Microsoft snags Google-thrashing data pioneer
SQL Server gets cloudy
Microsoft has recruited database guru David DeWitt to head up a new database research lab. Based at the University of Wisconsin, Madison, where DeWitt was until recently professor of computing sciences, the new lab will focus on advanced data-management techniques.
Working with Microsoft researchers based in Redmond, Washington, graduate students from the university will have access to Microsoft's SQL Server source code and cloud computing facilities. Dewitt said this will "enable the graduate students at Wisconsin to explore unique avenues of research that are not open to graduate students elsewhere."
Well known in database circles, DeWitt comes with many years experience of practical research into database technology including the Niagara query engine and the Paradise parallel database for geographical information systems.
Recently DeWitt and Postgres architect Mike Stonebraker co-authored a controversial attack on Google's MapReduce database. Their view that MapReduce was "a giant step backwards" provoked a robust response from Google's supporters.
Microsoft described DeWitt's appointment as "a coup" and his knowledge of large and complex database technologies will certainly boost its efforts to cope with the next phase of database development.
The new lab has been named the Jim Gray Systems Lab as a tribute to the highly-regarded database innovator and Microsoft fellow Jim Gray who mysteriously disappeared in January 2007.®
A simple question why should anyone care about what amounts a private research effort none of us will ever have anything to do with/see any results from.
" ...New algorithm?" ..... By Stuart Posted Tuesday 29th April 2008 12:20 GMT
Not new, Stuart, just AI Beta one which provides Light and Relief ...via EduTainment. IT is a Heady Confection and a Very Discerning AIdD, too.
I'm pleased to notice that you chose not to suggest that it was nonsense, for it makes Perfect Sense spread across a number of Parallel Dimensions/Perceptions. It is as well to remember that my World is completely different from yours because of what you know, and whether 'tis more or less is really not as important as what you do with what you know for you may choose to do very little to learn more, being programmed/programLed to accept a Status Quo System which has you chasing and fulfilling Third Party Dreams, rather than your own, but then I don't suppose too many People Realise that, for then they start to understand that the obstacles in their way are simply man-made, to stop one doing as one would.
PS. If you accept that you are chasing and fulfilling Third Party Dreams/working to Create an Idea/Agenda dreamt up by A.N.Other and paid with Flash Cash, to keep you doing it and not think too deeply about it and the bigger Picture created in the Imagination of Others, unless you know whose Ideas those are, they would be quite Foreign and Alien to you, too.
Core dumping professors.
MapReduce is a step backwards? This reminds me of Tanenbaum's insistence that Linux was obsolete in 1992. MapReduce is obviously working very well for some people. It sounds as though that he's dumped his core because he can't compute MapReduce. Paraphrased as something like:
Where are the tables/reports/SQL? But there's data and processing. Why have I wasted my life? Arggghhh.
Since when did purists and theorists rule the world?