AS THE MARKET BECOMES MORE SOPHISTICATED, AUTHORS HAVE HAD TO CATER FOR THE NEEDS OF NEW GROUPS OF USERS. IF HACKING IS YOUR PASTIME OR MACHINE CODE IS YOUR STAPLE DIET THEN THERE ARE BOOKS FOR YOU. ON THE OTHER HAND, IF YOU PREFER A GENTLE INTRODUCTION TO BASIC OR SUPERBASIC YOU WILL FIND THAT WHOLE SHELVES HAVE BEEN DEDICATED TO YOUR NEEDS. JOHN GILBERT SORTS THROUGH THE MOUNTAIN OF COMPUTER LITERATURE WHICH HAS KEPT USERS HAPPY DURING THE PAST YEAR.

Although few authors have produced computer books which have reached the best seller lists, technical anthors must be amongst the most prolific writers in the world. A literary mountain was produced in 1985 alone, in order to further the knowledge of Sinclair computer owners and the popularity of Sinclair machines.

One author who almost reached the non-fiction best seller lists was Hugo Cornwall with The Hacker's Handbook, a slim but explosive book from Century Hutchinson. Hugo Cornwall still remains an enigma, even his name is a pseudonym.

The Hacker's Handbook is a skillful mixture of personal experience, stories from secondary sources and practical information. Indeed, the way in which the book has been written leads me to suspect that Cornwall is a journalist as well as a hacker, or that the book was ghost written by someone from Century Hutchinson.

The practical information in Cornwall's book nearly landed him - her? - in trouble. It begins harmlessly enough with an introduction to innovations in communications technology. Once past that, and some innocuous remarks about his programming experience, Cornwall starts getting technical - and goes way beyond the definition of baud rate.

Although Cornwall gives information on communication protocols and how to go about finding passwords for various types of system, he is careful to point out that hacking is done more as a sport than as a criminal action. If it were more than that, banks would be defrauded daily, and our security installations would be broken into every evening.

The same point is made in a book written by Geoff Wheelwright and Ian Scales, published by Longman. It too is called The Hacker's Handbook, although it takes a slightly different angle on the subject.

The authors have been careful to steer away from controversy. They deal specifically With the Spectrum and how it can be connected to systems such as Micronet 800 and Prestel. An overview is given of both systems together with information on how to connect your computer to a VTX 5000 modem.

The final section of the book takes a look at electronic mail and bulletin boards. Telephone numbers are given to allow access to boards which are open to the public. Although Wheelwright and Scales are not as revealing as Cornwall about the secret world of the hacker, their introduction to the subject is practical and interesting.

Another subject of major interest to the computing public, especially QL owners, has been the secrets contained within the QDOS operating system which has been so highly vaunted by several books. There are no fewer than three titles on the market which jostle for acclaim as the best book about QDOS. They are The Advanced QL User's Guide by Adrian Dickens, from Adder; The QDOS Companion by Andrew Pennel, from Sunshine; and The QL Technical Guide by Tony Tebby and David Karlin, from Sinclair Research.

The books from Dickens and Pennel are very similar in structure. The first to appear was the one from Adder. Dickens managed to obtain the QDOS documentation from Sinclair Research and put together a book using it. The result is, none the less, impressive and, despite the mistakes, provided an introduction to the 68008 chip and operating system. Each QDOS trap was listed and examples of its uses were given.

As well as discussing memory management, graphics, sound and file processing, Dickens also included a chapter in which he showed how SuperBasic worked. Detailed appendices showing the memory map, full 68000 instruction set and microdrive formats were also provided.

Pennel's book differs only slightly from the advanced guide. It arrived on the market four months after Dickens' book and the result was a text full of examples which had obviously been created by the author.

The companion is structured as a graded introduction to multi-tasking, Input/Output, device drives and the QDOS utilities. Information is also included on the 8049 second processor, which acts as a control for the keyboard and sound, as well as an explanation of how to create external ROMS.

The third book, from Sinclair Research, resembles the QL Advanced User Guide but was rather a disappointment. It provides all the information that you would require to program QDOS but its style is clinical, with almost no examples to illustrate the wealth of knowledge given by the authors. It is a humourless read and, even as a reference book, lacks depth. The one or two paragraphs used to describe each trap, exception or interrupt are barely enough. Tebby and Karlin could at least have included a section detailing an example of QDOS in action. As it stands, the authors expect readers to share their competence - not a healthy assumption for any technical writer to make.

The Psion business packages, bundled with the QL, did not escape the attentions of authors last year. Three companies brought out books which cater for the business or home user who wants to put Quill, Archive, Abacus or Easel to work.

Hutchinson was the first onto the scene with a series of QL handbooks. The series contained one title aimed at Quill owners and one aimed at Archive owners.

Word Processing with the Sinclair QL, by Mike O'Reilly, is competently written, although it is long-winded and says little. It shows how to power up the QL, how to load in Quill and how each of the commands work. It even goes as far as to show how a letter can be written and reports drafted. Unfortunately it does not go much further.

Database Management on the Sinclair QL, also by Mike O'Reilly and also from Hutchinson, is much the same as the introduction to Quill. It does little better than the official Sinclair manual, a damning fault, as that manual leaves a lot to be desired.

The Hutchinson books are not the best on the Psion packages, and it is left to the series of Psion books from Century Communications to fit neatly into that niche. There are four books; one for each of the packages. The book on Quill, written by Clare Spottiswood, marked the series as a winner. Even Sir Clive Sinclair stuck his oar in and said that it was one of the most fab things he had ever seen. The simple style, combined with diagrams, cartoons and many useful examples, makes the book a joy to read.

Other titles in the series are similar in style but have different authors. QL Archive, by Ian Murray, provides the easiest and most interesting explanation of the database package ever committed to print. It is a delight to read, and the examples provided can be practically useful. Although it arrived late on the scene it should be very successful.

Sunshine also provided a book on the Psion packages, although it combined all four and was different in approach from the titles from Hutchinson and Century. Quill, Easel, Archive and Abacus on the Sinclair QL, by Alison McCallum-Varey, may be a bit of a mouthful, but it achieves its aim easily and concisely. The book Shows how the four packages can be used together in a business environment. It also shows how to import and export information from them and which of the programs are compatible for each operation.

It was the first book to include comprehensive examples which were worth entering into the Psion packages and which stretched them to their limits. It also showed that the author had an understanding of the QL, while her colleagues in the computer publishing industry showed only their ignorance of the subject.

To be fair, Sinclair Research was to blame for much of the ignorance which authors exhibited after the launch of the QL. Boris Allan is the classic example of a writer so keen to get a book out about the new machine that he forgot the cardinal rule. Wait until a finished product arrives. His book, The QL Companion from Pitman can still be found on the bookshelves of WH Smith and Boots.

Anybody who buys it in the hope of learning something new about SuperBasic will be disappointed. It was written with the aid of a pre-production manual. That would have been alright if Sinclair had not decided to change SuperBasic. The language has gone through three transformations so far.

He is not alone, however. Other authors and publishers have made the same mistake. One such disaster was the QL User Guide, written by Lionel Fleetwood and published by Sigma Press. The author's object was to produce a book about SuperBasic and the Psion bundled packages. He also took his information from the manual and most of his examples were approximately half a page long - or one page when he was at his most impressive.

The best was yet to come. Fleetwood had obviously intended to write a section on the 68008 chip and QDOS but never quite got around to it. His publisher obviously expected it, as early copies of the book had a slip of paper stuck over part of the back cover. It hid the publisher's blurb about the 68008 section which was to have appeared in the book.

Hutchinson also tried its best to go to the rescue of knowledge-parched SuperBasic users. The company managed to produce a series of five books in what seems to be a record time of two months.

The first in the series was Introducing the Sinclair QL by Garry Marshall. It gave simple explanations on how to plug in and switch on the machine, followed by a resume of SuperBasic and an equally short section on the four Psion packages.

The other five books - mentioned at the end of this article - ran along similar lines. The authors wrote about the same subjects, but in a different style.

Fortunately, Hutchinson boosted its reputation for QL books by launching another five books, six months after the first part of its series. They dealt with graphics, machine code, business uses and database management.

Machine Code Programming on the Sinclair QL, by Martin Gandoff was the best of the five titles. It provided one of the most readable introductions to the art of 16-bit 68000 programming. Unlike many of the other books which tried to cover the subject, such as QL Machine Code from Melbourne House, it dealt specifically with QL hardware and software. The most important chapter of Gandoff's book was the section on exception processing. It showed how to invoke QDOS routines and was followed by an explanation of multi~ tasking which even a newcomer to machine code programming could grasp. It is one of the most outstanding books of 1985.

Using Graphics on the Sinclair QL, by prolific and blunt-styled Garry Marshall, provided another exceptional contribution to the QL book scene. The text deals only with SuperBasic graphics commands but it does explain the different types of screen co-ordinates, windows and scaling windows. The examples provided are still worth running and provide some of the simplest but most effective graphics displays you can obtain on the QL.

The other three titles tied the series of ten books together. Profiting from the Sinclair QL, by Barry Miles, showed how to use Abacus and Database Management on the Sinclair QL - Mike O'Reilly performed the same service for Archive. Neither of those two titles expanded much on the manual but they did manage to score points on style. The Psion manuals for both packages were sketchy and uninspired.

Finally, Making the Most of the Sinclair QL, by Dick Meadows, was a repeat of introducing the Sinclair QL with a few more examples.

Unfortunately, the enthusiasm which publishers had for the QL did not pay off. Only 50,000 QLs were sold worldwide during the year of its launch. As a result, the QL publishing market is not a happy place at the moment.

One area which is looking healthy is that of artificial intelligence. The most notable contribution for the Spectrum came from Keith and Steven Brain - no puns please.

They showed techniques by which the computer could learn simple games, understand English sentence input and construct intelligent replies. Most of the information described algorithm methods and not heuristics. That means the programs which they developed in the book use a method which the computer steps through every time. An heuristic is a method by which the computer learns from its mistakes by trial and error, and a little more about such techniques would have been appreciated.

One book which did deal with heuristics was The Creative Computer by Donald Michie and Rory Johnston. It is a general computer book which publisher Penguin/Viking pointed out was unlike anything which had ever been printed. That pronouncement sent reviewers into a flurry trying to obtain a copy.

The premise of the book is that it is possible for computers to take in data and, from that information, produce new data. The authors show examples of intelligent systems, most of which accept knowledge from professionals such as doctors or architects. Those computers can then use that information to diagnose a disease or even find an oil field.

It becomes obvious that all Michie and Johnston are talking about are relational databases, commonly called expert systems.

Nobody wrote a book on true artificial intelligence last year, even though some authors tried to disguise their books as such. True AI is bound up with all sorts of philosophical questions, and consciousness must also be linked with the subject.

Trends within the computer book publishing industry are similar to those within the field of artificial intelligence. Few people have said anything worth listening to. The trend in formula books started in 1984 and continued into 1985. It is still with us and, as a result, computer authors have turned out copies of each other's books, texts which say the same thing in a different way.

Unfortunately, for both publishers and authors the public has caught on. It is no longer willing to put up with a standard of publication which is no better, and sometimes worse, than the pulp fiction of the 1920s.

If both the computer industry and book publishing industry are going to survive then authors had better start coming up with new ideas. There are many fields which have not been covered this year. They include the use of computers as control devices, how to go about connecting your machine to the outside world using telephone or radio links, or even the new graphics science of Fractals. None of those subjects was covered last year, but each expands the use of the Spectrum and QL and the interests of their owners. It is about time publishers thought less about making money on formula books, and more about continuing the interest of those who use computers.