Back in the old days that were so very good for IBM Corp, IBM-watching was a fairly straightforward activity once one had learned the rules: whenever one received a piece of intelligence or gossip about IBM’s plans, one passed it through a simple verification test: does it makes sense, is there an IBM precedent for it, would the top brass allow it, have there already been hints along those lines. If it passed all those tests, one could go public with it with reasonable confidence. If it failed one or two of the tests, but passed the others, then it was likely to be a piece of wishful thinking or deliberate disinformation put out by one of the groups within the company that was relatively low in the pecking order, and which was regularly squashed by the old Data Processing Division or its successors. It may seem amazing now, but the General Systems Division, which was responsible for the predecessors of the AS/400 and System/36, and such sundry other items as the Series/1 minicomputer, was one of the parts of the company that was regularly squashed by the all-powerful overweening mainframe group. Then it all fell apart: John Akers embarked on his root-and-branch reforms, and suddenly nothing was predictable any more, none of the old tests worked, and IBM watchers were left groping in the dark, waiting for the company to settle down again so that they could start formulating new rules.

Dead stupid

That process is far from complete yet, but one throwback to the old days has surprisingly returned under the shining new Louis Gerstner regime: factional in-fighting. Our reference in the headline to somebody up there is not intended to refer to the Almighty, even though He may well have His doubts, but rather to one of the Armonkeys that still get the final say on every development. It is easy to believe that Mr Gerstner and his cronies from the RJR Nabisco and American Express Co days have only a weak comprehension of the computer industry, but they would have to be dead stupid not to realise that to launch a new product when it is already smeared with the tar of failure is a certain way to kill it stone dead – and that is exactly what Richard Thomain appears to be about to do with the Power Personal Systems. They will be launched on October 18 sans anything that could be called an attractive personal computer operating system – that won’t be ready until the middle of next year. And for once we don’t have to stick our necks out with cries that the Emperor has no clothes – analysts have done it for us already. PowerPC systems running AIX and Windows NT might appeal to limited numbers of workstation users attracted mostly by the low prices – starting at $2,500, analysts told Infoworld – but they will not appeal to personal computer users. Instead, they will be desktop boxes running server and workstation operating systems – a combination sure to bomb in the market, analysts told the paper.

Destined to fail

If IBM were to announce a PowerPC-based desktop without a mainstream operating system, we feel that product would be destined to fail, Randy Giusto, an analyst with BIS Strategic Decisions, in Norwell, Massachusetts told the paper. Who would buy a PowerPC-based system to run 16-bit-based applications under NT [at 33MHz] 80486 performance? asked John Dunkle, president of Workgroup Technologies Inc, a research firm in Hampton, New Hampshire. You don’t get second chances, said Richard Zwetchkenbaum, an analyst with International Data Corp, in Framingham. If IBM comes out with its first [PowerPC personal computer] product and it’s labelled a loser in the first few months, they may never be able to re-establish their momentum. There is copious evidence for that view – Series/1 may have eventually made back IBM’s investment in it, but it was always regarded as a poor, crippled, misshapen little thing, and the 8100 may have been widely used at its demise, but it never fulfilled its intended role and was disparaged throughout its life. Analysts are suggesting a low-key launch with the company stating clearly that the mach

ine is not expected to sell beyond a few enthusiasts until the desktop operating system is ready, but so much pre-publicity and hype surrounds the machines that that will be enough to create a whiff of failure over it on its own. The only way to save what looks a desperate situation would be to hold the launch until OS/2 for PowerPC is ready, and then launch configurations originally intended to sell for $3,000 at $1,000: at the time of launch, they must be configured with everything a user expects in a $3,000 Pentium machine and be priced not a little cheaper but dramatically cheaper, and treated by IBM as an investment in the future. But enough of the old IBM survives that it will never do that: if it wasn’t prepared to build a System/36 CPU into a low-end AS/400 box and price it at the baby AS/400 price in order to win over all those 36 users that were dragging their heels, it will never try to win over Windows users with loss-leader pricing on PowerPC boxes. And at that point, the only logical conclusion to be drawn is that one of the Armonkeys thinks the idea of building personal computers around the PowerPC was always a rotten one, and is determined to see the things killed off as soon as possible, so as to prevent IBM investing the same kind of large fortune that has shored up OS/2 in developing its new concept in personal computing further. Whoever this eminence is, he or she is getting plenty of help from senior vice-president and head of the Personal Computer Co Rick Thomain, who has been telling Louise Kehoe of the Financial Times that while PowerPC technology has capabilities that don’t exist in the current technology in terms of speed, there are obvious negatives for PowerPC because it does not have an established base of application software, and indicated that in contrast to the enormous preview splash at Comdex/Fall in Las Vegas last year, the Power Personals will not now be heavily promoted at all.

Tatters

All of which could still prove a little embarrassing, since IBM Microelectronics has been boasting of the well over 1m PowerPC 601 chips it has made, and there’s no way that Apple Computer Inc can eat all those before it moves onto the newer versions of the RISC, and IBM now looks as if it has another PCjr on its hands. Apple can hardly be pleased but the one that must be unhappiest at the latest turn of events must be Motorola Inc, staring at a RISC strategy that was intended to get it out of the 88000 hole but is now itself in tatters. About the only way Motorola could salvage a desperate situation is to persuade Apple to sell it a licence to Macintosh System 7 and launch its own line of multi-operating system PowerPC-based personal computers, and do it at the same time IBM launches next month while vowing never again to be seduced into a joint venture with IBM in which it is not in full control.