Sunday, March 02, 2008

Data In, Data Out

An important emerging web development initiative concerns "data portability", or how openly and transparently a website or application consumes and provides data. In a world of mashups, web services, and walled gardens like Facebook or MySpace, it is becoming an increasing concern for a variety of reasons.

To begin with, users want to know that the information they enter into a website is theirs to manage, share, and remove over time.  This matters for convenience (so they don't have to reenter it all the time) but also from a privacy perspective.

A famous recent example saw Robert Scoble temporarily banned from Facebook for breaking its terms of use by running a script to access his social graph (contacts). This focused a great deal of attention and debate on who "owns" this sort of information - the person entering the data, the website that hosts it, or the individuals whose information is being stored.

Developers also care about this as we are all tired of reinventing the  wheel every time we want to share data with another application. A major consequence of the recent wave of web development trends is the increasing importance of application integration and data integration standards. These days, "No Web App Is An Island".

To help address these needs, a community is coalescing under the banner of the DataPortability Group. Their website is located at and discusses the issues in detail. They define portability as

both physically moving data or simply porting the context in which the data is used

Their effort involves identifying and evangelizing existing data portability standards, rather than creating new ones. They also hope to encourage a trust framework that will benefit vendors and consumers.

The rapid adoption of this initiative shows how quickly things move in the IT world, if proof were needed. The project was first founded on October 11, 2007, and is already gaining a great deal of steam. The website has an interesting timeline showing how quickly support is building:

Some recent events of note included Google, Plaxo, LinkedIn, Twitter, Flickr, Six Apart and Facebook joining the workgroup; Google, Yahoo, IBM, Microsoft, and VeriSign joining the OpenID Foundation board; and MySpace launching its Open Developer Platform.

Obviously most of the activity is driven by the Social websites as they have the most to win / lose. Issues of personal privacy and trust are crucial to their continued popularity (the Facebook Beacon PR disaster is a prime example of this). 

Nonetheless, I imagine within a few years every major Software vendor will have the DataPortability-compliant tickbox in their sales material, or their shareholders will demand to know why not. Case in point: Blogger, owned by Google, now supports OpenIDs on its blogs (such as this one), and Yahoo users can use their Yahoo! accounts as OpenIDs.

Interestingly enough, some of the standards the DataPortability group advocates are the ones that might lead to the long-envisioned semantic web - namely microformats, which can potentially add machine-readable "context" to data.

It's an interesting space to explore, and I plan to blog about it as I learn more.

No comments:

Post a comment

Note: only a member of this blog may post a comment.