James Senior at Microsoft: Like a kid in a candy store

"We all do agree the world is not flat, right?"

"We all do agree the world is not flat, right?"

  • Comments 1
  • Likes

At the recent BI conference that my colleague Ben has been shouting about recently, one of the Keynotes featured SQL Server 2008 and Virtual Earth.  Why these two products? Well, because SQL Server 2008 will be featuring new types of data - in particular geographic data.


Ted Kummert, VP of Data Platforms at Microsoft gave the keynote and he showed a nice Virtual Earth app the pulled on data from SQL Server 2008 to provide a visualisation layer of sensus information in a particular area.  On the data side, what's interesting is that they are going to map two different models of the earth.  The first being geometric (i.e. flat) and second being geographic (i.e. spherical - like the earth).  Each method has it's own advantages and limitations and Ted explains more in this interview with Directions Magazine:

The "Flat Earth" (planar) data type represents geospatial data which has been projected from its native, spherical, coordinate system into a plane (we all do agree that the Earth is not flat, right?). As such, the geometry model necessary to support planar data is much different than its counterpart for supporting the "Round Earth" geodetic model. The Flat Earth data type (GEOMETRY) will feature a traditional planar model with all of the advantages and limitations. The Round Earth data type (GEOGRAPHY) will operate on an ellipsoidal model in which the Earth is defined as a single continuous entity. The geodetic GEOGRAPHY type will not suffer from the singularities (international dateline, North and South Poles, map projection zone "edges," etc.) which causes data handling problems for the planar GEOMETRY type. Without map projections to worry about, the typical user can concentrate on the spatial problem at hand. With the growing use of GPS-like devices (making room for the expected follow-on technologies to GPS…) and the growing availability of data in geodetic coordinates (i.e. WGS84-based latitude and longitude coordinates), we felt that our geospatial offering should recognize the inherent difference between the two Earth models and not try to co-mingle them.

See also discussion on this topic from Isaac Kunen of the SQL Server Spatial engineering team at his blog.

This is pretty interesting stuff.

Technorati Tags: ,
  • Joy of Tech has some insight into why St. Steve sucked at his recently-concluded WWDC..... Original link

Your comment has been posted.   Close
Thank you, your comment requires moderation so it may take a while to appear.   Close
Leave a Comment