Database Reference
In-Depth Information
object. So, a GIS would have a layer of line objects that might represent roads; another
of polygon objects that could represent forestry, lakes, and urban areas; and perhaps
another polygon or raster layer representing different soil types. Analysis could be
performed using the GIS to answer questions such as, “Which areas of forestry con-
tain a particular soil type and are within 10 miles of a lake or river?”
All of these early GIS were also entirely proprietary in terms of their implementa-
tion. Each had its own data formats and structures, each being mutually incompatible
with all the others. This was a time when it was also not unknown for organizations
to develop their own entirely specialist GIS for their own particular purpose, so a
local authority might develop a GIS to manage a land terrier, a utility company might
have developed its own solution for asset management, and so on. The 1980s and
indeed the early 1990s were not noted for interoperability.
By the early 1990s and with the introduction of the IBM PC (personal computer),
the GIS was becoming more affordable, although it still tended to reside largely
within GIS departments. During this period, the functionality rose, and the GIS
transitioned from workstations to PCs. The market penetration of GIS also widened
as vendors developed more lightweight GIS. ESRI, with its industrial-strength GIS
ARC/INFO, launched a product called ArcView, designed for the PC market, which
was compatible with ARC/INFO but had comparatively limited functionality and
was primarily aimed at visualizing GI. The early to mid-1990s also saw rising inter-
est in GIS from the main database vendors such as ORACLE, IBM, Informix, and
Ingress. They all developed spatial indexes and extensions to the SQL (Structure
Query Language) to enable spatial queries and simple GIS functions to be per-
formed. The establishment of GI capability within large-scale database systems not
only increased the volume of data that could be held but also meant that GI started
to be managed in the same way as any other corporate data.
The early 1990s also saw the development of new GIS based around the
object-oriented (OO) design and programming paradigm, two such examples being
the Smallworld GIS and GOTHIC from Laserscan. These attempted to break free
from the idea of GIS layers where the main driver was the type of geometry. This
meant that they tried to more closely model the nature of a real-world object. In these
models, topology was as important as geometry. However, such systems have as yet
failed to gain widespread appeal, in part because of the investment companies have
made in the more traditional GIS models. OO GIS have been sidelined into niche
markets; for example, Smallworld specializes in networks, largely for utility compa-
nies, and Gothic serves digital map production systems.
As the capabilities of GIS advanced, so did take-up increase. Today, GIS is firmly
established in specialist GIS sections of central and local government and in many
commercial companies. It is used in a wide range of applications, from the tradi-
tional uses such as preparing maps, to a wide range of applications covering every-
thing from mineral resource discovery to assessing insurance premiums and risk.
Beyond GIS, the use of GI also has grown in industry and government as a means
to integrate different data using addresses (most usually postal addresses). The reasons
for such integration are many and varied and include integrating data following com-
pany or government departmental mergers; fraud detection and credit checking; mar-
keting and general development of a demographic picture; public health awareness;
