Geospatial support, SQL Server 08 vs MongoDB?

MongoDB has lots of street cred especially since FourSquare uses it. MS SQL Server 2008 R2 also has Geospatial support.

Which DB is easier/better for doing GPS-like search? e.g. k-nearest points around po开发者_开发百科int X,Y?

If the only Geospatial function you need is to find k-nearest points around point X,Y then any old database will do. Just use Haversine formula, and it's been implemented in a bunch of languages.

Simple Geospatial Queries with MongoDB

http://blog.mxunit.org/2010/11/simple-geospatial-queries-with-mongodb.html

Much easier to use then SQL Server! I like it.

To make a long story short, SQL Server is way more robust than mongo with geospatial support. However, if you are just storing points on a map and want to calculate distances, mongo is more than adequate.

MongoDB supports Geospatial points. SQL Server supports geospatial objects of an arbitrary number of points.

Either solution will be fine for your geospatial needs, so its more about your data model, scalability, and what database your comfortable with.

You might also want to look at PostGre. GeoSpatial data processing is where PostGre excels in. Can't really tell you about Windows performance however