Difference between revisions of "invisible-data-considered-harmful"

From Microformats Wiki
Jump to navigation Jump to search
(→‎invisible metadata failures: adding Wikipedia persondata)
(→‎invisible metadata failures: PERSONDATA has been deprecated! Hallelujah!)
Line 31: Line 31:
* Wikipedia's [https://en.wikipedia.org/wiki/Wikipedia:Persondata persondata] template
* Wikipedia's [https://en.wikipedia.org/wiki/Wikipedia:Persondata persondata] template
** Contains invisible data. Not necessarily updated when someone dies or when errors are found.
** Contains invisible data. Not necessarily updated when someone dies or when errors are found.
** [https://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28proposals%29/Archive_122#RfC:_Should_Persondata_template_be_deprecated_and_methodically_removed_from_articles.3F Deprecated on 2015-05-26].
== related ==
== related ==

Revision as of 09:45, 16 June 2015

<entry-title>invisible data considered harmful</entry-title>

This article is a stub.

Invisible data or metadata for that matter is undesirable. The general problem with invisible data is that being invisible, nobody sees when it's wrong. If something is wrong on a web page and it's visible, a visitor can see it, a page author (without technical abilities) can see it. Invisible data doesn't get seen and errors inside it are not found as easily as visible data is.

See Principles of visibility and human friendliness.

Also known as, "dark data", "meta tags", "side files".

invisible metadata failures

  • meta keywords
    • popular in the 1990s, eventually spammed, and rotten, ignored by Google, Yahoo, and other search engines.
  • meta ICBM
    • people move, don't bother to update their meta ICBM
    • people get it wrong (because it's not obviously visible)
      • swapping lat and long
      • "correcting" negative values to positive (or vice versa)
      • note clusters of sites in the middle of oceans or other open spaces that correlate with inverse (or negated) coordinates of actual cities
  • lang="en"
    • lots of templates, CMS's shipped with this default
    • people used them worldwide
    • now lang="en" is effectively meaningless/untrustworthy since tons of non-en sites all have it (from abovementioned templates etc.)
  • unintentional privacy violations through EXIF data
    • See privacypatterns.org: Strip invisible metadata
    • A number of people have unintentionally released their geolocation through its inclusion in EXIF data attached to photographs uploaded to social networking sites like Twitter. Twitter now strip EXIF data and Flickr allow users to remove EXIF data if needed.