The feature was announced on Wikimedia blog in early April 2015.

Timeline of earlier discussions[edit]

Legal note[edit]

I have received a few questions about the share-a-fact feature, so I wanted to share more details on how we think about attribution in a feature like this. The share-a-fact feature is designed to invite people to Wikipedia by making it easier for users to share bite-size facts from articles. The goal is to attract more readers editors on Wikipedia by providing an easily digestible way to share knowledge.

All media and text on Wikipedia is released under a free license or in the public domain so that it can be easily shared. The Wikipedia Mobile App's sharing feature lets users select a portion of text within an article to make a card with a reduced-resolution version of an image from the page in the background. This card also includes the CC BY-SA icons to inform others that the source article on Wikipedia is released under the CC BY-SA 3.0 licensed. The sharing feature adds text with a URL to the Wikipedia article that generated the card, so others can navigate back to the Wikipedia article for more information.

Media released under a CC BY-SA license may be shared and remixed under minimal conditions, such as providing credit in "any reasonable manner" (CC BY-SA 3.0, Section 4(c)). This includes providing attribution and other required information in a linked resource, a common and accepted practice under 3.0 that Creative Commons reaffirmed in version 4.0 of the license (see “Common-sense attribution” under What’s New in 4.0 on the CC website). In order to ensure that the requirements of the CC BY-SA licenses were being met by the share-a-fact feature, we spoke with Diane Peters, the General Counsel for Creative Commons. She confirmed that a URL to an article that includes easily accessible links to attribution information satisfies the CC licenses' attribution requirement. Providing attribution by including a URL is consistent with the practice on Wikipedia, as well.

It's important to note that the simple requirements in the CC licenses do not change other important rights available under the law, including fair use, fair dealing, or other rights, or the ability to freely share and remix material not protected under copyright, such as the public domain material or a de minimis use. Using a small version of an image and a short quote from the article would generally be considered fair use in the US, if it does not replace the original work and adds additional value. Using a short quote from an article may not be protectable at all under copyright, as a de minimis use, or it may be allowed under other copyright exceptions. If you are sharing CC licensed media in a country that does not have an exception similar to fair use, then you may need to keep the URL as attribution to follow the license.

The sharing feature encourages users to provide credit by including a URL to the source Wikipedia article that leads you to the author of the text quote and media in the card. When designing the share-a-fact feature, our goal was to automatically include a sufficient amount of information without adding too much text that detracts from the feature's overall usability. Although it's possible for users to delete the text URL on their own, the decision to embed more license details with a watermark is a question of design, and not a legal requirement, as the URL is easily accessible. The feature includes an appropriate amount of attribution information to meet our goal to make it easy for people to enjoy, share, and remix content from Wikipedia, consistent with the goals both free licenses and the Wikimedia mission.

We have received a lot of great feedback about how to improve the feature, such as including more license information in the image metadata (if technically possible). I am following the structured data project on Commons, and I hope that better license-related APIs and metadata will allow us to build new forms of attribution in the future. I also know that Moushira is interested in tracking the feedback and community norms on attribution to ensure that we consistently provide good license information. We will incorporate your feedback into our advice on new features, too. I appreciate everyone's thoughtful comments and suggestions, and I will follow up with some of the more detailed points on the talk page.

Best, Slaporte (WMF) (talk) 22:45, 21 April 2015 (UTC)[reply]

Open questions[edit]

  • Given the different legal interpretation of laws in different countries, how do we, as global community and movement, find a common ground that doesn't just keep the product owner, being WMF or a chapter, or other in legal compliance, but also meets satisfaction of a broader audience of contributors?
  • Given the challenges of accuracy of license data extraction from files, what are our other alternatives, or time plan towards fixing the problem?
  • The presence of a username on media, be it a card or an image thumbnail, eliminates conflict, and meets contributors satisfaction, can we align our design efforts to keep in mind that this information needs to be integrated in a way that doesn't clutter the product? — Preceding unsigned comment added by Melamrawy (WMF) (talk • contribs)

Is it possible to check for an association between people making an article contribution and using Share-A-Fact? In particular, people contributing to an article and then Share-A-Facting that same article within a half hour or so? Alsee (talk) 22:14, 20 April 2015 (UTC)[reply]

  • Copyright issues when sharing a "fact" with images: Attribution is lacking. For details, see previous discussion. Problems are namely:
    • Authors are not credited (neither for text nor for images)
    • The State Court of the German state Bavaria, LG MĂźnchen, decided in 2011 that an attribution by an external link is not a proper attribution and thus violates the CC-by-sa license terms. An author's consent to the indirect attribution by links to an image's description page at Wikipedia would not mean that the author would waive his right to a locally proximate attribution when his work is used at other websites than Wikipedia.
    • The German Higher Court of Cologne additionally decided in 2014 that cropping an image that was released under a CC license is only allowed if the modification is declared.
    • An often wrong license is attributed by the app (the generic CC-by-sa icon shown with each "fact" does not reflect the actual variety of media licenses at Wikimedia Commons)
    • Public domain content is falsely attributed with a CC-by-sa icon.
    • CC-by-sa requires at least an URI, in some language versions specified as the internet address, of the full license legal code; GFDL requires a local copy of the full license text.
    • Unauthorized sub-licensing of Wikipedia contents by "sharing a fact" to platforms like Facebook, Instagram, Twitter. For details, see previous discussion and a previous WMF legal notice here.
  • YouTube video by WMF doesn't credit authors either — Preceding unsigned comment added by Martina Nolte (talk • contribs)

What about images that have been removed from the article? Do you expect users of this tool to search the article history to find the original image or will you provide permenant links?
The structured data project on Commons is stuck due to the closure of the WMF multimedia team, WMDE cannot handle this on its own under the impression of the Freebase and VIAF migrations. Whether or not improvements in this regard will happen soon, is currently quite unlikely. Cheers, DerHexer (talk) 10:55, 22 April 2015 (UTC)[reply]

Good point. See this unaddressed early bugreport:
T75130 Skip images when required attribution is impossible Nemo_bis created this task (Nov 7 2014). bzimport added a project: MediaWiki-extensions-Collection (Nov 22 2014) and
What did legal write initially: "we should still try to display as much licensing information as we practically can, and consider what to do in cases where we can't. (e.g., if we can't parse the license information of an image, should we rely on fair use or should we just not show the image?)" --Atlasowa (talk) 05:24, 5 May 2015 (UTC)[reply]