A recent Network World article reports on Eric Schmidt, CEO of Google, touting his company's efforts to ensure that data stored on their system is portable. Schmidt defines portability as a user being able to take their data and move it to another service should they become unhappy with their current provider, including his own company.
I'm wondering if Google happened upon this conclusion after making data portable within their own suite of hosted applications. And if data that can move around seamlessly in their own application suites just happens to be portable to other services then the inverse might be true as well. Data on other services with portability to Google might then be useable amongst the range of Google's hosted application suites. A powerful way to lure potential users.
The next step in data portability is data virtualization. And by that I mean data being equipped with the smarts to present itself differently to different applications depending on how it is accessed. This makes the accessing applications run more efficiently. I often use the example of insurance data. The ways in which, and the applications used to access data by an actuary and a claims adjuster are quite different, even though it may be the same quanta of data.
There are some pretty smart folks that work for Google, that's a given. I'm sure that under this new data portability paradigm they are working on some pretty cool data virtualization methodologies as well.
And BTW, I promise never to use the words "paradigm" and "methodologies" in the same sentence ever again.