On September 11, IBM Corp moved the inevitable demise of the monolithic mainframe a giant leap closer when it decided to go for its time-honoured macho gesture and launched – nay boasted that it was launching – over 100 products on one day. The reaction was uniformly negative. Ridiculous. Ludicrous. Suicidally counter-productive. Those were some of the more printable reactions to the announcement and the journalists and analysts returned glumly to their offices and wondered how the Dickens they were going to make sense of the announcements to themselves, let alone their readers and clients. Inevitably, they mused despairingly, a lot will just get lost: there’s no way we can master such a mountain of indigestible information and present it comprehensibly.

Moaning Minnies

What a load of moaning minnies! IBM doesn’t make major announcements for the benefit of the IBM-watching fraternity: they are a personal matter between IBM and its consenting users in private? Quite right. IBM doesn’t need intermediaries to explain itself and its directions to users: that is what it has highly-paid sales representatives for. Well that might do for public consumption. But if the people that have all the time in the world to sit around and do nothing but reflect on IBM’s outpourings find that the task is beyond them, how is the poor user, who has to try to get to grips with all the new offerings in between the deadly serious business of running a large and demanding installation day-to-day, expected to begin to master such an intolerably onerous brief. Consider that IBM’s 100 product blitz came just a week after a major AS/400 announcement, and one that is of considerable importance to MVS users, who are beginning to look on baby AS/400s as appropriate machines to distribute in remote locations. What made the IBM self-indulgence even more unforgivable was that apart from the big disk drives, almost nothing in the announcement is deliverable until well into next year – and much of it is a full 12 months away. It may have suited IBM’s convenience to get the whole lot out of the way in one insulting expectoration, but it was emphatically not to the benefit of the people that in the last analysis put up the money for IBM’s entire payroll, the poor benighted user. A reasonable company would have confined the September 11 announcement to the new Summit mainframes, the new 3390 disk drives and the additional functionality extensions to Escon and such – that related directly to the core of the announcement. Even the new 9121s could have kept to another Tuesday and with them the two new smaller disk drive offerings. Throwing in AIX/ESA as well makes it clear that IBM regards Unix on the mainframe as a very bad joke that it hopes everyone will forget about before the leaves have fallen – and even that misfired because the only aspect of the announcement that was reported by the utterly bemused New York Times and the Financial Times was – Unix on the mainframe!

Data dustbins

It is axiomatic that complex software announcements Warehouses, Repositories and sundry other data dustbins, evolutions of Systems Application Architecture, AD/Cycle and such – have to be presented and assessed in a calm atmosphere unsullied by the need to rush something into print, the need to do the exacting sums on whether the data processing budget will actually run to a new Summit in June next year and whether the thing should be leased or bought outright or financed in some other way. Early November would have been soon enough to unleash those announcements on a slow-to-comprehend world – after all the networking announcements had been got out of the way on a mid-October Tuesday. Once again the suspicion – and it’s rife has to be that Systems Application Architecture as originally conceived is falling apart, IBM has no more idea of what people want and need in the way of software engineering that users themselves do, and that the company’s entire macro software strategy is in complete chaos, with there being no chance that all the weird and wonderful technologies

being brought in from third parties will ever work together or live up to their promise. That may all be untrue, but if that is the impression that the user community and the industry-watchers have come away with, it will be a self-fulfilling prophecy. While declaring 1988 the Year of the Customer, IBM has never treated its users with much respect, constantly pursuing architectural, product and software strategies that put tripping up the competition or derailing industry standards – like X25 – that IBM doesn’t want to support at the top of the list of priorities, manufacturing as cheaply as possible next (no reboot button on the original Personal Computer, a contemptuously inadequate keyboard buffer in contrast to every cheapo CP/M machine of the time), and the true benefit of the user nowhere. IBM has always taken the attitude that its mainframe customers will follow it down whatever route it maps out, will somehow get to grips with whatever it announces, no matter how difficult it makes that process, because they have no alternative: all you had to do to bring a stroppy customer to heel was to suggest that if that was the way he felt, well the number of the nearest Burroughs Corp sales office was to be found in the phone book.

Let 100 flowers bloom

IBM could make life as difficult as it liked for its mainframe users because there was nowhere else for them to go. But every additional layer of complexity that the company tries to build onto the wholly inadequate structure of System 360 and OS/360, the closer it hastens the day when with the best will in the world even its most loyal customers will admit defeat and ask wearily whether it all really has to be this complicated. And suddenly, it doesn’t. The siren call to the brave new world of open systems sounds ever louder, and slowly all the appendages that are lacking in Unix are being built in. There is no way that the typical four-square MVS user can yet think of shipping the iron to the scrapheap and downsizing to Unix – but there is already every reason for such users to consider implementing anything completely new under Unix and putting the mainframe applications on a care and maintenance basis until Unix catches up with everything that is needed. By its hubristic decision to let 100 flowers bloom on a single day, IBM brought that eventuality a giant leap closer.