User-agent adventures: flipboard

Flipboardis one of the most talked about apps of the last fewmonths. It presents itself as a social magazine, where users are to access portions of websites without having to access them on the browser, and instead get all the topics of a number of websites on one single place. Let’s call it a RSS reader on steroids.

Also, there is an alternative from Google called Google Currents, which does mostly the same. Both exist for both Andoid and IOS, but Flipboard is the incumbent as is was launched earlier.

From the websites’ point of view, both apps are a new way of users getting access to content, if things are done right…. And sometime they aren’t.

Flipboard started to offer full websites pages, stripped of formatting and advertising, which enraged the sites’ owners and managers. Fortunately, this was solved, and nowadays, Flipboard only makes available a snippet of the website, allowing the user to access the full site. Google on the other hand already knew the pains of doing this, and never had these issues.

What Flipboard still hasn’t learned, is how to handle web masters and site admins.

First, Flipboard tries to aggregate content through Facebook pages. No problem here. However, after getting to a site, Flipboard tries to index the whole site with a bot. Now this bot has some really weird user agent:


Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2) Gecko/20100115 Firefox/3.6 (FlipboardProxy/1.1; +

Now, let’s analyze this. For one, the source IPs Flipboard uses are from Amazon EC infrastructure, so it’s a little weird that is presents itself as an OSX hosts, which probably isn’t. Also, there isn’t any connection between user access to this content and this accesses (more on this later), which makes this a bot or a crawler instead of a proper proxy, and should be identified as such. So, also this obviously isn’t a Firefox access, this tends to screw web site statistics. Also, Flipboard should include user statistics information while crawling, just like Google does:

Agent: Feedfetcher-Google; (+; 3 subscribers; feed-id=5117030660366433681)


Second, when users actually access the full website content, at least on IOS and probably on Android as well, uses an in-app browser window, which is fine but itself but it needs to be properly implemented, which isn’t:

  • The referer field needs to be properly used. Google does it on Google+, even on the mobile version, and Facebook does it correctly as well, so why can’t Flipboars do it? Web masters need to know how a user access it’s content and where it comes from.
  • The user-agent isn’t properly filled as well, and on IOS used the default UIBrowser window user agent, which again doesn’t allow for proper identification. This is not easy but it can be done, even on IOS5.
In sum, if Flipboard is to fight Google, it needs to help itself by sharing it’s data with web masters and behaving properly, because Google already does it, the last thing Flipboard needs is to have sites blocking it.




Leave a Reply

Back to Top