Introduction

The Geodetic DataBlade module and the Spatial DataBlade module are server extensions that manage spatial data using different GIS technologies. This technical note describes the differences between these DataBlade modules and how to migrate data between them.

Back to top

Overview

The Geodetic DataBlade module and the Spatial DataBlade module are server extensions that manage GIS data in an Informix database. They use different core technologies that solve different problems and complement each other:

• The Geodetic DataBlade module treats the earth as a globe. Geometric operations are precise regardless of location.

  The Geodetic DataBlade module is best for global datasets and applications, such as satellite imagery repositories.

• The Spatial DataBlade module treats the earth as a flat map. Projecting the curved earth onto a flat map creates distortion around the edges of the map, so geometric operations can be inaccurate around those edges.

  The Spatial DataBlade module is best used for regional datasets and applications.

This section explains the benefits of managing this data in an object-relational database (ORDBMS) and introduces the core technology for both DataBlade modules.

Why use an ORDBMS for spatial data?

First of all, what exactly is a DataBlade module? And why on Earth (cough cough) would you want to manage spatial data in the database?

A DataBlade module is a server extension. Typically it extends the server to manage a new data domain, such as geospatial data. Informix's object-relational database technology lets developers extend the server with new:

• User-defined types (UDTs).
• User-defined routines (UDRs), which are functions and procedures implemented in C, Java, or SPL (Informix Stored Procedure Language).
• Indexing methods.
• Other access methods; for example, to access to external data stores.

DataBlade modules take advantage of the following features of an object-relational database management system (ORDBMS).

Uses standard SQL

DataBlade functions are called in SQL statements, a language with which customers are already familiar. Customers can start taking advantage of new features immediately; they don't have to learn a new language.

Simplifies client applications

Since DataBlade functions are callable from SQL, it simplifies client application programming.

Furthermore, client applications don't have to be relinked with new versions of the source code. After a new version of the DataBlade is installed, SQL queries automatically start using the new version.

Finally, if the client runs on one machine architecture and the database on another, the Informix server automatically performs any required conversion for transporting data between architectures. (OK, so DataBlade modules have to add some generic support code for the server to call, but even the DataBlade module code does not need to know the machine architectures.)

Manages complex data

The complexity of processing new data is pushed into the server, again simplifying application code.

Ensures Data Integrity

Since a DataBlade module runs as part of the server, it enjoys all the database services that ensure data integrity, such as:

• Concurrency control and transaction management.
• Backup and recovery.
• Any triggers that the customer may have created to enforce local rules.

Improves Performance

The database scales for multiple users far better than client applications do.

Side-by-side tests using the Spatial DataBlade module demonstrate that DataBlade module performance is measurably better than the exact same code running in a client application as you add more concurrent users.

Some of the features that help the database scale so well include:

• IDS buffers data in memory.

  Internally, the Informix server buffers database pages fetched from disk in memory. This means that if many users query overlapping data, the database can fetch results from the internal buffers and avoid expensive disk i/o. Database administrators can tune the size of those buffers (and can also tune many other related parameters, such as how many database pages should automatically be read ahead).

• Query optimizer reduces internal processing.

  Since DataBlade modules are tightly integrated into the server, the query optimizer can:

  • Determine if any indices are available to filter results.

  • Delay execution of expensive functions so that they execute on the fewest intermediate results possible.

• Just results are delivered.

  Just query results get delivered to the client, reducing network traffic.

• Random access to large data also reduces results delivered.

  An IDS 7.x blob ("dumb blob") must be completely read into memory before the data can be manipulated. Only a client application can manipulate a dumb blob; an SPL routine cannot.

  An IDS 9.x smart large object ("smart blob") provides random access to the data. The entire object does not have to be read; instead, the code can directly jump ("seek") to the interesting data, then clip out just the required data. A client application can manipulate smart blobs, but more importantly, a UDR written in C or Java can randomly access smart blobs. (An SPL routine can manipulate a UDT that contains a smart blob by invoking UDRs for that UDT.)

Geodetic DataBlade

The following description is adapted from the Informix Geodetic DataBlade Module User's Guide, chapter 1, "Concepts and Principles".

The Informix Geodetic DataBlade module is developed in partnership with Geodyssey, Ltd., of Calgary, Alberta, Canada.

The Geodetic DataBlade module takes its name from the discipline known as geodesy. Geodesy is the study of the size and shape of the Earth (or any body modeled by an ellipsoid, such as the Sun or the celestial sphere). The Geodetic DataBlade module is designed to handle objects defined on the Earth's surface with a high degree of precision.

To do this, it uses a latitude and longitude coordinate system on an ellipsoidal Earth model, or geodetic datum, rather than a planar, x- and y-coordinate system. This avoids distortions, inaccuracies, and imprecision that can be introduced using flat-Earth projection.

A geodetic datum is a reference system that describes the surface of the Earth, and it is based on an approximation of the general shape of the Earth by an ellipsoid of rotation (also called a spheroid). This is the three-dimensional shape described by an ellipse when it is rotated around one of its axes. The Geodetic DataBlade module defaults to the World Geodetic System 1984 (WGS-84) datum, but it supports over 175 geodetic datums and over 40 ellipsoid specifications, and also allows defining new datums.

The Geodetic DataBlade module's coordinate reference system uses geodetic latitude and longitude to describe locations relative to the Earth:

• The geodetic latitude of point P in the diagram above is the angle between the datum surface normal at the point and the equatorial plane.

• Longitude is the angle in the equatorial plane between the line a that connects the Earth's center with the Prime Meridian and the line b that connects the center with the meridian on which the point lies. A meridian is a direct path on the surface of the datum that is the shortest distance between the poles.

Geodetic latitude and longitude are always based on a specific datum.

Spatial 8.1 DataBlade

The Informix Spatial 8.1 DataBlade module is developed in partnership with ESRI. It is the successor to the ESRI SDE 3.0.2.2 DataBlade module, which is currently generally available. As of this writing, the Spatial DataBlade module is in beta test.

The Spatial DataBlade module uses planimetric (flat-plane) geometry, which means that it approximates the round surface of the earth by projecting the surface onto flat planes using various transformations.

To apply simple, flat-plane coordinates to the earth, first you apply a transformation to map the round earth's surface to a flat piece of paper. This is called a projection. You can't wrap paper around a ball without wrinkling or tearing it, so it always introduces distortions of distances, angles, and areas. To keep the distortion within bounds, the area covered by the map must be limited. World maps cannot be used for reliable angle, direction, distance, or area measurements.

The rectilinear coordinates (x and y) in a projected map are usually called Easting and Northing. Often, the theoretical origin (0,0) is arbitrarily put well out of the way of the map in question so that all the coordinates in the map are positive.

Contents

Back to top

DataBlade Features

User-Defined Types

Each DataBlade creates new user-defined types (UDT) that are organized into a type hierarchy, with many subtypes under a single supertype. The types themselves are slightly different for each DataBlade module, although there is some overlap (point, line, polygon):

Geodetic Type Hierarchy

Spatial Type Hierarchy

Each subtype inherits the functionality of the supertype.

Furthermore, the user can create a column that is of the supertype, such as GeoObject or ST_Geometry, then mix subtypes in that one column. Or the user can create a column of the subtype, thereby restricting the column to values of that one type.

External Formats

Both DataBlade modules support external, public formats.

Spatial

• ESRI Shape

  The ESRI shape format is public. The Spatial DataBlade module comes with utilities that load and unload shapefiles. Additionally, source code is freely available for accessing and manipulating shape files.

• OpenGIS

  The OpenGIS Consortium specified a data transfer standard for geospatial data. The Spatial DataBlade module provides functions that convert between a UDT value and the OpenGIS well-known text (WKT) and well-known binary (WKB) formats. The specific functions are listed below in the section on "Converting Spatial to/from OpenGIS formats".

Geodetic

• OpenGIS

  The section below on "Converting Geodetic to/from OpenGIS formats" describes how Geodetic support the OpenGIS formats.

• See the online release notes for additional supported formats.

Contents

Back to top

DataBlade Functions

This section lists Geodetic and Spatial functions side-by-side.

Functions that deal with Z values

Both the Spatial and Geodetic DataBlade modules allow you to store a double- precision value with each vertex of an object. This is intended for storing the altitude or depth of a vertex, but it can be used for many other purposes. Z values are not considered when testing the spatial relationship between two objects.

Functions that deal with Measures (M values)

The Spatial DataBlade module allows you to store an additional double- precision value with each vertex of an object. This is intended for storing the distance of a vertex from the beginning of a linestring, but it can be used for many other purposes. M values are not considered when testing the spatial relationship between two objects.

Functions that deal with AltRanges and TimeRanges

The Geodetic DataBlade module allows you to associate an altitude range (an ordered pair of double precision values) and a time range (an ordered pair of datetimes) with an object.

Contents

Back to top

Migrating Data between Geodetic and Spatial

The OpenGIS WKB and WKT formats let you migrate data between Geodetic and Spatial. Both DataBlade modules provide functionality for converting UDT values between the internal representation and WKB / WKT.

Converting Spatial to/from OpenGIS Formats

The Spatial DataBlade module provides the functions summarized below for converting UDT values between OpenGIS formats and the internal DataBlade format:

Converting Geodetic to/from OpenGIS Formats

Geodetic does not use functions to convert data to a specific format.

Instead, the GeoParam metadata table manages the data format for transmitting data between client and server. If the "data format" parameter is set to "OGC", then binary i/o is in WKB format and text i/o is in WKT format. (For specific details, see Chapter 7 in the Informix Geodetic DataBlade Module User's Guide).

Caveats

There isn't always a direct mapping between Geodetic and OpenGIS data types.

The table below (from the Geodetic on-line release notes) summarizes the mappings. The mapping arrow indicates the supported conversions between OpenGIS and Geodetic types. Bidirectional arrows signify a direct mapping. Unidirectional arrows signify a conversion which is only possible in one direction. For example, GeoCircles are converted to Polygons on output, but there is no conversion from an OpenGIS polygon to a GeoCircle.

Geodetic type	Mapping	OpenGIS type
GeoPoint		Point
GeoLineseg		LineString
GeoString		LineString
GeoRing		LineString
GeoBox		Polygon
GeoCircle		Polygon
GeoEllipse		Polygon
GeoPolygon		Polygon or GeometryCollection of Polygons
not supported		MultiPoint
not supported		MultiLineString
not supported		MultiPolygon
not supported		GeometryCollection of Points
not supported		GeometryCollection of MultiPoints
not supported		GeometryCollection of LineStrings
not supported		GeometryCollection of MultiLineStrings
not supported		GeometryCollection of GeometryCollections

Contents

Back to top

Glossary

Terms and acronyms used by this tech note include:

DataBlade Module

Server extension

IDS

Informix Dynamic Server

ORDBMS

Object-relational database management system

UDR

User-Defined Routine

UDT

User-Defined Type

WKB

Well-Known Binary, OpenGIS data transfer format

WKT

Well-Known Text, OpenGIS data transfer format

Contents

Resources

• Geodetic DataBlade Module: For more information on this product, please visit the Geodetic DataBlade Module product information page.
  
  
• Spatial DataBlade Module: For more information on this product, please visit the Spatial DataBlade Module product information page.
  
  
• Information about the OpenGIS consortium is available at http://www.opengis.org.
  
  

About the authors

Rate this article

Comments

Back to top

Table of contents

• Introduction
• Overview
• DataBlade Features
• DataBlade Functions
• Migrating Data between Geodetic and Spatial
• Glossary
• Resources
• About the authors
• Comments

Next steps from IBM

• Try: Informix Ultimate Edition
• Buy: Informix Ultimate Edition

Dig deeper into Information Management on developerWorks

Try IBM PureSystems. No charge.

Experience today!

---

Get started with the cloud-based trial and pattern development kit.

Special offers