By Dennis D. McDonald posting from Alexandria, Virginia
I’ve made some attempt to keep up with public discussions of DataPortability.org. I’ve had a suspicion that the project is experiencing the growing pains that technology industry standards groups sometimes experience when there is no single strong and deep-pocketed voice willing to weigh in, knock heads, and force progress along a single path. The groups involved are diverse, the problems being addressed have a great deal of variety, and progress appears to be more evolutionary and consensual than more narrowly defined problem areas.
Evolution does take time, as we know, especially when some participants come from a more collaborative open source background. That’s why I was interested to read this recently published post, What DataPortability Is/Isn’t. According to the list of bullet points discussed here, one thing DataPortability.org isn’t is Currently end-user focused.
What, you might ask? Isn’t the whole purpose of DataPortability.org to make it easier for users to control and manage their own data (whatever that means)?
Well, yes, but consider the technical nature of the problems being addressed. Possibly the best way to get a handle on what DataPortability.org is trying to accomplish comes from a reading of the current use cases (and background documents) it lists on its web page (this is a link to a Google Groups page). This is a beginning of a list of things the Dataportability specification will be designed to support. Just a cursory review clearly shows that working through the details, especially in an open and collaborative setting, is going to expose a lot of gnarly problems that will need a lot of discussion to resolve.
But why the reference to “isn’t currently end user focused”? I think the answer to that question gets at the core of what the group is trying to accomplish. That requires an understanding that standards aren’t the same as applications. Often the people who develop applications are in business and must balance customer requirements (and desires) against financial and commercial realities.
Which is not to say that efforts such as those sponsored by DataPortability.org are anti-user; far from it. But it does suggest that we need to be patient with a group whose efforts have to serve the interests of a very wide range of applications ranging from purely social to purely business. Is that really possible? We’ll see. I’m optimistic but I think it’s going to take a while.
Similar issues have arisen over at Google’s OpenSocial initiative (Google is now participating in DataPortability.org efforts). ReadWriteWeb’s Comment of the Day: OpenSocial is a “Social Cloud” For Developers, Not Users has an extended comment from theharmonyguy that I think encapsulates the nature and complexity of the challenge. Here’s an excerpt:
Currently each social networking site implements its own variation of managing a social graph - in other words, you have an application with social functions built on top. But with technologies like OpenID and OAuth, we may reach a point where social networking sites are built on top a distributed social graph. Essentially, the social aspects of things like friend connections become invisible to the user and simply another layer that people consider part of the Internet. Then we’ll have a true social cloud for users.
If you re-read the DataPortability.org’s current use case list and think about it, you’ll probably realize that we have a long, long road before we reach the conditions that theharmonyguy describes. Even if we solve the purely technical problems we’ll still need to address the various business, privacy, and ownership issues that DataPortability encompasses, and that’s where the real challenges exist.