Sinclair and the 'Sunrise' Technology: The Deconstruction of a Myth (11 page)

Chris Curry realized that if Science of Cambridge was to reap the advantages of launching a promising product into a virgin market, he was going to have to move fast. Williamson’s prototype had already been seen by a number of other companies, and there was always the chance that someone else would have the guts to snap up a new idea. Curry and Sinclair agreed that it was time to formalize an agreement with Williamson. Sinclair arranged a hotel conference with the engineer and contractual conditions were discussed and agreed. Science of Cambridge would license Williamson’s design, for which the company would pay a flat fee of £5000. The inventor would also receive royalty payments according to the number of units sold. Shortly after his meeting with Sinclair, Williamson received a contract detailing the licensing deal he had agreed at the meeting. This he signed and returned to the King’s Parade offices. All that was required for the deal to be formalized was for Williamson to receive a copy of the contract bearing Sinclair’s signature. It never arrived.

Williamson’s project has been chronicled in detail in an attempt to clarify the circumstances in which his prototype was produced and to facilitate an informed assessment of subsequent developments. By now the reader should be able to appreciate that Williamson was essentially selling an idea whose commercial value was bolstered by the existence of a working design based around an American chip. The fact that the prototype made use of Sinclair components is almost irrelevant, apart from showing yet again Sinclair’s recycling instincts and the fact that they were almost certainly the cheapest available components. Considering the relationship that existed between the two companies, Williamson is hardly likely to have argued that it was of any benefit to Science of Cambridge that the production of his machine required the purchase of components that might possibly be held in quantity by Sinclair Radionics. He was selling an idea for a product whose economic and technical viability was confirmed by the prototypes he built and demonstrated. Clearly any company that decided to market Williamson’s design would first take the precaution of costing each of the product’s components, and at the same time shop around for cheaper alternatives. Obviously, if a selective substitution of components could reduce production costs, it would be acceptable for the manufacturer to request the designer to make any reasonable modifications required by the inclusion of the new components. Such situations are common enough to merit a standard clause in most licensing agreements.

Such considerations hardly seemed relevant in the case of Williamson’s machine. There were only two elements of the hardware design whose replacement would effectively invalidate the brilliant software that established the uniqueness of Williamson’s work. The first of these was the Scamp chip itself, which defined the machine’s limitations and the way in which it could be programmed. The replacement of this chip would effectively mean the creation of an entirely different machine tailored to the idiosyncrasies of a new microprocessor. The second hardware feature on which the rest of the design depended was the inclusion of a standard calculator keyboard for the communication of hexadecimal input. As we have already mentioned, Williamson laboured over almost insurmountable problems in his determination to use an existing keyboard in his design. Ever conscious of production costs, he reasoned that a unique keyboard would require a unique production process. It would surely be cheaper to take advantage of an existing process that was churning out what could become a dual-purpose component. So a software design based around octal notation was eventually programmed on to the ROM (read-only memory) chips, which allowed hexadecimal instructions to be communicated via a standard Radionics calculator keyboard. Any decision to change the keyboard would mean the design of an entirely new input/output (I/O) program.

Williamson sat around waiting for the contract that was never sent. Finally, he received a phone call from an uncharacteristically edgy Curry. With a profusion of apologies, Christopher gallantly faced up to his responsibilities as the spokesperson for Science of Cambridge, and informed Williamson that the deal was off unless he was prepared to modify his original design. The engineer made it clear that he was prepared to consider any reasonable changes that Sinclair and Curry deemed necessary. It was explained that the calculator keyboard was to be replaced by one that made use of an economical membrane design, and that Williamson’s I/O software would have to be modified accordingly. Williamson explained why such changes were impossible since they demolished the foundations on which the machine had been designed.

Science of Cambridge had changed its plans in the light of an offer from National Semiconductors (NS), the firm that manufactured the Scamp chip at the heart of Williamson’s design. With a thoroughness to tempt tedium, we have established that because of the circumstances in which it was produced, Williamson’s prototype made use of components that originated from a variety of sources but which, for the most part, were unexceptional enough to be replaceable when necessity demanded the application of a little creative substitution. When Curry approached National Semiconductors to negotiate the purchase of the first batch of Scamp chips, the Americans turned round and made the latent millionaire an offer he couldn’t refuse. They offered to redesign the kit for the Brits free of charge. The idea was that the capabilities of Williamson’s prototype would be realized in a product built exclusively from National Semiconductor chips.

Such a deal made obvious business sense to both parties. The Americans would be able to sell a product that would succeed only to the extent to which it was used - which in the case of the Scamp chip was hardly at all. In addition, although the primitive microprocessor had turned out to be a lost cause with the folks back home, there was a chance that it might catch on in Britain if adopted by high-profile whizzkids like the boys at Science of Cambridge. Finally, a redesigned product which made exclusive use of National Semiconductor components multiplied the original unit sales offered by the Williamson design by the number of components required in the NS design. All it would cost NS was the price of a new design.

The final National kit used the same processor, the same instruction set if you like, the same display and had the same memory size. Apart from that the hardware design was not similar because National tailored it to the products they had available. They already sold kits over here and in the US. You have to remember that my idea had been to make it as cheap as possible — right down to using a Cambridge [calculator] keyboard. (Ian Williamson, Interview, 29 October 1985.)

The attraction of the NS deal for Sinclair’s company should be obvious. The logistics of manufacturing a product that requires components from a variety of sources are often unworkably complex. As the Sinclair team had learned from bitter experience, it needed only a single source to foul up and the entire production process could be brought to a standstill. Although delays in fulfilling orders were by now regarded as almost a defining quality of any Sinclair operation, any situation that diminished the risks of such difficulties would have been embraced with enthusiasm. It would clearly be easier to control the supply of components from a single source than attempt to plan production according to predictions concerning the reliability or otherwise of a large number of small suppliers. It would be reasonable to assume that the NS deal also provided Curry with an opportunity to negotiate a cut-price deal, since it could be argued that Science of Cambridge would be purchasing a large part of the American company’s range in the kind of quantities that justified job-lot rates. Finally, Chris Curry has something of a reputation for his ability to bank the fruits of a competitor’s weakness, and is unlikely to have pulled his punches when negotiating a deal that centred on a chip he knew National Semiconductors was desperate to promote in the UK.

As far as Science of Cambridge was concerned, the NS offer provided the company with an irresistible opportunity to make a fast and unusually respectable return on a relatively modest outlay. The only thing standing in the way of a straightforward deal was Ian Williamson. At this stage in the proceedings it would have been tempting to push to the back of the mind the fact that it was Williamson who had initiated the project and brought it to fruition.

From the security of life near the top of a national institution, Ian Williamson clearly feels that he can afford to be generous when looking back on his life in the Cambridge of the 1970s. These days he’s sufficiently sanguine to be able to dismiss his experiences as a young inventor and aspiring entrepreneur as simply the first of many tough lessons that make up a commercial education. It’s clear that Williamson has never regretted his decision to forsake the life of the inventor, which under the circumstances is hardly surprising.

Williamson insists that when Sinclair backed off from the deal they had agreed in favour of the NS option, his bitterness was primarily directed at the American company rather than the management of Science of Cambridge:

I was never particularly pissed off with Clive. If anything I think I was pissed off with National Semiconductors. You see, considering what NS were offering and the state of the company [Science of Cambridge] at the time, they [Sinclair/ Curry] were faced with an offer they couldn’t really refuse. I was pissed off with NS for having the sense to offer Sinclair a design which only used components which could be ordered from one place. (Interview, 29 October 1985.)

In June 1978 Science of Cambridge launched a microcomputer kit based around the National SC/MP chip. The machine was marketed as the MK14, and the features it offered were identical to those of the prototype Ian Williamson had demonstrated to Chris Curry the previous summer. In the words of Williamson, the launch of the machine and the familiar post-natal depression that followed bore all the hallmarks of a ‘typical Sinclair flop’. An inability to fulfil the first batch of orders inspired by the launch was, as usual, compounded by the effects of the extravagant advertising campaign by which it was preceded. Science of Cambridge had instructed National Semiconductors to produce a mere 2000 sets of components for the launch of the machine. It seems likely that this conservative launch stock was partly a reflection of the financial plight of the new company, and partly an indication of Sinclair’s doubts about the potential of the home-computer market. In any case, the punters’ donations flowed in to fill the corporate coffers, but little in the way of product flowed out of the company’s doors. In the age-old tradition of a Sinclair launch, supply immediately collapsed from the shock of attracting an encouraging level of public interest. Anyway, the first 2000 MK14 kits were no sooner packed than dispatched, after which eager customers were invited to kick their heels while the NS manufacturing machine was brought back to life, funded by the punters yet again.

In its short, sweet life as the UK’s first cheap computer kit, the MK14 design was sold into between 10,000 and 15,000 homes. The rights of the manual Williamson had written for his machine were bought by Science of Cambridge and the book included as part of the MK14 package. It is clear that the company’s principals were less than comfortable about Williamson’s reward for his labours and enterprise, since Sinclair felt obliged to fork out the princely sum of £2000 for the right to use the engineer’s documentation.

The unexpected success of the MK14 directly influenced product development at both Radionics and Science of Cambridge. At the former, it stimulated the development and design work on a more sophisticated home computer, a decision that, many years later, resulted in the appearance of the ill-fated NewBrain, and rather sooner, the ZX80. The Science of Cambridge followthrough was considerably more modest. The company churned out a small range of products that enabled the hobbyist to upgrade the basic MK14. There was the VDU Module (£33.75), which enabled sixteen lines of thirty-two characters to be displayed on a UHF television. To this was added the Cassette Interface Module (which allowed programmers to save programs to magnetic tape), the Prom Programmer, and a power supply to drive the expanded system.

According to the first editor of Personal Computer World magazine, Dave Tebbutt, the success of the M K14 was one of the major inspirations behind the decision to launch the UK’s first home-computing magazine. It was undoubtedly also a contributory factor in Chris Curry setting up Acorn Computers some time in 1978, thus continuing the entrepreneurial splintering that his mentor Sinclair had taught him. Indeed the first Acorn product, the System 75, was remarkably akin to an enhanced MK14.

There was a more direct sense in which the MK14 hinted at the shape of things to come. The early hobbyist who cobbled together the complete M K14 system had before him the heart of the kit’s legendary successor. The Scamp was ditched and replaced by the Z80 A chip; the easier BASIC language replaced hexadecimal numeric code and provided programmers with a less cumbersome method of communicating with their machines. But such factors aside, in every expanded MK14 the essentials of a ZX80 lurked, waiting to take the world by storm. Well, almost...

[5] THE ZX80: A NEW MEANS TO AN OLD END

One of the unshakeable tenets of microcomputing mythology insists that Sinclair’s ZX80 computer kit established the parameters of price, computing capacity and marketing stance that shaped the entire first generation of home micros. There are many who regard this as a laudable achievement, while others, like respected computer journalist and author David Ahl, refuse to regard Sinclair’s influence as a necessary good: ‘
Sinclair ZX80: with an unusable keyboard and a quirky BASIC, this machine discouraged millions of people from ever buying another computer
’ (Personal Computer World, October 1985.)

Although none would dispute the machine’s limitations, the ZX80’s launch is nevertheless regarded as heralding a new era of consumer electronics. This is not to say that the Sinclair machine was by any means the first personal computer to be marketed in the UK. In 1979, a number of US products were providing the foundations for a growing number of retailers specializing in computing equipment. For example, Tandy’s TRS-80 was on offer for just under £500, while bargain-hunters could pick up a Commodore PET for around £450. However, high prices ensured that the emphasis was on ‘personal’ rather than ‘home’ micros, and the majority of imports was aimed at the business and academic market.