Data Portability, Cloud Computing and PaaS

In a post from earlier today Read/Write Web pointed to this piece by Google alum Bret Taylor titled: “We need a Wikipedia for data”. His premise is that an open, public storehouse for “useful” datasets would unleash entrepreneurial creativity globally, leading to the creation of both social good and business value. I was beating a related drum back in January with a series of posts on the Data Portability Workgroup, albeit with a focus on user-specific data. Long story short, I think Bret’s absolutely right.

To take this idea one step further, Google’s recent announcement of their cloud computing initiative reminded me of an idea I heard recently from a Seattle-area entrepreneur: integrating useful datasets directly into a Platform-as-a-Service stack (i.e., data + development environment + runtime enviroment). This might be one of those ideas that sounds better in theory than it would be in practice, but I liked the simplicity of a vertical service stack in the cloud that enabled users to quickly build and ship apps that leveraged a growing pool of public data.

As Bret said in his post, this falls into the category of “things I want but want someone else to build”. It’s a hugely ambitious vision even for a company like Google to pursue, but it struck me as a plausible next step in the evolution of cloud computing.