`
`Exhibit 2005
`
`
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`VOLUME 1 : PRINCIPLES
`
`GEOGRAPHICAL
`INFORMATION SYSTEMS
`PRINCIPLES AND APPLICATIONS
`
`EDITED BY
`DAVID J MAGUIRE,
`MICHAEL F GOODCHILD
`
`AND
`DAVID W RHIND
`
`~•~ Longman
`::: Scientiiic &
`_... Technical
`
`Copublished in the United States and Canada with
`John Wiley & Sons, Inc., New York
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`Longman Scientific and Technical,
`Longman Group UK Ltd
`Longman House, Burnt Mill, Harlow,
`Essex CM20 2JE, England
`and Associated Companies throughout the world.
`
`copublished in the United States and Canada with
`John Wiley & Sons, Inc., 605 Third Avenue, New York,
`NY 10158
`
`©Longman Group UK Limited 1991
`
`All rights reserved; no part of this publication may be
`reproduced, stored in a retrieval system, or transmitted in
`any form or by any means, electronic, mechanical,
`photocopying, recording, or otherwise without either the
`prior written permission of the Publishers or a licence
`permitting restricted copying in the United Kingdom
`issued by the Copyright Licensing Agency Ltd, 90
`Tottenham Court Road, London W1P 9HE.
`
`Trademarks
`Throughout this book trademarked names are used.
`Rather than put a trademark symbol in every occurrence
`of a trademarked name, we state that we are using the
`names only in an editorial fashion and to the benefit of the
`trademark owner with no intention of infringement of the
`trademark.
`
`First published 1991
`
`British Library Cataloguing in Publication Data
`Maguire, David 1.
`Geographical information systems: Principles and
`applications
`I. Title
`II. Goodchild, Michael F.
`III. Rhind, David W.
`910.901
`
`ISBN 0-582-05661-6
`
`Library of Congress Cataloging-in-Publication Data
`Maguire, D. 1. (David 1.)
`Geographical information ssyyyystems I by D. 1.
`Maguire,
`Michael F. Goodchild, and David W. Rhind.
`p.
`em.
`Includes bibliographical references and index.
`Contents: v. 1. Principles- v. 2. Applications.
`ISBN 0-470-21789-8
`1. Geographical information systems.
`I. Goodchild, Michael F.
`II. Rhind.
`III. Title.
`David.
`G70.2.M354 1991
`910'.285-dc20
`
`91-3724
`CIP
`
`Set in Great Britain
`
`'
`
`[!JJ
`lfU\rif.~ffi
`\ -- [~]
`l.lBRA~-{
`
`·f..
`Q
`
`Printed and Bound in Great Britain at the Bath Press, A von
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`VOLUME. 1
`
`: PRINCIPLES
`
`Preface
`List of contributors
`Acknowledgements
`
`Section I Overview
`
`Introduction
`D J Maguire, M F Goodchild and D W Rhind
`
`~ 1. An overview and definition of GIS
`DJ Maguire
`
`2. The history of GIS
`J T Coppock and D W Rhind
`
`3. The technological setting of GIS
`M F Goodchild
`
`4. The commercial setting of GIS
`J Dangermond
`
`5. The government setting of GIS in the United Kingdom
`R Chorley and R Buxton
`
`6. The academic setting of GIS
`DJ Unwin
`
`7. The organizational home for GIS in the scientific
`professional community
`J L Morrison
`
`)-
`
`8. A critique of GIS
`R T Aangeenbrug
`
`Section II Principles
`
`'x. Introduction
`M F Goodchild, D W Rhind and D J Maguire
`
`xiii
`xvii
`xxvii
`
`3-7
`
`9-20
`
`21-43
`
`45-54
`
`55-65
`
`67-79
`
`81-90
`
`91-100
`
`101-7
`
`111-17
`
`vii
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`. VO.LUME 1
`
`: PRINCIPLES
`
`.
`
`(a) Nature of spatial data
`
`)I
`
`9. Concepts of space and geographical data
`A C Gatrell
`
`10. Coordinate systems and map projections for GIS
`DH Mating
`hl. Language issues for GIS
`A U Frank and D M Mark
`'X 12. The error component in spatial data
`N R Chrisman
`
`~ 13. Spatial data sources and data problems
`P F Fisher·
`
`14. GIS and remote sensing
`F W Davis and D S Simonett
`
`(b) Digital representation
`
`;{-15. Computer systems and low-level data structures for GIS
`Wm R Franklin
`
`%'16. High-level spatial data structures for GIS
`M J Egenhofer and J R Herring
`
`17. GIS data capture hardware and software
`M J Jackson and P A Woodsford
`
`119-34
`
`135-46
`
`147-63
`
`165-74
`
`175-89
`
`191-213
`
`215-25
`
`227-37
`
`239-49
`
`viii
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`VOLUME 1
`
`: PRINCIPLE.S
`
`18. Database management systems
`R G Healey
`19. Digital terrain modelling
`R Weibel and M Heller
`·20. Three-dimensional GIS
`J F Raper and B Kelk
`
`251-67
`
`269-97
`
`299-317
`
`319-35
`
`337-60
`
`(c) Functional issues
`21. The functionality of GIS
`D J Maguire and J Dangermond
`22. Information integration and GIS
`I D H Shepherd
`23. Cartographic modelling
`CD Tomlin
`24. Spatial data integration
`R Flowerdew
`.X 25. Developing appropriate spatial analysis methods for GIS 389-402
`S Openshaw
`26. Spatial decision support systems
`PJ Densham
`27. Knowledge-based approaches in GIS
`T R Smith and Je Yiang
`
`361-74
`
`375-87
`
`403-12
`
`413-25
`
`ix
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`..
`
`VOLUME 1
`
`: PRINCIPLES
`
`(d) Display issues
`28. Visualization
`B P Buttenfield and W A M ackaness
`29. Computer name placement
`H Freeman
`30. Generalization of spatial databases
`J-C Muller
`
`(e) Operational issues
`31. GIS specification, evaluation and implementation
`A L Clarke
`.)1 .32. Legal aspects of GIS
`E F Epstein
`33. Managing an operational GIS: the UK National On-Line
`Manpower Information System (NOMIS)
`M J Blakemore
`-'* 34. Spatial data exchange and standardization
`S C Guptill
`
`Consolidated bibliography
`List of acronyms
`Author index
`Subject index
`
`427-43
`
`445-56
`
`457-75
`
`477-88
`
`489-502
`
`503-13
`
`515-30
`
`531-591
`593-598
`599-613
`615-649
`
`X
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`THE HISTORY OF GIS
`
`J T COPPOCK AND D W RHIND
`
`Computer-based GIS have been used since at least the late 1960s: their manual·
`predecessors were in use perhaps 100 years earlier. Acknowledging the paucity of
`well-documented evidence, this chapter describes tlte background to the development
`of such systems, stressing the context in which such development took place, the role
`of organizations and individuals where this can be ascertained, and the applications
`which the systems were intended to meet. A broad definition is taken of GIS so as
`not to exclude any significant developments; computer mapping systems of all types
`(including those with line-printer graphics, the forerunners of contemporary raster
`systems) are included.
`It is demonstrated that most, but by no means all, of the early developments
`originated in North America. The roles of key organizations such as the US Bureau
`of the Census, the US Geological Survey, the Harvard Laboratory for Computer
`Graphics and the Experimental Cartography Unit are described and the activities of
`the commercial sector are exemplified by a case study of Environmental Systems
`Research Institute. Reasons are suggested for significant international differences in
`the development of GIS, such as the attitudes to ownership of data and the perceived
`role of the state. It is concluded that several stages of evolution of GIS can be
`defined. These overlap in time and occur at different moments in different parts of
`the world. The first, or pioneering age, extended from the early 1960s to about 1975;
`in this, individual personalities were of critical importance in determining what was
`achieved. The second phase, approximately from 1973 until the early 1980s, saw a
`regularization of experiment and practice within and fostered by national agencies;
`local experiment and action continued untrammelled and duplication of effort was
`common. The third phase, running from about 1982 until the late 1980s, was that of
`commercial dominance. The fourth (and current) phase is one of user dominance,
`facilitated by competition among vendors, embryonic standardization on open
`systems and increasing agreement on the user's perception of what a GIS should do
`and look like.
`
`INTRODUCTION
`
`A variety of information indicates that the field of
`GIS has expanded rapidly in recent years (see
`Maguire 1991 in this volume). From where did all
`this business and the resulting jobs arise?
`Unhappily, we scarcely know. GIS is a field in
`which history is little more than anecdotal. To
`rectify this, a search through the archives of
`
`government departments and agencies would
`certainly help. As yet, however, few organizations
`have given any thought to formalizing the history of
`their involvement in GIS and at least one major
`player (Ordnance Survey; see Finch 1987) has
`refused to let its detailed records be examined by
`external researchers. Less certainly, the records of
`computer hardware and software companies could
`also be a source of relevant information but no such
`
`21
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`J T Coppock and D W Rhind
`
`material has been uncovered. Unfortunately for
`those writing the history of GIS, neither staff of
`commercial companies nor government officials
`have a tradition of writing books or papers on their
`experience of an emerging technology. Research
`staff in government or private sector research
`organizations are exceptions to this rule but, even
`for them, writing papers for the benefit of the
`scientific community at large has a relatively low
`priority. As far as is known, the only official attempt
`anywhere to provide a broad overview of the field as
`a whole is that given by the Report of the
`Committee of Inquiry into the Handling of
`Geographic Information (Department of the
`Environment 1987; Rhind and Mounsey 1989).
`The main source of information, with all the
`risks of partisan bias, remains researchers in the
`academic community. In reality, however, even the
`numbers of academics working in this field were
`quite small until the expansion of the last decade.
`Moreover, as Chrisman (1988) and Rhind (1988)
`both testify, those active in universities in this field
`in the early stages of the development of GIS were
`often outside the formal academic career structure
`and were so heavily involved in project work that
`they had little time or inclination to write papers. In
`any case, at the beginning there were no obvious
`outlets for publication in a topic that was seen as
`marginal to a large number of interests; Rhind's
`(1976) report, for instance, may well be the first
`example of a record of GIS conference papers
`which were described as such in a mainstream
`academic publication. While the advent of specialist
`GIS conferences (often disguised by use of other
`titles such as AUTOCARTO) provided one
`publishing mechanism from 1974 onwards, the early
`conference proceedings were intermittent and were
`not easily accessible to those who had not attended
`the gatherings. We do not believe this postulated
`paucity of recorded history represents
`incompetence on our part: a correspondence
`prompted by the editor of Photogrammetric
`Engineering and Remote Sensing, for example
`(Marble 1989; Tomlinson 1989), generated great
`controversy and revealed a lack of documentation
`on the first use of GIS in the refereed literature.
`Finally and most crucially, the content of any
`history of GIS depends in large measure on the
`definition adopted. A strict definition, as a
`computer-based system for analysing spatially
`referenced data, would greatly restrict the field
`
`22
`
`because, with the major exception of the Canada
`Geographic Information System (Tomlinson 1967),
`this was not a common feature until the 1980s. A
`more general interpretation, as any system for
`handling geographical data, would greatly widen the
`field and hence enlarge the number of contributors.
`Such a definition would embrace, not only the
`whole field of automation in cartography (which
`was often the precursor to any involvement in GIS
`and provided, in terms of computer-generated
`graphics, the most common form of output for most
`early systems), but also many general-purpose
`statistical and database packages capable of
`handlingx,y,z point data. Formal definitions of GIS
`are not, therefore, of much help and relatively little
`reliance is placed on them in this book as a whole.
`In any event, the field evolved not from some ex
`cathedra definition of the subject but through sets of
`interactions. The main backgrounds of those
`involved have been cartography, computer science,
`geography, surveying, remote sensing, commercial
`data processing, mathematics and statistics. The
`purposes to which the systems have been put
`include environmental protection, urban and
`regional planning, land management, property
`ownership and taxation, resource management, the
`management of utilities, site location, military
`intelligence and tactics, and many others- as later
`chapters in this volume testify. The field has
`developed, then, from a melting pot of concepts,
`ideas, practice, terminology and prejudice brought
`together by people from many different
`backgrounds, interacting with each other often on a
`chance and bilateral basis in the early days and
`normally proceeding in blissful ignorance of what
`was going on elsewhere. The essence of GIS is thus
`its multidisciplinary character, with some at least of
`those involved in developing this technology having
`little previous involvement, or even interest, in the
`handling of geographical data as such (see Maguire,
`1991 in this volume for further discussion of the
`definition of GIS).
`This review of the history of GIS is inevitably a
`consequence of the authors' accidental exposure to
`early developments and their own set of value(cid:173)
`judgements; different views certainly exist, such as
`that manifested in Cooke's portrayal of the
`genealogical structure of geoprocessing systems in
`general (Fig. 2.1). In particular, it is suspected that
`the role of those who did not contribute to the
`formal literature has been underplayed, especially
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`those working in the military. While regrettable,
`this is probably unavoidable: history very often
`consists solely of what has been written down.
`
`THE GRASS ROOTS EVOLUTION OF GIS
`
`What seems clear is that there were many
`initiatives, usually occurring independently and
`often in ignorance of each other, concerned with
`different facets of the field and frequently
`originating in the interests, often disparate, of
`particular individuals. Like the reality (as opposed
`to the reporting) of scientific research, there was no
`strictly logical progression towards the development
`and implementation of GIS, but rather a mixture of
`failures, set-backs, diversions and successes.
`Inevitably, more is known about the successes than
`about the failures which, according to both
`Dangermond and Smith (1988) and Tomlinson
`(1988), have been numerous and often attributable
`to bad advice, ignorance and a determination to go
`it alone. This is unfortunate because failures are
`often as illuminating as successes, if not more so
`(Giles 1987). What also seems clear is that
`particular individuals and institutions played key
`roles, acting as examples or as sources of expertise,
`advice and often skilled personnel; since these
`contributions are now better recorded than is the
`generality of progress, this account will tend to
`emphasize them, particularly those of Howard
`Fisher in the Harvard Laboratory for Computer
`Graphics (LCG), Roger Tomlinson in the Canada
`Geographic Information System (CGIS) and Jack
`Dangermond in the Environmental Systems
`Research Institute (ESRI) in North America, and
`David P. Bickmore at the Experimental
`Cartography Unit (ECU) in the United Kingdom.
`Many others played significant parts (e.g. Tobler
`1959; Nordbeck 1962; Cook 1966; Hagerstrand
`1967; Diello, Kirk and Callander 1969 and Boyle
`(see Rhind 1988)), but these four have been the
`subject of particular articles in a special and
`invaluable issue of The American Cartographer
`(Tomlinson and Petchenik 1988). Fortunately, these
`individuals seem to typify the interests, attitudes
`and commitments of those working in the vintage
`era of GIS from the late 1950s to the end of the
`1970s.
`The motivations for developing GIS or
`
`The History of GIS
`
`components of such systems have varied very
`widely. They have ranged from academic curiosity
`or challenge when faced with the possibility of using
`new sources of data or techniques, through the
`desire for greater speed or efficiency in the conduct
`of operations on spatially referenced data, to the
`realization that desirable tasks could be undertaken
`in no other way. The last was undoubtedly a
`powerful motive in two key developments which are
`discussed in more detail below- the Oxford System
`of Automated Cartography and t~e Canada
`Geographic Information System. It was the
`experience of publishing the Atlas of Great Britain
`and Northern Ireland (Bickmore and Shaw 1963)
`and the criticisms this attracted of being out of date
`and unwieldy that convinced D.P. Bickmore,
`probably in 1958 but certainly no later than 1960,
`that only the computer could provide a cost(cid:173)
`effective mechanism to check, edit and classify data,
`to model situations and to facilitate experiments in
`graphic display (Rhind 1988). Similarly, it was the
`impossibility of analysing maps of East Africa at an
`acceptable cost that first led R. Tomlinson (1988) to
`think of a digital approach. A calculation made in
`1965 indicated the need for some $Can 8 million in
`1965 prices and a requirement for 556 technicians
`for three years in order to overlay the 1 : 50 000 scale
`maps of the Canada Land Inventory; this
`unacceptable level of resources acted as an
`incentive to develop a more automated approach.
`It was, of course, the advent of the digital
`computer and the order-of-magnitude decrease in
`computing costs every six years over a 30-year
`period (Simonett 1988) that made such alternative
`digitally based approaches viable. It is interesting to
`note, however, that not all early work used the
`digital computer. Thus perhaps the earliest attempt
`to automate map production, the preparation of the
`Atlas of the British Flora, employed a modified
`punch card tabulator to produce maps on pre(cid:173)
`printed paper from cards on which had been
`punched the grid references of recorded
`occurrences (Perring and Walters 1962). Although
`this approach was not repeated and Perring (1964)
`later recognized that the analysis of voluminous
`data could more easily be undertaken by computer,
`it anticipated the widespread mapping in the late
`1960s by line printer. It is also interesting to note
`that Perring was a botanist, with no training in
`cartography, who was faced with the task of
`providing 2000 maps from data that had been
`
`23
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`J T Coppock and D W Rhind
`
`Schweitzer
`
`Fig. 2.1 An individual perception of the genealogy of geoprocessing in the United States (Pers. Comm.
`Don Cooke, 1990). Circles are 'places', i.e. companies, government agencies, universities, etc.; rectangles
`are ideas or concepts, often embodied in a software package or database; directed lines show direct or
`indirect migration or influence in a number of different ways. Examples of flows or lack of expected ones
`include:
`
`24
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`The History of GIS
`
`?
`
`© Donald Cooke 1988
`
`• Harvard Labs influence on GIS vendors (Morehouse to ESRI, Sinton to Intergraph; Odyssey to
`Synercom)
`• DIME was independent from the SACS (Small Area Census Studies)
`• the diagram suggests that the USGS and the US Postal Service had very little influence on most
`developments.
`
`25
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`J T Coppock and D W Rhind
`
`recorded on punch cards. His initiative also
`illustrates an aspect to be repeated in many later
`projects where the application of technology was
`driven by an urgent need of the users, that such a
`task would have to take advantage of the best
`available technology- whatever its limitations(cid:173)
`rather than await the ideal solution; it was also
`similar to many later applications in that it was a
`'one-off' development which, having served its
`purpose, was not taken any further. Slightly later
`work (around 1967) by Bertin in Paris involved the
`modification of IBM 'golfball' typewriters driven
`directly by punch card readers to produce
`proportional symbol maps.
`It is also clear that it was in North America that
`most of the significant early developments in, and
`applications of, GIS and related technology were
`made. By the early 1980s, Tomlinson (1985)
`estimated that there were probably more than 1000
`systems in North America, a figure that must have
`represented a very high proportion of the systems
`then existing in the world as a whole. The bulk of
`this account will accordingly focus on North
`America, with later references to the United
`Kingdom and other European countries and to
`developments elsewhere in the developed world. It
`is only in the late 1980s that any significant
`developments have occurred in developing
`countries and then often through the aid and
`encouragement of developed countries (see Taylor
`1991 in this volume).
`
`THE NORTH AMERICAN SCENE
`
`Aangeenbrug (pers. comm. 1990) has argued that
`the earliest antecedents of GIS in the United States
`can be traced back to the University of Washington.
`In the 1950s, both geographers (notably Garrison)
`and transportation engineers (notably Horwood)
`developed quantitative methods in transportation
`studies. Garrison's colleagues and students included
`Berry, Tobler and Marble; Horwood's included
`Barb and Dueker (see Dueker's important 1974
`paper). Much of the original leadership of the
`Urban and Regional Information Systems
`Association (founded in 1963) and that of other key
`bodies was derived from or directly influenced by
`this group.
`By the early 1960s, at least in North America,
`
`26
`
`large mainframe computers were becoming widely
`available. In 1964, IBM introduced its 360/65
`computer, with a processing speed 400 times faster
`and a memory 32 times as great as its predecessor,
`the IBM 1401 (Tomlinson 1985). These machines
`were employed primarily for one of two very
`different purposes: for routine administrative and
`data management tasks in business and government
`(such as pay-roll, stock control and record keeping
`of various kinds) and for scientific applications
`inyolving extensive computations, notably in
`chemistry, mathematics and physics. There was
`inevitably a good deal of discussion in government
`departments and agencies about the possibility of
`applying computer technology to handle numerical
`data, especially where these were already in
`machine-readable form, as with many censuses,
`where punch-card technology was widely used. In
`1965 the US Bureau of the Budget compiled an
`inventory of automatic data processing in the
`Federal Government, in which it noted the
`significant use of computers to handle land use and
`land title data (Cook and Kennedy 1966). The
`following year, a conference on a comprehensive
`unified land system at the University of Cincinnati
`was advised that a system must be designed such
`that it obtained the maximum benefit from
`electronic data processing equipment (Cook 1966).
`The conference also heard that the District of
`Columbia already had a property data bank, which
`could be searched, updated and retrieved, and that
`Nassau County in New York would be the first to
`provide fully-automated access to records of land
`ownership.
`The significance of the developments at the US
`Bureau of the Census, stemming directly from its
`need for automated address matching, is difficult to
`overemphasize. This need arose from the
`predominantly mail out/mail back nature of the US
`census and the requirement to produce area based
`tabulations from records whose only geographical
`reference was the postal address. An early advisory
`committee on small area data included Garrison
`(see above), who urged a development project to
`test automated data linkage procedures. A director
`hired to run the test, Caby Smith, recruited a team
`which included Corbett, Cooke, Maxfield, White,
`Farnsworth, Jaro, Broome and others who appear
`elsewhere in these pages. The first demonstrations
`of address matching, computer mapping and smal~
`area data analysis were provided through the 1967
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`New Haven Census Use Study (USBC 1969-73).
`Subsequent studies elsewhere in the United States,
`the launch of the DIME workshops in 1970 and the
`development and widespread distribution of
`ADMATCH (address matching software) all had
`major impacts upon government and academia in
`the United States. Indeed, the Census Use Study
`also sponsored the First International DIME
`Colloquium in 1972, leading to the creation of the
`Segment (later re-named as the Spatially)
`Orientated Referencing Systems Association (or
`SORSA), an organization which still holds
`international conferences.
`Increasing availability of computers in
`universities was undoubtedly instrumental in the
`development of the quantitative revolution in
`academic geography in the early 1960s (James and
`Martin 1978; Hudson 1979), particularly in the field
`of spatial analysis (a term which was in general use
`by the late 1960s- see Berry and Marble 1968), with
`its emphasis on the statistical treatment of
`geographical data and on modelling. However,
`these applications, despite their potential relevance
`to handling geographical data, had little interaction
`with computer mapping, primarily because the
`statistical methodology was largely aspatial. One
`exception is a paper in an edited collection on
`computers in geography which related modelling to
`a crude cartography using the line printer (Rushton
`1969). It is only in the middle and late 1980s that
`successful attempts have been made to develop
`closely coupled spatial statistics and 'geographical'
`displays.
`Computers in the 1960s had, in general, no
`explicitly graphical facilities, usually operated in
`batch mode and were very expensive by today's
`standards. Despite this, Tobler (1959) had early
`recognized their potential for automating
`cartography, as had Nordbeck (1962) in Sweden.
`There were, indeed, developments in automating
`cartography in several national agencies concerned
`with mapping and in military establishments which
`could afford equipment that was prohibitively
`expensive to others. The US National Ocean Survey
`was creating charts on a Gerber plotter for the
`production of 'figure fields' or matrices of depth
`values and such organizations as the Aeronautical
`Charting and Information Center at StLouis, the
`Rome Air Development Center and the Central
`Intelligence Agency were active in aspects of this
`field (Diello, Kirk and Callender 1968; Tomlinson
`
`The History of GIS
`
`1972). By the end of the 1960s, map production
`assisted by computer appears to have become
`widespread; for example, the Canadian
`Hydrographic Survey had automated display
`facilities in operation and Surveys and Mapping had
`embarked on a programme to apply automated
`cartography to the 1 : 50 000 series in Canada. In the
`main, however, the aim in computer applications in
`national mapping agencies was to mimic manual
`methods of production and so to produce maps that
`were virtually indistinguishable from their manual
`counterparts. Little information appears to be
`available on the extent to which these methods were
`cost effective, although Tomlinson (1985) suggests
`that the high cost of hardware placed them at a
`disadvantage in competition with manual systems:
`continuing evaluations of costs by the Ordnance
`Survey in Britain, for example, did not find
`automated approaches to map production as a
`whole to be cost effective until the 1980s. Unlike the
`situation in Britain, where a digitizing production
`line was in operation from 1973, the Topographic
`Division of the United States Geological Survey did
`not implement plans to automate the production of
`topographic maps until the start of the 1980s- a
`severe handicap to the development of many
`geographically-based information systems in the
`United States.
`· An entirely different approach to the
`automation of cartography was adopted elsewhere,
`notably in the universities, using the standard line
`printer as a mapping device. In cartographic terms,
`the results were crude, but this was not the point;
`the aim was to produce maps quickly and cheaply so
`as to display the characteristics of the data
`(especially statistical data for census tracts and the
`like) and to undertake simple analyses of such data
`by relating different parameters. It was here that
`Howard Fisher made a significant contribution and
`this approach found ready applications in landscape
`design, in urban and regional planning and, to a
`lesser extent, in resource management.
`
`The Harvard Laboratory for Computer Graphics
`
`Fisher was not a cartographer but trained and
`practised as an architect. He had begun work on a
`computer mapping system in 1963 while at the
`North Western Technical Institute (Schmidt and
`
`27
`
`Bradium Technologies LLC
`
`Exhibit 2005
`
`
`
`J T Coppock and D W Rhind
`
`Zafft 1975). On his retirement, he succeeded in
`obtaining a grant from the Ford Foundation to
`develop this work and, after making unsuccessful
`approaches to Chicago and Northwestern
`Universities (both strongholds of non-spatial
`computer applications to the analysis of
`geographical data), established the Laboratory for
`Computer Graphics (a title subsequently
`lengthened by the addition of 'and Spatial
`Analysis') in 1965 in the Graduate School of Design
`at Harvard University- from which he himself had
`graduated. There he built up a team of
`programmers and others to create a mapping
`package (SYMAP) which used the line printer as a
`mapping device and was capable of producing
`isoline, choropleth and proximal (Thiessen polygon
`or Dirichlet tessellation) maps. The package was
`easy to use by the standards of the day, particularly
`in relation to data for census tracts, incorporated
`default options when nothing was specified by users
`and was widely distributed. In addition to many
`pirated copies, over 500 institutions acquired
`SYMAP (Schmidt and Zafft 1975; Chrisman 1988);
`half of these were in universities, with the
`remainder equally divided between government
`agencies and private institutions. Copies were
`acquired not only in North America but also in
`Europe and elsewhere and the manual was
`translated into several languages, including
`Japanese. A subsequent program, CALFORM,
`which produced higher quality choropleth maps by
`pen plotter and reflected the increasing (if still
`sparse) availability of these plotters, seems to have
`had less success although it too was a pioneering
`effort. SYMAP was important as the first widely
`distributed computer package for handling
`geographical data. It introduced large numbers of
`users to the possibilities of computer mapping; it
`was the precursor, and possibly the progenitor, of a
`large number of other programs using the line
`printer; and it found a wide range of applications
`particularly through the connection between the
`Harvard Laboratory and landscape architects in the
`Graduate School of Design, notably C. Steinitz and
`his associates- one of whom, D. Sinton, produced a
`cell-based program (GRID) which permitted
`multiple overlays of data. Somewhat surprisingly,
`the appointment of a theoretical geographer, W.
`Warntz, to succeed Fisher as Professor of
`Theoretical Geography and Planning and head of
`the Laboratory in 1969, had little effect on the work
`
`28
`
`and apparently stimulated little interaction between
`quantitative geography and computer mapping.
`The Laboratory generated a wide range of
`contracts which, after the expiry of its grant from
`the Ford Foundation, became the main source of
`finance, along with income generated by the sale of
`mapping packages. It never developed a teaching
`programme (which might have prolonged its life)
`and thus only directly added a few new
`professionals to the field, although it did organize a
`highly significant symposium on topological data
`structures in 1977 and hosted influential Harvard
`Computer Graphics Weeks between 1978 and 1981.
`It also attracted at various times talented individuals
`who contributed in many ways to the development
`of computer mapping and, by extension, to
`geographical information systems. Among these are
`N. Chrisman, J. Dangermond, G. Dutton, S.
`Morehouse, T. Peucker and D. Sinton, several of
`whom contributed to the design and construction of
`ODYSSEY, arguably the prototype of
`contemporary vector GIS (Chrisman 1988).
`Unhappily, the subsequent history of this system
`was characterized