Simon Says – Musings of a Technology CTO

This week I was introduced to the web site age-appropriate rating system Age-Label, sponsored by OMK in Germany. Proposed as a standard for self-regulation of web sites, it allows owners to insert a small xml file “age-de.xml” in the root of their websites which defines the appropriate age ratings of the site, or subsections of such. I dug deep into the system and did some trawling across the internet to find out how used it is.

You can read an English translation of the standard online.

It would seem like a good idea – instead of relying on a third party to analyse the content of your site and make a determination on what age groups it’s appropriate for, web site owners can define it for themselves. The XML file also allows you to specify different sections of your website for different age readers.

Of course, this requires some appropriate technology on the readers device to look for, interpret, and act on the age-de.xml file – but if you imagine a world where the majority of sites are (honestly) tagged, and browsers use the xml data, and parents set the browsers with the appropriate age information, we could indeed go a long way towards protecting minors from inappropriate content.

Let’s take a look at one of these files – this one can be found at disney.de/age-de.xml

<age-declaration>
  <ageblock-basic>
    <age-issuer>www.fsm.de</age-issuer> 
    <last-change>2014-03-17</last-change>
    <country>de</country>
    <label-version>1.0</label-version>
    <revisit-after>7days</revisit-after>
  </ageblock-basic>
  <ageblock-labeltype>
    <xmlfile>true</xmlfile>
    <httpheader>false</httpheader>
    <htmlmeta>true</htmlmeta>
    <label-z>false</label-z>
    <default-age>12</default-age>
    <alternate age="0">http://www.disney.de/disney-junior/index.jsp</alternate>
    <alternate age="6">http://www.disney.de/disney-spiele/</alternate>
    <alternate age="12">http://disney.de/</alternate>
    <alternate age="16">http://disney.de/</alternate>
  </ageblock-labeltype>
  <ageblock-labeltype-definition>
    <labeltype-htmlmeta-definition>
    <label class="default">
      <min-age>0</min-age>
      <default-age>12</default-age>
    </label>
    </labeltype-htmlmeta-definition>
  </ageblock-labeltype-definition>
</age-declaration>

You can see it’s pretty “human readable” – four sections indicate three possible disney pages to go to depending on the age of the viewer – 0-5 years goes to one, 6-11 to another, and 12+ to the main site.

To work of course a number of facts have to be true :

1. The site owner has to implement the age-de.xml file
2. The site owner has to implement the age-de.xml file truthfully
3. The visitor needs their browser to act on the age-de.xml file
4. The visitor’s browser needs a truthful age-de.xml rating

Taking them one by one to understand the impact of this rating system, I wrote a simple script to trawl the Alexa Global top websites looking for the presence of the age-de.xml file.

The results were a little disappointing.  Of the top 500 most visited global sites, only 6 supported Age-Label

go.com, twitch.tv, steamcommunity.com, t-online.de, bild.de, gameforge.com

Given two of these were German specific .de domains, and that this initiative is sponsored by a German organization, I felt perhaps the Global500 was too general a list of sites to check, so I re-ran my test against the top 500 websites visited by people in Germany (again, using Alexa data).

The results were a little more encouraging – 38 of the 500 sites most visited by people in Germany supported age-label.

t-online.de, bild.de, twitch.tv, nowtv.de, computerbild.de, sport1.de, freenet.de, giga.de, vodafone.de, arcor.de, sky.de, joyclub.de, steamcommunity.com, quoka.de, moviepilot.de, maxdome.de, markt.de, gamestar.de, computerbase.de, songtexte.com, n24.de, bigpoint.com, prosieben.de, myvideo.de, gofeminin.de, golem.de, gameforge.com, netzwelt.de, spin.de, spieletipps.de, ran.de, clipfish.de, sat1.de, pcgameshardware.de, buecher.de, battlefield.com, kino.de, mydirtyhobby.com

Remember, we can see how these sites rate themselves simply by looking for age-de.xml in their roots – so gameforge.com/age-de.xml returns a file restricting a number of games to age 12 and above, and sky.de/age-de.xml restricts a number of TV show microsites to specific age groups.

The authors of the age-de.xml standard propose that this system could be extended to go global, with each country maintaining the appropriate age.xx-xml file, and further an age.xml file used to tie everything together.

Unfortunately, scanning the Alexa Global 500 for age.xml files does not yield any results – it seems no popular site site are using it for certainty.

Next, let’s look at the technology needed on the visitors device

Firstly of course, no mainstream browser supports Age-de.xml files natively – you need plugins or third party tools.  There’s a list at http://www.age-label.de/tools where you can find a number of companies who collectivly offer filtering systems for PC, Mac, iOS, Android, and even a dedicated wireless router with built in technoilogies.

Readers outside of Germany will find unfortunately that these offerings are targeted at German speakers only.

Though self-certification seems like a good idea, this evidence indicates that third party web site analysis from companies who mine the web and categorize sites is still the only realistic way of applying content restrictions to protect minors.

Without forcing every website owner to tag their content, and having penalties for not complying (or tagging incorrectly) – how can initiatives like Age-Label be effective?

My suggestion would be for the major search providers to prioritize tagged sites?