Microsoft's new "Dallas" data catalog for Windows Azure, announced this morning, aims to make big sets of public and commercial data available to developers using the company's cloud computing system. Ray Ozzie, the company's chief software architect, called it a "game-changing" subsystem of Azure during his keynote address this morning at Microsoft's Professional Developers Conference in Los Angeles.

Amitabh Srivastava, a Microsoft senior vice president and one of the fathers of Windows Azure, said in a phone interview afterward that one key to the service is the ability to make a wide variety of public and commercial data available in a uniform way, so that it can be used more easily by applications built on Azure.

Although much of the data is available in a variety of places now, it comes in a variety of forms and formats, making integration tough.

"There's so much data from NASA on space, geological data from GSI, and other things like that," Srivastava said. "If we make that available, along with almost an infinite amount of computing, and all of these interesting data sets are available in a uniform way, you don't even have to have any computers, you can buy a thousand nodes (in Azure) and then you go and analyze this data, you can publish it, you can sell your app. The possibilities just become immense."

The company issued a technical preview of Dallas this morning, with initial sets of data from the Associated Press, Citysearch, DATA.gov, ESRI, First American Corp., infoUSA.com, NASA, National Geographic, NAVTEQ, RiskMetrics Group, the United Nations, WaveMarket and Weather Central Inc.

"By delivering data as a service, our aspirations are that Dallas might catalyze a whole new wave of remixing and experimentation by developers -- an opportunity for innovation that's uniquely unlocked by the cloud," Ozzie said at Microsoft's developer conference and via webcast.

The company used the term "marketplace" for the service, reflecting the fact that some providers will ultimately charge for the use of their data in the applications.