LAS VEGAS--While Internet Explorer 8 demos, Silverlight progress and a Monkey Boy reprise from Steve Ballmer captured much of the attention at Mix 08--it was a database announcement that could be the sleeper announcement of the show.
What Microsoft announced was a database-in-the-cloud service where Web developers can store their data. Those attending the Mix show here were able to sign up for a beta test that is set to begin in three or four weeks, with a final version aimed to be launched by the end of the year, according to Dave Campbell, a technical fellow in Microsoft's SQL Server division.
"We're taking SQL Server we're pulling pieces of it apart and we're putting it back together," Campbell said in an interview. One of the challenges is taking software designed to run highly reliable servers and storage and turning it into a commodity service.
But, Campbell said, that's also where the opportunity lies. "In this world, dumb and fast rules."
"If you give up a tiny bit in terms of the degree of consistency in the architecture, you can get tremendous resiliency and scale, but you want to retain enterprise-class quality around data service," he explained. Factoring data into classes of data or tiers is one way to improve the class of service using commodity gear, he said.
What Microsoft is doing in this case with its database is also the kind of thing it makes sense to imagine the company doing with a variety of "building block" services.
"SQL Data Services is a building block for Microsoft's longer term vision of a services fabric for developing and deploying applications," Campbell said. "Imagine at some point a version of Visual Studio with a services palette in the toolbox and wiring up and composing services."
Illuminata analyst Gordon Haff said that it's important for Microsoft to be a player in this area.
"A lot of computing and storage is going to be moving online," he said. "For them not to get into this game would be suicidal at some level."
Campbell downplayed the notion that Microsoft was competing with Amazon.com's S3 service.
Haff said that there are some differences, since Amazon targets largely unstructured data and Microsoft is aimed more at structured and semi-structured data, but said it's not like the two don't overlap.
"They are targeting a somewhat different market, but everyone competes at some level," he said.
Campbell wouldn't get into how Microsoft would price the service, but said it some type of usage model that takes into account both the amount of data stored in the cloud as well as the bandwidth used in transferring information.
CNET News.com's Dan Farber contributed to this report.