Refractions Research is pleased to announce the 0.5 release of PostGIS :
Geographic Objects for the PostgreSQL object-relational database server.
PostGIS/PostgreSQL is a spatial database much like ESRI's SDE and
PostGIS/PostgreSQL includes the following functionality:
- Simple Features as defined by the OpenGIS Consortium (OGC)
- Support for Well-Known Text and Well-Known Binary representations of
- Fast spatial indexing using GiST
- Geospatial analysis functions
- PostgreSQL JDBC extension objects corresponding to the geometries
- Support for OGC access functions as defined by the Simple Features
The 0.5 release of PostGIS includes support for the MapServer internet
web mapping server from the University of Minnesota
(http://mapserver.gis.umn.edu). Download the latest CVS version of
MapServer to use PostGIS as a internet web mapping backend.
- New functions
- X(), Y(), Z()
- AsBinary() + XDR and NDR variants
- New Objects
- SPHEROID(<NAME>,<SEMIMAJOR AXIS>,<INVERSE FLATTENING>)
To be used with the length_spheroid functions for accurate
length calculations on lat/lon data. (Thanks to Geographic
Data BC and David Skea)
- Minor bug fixes
- Internal Functions
- Extra constructors to make geometry manipulation easier
- Structural Reorganization
- Broke postgis.c up into four new files
postgis_debug.c -- debugging functions
postgis_fn.c -- generic functions (like length())
postgis_ops.c -- operators and indexing functions
postgis_inout.c -- type support functions and data conversion
[postgis] Re: [mapserver-users] PostGIS 0.5 Released
Deni Suwardhi wrote:
> Have you develop it with topological structure?
Depends on what you mean by "topological" :) It more topologically rich
than the default spatial objects in PostgreSQL: polygons have holes,
aggregate geometries are supported, geometry collections are supports.
And of course there are functions like area and length and so on which
recognize that they are GIS objects that can support GIS-style
calculations. The OGC Simple Feature Spec is our roadmap for now.
As for more complex things, like coverages and other winged-edge
topologies, no we have not done anything with that kind of thing yet.
Which is not to say, however, that one could not whip up a coverage or
network model just using RDBMS capabilities and properly indexing the
data. What needs development, I think, are the kinds of functions to
automate the kinds of queries one usually makes on a coverage or network
My take, anyways. I'd be interested to know what others think...