Sinclair and the 'Sunrise' Technology: The Deconstruction of a Myth (13 page)

Although condemned by reviewers and users alike, savings in manufacturing costs ensured that the membrane design would become a consistent feature of the ZX range, modified but never abandoned. According to Tony Tebby, one of the QL’s designers, the development of the Quantum Leap machine three years later saw an inexplicable decision to stick with the unpopular, non-standard design which was clung on to with a determination verging on perversity. Given that the QL was to be marketed as a business machine and should thus offer a quality keyboard suitable for wordprocessing, it seems incredible that the company should have abandoned alternatives in favour of an enhanced, but still membrane design. A number of Sinclair’s R&D team have indicated that Sir Clive himself vetoed the use of standard keyboards, insisting that the membrane design was an intrinsic part of the image of a Sinclair computer. Whatever the reasons, apart from initial savings in manufacturing costs to Sinclair Research, the only group to benefit from this policy has been the peripheral manufacturers, who have provided a wide range of alternative keyboards.

Although the basic ZX80 was relatively cheap, the realities of such economy can be judged only in the light of the final cost of a halfway useful machine. One of the most significant weaknesses of the basic computer was that it arrived with only IK of RAM - not nearly enough memory to enable owners to write serious programs. With this in mind, Sinclair developed and marketed memory expansion peripherals, which became available just after the launch of the kit. By mid-1980, Chris Curry had co-ordinated the launch of Acorn Computer’s second product, the Acorn Atom microcomputer. The assembled version of the machine was priced £330 and equipped with 12 K of RAM. Now although at first glance the Sinclair product appears to offer a far better deal, such impressions are dispelled when the ZX80 is upgraded to meet the Atom’s modest memory specifications. The costing is as follows:

1 assembled ZX80 (with 1K RAM) £99.95

1 power supply £8.95

1 1K memory chips @ £16.00 £176.00

4 memory boards @ £12.00 £48.00

total £332.90

Apart from offering a larger memory than the ZX80, the Atom also arrived with a full-sized keyboard, a floating-point maths package (as opposed to the integer-only capacity of the ZX80), a potential for colour graphics and sound capabilities. Thus although the ZX80 was notionally cheap, it was also only notionally a computer when one considers the capabilities of its rivals.

When compared with subsequent Sinclair development programmes, that of the ZX80 appears smooth, fast and relatively untroubled. However, it must be remembered that although work on the ZX80 started only in May 1979, the development of the Radionics computer that would become the NewBrain had been progressing since July 1978. Since the R&D work on both machines was carried out by much the same personnel, it’s clear that the ZX80 programme benefited from a year of Radionics experience. The incestuousness of the two projects is highlighted by the fact that today few of the engineers involved seem capable of remembering who they were working for during this period.

In April 1979, Clive Sinclair arranged a meeting with John Grant of Nine Tiles to discuss the development of the successor to the MK14 kit. At that time, Grant was working on the software for the Radionics computer, and Sinclair made it clear that the new Science of Cambridge project would be shaped by many of the decisions defining the machine that would later become the NewBrain programme. The Radionics machine was designed around the Z80A microprocessor (as opposed to the MK14’s SC/MP), and because of the development team’s familiarity with the chip it was decided to use the Z80A in the Science of Cambridge machine. Grant recalls that Sinclair’s brief to Nine Tiles was mainly concerned with ensuring that software development was tailored to the limitations imposed by the components he had selected. At all times the design of the ZX80 was driven by the goal of producing a computer that broke the £100 barrier, yet still returned a comfortable profit. The product’s capabilities were of secondary importance.

It’s clear that Grant’s involvement with the development of the ZX80 was not inspired by expectations of significant financial gain. The feeling at Nine Tiles was that the creation of a mass-market microcomputer was in itself an exciting project, and one with which the company was interested in being associated. Given that the estimated R&D costs for the entire ZX80 project are generally agreed to have just about reached the five-figure mark, Grant’s cut of such a budget would have offered his company little more than pin money. Once again, it’s worth emphasizing that a major percentage of the ‘creative’ design work on this new Sinclair product was not performed by the company itself, but contracted out and defined by an unusually nebulous brief.

The month following Sinclair’s initial meeting with Nine Tiles was particularly fraught for him, since in May 1979 the NEB announced its plans to sell off Radionics’ television and calculator interests. While Sinclair was busy penning his resignation from Radionics, work started at Nine Tiles on the development of the ZX80 software. It has often been argued that the success of the early Sinclair machines played an important role in establishing BASIC as the resident computer language for the majority of home computers. In retrospect, this situation has been regretted by many in the industry, since it is generally agreed that BASIC is only moderately successful as a learning tool and positively obstructive to the development of serious programs. However, because BASIC had been selected as the Radionics computer’s resident language and full documentation was readily to hand for the ANSI Minimal BASIC dialect, Sinclair instructed Nine Tiles to prepare a similar implementation for the Science of Cambridge machine. Grant remembers suggesting that a more flexible language such as Forth might offer more progressive facilities to the new programmer, but, since such an approach would have required a longer development programme, the possibility was never seriously considered.

Nine Tiles’s work on the ZX80’s software is generally hailed as a triumph of ingenuity over primitive resources. Given that Grant and his team had only 4K of ROM into which to squeeze the machine’s operating system, editor and BASIC interpreter, the product of their labours set new standards for concise programming. Another unusual quality of the company’s work was that it was completed more or less on time, an event almost without precedent in the world of R&D. The bulk of the ROM was written in the months of June and July, but the resultant code required 5K for its storage. Thus August 1979 was spent trimming the code to fit the ZX80’s4K ROM restrictions.

While the ZX80’s software development can be chronicled in detail, a shroud of mystery hangs over the design of its hardware. Even at the time, Grant recalls a ‘cloak-and-dagger’ aura to everything associated with the machine’s hardware. One possible version of the story is that Mike Wakefield, who at the time was working for Newbury to design the NewBrain’s hardware, may have assisted Science of Cambridge. The hardware was not completed by the August deadline. Some participants suggest that Wakefield simply hadn’t managed to design and build the circuits required, others that Newbury was threatening to cause trouble over his participation in the Sinclair project. All sources appear to agree that, by the end of August 1979, the ZX80’s hardware had been handed over to the redoubtable Jim Westwood who, reliable as ever, finalized the work by the end of October.

The events of 1980 must have been extremely gratifying to Clive Sinclair. Having resigned from Radionics in July 1979, Sinclair took the £10,000 golden handshake offered by the NEB and concentrated all his efforts on carving out a future for his new enterprise. By August, still desperately short of working capital, he reluctantly parted with his vintage Rolls-Royce and sold his house. Undoubtedly, the autumn of 1979 must be regarded as one of many make-or-break points of Sinclair’s career.

The launch of the ZX80 heralded a turning point in Sinclair’s fortunes. In the eight months following the first appearance of the kit at the Wembley computer fair, Science of Cambridge sold 20,000 units into a virgin market. Having decided that in-house production had led to an overall inflexibility at Radionics, with the ZX80 Sinclair initiated a policy of ‘subcontracting everything that could be subcontracted’. The early machines were put together by a small electronics company in St Ives, but before long production was shifted to the Timex factory in Dundee. This move to Scotland marked the beginning of what would become an important relationship between Sinclair and the American watch manufacturer, a cooperation that would prove more enduring than Sinclair’s commitment to home computers.

The contract between Timex and Science of Cambridge was the realization of a significant new strategy for Fred Olsen, the ‘Norwegian Howard Hughes’ and the tycoon behind the privately owned and intensely secretive Timex corporation. Myron Magnet, writing for Fortune magazine (8 March 1982), explained the problems facing the Connecticut-based company at the beginning of the 1980s:

Timex fell behind technologically as watches became digital in the seventies: unit sales stagnated, market share declined, and profits dwindled to virtually nothing by 1979. So Olsen has reason to diversify out of the mechanical watch business that has long been Timex’s mainspring.

The arrangement worked well for both companies. Although in 1980-81 the production of the Sinclair machines could hardly have generated enough revenue radically to improve the crisis at Timex, by 1982 the relationship between the two companies had reached the point where Sinclair technology was to be licensed by Timex and marketed under the watchmaker’s name in North America. Although the deal turned out to be a disaster for Olson’s company, Sinclair s comments at the time underline the importance he placed on the link that was forged with the ZX80’s production:

I think that Timex will be making more money out of computers than watches within the next five years ... It will be a $1-billion-a-year business for them and $50-million-a-year for us. (ibid.)

The lack of resources at St Ives and the production delays incurred with the shift to the Timex plant in Dundee ensured that the public suffered the usual delays associated with a Sinclair launch. As an early example of the type of complaint against Sinclair that would soon become a standard feature in the letters pages of the computing press, we’ll take the case of D. J. Harper. Clearly the kind of hi-tech enthusiast the company should have been courting rather than ostracizing, the youthful Harper dispatched his cheque to Science of Cambridge in February 1980 and heard nothing for five months. Although Harper was unnaturally patient, Sinclair’s announcement from the Las Vegas Consumer Electronics Show that Science of Cambridge were to market the ZX80 in the States proved too much to take. A copy of his plaintive letter was sent to Computing magazine:

Dear Mr Sinclair,

I was interested to see that you intend marketing the ZX80 in America. Perhaps before attempting to deliver to the States you could try delivering to Colchester, Essex, UK, a distance of 40 miles instead of 4000 ...

Aside from delivery problems, by the standards of subsequent Sinclair launches that of the ZX80 was relatively trouble-free. Primarily because of the relatively simple hardware design and the efforts of an unusually conscientious software team, the percentage of returns was the lowest of virtually any Sinclair product - official sources put the figure at around 1 per cent. By and large, the first purchasers of the ZX range seemed to have been satisfied with their investment in the new technology. There were occasional complaints about the keyboard, and everybody agreed that the ZX80, like its immediate successors, was prone to overheating. Author Tim Hartness, in nostalgic mood, recalls:

I’m not sure that programming will ever be like the ‘old days’. It may sound silly, but I used to enjoy finding out about the ZX80 while I balanced a frozen milk carton precariously on top to cool it down! (Your Spectrum, May 1985.)

In September 1980, Science of Cambridge released a 16K RAM pack which enabled owners to significantly expand their machine’s memory without the problems and expense of multiple chips and expansion boards. At £49.95 the RAM pack was considerably better value than the earlier expansion options, but unfortunately its decidedly clumsy design generated a new variety of problem. Although the new peripheral simply plugged into the back of the ZX80, thus avoiding the tortuous and unstable construction required by the earlier option, the RAM pack was dangerously top-heavy and had a habit of falling out of its socket. As far as the user was concerned, this failure of design was disastrous, since hours of programming could be lost if the RAM pack chose the wrong moment to break loose. However, even in the infancy of the micro boom Sinclair’s customers proved themselves to be both tolerant and resourceful. They resigned themselves to the inevitable, and solved the problem with unsightly gobs of Blu-Tack or chewing-gum.

Enjoying the advantages of little or no competition, consumer tolerance born of the pioneering spirit of the times and a general ignorance about what to expect from a computer, Sinclair’s company was able to emerge unscathed and in profit despite the unnecessary delays and the thoughtless design of the ZX80. Such conditions prevailed, and protected Sinclair Research, up to and beyond the launch of the Spectrum. While initially enabling the company to consolidate its domination of the market - and at the same time encourage low standards within the home-computer industry as a whole - the enormous success of the ZX range encouraged Sinclair to believe that the company had a God-given right to treat its customers in a manner that would have spelt commercial suicide for a manufacturer in any other industry. Reflecting on the declining fortunes of the Sinclair empire, computer journalist David Ahl made the following prophecy about the logical consequences of such policies:

Sinclair products are highly innovative, interesting and cheesy. In the long run, the lack of quality and utility, and a cavalier approach to customers, will spell doom for the company. (Personal Computer World, October 1985.)