Hardcore Volume 4 No. 3 1984 06 BASUG GB

hard core THE JOURNAL OF THE BRITISH APPLE SYSTEMS USER GROUP Tune i9s4 kl VOLUME 4 No. 3 Page 2 HARDCORE - The Journal of the British Apple Systems User Group June 84 Services ]} COLONIAL WAY, WATFORD, HERTS TEL. WATFORD (092. Call in now for a Macintosh demonstratior A special purpose language for multiple choice tests and quizzes, questionnaires and Computer aided learning (CAL) ~ £50 + VAT s MICKIE progiams can be written by people with no previous experiance of programming Many he neither the inclination nor the f general purpose languages sui # MICKIE can be used by pe remember format, designed sp time user who does not want E absolutely necessary about computers and computer languages & Originally developed for medical history taking, MICKI jatad in hospitals Schools and Nd Drovides comprehensive, has been @ 1@ Dr Chnstopher Evans The Mighty Micro’ and Tne Mating of tne Micro} Full Apple implementation trom SYSTEMICS LIMITED 21 23, The Bridge, Harrow, Middlesex HA3 SAG. Yat: 01-863 0079 FY ice se rxaeman ot ne nana Prva! tocar TEE BASF DISKS - individually numbered - 5.25" single sided, single density - suitable Apple J{,//e & // - box contains 10 disks 1 BOX £ 15.50 NO 10 BOXES £140.00 EXTRAS am Computer td 37 pe PAY VAT, POSTAGE & PACKING Send order & remittance to: SYSTEMICS LIMITED, 21-23 THE BRIDGE HARROW MIDDLES#X HA3 5AG 01 863 0079 Access & Barclaycard accepte: ] June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 3 hardcore THE JOURNAL OF THE BRITISH APPLE SYSTEMS USER GROUP P.O.BOX 174 WATFORD WD2 6NF BASUG Ltd. is a non-profit-making company limited by guarantee. Basug Committee Bob Raikes Chairman Norah Arnold Secretary Roger Gear-Evans Treasurer Administrator Software Library Meetings Fran Teo Jim Panks John Rogers David Bolton & Quentin Reidford Communications Keith Chamberlain Lit. Library & Membership Alick Elithorn Medical SIG Yvette Raikes Hardcore Editor Jen Bettle Peter Blair & Jim Panks Assistants HOBIE ECONO U IEEE EERE BASUG Telephone No.: Bracknell ((jiii#)) #02 Please note: This number is administration ONLY. Bulletin Board: |() He, Founder members of the Association of Computer Clubs. COPYRIGHT (C) - The contents of this journal are copyright of the British Apple Systems User Group and /or the respective authors. However permission will normally be granted for non-commercial reproduction by user groups affiliated to the International Apple Core, provided the author and source are properly credited. The opinions and views expressed are those lof =the various contributors, and are not necessarily supported by the British Apple Systems User Group. CONTENTS 4 Editorial Chairman's Corner 5 Artificial Intelligence Leo Crossfield 14 Apple //c 15 Competition 16 Format 80 : Enhanced Chester Kemp 19 Software Library 20 More on the ITT 2020 Andy Holderness 21 Lonely Apples 22 Sercom Serial Card Ewen Wannop 26 Applewriter I Strikes Back Roger Mather Selwyn Ward Richard Welch Norah Arnold 29 Eamon Update 30 Book Review 32 Education 35 Beginners' Page John Sharp 36 Apple Prices 38 Daisy Wheel Printers Bob Raikes 40 Local Groups 41 Exhibitions Readers' Letters 46 Diary 3% INTERNATIONAL APPLE CORE MEMBER OF THE INTERNATIONAL APPLE CORE HARDCORE is produced using Applewriter J, and printed on an APTEC Flowriter with 4 Madeleine Daisy-wheel Front Cover: Another Norah Arnold bit image masterpiece printed on a Kaga KP810, Reproduced by kind permission of Data Efficiency Ld. Page 4 HARDCORE - The Journal of the British Apple Systems User Group June 1984 Editorial Many thanks to those of you who have replied to our printer cards survey. We expect to have the results in the next issue. If any ‘of you just hadn't got around to sending in, please do it now. The more data we have, the better the final analy: exhibition there was a demonstration of Appleworks. Several children watched the whole thing from beginning to end with great interest. Who says children only want to look at games? At a recent We recently reviewed a book called "CP/M and the personal computer". The publishers told us the price was £10.95 but on two separate occasions this book has been seen in shops at £19.95. It seems there was a typing error in the publisher's information so if you are intending buying this book be prepared to pay the higher price. While on the subject of books, we are on the lookout for people to review books on interfacing and generally connecting the Apple to the outside world. It seems that few people know about this area so if you are one of the few (even if not an expert) and would like to review a book on it, please let me know on iif #fiH2iL This issue sees the start of our series on Artificial Intelligence. Next issue it will deal with Expert Systems. It is a complex but extremely fascinating subject. You may find you have to re-read the article a couple of times but it is well worth the effort. Contributions to Hardcore are always welcome. Obviously, we prefer these on disk (unless it is just a short note) but if you can't manage. that, we will accept the written word. Don't let a mere technicality stop you writing. If you need advice just ask. I look forward to hearing from you, $.1.G.s Spreadsheet SIG We are starting up a Spreadsheet SIG. Anyone interested in this, please contact Fran Teo ‘on Bracknell (itil) HEH, Chairman’s Corner Those of you who have been members of BASUG for any time will know the name of Tony Williams. One of the first members of the club, Tony has worn many hats. I first met him when he was membership secretary. One Sunday meeting at Parkstreet was punctuated by the Apple turning disks on and off as a sort was done on the database. I think the program being used was ‘Filing Cabinet’, a Basic progcam from the Apple contributed software library, and the sort took upwards of 4 hrs. The software on the Apple has come a long way since then. Tony had always been involved with Hardcore, and in mid-1982, he took over as Editor. His inimitable style (and the fact that he knew about grammer an' fings), not to mention his daisy wheel printer marked his time as Editor. The pressure of making a living finally told, and after a spell as treasurer, he became Press officer. Sadly the pressure of work has meant that he .has had to resign from the committee where he will be greatly missed, I'm sure that all will join me in wishing him success and thanking him for his great efforts for the club (Will someone please tell Tony, as he won't see this, his membership is overdue!). On the Apple front it has just been announced that the development team for Apple /// has been disbanded, This must mean that the model will be allowed to die a natural death. The introduction of the 128k Ile and Prodos, which allows a Profile hard disk to be used, means that most people who might have needed a /// will buy a //e. The /Ic's introduction is an exciting event, and confirms Apple's commitment to the 6502 machine. Rumours in the press of a //x, 8/16 bit machine later in the year, or early next, look very promising too. BASUG also moves forward, welcoming 2 new members who have ordered Macintoshes, although the latest is that availability in the UK is going to be further delayed, There seems little doubt that this will be a great machine, but unless they are in the market, the battle against IBM will be made much harder. (It would be interesting to know the comparative world wide sales for boards to make Apples ISM compatible and vice versa): Has anyone tried one of these MS DOS boards for Apples? June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 5 Artificial Intelligence by Leo Crossfield Department of Educational Studies University of Surrey April 1984 INTRODUCTION. This article gives an overview of artificial intelligence (Al) with the aim of offering a general background to what Al is and the nature of the academic environment in which it developed. A number of problems are also covered which are associated with the nature of ‘intelligence’, and have been used by some of the critics of Al to undermine its research claims. The reader is offered this background so that future articles can concentrate upon Al techniques and developments within an historical context. There is also a need to outline the tradition that Al has developed from, for the language of the Al community often creates a dis-chord in some humanists who tend to cry with indignation when Al researchers explain the behaviour of their programs in mentalist terms. They suggest that the psychological vocabulary such as ‘understanding’, ‘knowing’, ‘intentionality’ and ‘the capability of learning’, are cognitive activities which are specific to the domain of living systems. With the criticism being that to use such a vocabulary to describe the behaviour of a machine is to unnecessarily anthropomorphise the machine. The emphasis of the machine as a cognitive system strikes many dis-chords but it was not too long ago that the dominant paradigm of behaviourist psychology (which was headed by the American psychologist Skinner) offered a mechanistic view of psychology which refused to interpret behaviour in terms of conceptual schemes. This paradigm had a powerful and firm grip on the human sciences, suggesting that all mental activities could be explained by behaviour and that any mentalist term was at best a shorthand label. Therefore the behaviourist tradition did not allow talk of unobservable cognitive activities, but rather offered a mechanistic view that living entities could be observed as black boxes which had only input/output observables; namely behaviour. Consequently this paper, as well as introducing the reader to the development of Al, also introduces a number of developments which assisted the move away from the behaviourist tradition. The need to speak of ‘understanding’ as an intelligent activity was paramount to the early traditions from which AI developed. WHAT IS ARTIFICIAL INTELLIGENCE?. Artificial intelligence could be defined as a research effort which attempts to understand the human cognitive processes by developing models in the form of computer programs. Minsky (1968) terms Al as "...the science of making machines do things that would require intelligence if done by men". The term ‘artificial intelligence’ was first used by McCarthy in the early 1950s, and Since that time the actual term has created many problems because the concepts of ‘artificial’ and ‘intelligence’ are by their very nature problematic. A_PROBLEM OF SEMANTICS. The dictionary definition of the _ term ‘artificial’ suggests that it is “made by human skill; insincere, shallow, not genuine nor natural". As Simon (1969) suggests "our language seems to reflect man's deep distrust for his own products". The concept of ‘intelligence’ has been a continuous source of argument in the human sciences. With both philosophy and. psychology investigating notions which are generally assumed to be related to intelligence, such as intentionality, under- standing, belief, morality, knowledge, learning, teaching, wisdom, etc. Consequently, the concept of intelligence is loaded, for we all feel that intelligence is the essence which separates mankind from the animal kingdom, Intelligence therefore allows mankind to take a special _ prestigious position on this planet for it has been the basis from which mankind has been able to develop complex cultures, sciences and philosophies, and yet, it’ is also the base for all of man's, social, political and prejudicial problems. . It is, of course, impossible to define the nature of intelligence in one paragraph, but Page 6 HARDCORE - The Journal of the British Apple Systems User Group June 1984 it is important to outline that our intuitive understanding of what constitutes intelligence is often insufficient. For we rarely consider the act of walking and the act of problem solving as both constituting intelligent behaviour, but we would consider the act of walking as an _ intelligent activity if it was driven by some form of purpose, for instance, to move across a room to make a telephone call. Yet the problems that are involved in the co-ordination of basic motor skills are a fundamental area of interest to many Al researchers. Perhaps the problems that are inherent in the nature of intelligence are created by the generality and undefinable nature of the term. But although intelligence operates on many levels, it can perhaps be best defined as any action which is governed by intentionality. And we then find ourselves in the realm of philsophy. Some philosphers have taken this argument to suggest that machines, as symbol manipulating mechanisms, can never have intentionality no matter how Cleverly programmed, Searle's (1980) arguments are specific on this point, and Hartley (1980) has suggested that it does not help us to speak of intelligence in machines, but rather we should speak of ‘competence’. His argument suggests that using the term tintelligence' in machines is counter productive for it creates too many issues that are unsurmountable. = =The — term ‘competence! being a more neutral term which can be sensibly applied to developing machines which model human areas of competence. Also, problems which arise by using the term of intelligence as intentionality raise the issue of mind. Is it necessary to have a mind to be capable of exhibiting intelligent behaviour? On an intuitive level one would perhaps suggest that it is, but then we are deeply embedded within the philosophical tradition of the dualist theories of mind and body; and more recently mind and brain. The arguments in this area suggest that the concept of mind is a pre-requisite for intelligence, and exists as some form of spiritual entity which is somehow in control of the brain; counter arguments suggesting that what we intuitively consider as mind is nothing more than the consequence of the brains complex activities. A reasonable analogy for this counter argument has been put forward that mind can be considered as a process which relates to the brain in a similar manner to digestion relating to the stomach, We can see from this brief and general discussion that the term ‘intelligence’ can take us quickly into deep philosophical discussion, perhaps never to re-emerge, let alone do any interesting research, But we are still left with a problem of semantics. At this stage it is sufficient to suggest that an analysis of the terms ‘artificial’ and ‘intelligence’ does not really produce any form of understanding of what is encompassed in Al research, The early founders of Al, perhaps fell into the same trap as the early psychologists who attempted to devise a simple single test of intelligence. Similarly, some of the founders of Al hoped that it would be possible to develop a large program in a massive computer which could become intelligent in a manner similar to man. I have attempted to outline a few of the countless arguments which surround the nature of intelligence which I feel offer some reasons why perhaps the term ‘artificial intelligence' can be considered as an unfortunate terminology for a science which is growing in importance and which will inevitably effect many diverse disciplines. Western civilisation is moving at an exponential rate into a world which is dominated by Information Technology (IT), and within this context I suggest that the research area of Al will inevitably have profound effects upon our future. This paper so far has offered a very brief introdution to some of the philosophical difficulties which surround the AI community - namely the difficulty of semantics and terminology. A suggestion has been made for the importance of Al research in the future, but to balance this initial introduction it is necessary to look at the history of events which gave rise to the development of nce. this new s THE BACKGROUND OF Al. The emergence of cybernetics, in the 1940s, used general engineering control principles in an attempt to seek the general and fundamental principles of both animate and inanimate systems. Rosenblueth, Wiener and Bigelow (1943) argued that a simple goal of a system could be termed as goal directed behaviour and purposivenes of action, and that the adaptive behaviour of any system June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 7 could be realised in feedback mechanisms. The papers that were written at this time were not without their critics, for Rosenblueth, Wiener and Bigelow used mentalist concepts such as ‘intention’ to speak about adaptive man made systems, Suggesting that some machines are intrinsically purposeful, using a torpedo with a target-seeking mechanism as an example. Theretore, the early interest in control theory and servo-mechanisms allowed the early cyberneticists to use general engineering concepts to study the fundamental structure of different kinds of systems, They suggested that systems which have purposeful active behaviour can be sub-divided into two distinct _ classes, systems which utilise feedback (teleological systems) and systems without feedback mechanisms (non-teleological). This suggested that only systems which could utilise feedback control mechanisms, whether man made or not, could exhibit adaptive and intelligent behaviour and purposiveness of action, Although cybernetics suggested that the fundamental need tor an adaptive system was feedback, they also suggested that for the feedback to constitute a control mechanism it had to focus on the feedback of information rather than energy (energy historically being a centre of interest to the natural sciences). They therefore suggested that the feedback of information was necessary for a system to guide and adapt its behaviour, as well as allowing such systems some form of predictive behaviour. They suggested that the discontinuity of observable behaviour, in comparing humans and high mammals, is mainly duc to the fact that animals are limited to instinctive predictive behaviour, whereas mankind is capable of both instinctive and intellectual predictive behaviour. The Rosenblueth, Wiener and Bigclow paper offered a further dimension which laid down the foundations for Al; namely that it was possible to make comparisons beween living organisms and machines. They suggested that there were broad classes of behaviour which were the same in machines and living organisms. This emphasis suggested that one could view high order similarities between living organisms and machines rather than seek what was different. This high order similarity was to view both man and adaptive machine as symbol manipulating (information processing) systems, which gave rise to a move in psychology with some psychologists (for example Craik, 1943) tending to develop a model of man as an ‘information processor’, Cybernetics had a profound influence on many disciplines throughout the 1940s and 1950s; perhaps because if started as a truly inter-disciplinary group of people from such diverse areas as Psychology, Engineering, Mathematics, Philosophy, and Physics. Indeed it developed out of the thoughts of theoreticians and scientists who shared similar basic beliefs, but who offered cach other different views and experiences. THE TURING MACHINE, The development of models which saw the human as an information processing mechanism had foundations in the early work of the brilliant British mathematician, Alan Turing. Turing's (1936) early hypothetical work on finite and infinite state machines suggested that it was possible for a simple symbol manipulating machine to use a tape to read data, store symbols and output its computation. This hypothetical machine ‘The Turing Machine’ was limited to three activities, 1. Read the symbol that is stored in the tape position that is currently being scanned. 2, Write over the symbol, which is being scanned, with another symbol. 3, Move the tape one position to the left or to the right. It is important to note that the tape is Potentially infinite for extra tape can be added if need be. The ‘Turing machine is a theory of computability with the machine — being determined by a finite set of instructions. Each instruction therefore specifies the present state of the machine, the symbol to be scanned on the tape, the action to be carried out and the next state of the machine. It was concluded that a Turing machine could in effect become a ‘Universal Turing Machine! by reading data from the tape, and imitating the machine's function which is represented by that data. Johnson-Laird (1983) suggests that. "A digital Computer and Page & HARDCORE - The Journal of the British Apple Systems User Group its program stand in much the same theoretical relation to one another as do a Universal Turing Machine and a tape representing a speci Turing machine." Therefore Turing offered a thesis which showed that it was practical to suggest the possibility that a universal machine could be designed which could change its function based upon information which it read from the tape. One must recognise that this early work perhaps led to the initial design of the modern digital computer. It is also interesting to note that Turing was one of the first people to actually consider the developments of computer programs as constituting the development of machine intelligence. Indeed, his paper (1963) outlines the foundations for the famous test ‘The Turing Test' which he devised to test whether a = machine — truly —_exhibits intelligence. During this period von Neuman designed an architecture for a computing system. Although this architecture was initially built with valve technology, itis still the architecture which is used by modern computers; although the Japanese fifth generation computer project is attempting to move away from this architecture towards parallel processing machines. To summarise, this entire area encompassed a complex series of events. The emphasis on people as information processors was not without its critics for it offered an environment in which mentalist concepts could be used to refer to machines. The development of this mentalistic language, although offering a constructive and powerful tool for describing the behaviour of machines, opposed the basic tenet of the behaviourist’ tradition that dominated psychology. Behaviourism had attempted to break from the non-scientific, philosophical _ traditions that psychology had grown from; suggesting that there are no mental worlds, and offering a mechanistic approach _to psychology. This approach was dominated by the need to be 'scientific' and ‘objective’, and suggested that all human characteristics could be explained in laboratory conditions; although a great deal of this early research was done by observing the behaviour of lower animals. June 1984 Traditionally all psychological use of a mentalist language had developed _ from ‘idealistic! philosophy, but it is ironic how cybernetics had re-presented these early notions on the basis of ‘mechanistic! and engineering terms. It was this period which could perhaps be considered as the birth perlod for cognitive science. It is also interesting that in Britain the growth of cognitive psychology, which eventually broke the hold of _—the behaviourist tradition, had —_ progressed alongside the developments of computer technology which came to offer, for the first. time, a general-purpose tool for cognitive research, Theories of the cognitive processes began to develop too fast to be built into ordinary mechanisms, but gradually the flexibility of computer programs offered the basis for far more flexible experiments to take place. THE EMERGENCE OF Al. Artificial battleground of the early 1950s when the intelligence emerged from this first general purpose computers became available. Newell, Shaw and Simon (1956) began a number of serious programs to experiment with problem solving. This move was the first breakaway from cybernetics which had attempted to _ theoretically formulate basic principles of learning and behaviour. Consequently, as Minsky (1968) suggests, cybernetics divided into three main avenues: 1) The continuation of the cybernetic search for fundamental principles for self organising systems. This approach can be seen in the work of Ashby (1956) and Pask (1976). 2) The second avenue was an attempt to build working models of human behaviour. This work attempted to replicate the human as accurately as possible. 3) The third approach was Artificial Intelligence which attempted to build machines which are intelligent in their own right. It is this third avenue which is the main realm of current Al research, although the second avenue which is often termed ‘hard’ AI still has its support. STRUCTURED BASIC forApplell+ & Apple lle At last you can buy a true Structured Basic interpreter (this is not a pre-compiler) which has all the benefits of Pascal plus many other features, runs all existing Applesoft programs and uses DOS 3.3 as its operating | sie ‘and avoid system. The perfect language | PBageraeaecbit ny ag Sap cummed Gy for novices and professionals | Disposal of Variables Unique ‘RE- alike. LEASE’ command allows memory oc- cupied by unwanted arrays fo be re- i) Take a look at these claimed fr uso. na atures, onal Graphics Commands p lppe ticle tid Graphmode, Page, Hires, Lores, 3 | Superimpose, Textmode, Mixed, Full, ProcedureNames Subroutines called | Fillwith, Screen. by name as with Pascal. No line rofer- | Passed Parameters Variables may ences needed. | be passed to procedures as argu- Structured Commands RE- | ments. - PEAT .. UNTIL, WHILE . . ENOWHILE, | Compatibility Virtually all Applesoft/ FOR . . NEXT, | DOS 3.3 programs will execute without IF... THEN. .ELSE.. ENDIF | modificaton. | Advanced GASE Statement IF(expl), | 'Nterpretive In nature All 32new com- (expn) .. Caso 1..Casen..Else.. | Tandsmay be usedin immediate mode, End if | entered, sted and xooutedin programs Disk Procedure Libraries Proce- just like other Basic ands. dures not resident in the program are _| Improved Error Handling =ONERR, automatically read from disk and added | ERRSTART, ERREND. Procedure Overlaying Memory oc- | Available from all cupied by procedure called from disk | 1). pope atacenared ordi U-M dealers at £49.95 Uses when procedure is finished. Local Variables Ustsot variables may | aes be deciared for local use by each pro- | Srectured Basic was wrinen Buctans be deciared for loca ach pro- | Sc Conn OW and scan 9 summa (J- MICROCOMPUTERS PRESENTS <m7mmtt PERSPECTIVE DRAWING FS Ry) N This unique package = t J allows you to create perspective drawings on ] the screen of your Apple. Makes clear the principle ee of perspective drawing. A LDF: FS comprehensive set of cueene to plotters, graphics printers tc. Available from all U-M dealers at £30.00 Prices exclude VAT and delivery & UMICROCOMPUTERS Telex 629279 UMICROG Apple, trademark of Apple Computers Ine Page 10 HARDCORE - The Journal of the British Apple Systems User Group June 1984 THE DEVELOPMENT OF Al. The nature of the learning process has been of fundamental interest to the workers in both cybernetics and Al. Wiener (1948) spoke for cybernetics suggesting that all self adaptive systems must have the ability to learn, The following statement has been a fundamental focus of cybernetics: "Can man-made machines learn and can they re-produce themselves?" (Wiener 1948 p.170). Also Al has taken this tradition with the emphasis of many Al programs being on the need for the program to develop better reasoning strategies on the basis of experience as well as increasing its base of knowledge. McCarthy, in 1956, realised the potential power of computers and with a number of colleagues he organised the first Al conference at Dartmouth College in New Hamshire, USA. Just before this conference Newell, Simon and Shaw had been working on a program which modelled human problem solving called "The Logical Theorist" which had been designed to discover proofs from theorems in logic. Although this early Al program did not find any theorems that had not already been logically proved, it did find some elegant proofs of existing theorems. So, although computers had been mainly designed to aid researchers in disciplines which had inherent mathematical problems, such as Physics, it began to be used for non-numerical computations. McCarthy developed a language called LISP which allowed symbols to be manipulated in the form of lists, One of the major strengths that this language has over other formal languages such as Fortran, is that there are no differentiations made between program code and data. Consequently, programs which are developed in this language can analyse their own functions and re-write them if required; this facility has obvious and fundamental importance for the design of programs which need to adapt and learn. With this development, a number of programs were written to manipulate symbols in the form of problem-solving and pattern recog: well as modelling simple learning tasks. Artificial intelligence therefore developed out of two distinct groups, namely, mathematicians and computer scientists on the one hand and on the other researchers from such fields as cybernetics, neurology, biology and psychology whose main area of enquiry was the investigation of the Structure of biological systems. The early work in Al attempted to look for general mechanisms of learning by developing systems which could adapt to their environment without any prior knowledge. Unfortunately most of the attempts failed with the attempts that succeeded being seen as trivial. This period also emphasised the early Al researcher's beliefs that general mechansisms could be developed. It was during this early period that Newell and Simon (1963) developed their ‘General Problem Solver’ (GPS). Throughout the 1960s, the attention of Al researchers moved away from learning towards knowledge based problem solving and natural language understanding. Samuel (1963) developed a checker playing program which developed a better game on the basis of experience. But it was mainly Winston's research (1970) which renewed the Al community's interest in learning mechanisms. This new trend has tended to accept the fact that the learning process is a complex and difficult process. Consequently, Al research has moved away from attempting to discover general principles of learning and problem solving towards a more specific approach. This more recent approach concentrates upon simple learning problems on the one hand, and on the other, the development of large knowledge bases of domain specific knowledge. There was therefore a move away from attempting to discover general models to the more modern approach in developing systems which are intelligent in @ small domain. This major progression of Al has culminated in the current research on expert systems. Of course, the area of expert systems can be seen as the main area of success that Al has achieved to date. But, although the outcome of Al research, on the surface, appears to have achieved few practical outcomes in the form of products, it has achieved a great deal in developing tools which will aid future reseach in this area. Tools such as LISP, PROLOG, POPLOG, and POP-2 offer Cognitive research powerful programming environments and the Al techniques which are associated with knowledge representation and elicitation offer powerful methodologies for future work in the modelling of the cognitive processes. June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page II It is perhaps this background that has tended to influence governmental attitudes towards Al research in Britain, The entire nature of this research area is fundamentally difficult because of its expensive, long term nature and the fact that it is true pioneering research which is still in its infancy and therefore cannot offer final guarantees. THE POLITICAL BACKGROUND OF Al RESEARCH IN GREAT BRITAIN. The field of artifical intelligence, _ in Britain, has suffered a great deal from governmental policies. The recent recommendations of the Alvey Committee (1982) for the British government to encourage work in Al research has created a new focus. These recommendations have been made in an attempt to stop Britain from falling behind the recent innovative spirit of the Japanese fifth generation computer project. This new attitude is a major political change in Britain _ for, historically, the Science Research Council commissioned Sir James Lighthill (1972) to produce a report on the developments in Al: The Lighthill report was highly critical of Al research in Britain and consequently the repercussions of this report were that a number of valuable and active British Al research groups were cut. This led to a number of eminent Al researchers leaving Britain to sct up Al research units in countries which offered them the appropriate support; such as the USA. The Japanese fifth generation computer Project has had a major effect upon attitudes towards the development of intelligent machine research, The Japanese proposals were preceded by an exhaustive survey of international academic research in Al; the major emphasis being found in the USA. But interestingly, the language that they decided to use was primarily European based, This language is PROLOG, which is based on symbolic logic and was first designed to formalise natural language and human reasoning (Clark and Tarnlund 1982), The aim of the Japanese project is to create knowledge based = systems = which = are intelligent, in an attempt to achieve man-machine interfaces that are natural to man. The general aim is to pull together both software and hardware projects in an attempt to implement inference generation, learning, and pattern recognition (both visual and aural). Also to develop software which is intelligent enough to write its own programs with an intelligent interface which could elicit. the requirements of — the software from a human, The British government commissioned the Alvey Committee in 1982 to write a report in response to the Japanese initiative. The Alvey report (Alvey, 1982) made a positive response emphasising the need for a nationwide collaborative research and development (R&D) initiative into the study of intelligent knowledge based systems {IKBS). They suggested that such systems should be capable of handling large, incomplete, uncertain or rapidly changing stores of knowledge and have heuristic inference procedures to flexibly access, maintain and generally exploit the stored knowledge. Also the recommendation suggested that such systems should have intelligent front-end interfaces which can communicate with the user of the system in some form of natural language. Sparck Jones (1982 p.6) offers a structure for an IKBS as USER INTERFACE INFERENCE ENGINE KNOWLEDGE BASE VIRTUAL MACHINE A further consequence of this report was that the British government have now realised the implications that Information Technology (IT) holds for education. Con- sequently the Social Science Research Council commissioned Sage and Smith (1983) to look at the area of IT in education. The Alvey committee had identified four ‘enabling technologies': 1) very large scale integration (VLSI) 2) software engineering 3) man-machine interface (MMI) 4) intelligent knowledge based systems (IKBS) and Sage and Smith (1983) suggested that Page 12 HARDCORE - The Journal of the British Apple Systems User Group June 1984 items 3 and 4 were particularly relevant to the area of micro-computers in education. They suggested further that although the Alvey Committee had directed their attention towards “the significance of understanding...(the) suCognitive processes and the orgai jon of knowledge for the design of expert systems..." (Sage and Smith, 1983 p.11) it had paid little attention to the need for applying this area of intelligent systems to Its use in an educational context. The recommendations of the Sage and Smith report also suggested the need for a collaborative R&D initiative which should mect two particular areas of need "se an information base for the planning of curricula and - a research base for what might be called an ‘education industry" (Sage & Smith, 1983, p.15). Without going into too much detail, Sage and Smith recommended that many areas of Al research were relevant to education and cite Cooper (1980) “Perhaps the real challenge now is to combine the collective expertise of the computer scientists, the cognitive psychologist and the subject matter expert to develop true models of knowledge bases and acquisitive skills for different topics. Our systems need to become intelligent’, flexible and capable of learning themselves in order to be truly valuable to the student", To summarise, there has been a mas: change in the attitude of the British government from cutting research activities which were long term and which attempted to find answers to difficult areas such as the nature of 1) natural language understanding & processing 2) knowledge and how it can be represented 3) the learning process in general 4) perception and pattern recognition 5) automatic programming to actually supporting such research. CONCLUSIONS. To conclude, the relatively short history of Al has seen many changes both in its approach to its area of enquiry and in attitudes towards it as a research initiative. It has been shown that the ficld of Al can be generalised as an inter-disciplinary research community that is interested in using computer technology to model tasks that if carried out by a human being would be considered as intelligent activities. Therefore the basis of such research is pragmatic in as much as. it uses computer programs as working hypotheses, but such research is generally long term, and due to the complex nature of this field of enquiry, cannot guarantee success. Artificial intelligence can therefore be generally summarised as attempting to model and understand the areas in which philosophers have ~—been intellectually debating for centuries; namely the nature of the — cognitive processes. REFERENCES. Alvey (1982). The Report of the Alvey Committee. A_ Programme for Advanced Information Technology. Industry, October 1982. Department of Ashby, W.R. (1956). An Introduction to Cybernetics. Pub: Methuen and Co. Clark and Tarnlund (1982). (Eds.)_ Logic Programming. Pub: Academic Press. Cooper, A. (1980). Computer Based Learning at the OU and the CICERO system. In Lewis, R. and Tagg, E.D. (Eds.) Computer Assisted Learning, scope, progress and limits. Pub: North Holland, pp 55-64, Craik, K. (1943). The == Nature of Explanation. Pub: Cambridge University Press. Hartley, (1980). The Competent nical Report MCSG/6. Dept Science, Brunel University, Computer. of Computer Uxbridge, Middx. Johnson-Laird, PAN. (1983). Mental Models, Pub: Cambridge University Press, Lighthill, Sir James (1972), Artificial Intelligence. Report to the Science Research Council. Minsky, Ms (1968). (Ed) Semantic Information Processing. Pub: Massachusetts June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 13 Institute of Technology. Newell, A. and Simon, H.A. (1963). GPS, a The Blackboard Program that Simulates Human Thought. In Feigenbaum, E.A, and Feldman, J. (Eds.) ance phics printer card vet Computers and Thought. pp 279-297. Pub: McGraw Hill 1963, Pask, G. (1976). Conversational Techniques in the Study and Practice of Education. British Journal of Educational Psychology.46. pp.12-25. Rosenblueth, A., Wiener, N. & Bigelow, J. (1943) Behaviour, Purpose & Teleology. In Philosophy of Science vol, 10, pp 18-24. 1943, Samuel, A.L. (1963), Some Studies in Machine Learning Using the Game of Checkers. In Feigenbaum, F.A. and Feldman, J. (Eds.) Computers and Thought. pp 279-297. Pub: McGraw Hill 1963. Sage, M. and Smith, D.J. (1983), Microcomputers in Education : A Framework for Research, Pub: SSRC, London, 1983, Searle, J.R. (1980). Minds, Brains and Programs. In Behavioural. and Brain Sciences. 1980, Vol 3, pp 417-457, Simon, H.A. (1969). The Sciences of the Artificial. Pub: MIT Press 1969, Sparck-Jones, K. — (1982). _ Intelligent Knowledge Based System : Papers for the Price £99.00 plus Alvey Committee. Executive Summary Working paper. The Computer Laboratory, University of Cambridge, Epson FX-80/100 Editor Turing, A. (1936). On Computable numbers. In Proceedings of the London Mathematical Society. Series 2, 42, pp 230-65. Turing,A. (1963), Can a Machine Think. In Feigenbaum, F.A. and Feldman, 3. (Eds.) Computers and Thought. pp 11-39. Pub: McGraw Hill 1963, Wiener, N. (1963), Cybernetics : or Control : and Communication in the Animal and the £19.95 + VAT Machine. Pub: MIT Press 1963, AY Winston, P.H. (1970). Learning Structural Des- G, SER criptions from Examples, In Winston, P.H. (Ed) The Psychology of Computer Vision. Pub: McGraw Hill, 1975. pp 137-209. computer Ph limited Page 14 HARDCORE - The Journal of the British Apple Systems User Group June 1984 Apple //c Apple //c - A First Impression. With the announcement on 24th April of the new //c, Apple seem to have done it again. Not only have they managed to compress a //e down to a size ot 30cm long by 29cm wide and 6cm high but they have also squeezed into this space the equivalent of 2 serial cards, an 80 column card, 128K of user memory (RAM) and a 5.25 inch (143K capacity) disk drive. The //c is a portable machine - weighing only 7.5 Ibs it has a built in handle that also doubles as a tilting mechanism which gives a more comfortable typing angle. Apple say that the machine was designed primarily as a business computer to compliment their He systems, the majority of which, in the UK at least, are to be found in offices and frequently connected to hard disks and networks which would make it difficult to move the equipment around. Besides the drastic reduction in size the outstanding difference between the //c and its predecessors is that there are no slots at the back of the machine for plug in boards. This will not (in most cases) worry any potential buyer as it already contains more built in "add-ons" than most users ever aspire to. The back of the machine contains the 6 output ports, power supply input connection and an improved on/off switch. As is now standard with Apple products, the ports now have icons (pictures) of their functions rather than the printed word against them. Two of the inputs are a small 9 pin socket for the Apple mouse, joystick or hand controllers and a 15 pin socket to be used for a television or an RGB monitor. This output can also be used for an 80 column by 24 line, flat screen liquid crystal display which is expected to be available around September at an estimated cost of £500 - £600, A video monitor output and an external disk drive socket are also included as standard as are 2 serial ports, one for a modem to enable communications between computers and the other for a printer or a plotter. As you may have noticed with the Macintosh, the company no longer uses parallel interfaces for printers etc. on any of their new products, The processor is a CMOS version of the 6502 called the 65C02. While it is an improvement on the original chip, software that ran under the 6502 will still run without any problem, There are some programs however that will not run on the //c and generally these are the ones that are dependent on a piece of add-on hardware within the machine. Music and sound boards are a case in point and CP/M programs which rely on a plug-in 280 processor PCA is another. Other features of the //c included as standard are 128K RAM (user memory), 16 colour graphics, double high resolution (560 x 192 pixels), a modulator and a buil speaker with headphone jack and volume control, The full sized, 63 key keyboard is almost thé same as found on the //e and felt very comfortable to use. Above the keyboard are the reset button, a 40/80 column switch and a UK/USA character option switch. To the right of these are 2 lights, one for power on indication and the other shows if the built-in drive is being accessed. Apple say that to use the 80 column option a monitor should be used rather than a television. Add-ons available include a purpose built, 9 inch green screen monitor, the now famous mouse with the mouse paint program, an external disk drive, a serial printer (imagewriter) and a carrying case. The equivalent price ot a //e with all the add on boards required to bring it near to the same specification as the new //c is approximately £1481 whereas the new machine costs £1063.75 (VAT included), Why then would anyone still want to purchase the //e? Part of the answer of course lies in the second paragraph of this article. If you use an Apple for business applications the chances are that you will require a hard disk or will want to run CP/M. While this is quite possible on the //e and its predecessors (with add-on capability), the /ic has not been designed with this area in mind. In other words it is not as expandable. At the other extreme, integer is no longer catered for and neither are cassette tape based programs. Many people within the scientific community also use Apples as controllers for experiments using A to D and D to A boards and the like. This sort of work cannot be done with the //c. In other words you pay your money and you take your choice - expandability or portability, bearing in mind that what is offered as standard on the //c is, in many cases, more than adequate for the average user. June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 15 Competition A prize is on its way to the winner, Dave Miller of Finchley, London N3, who sent the following answer: "The reason why the program gives the unusual result the first time is that the start of string space pointer is pointing to location $BFFF. Thus the first string read in is placed at the very top of memory. All By adding nothing to A$ its contents are effectively moved down from the top of memory since the new value of A$ (to the right of the equals) is regarded by Applesoft as different to A$'s old value (to the left of the equals) and so piles down from the top of memory in the same way all other new strings pile down (it's slightly confusing). The fault seems to lie in the VAL function although I cannot say for sure what it is." subsequent strings ‘pile down' from ( this location, For some reason VAL doesn't function F properly when the string to be evaluated is at the very end of user RAM. The reason why this does not occur with DOS is that HIMEM is moved down so the first string is not placed at the end of user RAM. dL >| ) Let J R Systems unravel the computer tangle WE OFFER A COMPLETE SERVICE This can be illustrated by the Consultancy ——— What system and how much ? followi rogram aon ew "BOS Development —— If we can’t find it, we create it | 10 GET AS Instellation — — Fitting our system into your premises. 20 PRINT VAL(AS$) : ; 30 CLEAR Training A helping hand says more than words 40 GoTo 10 Maintenance —-— What happons the day after installation ? The program will now ; always give the odd Specialist. _ —— Mass data capture (including bar codes). result because CLEAR Tele-communications, removes all old Insurance strings thus each new value of AS will start from the top of user RAM. The following program will prevent — this problem — occurring: 10 GET A$ 20 AS = AS +" 30 PRINT VAL(A$) 40 GoTO 10 Are you making full use of your existing software ? Would you like to know the benefits of using packages such as Wordstar, Visicalc, BASE II, Flexware and others, then take advantage of our on-site training. Special prices on selected systems and complete installations. J.R. Systems Ltd. 2Rye Close Harpenden Tel. 05827-5100 Herts. ALS 4LD y Name . Position Company . Address Page 16 HARDCORE - The Journal! of the British Apple Systems User Group June 1984 Format 80 : Enhanced A Word Processor Review by Chester Kemp. To review a word processor is not as straightforward as it at first seems, because ultimately the value to the user is @ subjective thing: therefore what I have tried to do is to evaluate its capabilities by its own design standards, by the needs of a typical user, and also by my own subjective impressions. I have previously used a number of word processors on the Apple (such as Magic Window, Screenwriter Il, Applewriter 1.0 with Go Between for the Centronics 739, and Applewriter ID), and further to this | use Wordstar on a variety ‘of business micros in my professional life. My own writings consist of letters and articles at home, and lengthy reports and programs/ command files professionally. "Format-80 +: Enhanced" is a professional 80-column word processor package running on a 64k Apple //e or an Apple I+ with a 16k RAM card. It currently runs under DOS 3.3 (but a ProDOS version is expected to be out for the Slough "jamboree"). On the Apple Ie, there are no installation problems, you just boot the disk and you are up and running, but if you have a Il+ then you insert a supplied keyboard modification into the games socket to allow a shift-key facility. 1 ran the system with a Euro-plus Apple, U-microcomputers 16k RAM card, a Videx Videoterm with soft switch, and a Centronics 739 printer. Most 80-column cards are supported, and the boot program automatically identifies the appropriate board. (The Videx Videoterm does not support an inverse character set as standard, and an inverse chip can be cheaply provided by Elite Software as an optional extra). Entering text is simplicity itself, and the immediate error correction can be a plain destructive back-space with re-typing. Shift lock operation is standard on the //e but is equally easy on the I+ with shift-control setting it, and the shift key clearing it; thus it handles exactly like a physical typewriter. Tab stops are easily set and cleared, and movement to these stops are either by the Tab key (//e) or ctrl-I (I+). Margins, both left and right, are easily set and moved. It is quite possible to create a document with two columns of text on screen, and have them properly presented without printing and pasting. One of the prominent features of this package is the strong visual image on the screen, with information of the document name, page, line, column and type of action in progress; also you can see whether you are in upper-case, and where tab stops and margins are. Adding to the visual approach, this is a "what you see is what you get" processor, so that you know exactly what the printed page will look like whilst you are typing it. The authors (Mike Hardwick and Gordon Beckmann) have created an office typewriter environment, including the idea of working on sheets of paper. Each sheet of paper is a DOS filename (including Page number), yet there are facilities to handle all these sheets as a complete document, This approach means a shorter learning curve for typists (even if new to computers), and so an asset to the small businessman, The Escape key toggles between two modes of action ("Format" and "Enter" text), and these can be freely toggled at any time. As a visual aid, they use a different cursor for each of these modes. "Format text" has a number of powerful single-key commands (memorably named using the initial letter of the function) which allow insertion, editing and manipulation of text: fill-justitication (even right margins), the complementary compression of extra spaces, as well as centring text and aligning columns of numbers, The usual find, as well as find/replace functions are there, and these work either for the page, or for the full multi-paged document. Cursor control is straightforward and allows you to move to the start and end of text, plus the customary up, down, left, and right. The Processor recognises units of character, word, line, sentence and paragraph; these are used to move forward or backward, as well as to delete or space-fill areas. Blocks of text can be moved about a document by copying or deleting it into a buffer, and then subsequently inserting elsewhere. One feature 1 liked was a word count facility which allowed me to keep a track of my writing when having to provide articles by the word length (although, as it starts its count from the next whole word after the cursor, a document commenced at the very first character will be one short of the actual count). An impressive feature of Format-80 is the FORMAT-80 FOR THE APPLE ][ AND //e Format-80 is simple to use. Text entry is as easy as using a typewriter. Editing and formatting is achieved with single key strokes. “’D” for delete, “I for insert, “J” for justify, “C” for centre, etc. Easy to remember commands because they make sense. What you see is what you get. Format-80 performs virtually any editing and formatting function you can imagine, and displays on the screen the text exactly as it will print out. It supports all Apple compatible printers. Format-80 also includes a sophisticated mailing list, which comes complete with full sort and selection capabilities, The program resides entiraly in memory, including the mailing ist. All drives are free for text and mailing list data. It comes complete with a tutorial manual and a concise, easy to use reference manual, plus a handy user reference card, and is supplied on a standard DOS disc. The disc is not copy-protected. System requirements: 64K APPLE]| //e, and most Apple compatible machines. Disc Drive. 80 Column Card. (supports most 80 column cards) Also available as an upgrade to existing users of Format-80, including revisions to the reference manual, details of new features, and upgraded software. Ramview: 80 column board for the //e, upgradeable to 64K. PRICES Format-80: £129.00 Upgrade: £25.00 Ramview: £80.00 ELITE SOFTWARE COMPANY, 30, Guildford Street, Chertsey, Surrey, KT16 9PQ Tel: 09328 67839 Page 18 HARDCORE - The Journal of the British Apple Systems User Group June 1984 sophisticated print control, in particular Lengthy information is given on print for daisy wheel printers; alas, I only use a installation, and this is just as well, dot-matrix printer with Apples so I can only report on this aspect. There are capabilities of using full proportional type ‘and spacing (stating the aggressiveness of the justification), and also where the daisy wheel does not have a standard sequence of characters, you can set up the print installation so that it will correct for the hardware's shortcomings. Default settings are available for some printers to underline and embolden text, but I would like the variety increased to handle dot-matrix ones. Nevertheless, 1 was able to make much practical use of the print _ installation area, for Format-80 allows you ten special sequences (ctrl-0 to ctrl-9) which can be set to underline, double width, reverse line-feed, and half line feeds for super- and sub-scripts, as well as a variety of fonts. To print documents, you specify the number of copies required of either the page in memory, or a combination of page ranges and/or file names. The small business has a need for a mail-merge facility: this enables them to send out a standard letter apparently personalised by adding, for example, name and address. Format-80 has a simple database to handle this problem and allows up to 510 records of 256 characters each, Given the relevant information, you could make a selection from the database, so that your mail-shot could be to all people in Manchester who are not accountants. This was actually their own example (in a properly conceived database), but it does suffice to demonstrate the logic and power ot the search capabilities for a small business. A non-standard formatted disk holds a single database. The editing facilities for this list are basic, but in practical use this ought not to be of any real significance. Format-80 comes complete with an excellent manual in an AS ring-binder, plus a handy reference card. The manual commences with a short 17-page tutorial broadly covering most points, and together with an eight minute automatic on-screen demonstration program, 1 think it conceivable that this could almost be enough for anyone to lose their fear of computers and actually use the word processor! Next comes the reference manual: this is very clearly written and presented, giving useful examples of each command in action - one just cannot say better than that. because I had to read it a few times to get to grips with some of the points; though the instructions are accurately given, I do wonder whether they could be expressed more clearly. On one point they are in error by omission; the header and footer installations require the "Start Pagination" option to be No-Skip if they wish a heading to be printed on the first page, even when no page numbering is required. ‘The _ print installation also allows you extra chores, so that you can create a document of 3 pages of 80 lines, and then print it as 8 pages of 30 lines, with primary and secondary headings and footers inserted as needed. A document can be re-vamped to change length ‘of pages by the Quickscan facility, which can be useful for irregular but frequent Changes of page length. Text is stored as a number of lines, each completed by a carriage return, and a file can be up to 80 lines of 80 characters. Paragraphs are intelligently recognised so that if you have the odd formula written on a separate line in the middle of the page, it will realise that it is still part of the same paragraph. Occasionally, this is a nuisance: if you indent the start your paragraph by more than six spaces, it will not recognise the change. Another quirk is that the user-defined ctrl-0 is stored as ASCII 00 and translated by Format-80 at print times however, this is a problem to spelling checkers, for the 00 tells it the file has ended! Advice: use it sparingly. The booting program takes about a minute, and this can be speeded up to a quarter of minute using Diversi-DOS or Elite's beautifully designed Wildcard Plus; this speed increase is only ‘on the boot though, for the program handles text read/writes efficiently and speedily. The program is NOT copy protected therefore you are immediately advised to make back-ups. The complete program, including the database section, is all resident in RAM, so that if you only have one disk drive, you will get full use of a data disk, It will support two floppy disk drives, as well as most hard drives. It correctly recommends a video monitor rather than a TV’ screen, The /le's keyboard and design is fully used, but the Ils has a hidden snag in the keyboard modification: if it is installed and co-resides with games paddles with a two-way adaptor (such as EZ-Port), then your arcade June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 19 games will not function properly if they use the triggers. By the by, this is not the same as Format i (commonly sold in the USA), although I believe its roots are in the earlier and un-enhanced Format-80, 1 have previously been stuck between Go-Between's ability to handle the Centronics 737/ proportional font and the powerful flexibility of Wordstar, I still prefer that proportional font, but miss the on-screen visual control in Go-Between. Wordstar has the edge over Format-80 in general, but it has some negative points when running on an Apple: you lose space on disk because the program needs to be co-resident (because it overlays into RAM as It needs it), you are constantly thrashing the disk drive, and it also costs considerably more (especially if you also have to buy the Z80 board and CP/M). Wordstar will not run on the new //c but Format-80 will, For me, Format-80 is an _ excellent product by any standard (any quibbles being quite minor), and thus it displaces the alternatives. 1 would strongly recommend this to anyone, but in particular to small businesses, and to anyone not yet committed to the eccentricities of a particular word processing package. The program is written, supplied and well supported by Elite Software Co, (30 Guildford St, Chertsey, Surrey KT16 9PQ), although it is available trom most dealers. It currently sells at £129.00 (+ VAT = £148.35), and an upgrade from old Format-80 to take advantage of its new features and standard DOS, is £25.00 (+ VAT = £28.75), My thanks are due to Elite Software for providing a copy of Format-80 for this review, and Mark Whelan for answering my many questions, as well as the tolerance and helpfulness of the authors. Software Library Many thanks to all those who have offered to help in the Software Group. You will soon be getting something to work on. Any further offers will be welcomed. Just ring the Software Librarian or drop a line to hi the PO Box. Small Ads Software/Hardware for Sale Information Master (Database). Apple Doc (Aid to development and documentation of Applesoft programs). Macro-sced (Macro builder, Screen editor utilities). FDos Creator (Fast DOS). Symdis (6502 Disassembler). $-C Macro Assembler. Visicalc, Tasc Compiler. The Graphics Magician (Graphics System). Snapshot Il (Copy System). Graftrax Roms for Epson 100 Ill (Two char. sets + graphics). Type right lower case adaptor August '82). (Hardcore Offers to Chris on (iii 8.30 pm. ALP Daily until Perrrrrrrtritttitt er Epson MX80 FT/3 Printer £ 250 Digitek Printmaster Parallel Interface k 45 Any reasonable offers considered. Alan Russell : Tel: Bath (isi) BHP evenings & weekends. LODO UH UIE EERE EERE EE Apple System without disk drives but with > monitor, 16k ram card, colour card and parallel printer card. £450 o.n.o. Tels (eat natin, EERE EER EERE EERE EER EERE ERE ITT 2020 with Apple roms and large quantity of software, £195, Contact: R. H. Barnard. Hach aTRNe HANES (day) or tity (evening). See p. 40, for more. Page 20 HARDCORE - The Journal of the British Apple Systems User Group June 1984 More on the ITT 2020 by Andy Holderness Following on from Mike Siggins' article on the ITT in the February Hardcore, I thought a few more points, based mainly on bitter experience might be worth while. 1) CP/M. As mentioned in the article, the Softcard can be altered to work on the ITT, but at a cost. Peanut Computers are currently selling a Z80 card for the Apple Il at £44.50, but ask some £25 more for an ITT version. The following fix works: on the Softcard and the Digitek Z80 card (as usual with all fixes based on taking a soldering iron to your machine - it's all strictly at your own risk). All I can say is that it worked fine on the card that I use. Locate the 74LS107 IC at the top of the second column of ICs from the left (card facing you with the edge connector to the bottom). Remove this IC from its holder, and carefully bend up pin 12. Solder a length of wire to this raised pin, sufficiently long to reach over the card to the back of the edge connector. Replace the IC in its holder (pin 12 no longer enters the socket, of course), then bring the free end of the wire over the card and VERY CAREFULLY solder it to pin 37 on the edge connector (extreme care is needed here, as there is no proper connection to this point on the edge connector). The card can then be placed in any vacant slot, and CP/M booted. In the case of the Digitek card fitted to my own machine, a number of small capacitors had to be disconnected before all was completely well = if you are modifying *one of these cards, they are the ones that don't look as though they should have been there in the first place! Some of the cheaper cards do not have a full set of connections on the edge connector on the card. In this case, it will be necessary to find an appropriate point at which to tap the 8.8Mhz signal on the main board (soldering on the back of the board where the expansion sockets are fitted would seem to be the easiest place). Following on directly from the above fix for the Z80 card, I have not, so far, found any CP/M software that will not run on the ITT, with, of course, the exception of the HGR statements in the Apple version of Mbasic. These, like those in Applesoft, need either the Jailbreak fitted, or the hard-wired graphics fix, Having said that, however, I have never actually used the Mbasic hires routines. 2) Applesoft/Palsoft. To use Applesoft basic, I have replaced not only the DO and FO roms, but the E0 also. | do not know the precise difference between Applesoft and Palsoft at this point, but note that the Apple master diagnostics disk refuses to run with the Palsoft EO in place on the motherboard. 3) UCSD Pascal/Fortran. In order to use the full resolution of the ITT screen under Pascal, you will need to buy a copy of disk Pl from the club library. This is a version of APPLE! issued by ITT with revised Turtlegraphics routines in it to address the 9-bit graphics. Unfortunately this disk cannot be used directly with the current release of Apple Pascal (1.1). In order to use the routines, use the LIBRARY program on APPLE3 to transfer the Turtlegraphics units from SYSTEM.LIBRARY on PALI (the ITT disk from the library) into a temporary file, then use these units to replace those in the SYSTEM.LIBRARY on APPLE! (copy disks only, of course) The numbers of the intrinsic units are the same, so object code compiled with the Apple Turtlegraphics will run ok, albeit only addressing part of the screen. These routines can also be incorporated into the SYSTEM.LIBRARY file on FORT! in the Apple Fortran system. (In this case, of course, FORTLIB.CODE is used to make the transfer. The fix for upper/lower case entry in my letter in the October 83 Hardcore, by the way, also allows the ITT to produce left and right square brackets as per the Pascal manuals. June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 21 4) Aztec C. A similar fix is also needed to make Aztec C, from Manx Software, run on the 2020. In this case, however, the patches have to be applied direct to the disk, as although source code for’ the various device drivers is supplied, the Unix-like operating system that the C shell sets up requires all commands to be entered in lower case. As the shift toggle is once more at $C063, it is entirely pot-luck whether you have lower case of not when the system is booted, or you will retain it long enough to make the system accept the changes. The patches are as follows (on a COPY of Disk | - startup): Patch the following to $60 from $63 Track $13, Sector $0B, Byte $2E $18 SOE SAA $19 SOE SAA SOE SAA Once these changes have been made, the disk can be booted (once DOS has been added), and the device drivers can be rewritten &s you ChOOSErerserees 5) Transforth and Graforth, Two other languages tested - both from Insoft - are Transforth andGraforth, Surprisingly, Graforth ran (with the usual 9th bit provisos), but Transforth would not. It appears that the latter relies heavily on the graphics routines built into Applesoft and hence requires it to be present. 6) Others - misc. I have just attempted to test Softerm Il on my ITT, only to find that it cannot be run - this is one of those systems which requires a hardware adaptor in order to run; in this case a keyboard extender, which appears to conflict with the internal architecture of the 2020. I say appears, as the code in question is in an EPROM, and I have no means of analysing it, and there is no source code information provided in the package. Surprisingly the Mountain Computer CPS card is completely compatible with the ITT. This was fitted to the machine exactly as per instructions, and has run perfectly since. 1 use this system extensively for communications, and find this card, in conjunction with the Ascii Express Pro. software, reviewed recently, to be almost ideal. 7) Here We Go Again Dept. I have just bought a copy of the new version of GPLE (with the Pro-Dos files on it), only to find that the ctrl-a shift wouldn't work. Usual problem; the toggle was at $C063. Disk patches are: Track ut Sector $0C Byte $2c 21 $05 SFF $22 $04 $26 Each of these locations should be altered from $63 to $60. Limited experimentation with the Pro-Dos files on the disk seem to indicate that the shift toggle works fine here, but there seems to be some sort of problem with the ctrl-v video — toggle. -—_—Investigations proceeding... In conclusion, 1 would like to ask the following of any ITT owners:- a) Have any of you implemented a lower-case mod? b) Does the Videx Softswitch work? c) If you find incompatible software/ hardware, write to Hardcore about it, so we will all know what works and what will not! Lonely Apples The following member would like to contact other Apple users: Mr. G. W. Ferguson HB Es, ad. a. Es Tahu Haul Syathie preety Chris would like to get in touch with people who use Mountain Hardware's ROMPLUS+ card, He can be reached on ((fi\!) le daily until 8.30 pm. Page 22 HARDCORE - The Journal of the British Apple Systems User Group Sercom Serial Card A review by Ewen Wannop. First we had the Apple computer, then we had the peripheral cards. Without those eight slots tucked away inside our machines, where would the Apple be now? And what would we have been able to do with it? It was a far reaching decision indeed to incorporate them into the original design. Most of us have at least the disk card in Slot 6, and possibly a printer card in Slot 1. We might have an extension Ram card in Slot 0, but how many have a put a serial card in Slot 2? Communication between machines has grown in the last year to a point where we can all benefit from the ability to talk to another computer. The printer card enables one-way communication with a printer or other machine placed up to a few feet away, but a serial card opens up a whole new world, It can talk and listen to any make of computer that can support RS232 or RS432 serial protocols, and that means nearly every computer there js! It talks over a line that may be 20 feet or more, that at its least need only be three wires and, by means of a Modem, can talk over two wires to the whole world (though the phone bills will be equally impressive). There is a growth of communication using modems amongst the personal computer world, both for business use, data transfer, Prestel, Micronet and the bread-board message boards. Basug's own excellent board run by Quentin Reidford in Sheffield is now used by many Apples round the world. All of these techniques require in addition to the modem, a serial card in the Apple that can talk the language of Baud rates, Bit lengths and Parity. If you wish to access Prestel and Micronet then you really need a card that will support split baud rates, in order to be able to receive at 1200 baud and send at 75 baud on the back channel, There are very few serial cards that will support split baud rates, but now we have, all the way from New Zealand, a card that not only does that but has a built in dumb terminal program, allowing half and full duplex transmission, and the ability to be set to either Terminal or Modem settings on its RS232 cable. The now common Apple Super Serial card will do all of this except for the split baud rates, but at a price that is nearly half as much again! The Card. The Sercom II serial card from Computer Shop Ltd. comes in a padded box well protected from the rigours of the Post Office and is constructed to the high degree we now expect from most Apple peripherals. The board itself has a total of 12 logic chips plus a 6850 serial chip and a 2716 EPROM holding driving routines. There is a transistor which switches the current to the input buffer from the Apple bus, so we should be assured of no complications from that end from slow chips holding the data bus. On the top of the card, easily accessible while the card is in the Apple, is a six pole DIL switch that controls the baud rates to the transmit and receive functions of the chip. A 26way IDC plug and ribbon cable are connected to a free RS232 female 'D' plug. Two square ‘headers! may be turned to change the sense of the receive/transmit data, and the DTR/RTS line. What it can do. The card supports baud rates of 75/150/300/ 600/1200/2400/4800/9600, and these may be in any combination for transmit and receive. As a standard 6850 serial chip is used, all the usual bit sequences can be set up from software, though the manual does not tell us how. Interrupts are supported by the card, but not as the card is delivered. I found three blank spaces on the card for a transistor and two resistors. When fitted with suitable values I was able to get interrupts from the card, but only after a bit of detective work could I discover the address 1 had to poke to enable the interrupt output. Why the interrupts had not been wired up in the first place 1 do not know, and why was there no mention of it in the instructions that came with the card? The software. The card closely resembles the CCS serial card as far as software is concerned and all the terminal programs I tried the card out with ran first time when set up as a CCS serial card, These included the two versions of Ascii Express and Visiterm. However, June 19846 June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 23 —_—_—_e built into the board is a simple terminal program that can be accessed by issuing control codes. This allows full or half duplex operation, and can be controlled from either the Apple keyboard or from outside the Apple by sending into the board the correct sequence of control characters. How- ever 1 was disappointed that the instructions only told us that this would support lower case characters in inverse video. An address has to be poked to do this as upper case only is set on entry. The instructions tell us the value to get lower case in inverse video, but did not mention that if you poke this address with $00 you will get lower case normally, After setting the card to 1200/75 baud and enabling the interrupts, | was able to use the card with my own Prestel/Micronet software that uses a serial card rather than the games port that Micronet's own software uses. I like to keep my games paddles in there! The split baud fates are a much more satisfactory way of getting the two baud rates required than flipping constantly between the two whenever you send a character as is necessary with the Super Serial Card. After the good points come the bad ones. I have already mentioned some of them, that I had to put in the components needed to enable the interrupts for instance and that they did not mention how to get true lower case from the terminal mode. The split baud rates are a great plus for this card, but they must be set by means of the DIL switch; this means turning off the machine if you do not want to run the risk of the board slipping in its socket and blowing all your chips. It is not impossible to have had them set up by writing to a unique address, this would have allowed a change in mid stream without the danger of opening your Apple. The ribbon cable from the board is just long enough to extend to the back of the computer, I would have liked it to be longer and stretch to my modem, but this seems to be the length that all the serial cards think is needed. It is quite expensive to buy two 'D' plugs and ribbon cable to extend this any further! While looking at the back of the board, I noticed a large number of jumper points, all labelled, that obviously allow various options to be achieved, but on turning ta the pamphlet - I hesitate to call it a manual - the only ones mentioned are those that enable you to replace the 2716 with a pin compatible Ram chip. The instructions are very basic indeed. What they do cover they do fairly well, but the technical information is limited to one side of a page. It mentions nothing about the interrupts or how to achieve them, or any information other than the addresses of the ports that might be of use to a programmer. There are only 15 pages of instructions; they have been properly typeset however but it surely could have been possible to have made a more workmanlike manual than this. I would rather have seen what is now quite common, instructions printed on an ordinary daisywheel and photo-reduced, and had a full set of information than this rather meagre booklet. To sum up. An excellent card, giving good value for money, capable of running from all the terminal programs on the market, but let down by its meagre manual, and the need to add components to make it support interrupts. | hope the manufacturers can put these gripes to right, although even so, I recommend this card over the more expensive Super Serial Card for communication use. It will cope with all but the most obscure configurations and uses that the latter card supports, though for all the things that I wanted the card to do, I found none that it would not support. My thanks go to Data Efficiency Ltd, for lending me the card for review. 2 Rye Close Herts. ALS 4LD ‘Tal, 05827-5100 Page 24 HARDCORE - The Journal of the British Apple Systems User Group June 1984 FIRST CHOICE FOR PACE @apPLE ACCESSORIES @ ALL ITEMS FULLY GUARANTEED FOR ONE YEAR@IMMEDIATE DELIVERY @ FREE POSTAGE & PACKING WITHIN UK IDEVISE for Visicalc users PRINTERIVTENPACECARDS 444, OS SIDEVISE allows APPLE Il, Il+ and lle users to print their VisiCalc spreadsheets sideways. No longer does the width of the printer limit the 92 NEW CROSS STREET, of your spreadsheet. gs aly plus VAT BRADFORD B05 88S Tel (0274) 729306 PACE 0274 729306 2 SYSTEMS EXPANSION 1 BACE SOFTWARE SUPPLIES LTD. ™™ § = 1 s2new Cros Steet ost Boning, trsB05 wf | cen taomrnon at wovst I i= —— {! —_.— pt 1}— tl i . {it = +—1 UT crctse iy cheaue payable to PACE SOFTWARE # 1 SUPPLIES LTO. _ I I or use with L)it+ [te ! f | I I L 4 LOOK AHEAD WITH .... VISION 8O VISION 80 Simply the most versanin BO column card available for the Apple ily ile is 9x11 character set mateix is the largest and sharpest available, giving @ crisp cle display. On-board communications software allows the Apple to act as a terminal with a wide range of databases and a number of differant machines, including Apple to Apple communications. © Character dot matrix is 9x 11 which provides three lines of dots for lull deecenders on lower case letters. The full ASCII set of 128 characters is provided. © Shift key dows what you Would expect it to do. and also has a ahitt lock facility and a visual LED shift lock ndicator © Character set also includes twelve additional keys Ot normally available on the Appia keyboard © This card includes a built in softswitch ie no cable changes necessary when switching between 40/80 columns and graphics © The Duilt in communications software driver gives your Apple the ability to be used as a true interactive intelligent terminal to mainframe computers or communications facilities. Fully compatible with CCS Serial cards and Apple communication cards ete @ The Vision 80 typetace is of an attractive appearance and is highly legible due to its large 9x 11 character font. © It supports all Applesoh commands including the text Window ie Home, Text, GR, HGR, HGR 2, Tabete. It has inverse and normal display ie Highlight ang Lowlight in CP/M and Pascal @ itis possible to change the cursor choractortocither a block cursor or an underscore cursor. The speed of ‘cursor blink can be altered and itis also possibie to fe define the character set with your own personelised font. The card comes complete with demonet utilities disc ana is simple to install. W also includes a comprehensive users manual © For use with Apple Ilr, lle Supports OOS, Pascal and CP/M Software. £185 «var VISION AWII For users of the Apple Writer Il word processor, the Software on this aise automatically carries out al necessary prebout procedures to display Apples word processor output in 80 column format £18 vw VISION 128/256 The Vision 128/258 Ramcard is an inexpensive means of providing up 10 256 Kilobytes of additional RAM ‘main storage for the Apple Il i le. The card can be used as a fast scraichpad, supplementing the local storage af data arrays tor application programs that can utilise this facility directly (eq. Visicaic etc), alternatively with appropnate utility software support, the card c 'be ‘masked’ to appear as 0 fast disc storage unit, The Vision memory expansion is available as a basic 128K card which can be upgraded to 256K RAM as fequired. The card, which is fully compatible with the Vision 80, it fully butfored allowing lower power ONsuMptiON and gwving greater reliability The card iso features 6 L.E.O, indicators to indicate the current bank selgcted and read/write enabled. £240 «var ERE IS youn OR apple UR VISION VCE This praboot dise tor use with the Vision 80 allows VisiCalc users t0 see 3 full 80 columns displayed on the monitor, it so makes use of the extia memory available in the Vision 128 (upto 143K in al £39- va: VISION COMBO PACK Save £40 with this Combo Pack containing: Vision-80 80 Colurnn Text Card: Vision-128 128K Memory Expansion Card; Vision-VCE VisiCalc Expander £424.00 Available from your local dealer or direct from: Sole European Distributors PACE SOFTWARE LTD. 92 NEW CROSS STREET, BRADFORD BD 8BS. Tel. (0274) 729306 Telex 51564 Page 26 HARDCORE - The Journal of the British Apple Systems User Group June 1984 Applewriter | Strikes Back by E.R.Mather There must be a lot of folks like me who overlap isn't very helpful. So, 1 relocated have never managed to afford Applewriter II and have struggled on with I for a good number of years until it has become a matter of principle never to change! Applewriter 1 for my use, 1, You can't do multiple copies. 2. It's very tedious going to and fro from Editor to Printer to correct yet another spelling mistake, 3. You can't send control characters to the printer to change its print type (I use an Epson MX-80), Apart from these factors AW I adequate for my needs. has three big disadvantages is quite Criticisms 1 and 2 have been solved for quite some times there have been articles in Hardcore about multiple copies (February 1983), and since I bought Diversi-Dos and a 16K RAM card I have been able to use lan Trackman's Switcher routine (April 1982). Dare 1 say that Ian's program gave me lots of headaches, as I'm not too good with machine code, until I discovered that he'd got high bytes and low bytes interchanged. Problem 3 was solved by reading an article in the Spring ‘Orchard Computing’. Messrs Walter Wilson and Nicholas Brown have written a patch to the printer program which allows you to send control characters directly to the printer, thus changing the print mode. The crucial character is ~* (esc + shift n) in Applewriter 1. Any lower case letter following this is converted to a control character, and any capital to esc + letter. Thus to switch to enlarged printing type “n, to use emphasized type “F; the appropriate controlling codes are in the Epson manual. In addition, you can get at some of the odd characters by using ~ followed by a number. Their program was written to go into locations $1863-S18FA; but if your printer program is already patched to take account of multiple copies, this runs over the $1900 mark where the text files are loaded, the their program to begin at $300. Enlarged printing is automatically killed by the Epson with every carriage returnj Wilson and Brown used two special characters with * to switch on and off continuous expanded (enlarged) printing - and respectively. I managed to alter the program so that ctrl-n and ctri-t give continuous expanded print anyway. Step 1 - BEOAD your printer program noting its start address and length, call-151 to enter the monitor then type 12B8:A9 00 12BC:A9 03 this switches the printer output to the routine at $300 = BSAVE the altered printer file. Step 2 - still in monitor, type in the hex dump, or the assembly listing and assemble and then add 385:00 00 (this reserves 2 bytes for flags) 387:7B 7D 5B 5D 7C 7E 5C (these are the characters for the special symbols in the Epson manual - you can add others if you like, see the manual) BSAVE PRINT PATCH,A$300,L$8D (or as long as yours is). Step 3 basic,load hello and add a line PRINT PATCH. - ctrl-C return to get back to to BLOAD Note - Wilson and Brown do not annotate their program so I have added my own comments, hoping they are correct and understandable. The original authors give extra information about how to alter the program for other printers and how to translate capitals into esc+ctrl letter if required (the 2 NOP's at 356,357 are for this). June 1984 HARDCORE - The Journal of the British Apple Systems User Group Page 27 TREO ‘SOURCE FILE: PRINT PATCH2 sw-s> NEIT OBJECT FILE NANE 15 PRINT PATCH2.0BJ0 03008 0388: 386: 0382: 0383: 387: 0300348 0301229 80 0303385 FF 0333100 05 O335:A2 O1 O337:96 86 03 O33A:C9 74 033C:00 05 O33EsA2 00 0340386 86 03 0543:29 IF 0345110 11 0347309 5B O34 Aa 0350209 1B 0352220 7F 03 0355168 O3S6:EA 0357268 0358348 O359:AD 8b 03 O39C2FO Of O3SE68 O35F:48 1 ong 2FLcc EGU SFLEP EU APRIL EDU SPr2 EWU 6st EWU 7 PHA a ND 9 sta 10 PLA mM ND n PHA 13 itr) “ sta 15 oa 16 sta ” ity) 18 BNE W PLA 2» cn u Bee 2 BNE 23 INDEL LOA rT) STA 28 RIS 26 COMMAND DEC a PLA 28 cnr a wee 5) cn u Bes 2 one 3 BNE 4 Lon % sm 3e TEST CHP yu BNE 3 Lor ” om 40. NOT.T AND a BL AZ CAP HP ay acs “ CHP 45 Bet ab PHA ” LDA 8 ask " PLA 50 wor 3 NOP 52 PRINT PHA 33 LDA st Bea 55 PLA 56 PHA 8300 $385, $386 $382 $9383 $387 £980 SF {s7F 958 PRL $58 PR FLCC CORMAN C956 INDEL PRINT e601 Fue Fuee £860 ca £678 PRINT ES6E TresT £601 FLEP £674 NOT.T £800 FLEP esiF PRINT £658 PRI 641 SPECIAL £918 Pout FLEP. NOEKP FLAG FOR A CONTROL CHARACTER FLAG FOR CONT EXPANDED PRINT ADDRESS FOR PRINTER TABLE FOR SPECIAL CHARACTERS| CHECK IF PREVIOUS CHAR MAS * YES IT WAS SO BRANCH TT WASN'T SO CONTINUE 15 THIS CHARACTER A *? ves WO, SO PRINT IT SET CC FLAG TD 1 AND EXIT WITROUT PRINTING SET CC FLAG TO ZERO 18 IT LOWER CASE? NO, CODE TOO SHALL 15 IT LOWER CASE? WO, CODE TOO LARGE Is TAP? LT SET EXPANDED PRINT FLAB IsitaT? No CANCEL CONT EXP. PRINT CONV TO A CTRL CHAR AND PRINT 15 IT A CAPITAL? No, CODE 100 LARBE 18 IT A CAPITAL? NO CODE TOO SMALL CODE FOR ESC hasn't Trinder new tricks with a DMP this issue. Instead he sent us this. Peter It seems On second rather he's been converted! a good is it thoughts, AWicksesseee Page 28 HARDCORE - The Journal of the British Apple Systems User Group June 1984 036009 0057 cn ts00) 1S 17 CR? IBLOADPRINT PATCH2.OBI0 0362200 0838 NE NOEIP JCALL-151 364188 9 PLA 0385120 7F 03 40 ASR POUT 4385:00 00 7B 7D SR SD 7C 7E SC 036809 OE BT LOA ESE CODE FOR W O36A:D0 07 2 ‘BNE NOT.T #3800. 538D 0360368 63 MOEIP PLA 036005 FF GA PRT ORR SFE 0300- 48 29 80 8S FF 68 29 7F O3MFLAC 82 03 (PRL) 0308- 48 AS SA 8D 82 03 AS SB 0572:29 OF E80F ANYTHING GETTING HERE MUST © OS10- SD BS OS AD SS O DO OD sre BE SPECIAL 0318- 68 CF SE FO 02 DO 39 AF 0320- 01 8D 85 03 60 CE 85 OS 0377280 87 03 Str 0328- 48 C9 40 90 1A C9 7B BO OS7AIAG FE aE 0330- 27 C? 6E DO O5 AZ O1 BE O37C:4C 58 03 PRINT 0338- 86 03 CF 74 DO OS AZ 00 037205 FF FF 0340- SE 86 O3 29 1F 10 11 CP 0381120 02 Ct $0102 0348- SB BO 22 C9 41 90 23 48 0384380 m RIS O350- A 1B 20 7F 03 68 EA EA O3S8- 48 AD 86 OF FO OE 68 48 ‘Nt SUCCESSFUL ASSENBLY: NO ERRORS 0366- C9 OD DO O8 68 20 7F O3 0368- AP SE DO D7 68 OS FF 6c Tvping the numbers © to & after esc 0370- 82 O3 29 OF 86 FE AA BD Shift-N gives the printer codes Q378- 87 O3 AS FE 4C 5B 03 05 corresponding to the codes at the Os80- FF 20 02 Ci 60 90 00 7B end of the patch program. aa ARS RAS Lesa O.¢ 2.0 4. 6. 1.2 a 5.~ JBSAVE PRINT PATCH. A$S00.L$8D Here’s a sample to show the versatility of Applewriter 1. First, since the printer stays in the print mode that it was in when last used, iet’s return it to normal by using esc shift-N esc F (cancels emphasized mode).. esc shift-N T (cancels enlaraed mode).. esc shift-N R (cancels condensed mode).. SO now we are back in normal tyo: When switching on enlarged print you have to reset the right margin so vou don’t get a wrap around overlap (I alwavs have a PREL:ctr1-IGON:PREO in mv Hello program) so !RM40 sets us up for enlarged. Now this should be enlaraed. face if it ism*t. we can emphasize this 2s well with a simple Cesc shift—N esc —) which gives a heavy effect, let*s kill that one with the appropriate code, and new switch to enhanced mode since we are cenlarged plus condensed? , already in enlarged made, giving a (esc shift-N 0) will condense the enlarged and since a line full of this is 65 characters (I think) we can reset the RM eeaq on my Some of the printer codes operate only from where they are in a line, others change the whole of the line where they are, and I can never remember which is which so often we are a little different from expected. Switching off the enlargement (esc shift-N T) should Leave us in condensed print foreat and finally with a siaple esc shift-" R) we are (hopefully) back to normal. Although the control characters aren’t rpinted they are taken into account when counting for justification so the fill mode may look only left iustified in places. justify June 1986 HARDCORE - The Journal of the British Apple Systems User Group Page 29 Eamon Update by Selwyn Ward Many BASUG members will have already discovered "The Wonderful World of Eamon" courtesy of the BASUG software library. There were around a half dozen Eamon disks in the library, enabling keen adventurers to roam from the Beginners' Cave to the Lair of the Minotaur and beyond. For those that haven't already ventured into this particular section of the library, Eamon is an ingenious game system designed by Donald Brown and generously donated to the world. Unlike the bulk of conventional computer adventure games, Eamon does not involve hours of frustrating trial-and-error while attempting to best-guess the arbitrary words which the program will accept. The Eamon vocabulary is no broader than any other adventure, but the program makes it clear at the outset which words it will accept. The emphasis in Eamon isn't on puzzle solving. Instead the adventurer spends his time exploring a scenario and hacking, zapping or occasionally befriending any creature he encounters en route. All computer adventures seem to be able to trace their roots back to the fantasy role-playing game “Dungeons and Dragons": but with the honourable exception of Sir-Tech's Wizardry, no computer game gives a closer feel to the original "D&D" than Eamon, Most important, in Eamon, as in "D&D", players select a character to adventure on their behalf and that same character - if he survives - can move through any or all of the different scenario disks, There's no need to go into the detailed mechanics of the Eamon game system. The Player Manual was printed in full in Hardcore 2 (March 1981), and it is contained in full on the initial Eamon disk (which you must have in order to be able to run any of the subsequent scenarios). What members may not be aware of however is the sheer size of the Eamon range of scenarios. Because the original system was donated to the public domain and the program is perfectly listable many people have written additional Eamon scenarios. This was of course Brown's rationale in freely donating the system. Although there were only a handful of the =) Ba" Books 9 1-2-3 for Business 32 VisiCale borksheets s Ne Rodel won PROCESSING AMV about 1-2-3 Apple 11 Word Processing 14.35 How to use Supercale Complete Book of word Mastering Vistcale Processing and Graphics 13.45 Suoercale for IRM PC choosing 4 Mord Procer 11,00 SuperCale - The Book word Processing 11190 The Power af Latue The Power of Mult ip The Power of Supercate The Power of VisiPlot The Power of VistCale 1/2 Using 1-2-3, Visleate Dtskcutde Practical Wordstar Uses Mors Processing 22.00 VisiCale for apple 11 Text Processing with UNIT 95 VisiCale for the 18H AC Tenderfoots Guide to Word Processing oe Vistvord «by vistPrecs vorarr Training haa 4 Processing Buyers Nord Praceceing fost * Guide e Executive VisiCele for the tt fe: for the sopie 11 Mord Processing for Saal) Gusinesses Executive Superéate for the Osborne iSeries" by is ord Processing Primer i Hing on # Ricrocomputer Nord Processing In the Modern office Introduction to Vitieale patrtsing Mordstar made Cany (and eds) PREIPTen Rodel for Gus ne Tm ase E35 Shreastheet satbutre fron Nisicate 11.50 saan onoenon ASSOCIATES (14¢'3) ‘i ¥ PORIDGEWAT RAVER este Ne eteinen tintin tients 3 Ewa on nee seveonetanmyncntiann turotrg Faeec ase a) “no for the App SuperCele Home & Off ice Companion SuperCale Supersodele for Business The VisiCalc Aap} ications Book The VisiCale Book VisiCalc for Sctence L Engineering VisiGate Home & Office Conpanton VISICAIE Programe made Easy DATABASES Database Management Systens se TT Users* Gude tedase. 95 Primer Using dBase It Understanding dBase It MISCELLANEOUS TT Ue oo nople /1/ and Eattor de - sce} Introduction, Filer rc Page 30 HARDCORE - The Journal of the British Apple Systems User Group June 1984 many scenario disks in the software library, the number has dramatically increased as | have managed to track down to date a total of more than 40 different scenarios, plus several Eamon utility disks, The quality is inevitably variable; but with the possible exception of the apparently anonymously contributed “House of Ill Repute" (a soft porn variant!) there are no really bad scenarios, and there are many excellent ones. The themes of the various scenarios cover ancient mythology to Darth Vader, with some even set in the present day! Some are deadly seri while others include a touch of humour: "Quest for the Holy Grail" for example is based on the Monty Python film. Several of the scenarios introduce new improvements on the Eamon game system (including a "save" game in progress facility, abbreviated commands etc.), and "Lost Island of Apple" even introduces a non-text mapping element. If you have any new and original scenarios which you are prepared to make publicly available, get in touch with the software librarian. If you want some of the new scenarios or utilities, order them from BASUG as usual. Book Review Title: Build Your Own Expert System Author: Chris Naylor Price: £6.95 Paperback, 8x6 inches, 249 pages Publisher: Sigma Technichal Press Reviewed by Richard Welch If you have been intrigued by magazine and newspaper articles about expert systems and wish you could get to grips with the subject then look no further. Chris Naylor has written a book which cuts through the jargon to present the basic ideas in an easily understandable form. Not only is it packed with common sense and ood advice, it also contains listings of BASIC programs for the Apple II and Sinclair Spectrum microcomputers. There is very little machine-specific code (in the Apple listings especially) and it would be a simple matter to convert them for another machine, These programs should (in my estimation) occupy not more than 32k of RAM, and floppy disk drives are not essential. The first part of the book is designed to get the reader thinking about the whys and wherefores of expert systems, and their limitations. After dispelling any misconceptions the book continues with an introduction to probability theory and Bayes' Theorem. The middle section builds up in stages the framework of an increasingly sophisticated self-learning general purpose expert system. This system modifies its set of judgement rules according to the correctness of its answers. As new topics are introduced and explained they are incorporated in the expert system. Some of these topics are: parallel and sequential decisions, the description space, training the expert system and probability distribution functions. For the reader who wishes to have a ready made expert system with which to impress others, there is a listing of a medical diagnosis program. This uses a Bayesian inference method and asks only the minimum number of questions needed to arrive at the best diagnosis. Towards the end of the book there is a chapter dealing with some of the well established expert systems such as DENDRAL - used for determining chemical structures, and PROSPECTOR ~ used to help find valuable mineral deposits. Another _ interesting chapter discusses the languages to program expert systems. To round off the book there is a summary of the important technical terms, followed by a list of recommended books for’ those who wish to delve more deeply into this field. This book is written in a light, humourous style and the program listings are clearly printed, It should appeal to many hobbyists, especially those still wondering if the answer really is 42. Apologies to those whose review has not appeared in this Issue. We are short of space, but you should see your review in the next issue. ‘Weve taken the risk factor out of buying software! ‘So much so that we are prepared to refund you the full ‘purchase price on any of our Apple software products ‘within 21 days of purchase if you're not satisted with it! At present our range includes: ‘This wi print invoices and generate information. which can also be used by our Bookkeeper packago Invoicer will use standard pre-printed forms But we cis toctade @ smal lity program which wal preprint “forms” on blank paper Invoicer will hold a price list ‘and customer ist in memory (moiking for very fast jon with a core of fast moving lines and good ‘Chstomers) Much larger ists ean ako Be held on aise ems and customers not on existing lists can easily be ‘entered via the keyboard Bookkeeper ‘An essential package tor bookkeeping exercises, ‘ook with analysis by category Ist of Gesiom with dates, ote Rirchave Day Rook with Gnalysis, Cash Book, Pelty Cash Book, home accounts, ‘le. Bookkeeper taker care of VAT calculations and {lows you to ptoduce onaivsed and selected Squres as and when required, ie. management figures, otc access ’A superb Apple I+ le database manager (0d) Access hat all the rower cr versity you wes expec! nan Apple DBM, seguidles of pice Ms also ‘erat forinance twin on indexed tom in ‘Spproximately 3 seconds. ana wil remneve c recora srth complex sets of cntetig, such as AND, wldard Tange wnreh In ics thm Zi anconat Ths package ‘was previously avatlable or £199 9 Inchuding VAT and ‘ver ct that price users found iio ia good wate for money, Payrot This package takes care of Payroll calculations and record keeping for the smaller Pusiness ana will andl au eres our, tax codes including NT No ‘standard. reduced, to 100 employees, a3 well as larger organisations, for confidential proceting Only information which nas JUDGE FOR YOURSELF ‘changed since tne iast Payroll run need be entered. ‘employee, tax and National Insurance details aro Stored The Inland Revenue specications ate followed ‘accurately, no ertor of calculation has ever been Teported since the fis! version of Payroll wars tasted in 1980. Users have found that Payroll very easy to use veh verte Enowiedge of computers or poviell [Dfocedires necessary for successful operatic Bidition tool onucl Coleulations Payroll wal comput (gross pay and deductions fora given net pry Sick Pay) ‘Aftee standing program that will c