Casting light on Google's GData

In April, Google first publicly discussed Google Data API, or GData, which set off a round of discussion among an elite group of geek bloggers.

Now O'Reilly blogger Nathan Torkington has sparked some more discussion, following a recounting of his meeting with two of Google's well known engineers, Chris DiBona and Mark Lukovsky.

The idea behind GData is to provide a protocol for syndicating content, much the way ATOM and RSS do. One significant difference, though, is that GData allows people to query these Web feeds as well, according to Google's description.

Although it's very much aimed at hard-core programmers right now (Google has released a beta toolkit), end users could ultimately see benefits.

"Google provides APIs to let client software request information outside of a browser context...GData provides a general model for feeds, queries, and results. You can use it to send queries and updates to any service that has a GData interface," according to the GData page.

In his talk with DiBona and Lukovsky, Torkington uncovered another significant benefit: it makes it easy for Google to publish APIs to its Web sites.

"The big thing about GData for Google is that it's extremely simple to build into the server-side, so they can offer APIs very easily. This is important as they offer APIs for lots and lots of new stuff," Torkington wrote.

He went on to make comparisons to Microsoft's aborted Hailstorm initiative, which Lukovsky headed when he was still at Microsoft.

Google is already using the GData API for its Google Calendar. And the overall GData concept has been received positively by Jeremy Zawodny at Yahoo and Dare Obasanjo at Microsoft.

But the role of GData outside of Google is still unclear. Is this simply a nifty way for Google to drive more traffic to its sites or a technical approach with a more far-reaching impact?