ubuntu-webapps-bugs team mailing list archive

[Olivier Tilloy <olivier.tilloy@xxxxxxxxxxxxx>]

The default UA has been updated, and it now includes a placeholder token for hardware ID (see http://bazaar.launchpad.net/~phablet-team/webbrowser-app/trunk/view/head:/src/Ubuntu/Components/Extras/Browser/UserAgent02.qml for details). Note that by default this token is empty, I suppose it will customized by OEMs shipping actual devices.
I’m marking the bug fixed, feel free to re-open to continue the discussion or if you feel that this implementation is not enough to address the original issue.

** Changed in: webbrowser-app
       Status: Confirmed => Fix Released

-- 
You received this bug notification because you are a member of Ubuntu
WebApps bug tracking, which is subscribed to webbrowser-app.
https://bugs.launchpad.net/bugs/1252407

Title:
  Cannot identify device via proposed user agent string

Status in Web Browser App:
  Fix Released

Bug description:
  Issue: the proposed user agent string structure provides no support
  for inclusion of device model information.

  While recognising that this may have been debated to some extent as
  part of https://bugs.launchpad.net/webbrowser-app/+bug/1179596, it may
  be worth flagging some impacts of this. The usage of user agent
  parsing solutions is far wider in application than just the serving of
  content. Disclaimer: I work for dotMobi, who provide both server and
  client side device awareness capabilities, among other offerings.

  Content serving
  While RWD is seen as a good approach to serving web content across different device types, most leading organisations use a combination of server side and client side techniques to optimise performance and tailor first delivery according to the device capability (RESS is growing rapidly). Knowing the device capabilities before serving content is a key part of this. (Google serve a 4kB page to a low end phone, and a 600kB page to a high end phone. http://prism.mobiforge.com/?url=google.com. Extreme example, but not isolated by any means) 

  Web analytics
  Businesses are unable to identify which devices are visiting their websites, and accordingly are unable to optimise content for them. Where multiple devices are indistinguishable, the relative market penetration (or usage on the web) of one device model over another is unknown, impacting planning and business decision-making.

  Advertising
  The serving of ads to non-desktop devices is based on server side device identification. Both the targeting of ads, and the placement of correctly sized ads, are driven by server side identification of the device. Many businesses are entirely dependent on advertising for their business model, and mobile operators for example like to be able to provide upgrade offers to subscribers using less capable devices. With the proposed Ubuntu OS UA, this is no longer possible, since all devices appear identical on the server side.

  Emerging device categories
  Automotive manufacturers are embedding browsers into vehicles. A requirement they have is that web servers are aware that the context is specifically automotive, hence mobile/tablet is insufficient. They also want to know the model in their analytics reports. Needless to say, the number of different devices in which browsers are being embedded is growing rapidly. Should a TV be served mobile or tablet content? [Possibly it is not intended that the Ubuntu OS is for use in devices other than mobile phones and tablets]

  Finally, it is worth noting that Mozilla support inclusion of device identifying information in their UAs, although it is not their recommendation.
  https://developer.mozilla.org/en-US/docs/Gecko_user_agent_string_reference#Mobile_and_Tablet_indicators

To manage notifications about this bug go to:
https://bugs.launchpad.net/webbrowser-app/+bug/1252407/+subscriptions