Comparison of spatial data handlers

Boston GIS posted A comparison between these tools for handling spatial data:

  • SQL Server 2008 Spatial,
  • PostgreSQL / PostGIS 1.3-1.4,
  • MySQL 5-6

It is interesting that Manifold is mentioned as a viable alternative ... that is good after more than a year we threw flowers hoping to grow its popularity.

Although Manifold is not doing well with MySQL, and comparative does not consider Oracle, in which Manifold is well seated.

I admit that I have hardly done any translation, and for my interest in Manifold however if they are interested in the post, they can see it Complete in Spanish In its original source, since in the second stage of the post they show a list of the spatial functions of the different platforms.

Feature image
SQL Server 2008 Spatial

MySQL 5-6
PostgreSQL /
PostGIS 1.3-1.4
Operating system Windows XP, Windows Vista, Windows 2003, Windows 2008 Windows XP, Windows Vista, (not checked in 2008), Linux, Unix, Mac Windows 2000 + (including Vista and 2003, not checked in 2008), Linux, Unix, Mac
License Commercial - closed source Commercial open source (COSS), some GPL shares. FLOSS (PostgreSQL is BSD, PostGIS is GPL Open Source - commercial applications can be used but if changes are made that affect PostGIS libraries, you must return them to the community
Free GIS that loads data Shp dataloader for SQL Server 2008 developed by Morten Nielsen (not yet working with RC0) OGR2OGR, script Include shp2pgsql, OGR2OGR, QuantumGIS SPIT, SHP loader for PostGIS also developed by Morten using SharpMap.NET , There are others
Commercial GIS supporting it manifold, Safe FME Objects, ESRI ArcGIS 9.3 (in a last service pack) Safe FME Objects manifold, FME Objects, ESRI ArcGIS 9.3
Availability of drivers specifically for spatial components ? Not yet - SharpMap.NET eventually and probably built on a new ADO.NET 3.5 + GDAL C ++, SharpMap via OGR, AutoCAD FDO SharpMap.Net, JDBC postgis.jar included with postgis, JTS etc. tons for Java, GDAL C ++, AutoCad FDO beta support
Free desktop viewers and editors They will be built on SQL Manager, but not available in RCO and only for viewer GvSig OpenJump, QuantumGIS, GvSig, uDig
Desktop Viewers and Business Editors ESRI ArcGIS 9.3 Server SDE in the latest service pack, manifold, FME FME ESRI ArcGIS 9.3 Server, ZigGIS for desktop, manifold, FME
Tools for Web mapping - something like OpenLayers and other environments that support GML manifold, MapDotNet, ArcGIS 9.3 (in the last service pack), UMN MapServer see, MapGuide Open Source (using beta FDO driver) UMN Mapserver, GeoServer, MapGuide Open Source manifold, MapDotNet, ArcGIS 9.3, Mapserver UMN, GeoServer, FeatureServer, MapGuide Open Source (using beta FDO driver)
Space functions Both OGC SFSQL MM and Geodetic custom (more than 70 functions) OGC MBR (bounding box functions) some functions for spatial relationships, 2D only More than 300 functions and operators, no geodetic support except for poing-2-point in non-indexed distance functions, custom PostGIS for 2D and some 3D, some MM support in circular arrays and composite curves
Spatial Indexing
(According to some reports, Oracle also uses some R-Tree and can use quadtree IBM DB2 ... or something like that
Yes, 4 Multilevel Grids (BOL says its B-Tree based) R-Tree quadratic splitting - indexed only exists for MyISAM GIST - a variant of R-Tree
Real geodetic support, support for measurements along a spheroid.
It should be noted that Oralce has support for this
Yes, with some limitations No No
Shared accommodation Much Much Something, unless you have a dedicated server on Linux / windows you can accomplish many things

2 Replies to "Comparison of spatial data handlers"

  1. Cone mention in the post, it is a pity that the original document does not consider oracle in its comparative

  2. And Oracle?

    I do not understand how Oracle has not appeared in this comparison. It has much more advanced features than any of these three such as geocoding, Topology And raster storage.

    Of course in my opinion, PostGIS is (despite its slow development) the best alternative of geospatial storage, being the use of Oracle (I repeat, in my opinion) only justified in very specific cases because of its high price only suitable for pockets.

    By the way I think that another (new / old) player will come to the forefront of the geospatial databases: INGRES.

    We'll see.

Leave an answer

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.