Mr President, Ladies and Gentlemen
I am very pleased to have this opportunity to talk to you about Professor Bob Hopgood. I first met Professor Hopgood around 1968 when I was a young research student, and I visited the Atlas Computer Laboratory to see some of the pioneering work he and his team were doing in the field of computer animation. A couple of years later I was to meet him again, when he chaired a session at a conference where I presented my first paper. I can still remember how nervous I felt on that occasion. I feel similarly nervous today, but for a different reason - I have to talk to you about his work rather than my own! I am going to concentrate on some parts of his earlier career because I suspect that this is unknown to most of you, whereas some of you at least will be familiar with his more recent work.
Nowadays, computer graphics is so much a part of all computer systems that we tend to take it for granted. For example, at one extreme we find it used with simple accounting and business presentation packages on PC's, and at the other for the visualisation of complex scientific phenomena using powerful supercomputers. It has taken a huge investment of original thought, research and development to achieve this. We can say that computer graphics has matured as a discipline, and as we shall see, Bob Hopgood was one of the pioneers whose work has had a fundamental influence on this process. But although he is known particularly for his work in this field, he has also gained international recognition for important contributions to the wider fields of computing and computer science as well.
Following a couple of years in the RAF (the Royal Air Force - I thought I should explain that, as I am told that RAF means something different here in Germany!), Bob read mathematics at Cambridge University where he was a student at Downing College between 1956 and 1959. One of my colleagues in Manchester was a Cambridge contemporary and remembers Bob as a very bright student, one who stood out. Of course, in those days there was no such subject as computer science and all of the pioneers in computing started in other fields - engineering, physics, and mathematics, for example.
Upon leaving University, his first job was in the Applied Mathematics Group of the Atomic Energy Research Establishment at Harwell near Oxford. The laboratory was working on early developments in the application of computers for simulation.
One of his tasks at Harwell was to devise and implement programs for Quantum Chemistry calculations on early IBM computers. At that time there were almost no libraries of software for numerical analysis - we take these for granted today - so part of his job was to help write these. His work included the development of new, efficient and accurate methods for calculating multi-dimensional integrals using Bessel functions. Such calculations lie at the heart of many problems in science and engineering and so efficient solutions are rather important. Quite recently Bob showed me one of his notebooks from that period, describing the mathematics underlying his work. In this book the intricate calculations have been worked out long-hand, running over many pages, and it must have required enormous patience to get the right answers. It is a testament to the soundness and efficiency of his solutions that some of the library routines which he implemented then are still in use today, thirty years later. I think that is quite an achievement.
The period from 1959 to 1963 also marks Bob's first steps into the world of computer graphics. In fact, the name computer graphics doesn't seem to have been coined until 1964 and so he is one of a very select group working in that field before the rest of the world knew it existed! Bob wrote graphics software for visualising output from the simulations on a Benson-Lehner pen plotter. As a sideline he designed and implemented one of the earliest character font generators to allow the computer to plot text - principally so that he could append his name to the bottom of his output to prevent it going astray!
In 1963, he moved for a short period to the Atomic Weapons Research establishment at Aldermaston, which had taken delivery of an IBM STRETCH computer, and he became involved in an area which was to bring him widespread international recognition for the first time: compiler and translator writing.
The STRETCH computer had been delivered with an inadequate FORTRAN compiler, and the decision was taken to design and implement a better one internally. The compiler, known as S2, featured full block optimisation - an innovation at that time - and was developed by a team of five people in just four months, between 1st January and 30th April 1963 - surely a record in its day, and perhaps even to this day, for developing software of this complexity. (For comparison, I am told that another team took over two years to develop a similar compiler.)
Hopgood was a member of this team, and his personal contribution was the syntax analyser, which is central to the functioning of any compiler. His design was based on the use of production systems - a way of describing the allowable syntax of a language using a structured set of rules. This was a completely new idea in compilers and was the first time this approach had been used.
In October 1963, Hopgood moved to the fledgling Atlas Computer Laboratory whose remit was to provide a computing service for scientists and engineers based on a Ferranti Atlas, a machine designed at the University of Manchester.
The Atlas machine arrived at the Lab without any high level language compilers. Researchers at Manchester were developing an Algol compiler and Bob designed and implemented an input/output system for it, without which the compiler would have been practically useless! He also developed an improved version of the Manchester Brooker-Morris compiler-compiler - a novel tool for writing compilers. This was then used to develop of a wide variety of translators and compilers including languages such as LISP, SNOBOL and SIMULA. A far-sighted decision was to use the compiler-compiler to implement a number of preprocessors for the Atlas Algol system which allowed it to compile Algol programs from any of the other major computers in the UK. This novel step provided a unique level of program portability across machines. Program portability has remained a key issue for software developers to this day and was to become an important theme in Hopgood's later work on computer graphics standards.
As a consequence of this work Bob had become a leading authority on compiler design and implementation. In 1964, the UK Government's Department of Trade organised a conference on compiler writing techniques. Professor Hopgood was one of the experts who attended. The intention of the organisers had been that the results should be written up and published. In the event, none of the other experts bothered to write their contributions, and so over the next few years Bob wrote a complete book on the topic, published in 1969. Called simply 'Compiling Techniques' the book became internationally accepted as an authoritative guide and was translated into German, French, Polish, Czechoslovakian and Japanese. He doesn't know this, but it was one of the first books I ever bought about computer science. It has a direct and clear style of presentation which is a hallmark of all of his writing and which springs from his mastery of the subject. It is still used in teaching today. During the same period he also published several important papers on topics related to compilers and translators, for example, on efficient hashing techniques. Although I suspect he would be modest about these, they are really important contributions, used by many different software systems as well as compilers, and are part of the toolkit of any respectable computer scientist.
In 1966 and '67 he spent a year at Canegie-Mellon University in the USA continuing his work on compiler and translator writing. One of the things he did while there was to develop a display package for the Formula Algol system. This was designed to display complex mathematical formulae on the computer's screen. This was another example of an early innovation. Even today, many desktop publishing systems would have difficulty displaying such formulae.
It was upon his return to the UK that his interest in computer graphics, and specifically in computer animation began to blossom. The seeds of his interest had been sown while he was working on the S2 compiler in 1963, when a Stromberg-Carlson SD4020 film recorder had arrived at Aldermaston. He had made some early use of this device, but things really took off in 1968 when, thanks to Bob, the SD4020 was acquired by the Atlas Lab from Aldermaston.
The next few years were particularly productive, leading to international recognition for Hopgood personally and for the graphics group which he founded at the Atlas Lab. Building on the Atlas Algol system which he had helped to create, Bob invented a landmark graphics system called GROATS - GRaphical Output from the Atlas to the SD4020. This was the first of many systems with similarly weird names (SPROGS, PIGS, SMOG, FINGS, and so forth), and contained many important concepts which were to lay the foundations for later work. GROATS was designed specifically for creating films using the SD4020 and was one of only a handful of general purpose systems for computer animation in the world at that time.
At the heart of GROATS were several new ideas which were to have a lasting impact. Perhaps the most important was the introduction of the so-called transformation pipeline and normalisation transformations which have since become key concepts, and part of the formal definition of the GKS and PHIGS international graphics standards. In fact, GROATS had a much more powerful design than these later systems, and allowed coordinate systems to be defined recursively, with clipping and shielding defined at each level of recursion. This was to prove particularly valuable in applications such as computer animation where motion could be described conveniently with local coordinate systems.
GROATS contained many other innovations too numerous to describe here. Suffice it to say that GROATS and its successors were in several respects ahead even of the state of the art in computer animation today. If only more young researchers could find the time to read some of these early papers!
The availability of GROATS and its successor SPROGS led to a big increase in the use of the film recorder by making it easy to use. They were used to produce a number of educational films. Among these was a series of six films made for the Nuffield Foundation and issued by Penguin Books which are still shown today, dealing with chance and thermal equilibrium and illustrating the second law of thermodynamics. Another set of films was made for the Open University to illustrate some of its mathematics courses. Bob also made a film illustrating the animation system and also some computer science teaching films.
Slide of Bob speaking at UAIDE (general view)
During this period, he lectured and published widely in Europe and the United States on computer animation and its use for educational film making. In 1971 his group organised a one-day symposium on the topic which was reported in the Financial Times of London on 10th August, and I quote: "The subject of computer animation flared into prominence in the last fortnight when a one-day symposium was held by the Science Research Council at the Atlas Computer Laboratory in Didcot. This seems to be the spiritual home of the craft in Britain ..."
This was indeed true, and over the next few years, up until 1975, a number of other systems were developed and the Lab attracted many visitors who used the facilities to make films, as well as those generated by scientists working at the Lab itself. Visitors included animators such as John Halas and Stan Hayward, and Alan Kitching who developed the Antics System with help from Hopgood and his staff, as well as educationalists such as Juda Shwartz from MIT.
Just to give something of the flavour of this work I would like to show a short section of video. The first part is from a BBC television programme illustrating some aspects of computer animation and featuring the systems developed by the Atlas Lab group.
This is followed by a of film illustrating a technique called a hash table, used in compilers for storing the names of program variables. The aim is to find a technique for storing and retrieving names in a table which minimises the amount of searching required, when inserting a new entry or looking to see if it already exists. The film was made originally to illustrate a paper, written by Hopgood and a colleague, on an efficient hashing technique, published in 1972. The film is one of a sequence made by Bob Hopgood himself on the SD4020, with an accompanying soundtrack produced by a computer controlled sound synthesiser which the group had interfaced to one of their computers. The initial guess for locating an empty entry in the table is made using a hash function. If that entry is already occupied then a linear search is made for a hole. I should say that this is not the efficient method advocated by Hopgood in his paper, but it would take me too long to explain the more elaborate quadratic hash function.
*** Video shown here ***
Slides of ANTICS "snake"
In the video we saw an example of "in-betweening", in which the computer is given a number of key frames and it generates the intervening ones automatically. Here is an example, made with the ANTICS system, which shows that quite beautiful artistic effects can also be produced in static images by artistic application of this technique. This is a picture presented to Bob by the artist Alan Kitching as a gift for his help in creating the ANTICS system.
In 1975 a decision was made to abolish the Atlas Computer Lab as an independent entity and to merge its activities into Rutherford Appleton Lab, or RAL. By this time Hopgood was very highly regarded, as much for his management skills as for his international scientific reputation. Over the next few years he assumed greater and greater responsibility for the computing activities of the laboratory, ending up as Head of the Computing Division with a staff of 180 and an annual budget of 25 million DM. One of the casualties of the move was Hopgood's graphics group which was effectively disbanded by moving its staff into other areas of responsibility. This was indeed a cruel blow considering the international reputation of the group by this time. It is a good example of how technical excellence can be undermined by high-level political decisions.
Undeterred, a year later in 1976, Hopgood was helping to launch a new activity in the graphics area - the quest for an international standard for computer graphics. The start of this was an historic meeting held in France, known as Seillac I after the French Chateau where it took place. It was attended by many of the world's leading authorities on computer graphics and was where Hopgood first began to work with Professor Encarnacao. This was to lead later to various forms of collaboration between Darmstadt and RAL.
It is interesting to note that at that time Bob did not believe it would be possible to conclude an agreement on a standard. Indeed, his paper at the Seillac conference argued this. However, true to form, rather than adopt an unproductive negative viewpoint, he presented in his paper a number of ways in which one might attempt to define a standard. How fortunate we are that he became an active proponent of standards, and he continues to work on them to this day. By 1979 things had progressed to the point where two competing proposals were being considered, one from the USA and one from Germany. The German proposal - the Graphical Kernel System, or GKS - was eventually chosen as the basis of the proposed standard. There followed six years of "international debate" and development, involving painstaking and detailed technical design.
During this design phase, one of the key issues to be solved was to settle on a method for defining pictures which was portable across different kinds of hardware, and which could be addressed from different computer languages. It was here that Hopgood's work on GROATS was to form a fundamental contribution, leading as it did to the concept of multiple coordinate systems and normalisation transformations. This idea has since been adopted not only in GKS, but in a modified form in the PHIGS international standard, and in nearly all other modern graphics systems.
Hopgood's group developed one of the world's first implementations of GKS at RAL. This was very widely distributed, and was used as the main graphics system for a major network of machines dedicated to astronomy in the UK - the so-called STARLINK project. In 1983 he and David Duce, who is also here today, Julian Gallop and Dale Sutcliffe published one of the first books about GKS which did so much to popularise the system as a tool for research, for writing applications in industry, and especially for teaching of computer graphics. Bob's capacity to write clear and readable books continues unabated, and he and David Duce recently published one on the PHIGS standard, and I hear there are more in the pipeline. He has also helped to write and edit numerous other books. Today, GKS is supported by all of the world's major computer manufacturers.
He has also continued to work on the technical aspects of standards. Constantly striving to improve, he is currently one of the prime movers behind a completely revised version of GKS which brings a stronger mathematical basis and elegance to the new system.
In a short talk like this it is difficult to do justice to such a highly productive and original career spanning over thirty years. I will therefore briefly mention some other important contributions which Prof Hopgood has made, and hope that he won't be offended if I have forgotten anything of which he is particularly proud.
He is held in very high esteem by many computer scientists, especially in Europe and the UK, as well as other parts of the world, where his advice is frequently sort. He is active in projects to promote research and development in computer graphics within Europe, including continuing collaboration with Darmstadt under various European Community Programmes.
As Head of the RAL Informatics Division, and an Associate Director of the Lab, he is known and respected worldwide. He has been responsible directly or indirectly for a number of major innovations and policy developments. For example, around 1980/1981 he was responsible for persuading ICL to adopt the PERQ workstation, leading directly to the major switch to single user, desktop workstations in the UK, and which helped to lead the way for similar developments in other countries.
Slide of Bob opening first PERQ in UK
He was a member of the UK's Roberts' Committee, and drafted the papers which led to the establishment of the Alvey Programme for Information Technology, since superceded by the European Esprit programmes. The Alvey Programme gave a major boost to IT R&D in Britain.
Slide of EG'86 meeting!
He was one of the founder members of the European Association for Computer Graphics, and was for some years its Vice Chairman at a time when Professor Encarnacao was the Chairman. He has worked tremendously hard on behalf of the Association, often behind the scenes, helping with the organisation of conferences and technical workshops. His very broad range of knowledge and experience have been invaluable in helping the Association to build and maintain a very high professional and technical standard.
Slide of UAIDE (close up).
Since 1972 he has been a full Professor of Computer Science at Brunel University, having lectured there since 1967. He teaches - what else? - compiler and translator writing and computer graphics and supervises postgraduate students up to PhD level.
What is he like as a person? I know many people who have worked with and for Bob over many years. They tell a consistent story of a man who is a superb manager, someone who leads from the front and who who inspires those who work with him to produce high quality work. He is always willing to help and advise colleagues and friends, and especially younger research workers. His enthusiasm for good ideas and high quality work remain undimmed after more than two decades at the top of his chosen field. I can say with all honesty that for years he has been a source of inspiration and encouragement for me personally - he consistently sets a high standard for everything he does and provides a shining example for others to aspire to. If you want an illustration of this, I can tell you that his daughter Amanda is about to follow in his footsteps by studying computer science at Cambridge University! Who knows, in another twenty years we may be honouring another Hopgood.
But he is not a man who seeks the limelight. Time and again you can find examples of success which he has helped to create by patient, careful, thoughtful work behind the scenes.
Like all who rise to the very top of their profession Professor Hopgood's success is founded not only on outstanding ability but on dedicated hard work and some sacrifices, including long hours and travel away from home. He is happily married and has two daughters and a son. I am sure that Bob himself would agree that this family background has been a great source of support to him during his career. I am therefore particularly pleased that Mrs Barabara Hopgood is here today to see Bob receive this degree and to share in this honour.
In closing, I would like to thank THD and Professor Encarnacao in particular for inviting me to talk to you today. Like Professor Hopgood, Professor Encarnacao has risen to the very top of the computer graphics field, and has established here a research centre of world renown of which you can all be justifiably proud. It therefore seems particularly fitting, in view of their long history of cooperation, that THD should confer this degree on Professor Hopgood.
I am sure that all those who know Bob will consider this an honour which he truly deserves, and will wish to join with me in congratulating him and hoping that he will continue to lead and inspire us in the years ahead.