Martin Banks, Personal Computer World 07/87 - checked

July 1987

There is certainly one thing you can always say about IBM. Once it decides to stick its toes in the water, it somehow manages to stir it up so much that no-one can see clearly for months afterwards.

It did it once before, you will remember, with the launch of some system or other, the PC I think they called it. In April, they did it again with the announcement of the Personal System/2.

It has been interesting to read through all the information the company handed out on this machine (at the time of writing, I am still reading. Never let anyone tell you that IBM never says anything about new products. When the information comes it can be quite an arduous task to wade through it all). It has also been interesting to read some of the initial responses, especially from the compatibles and clone makers.

Olivetti, for example, promptly said it was `underwhelmed' by the announcements, and felt that there was nothing in the new Model 30, Model 50, Model 60 or Model 80 that they hadn't already done or couldn't clone quite easily. Compaq expressed similar opinions.

This may well be the case, but I suspect that we will find, when the waters have settled (at a time not totally unrelated to when the machines themselves start to appear) that IBM has actually moved the game somewhere else. It is possible that, despite keeping the PC/XT and /AT going as current products, IBM has kissed the old PC market a fond farewell.

And with that act, it will have also kissed the compatibles and clone makers good-bye as well.

The reason, if I am right, is quite simple. Compaq, Olivetti et al can clone the new machines as accurately as they like. With MS-OS/2, they will even have a very close resemblance to the currently promoted operating system. All that will be as nought, however, for they will not be able to clone one vital factor - the typical IBM customer.

Big Blue has always been after the big corporate marketplace, but over the last few years it has lost its sense of direction just a bit. The PC took it down a tangential route. It had sufficient importance to the corporate user, however, to remain a significant product for the company. But it was a digression, none the less. IBM also failed to develop a coherent environment across its product range. This meant users could not grow from a small IBM machine to a bigger one without major changes and upheavals.

This didn't matter when no other computer company offered such an environment, but when Digital Equipment announced the VAX family running the MVS operating system users suddenly had that option. An examination of their balance sheet, compared to IBM's, shows how popular the idea has proved.

But do you, as personal computer freak, want to know about mainframes and minicomputers? Well, these days perhaps you should, for this is what lies behind the IBM Personal System/2 family. The customers IBM wants to sell to want to buy IBM if they can. Now they can, for that is what these machines are all about. They are the new, far reaching fingertips of a coherent product family with mega-mainframe systems sitting at the top. That is what corporate customers want, and it is what the likes of Compaq and Olivetti cannot provide, even if they can clone the hardware.

And then, can Microsoft clone the software for them?

This may seem a strange question as it has been working with IBM to produce OS/2 and will be launching a generic version, MS-OS/2, when the IBM system comes out. This is not necessarily the important point, however, for I suspect that OS/2 may not be the important operating system for IBM.

I have written here before about how Unix could take over from DOS as the next generation operating system, and that perhaps IBM's own version, AIX could prove interesting. Well, I now think this is about to happen.

AIX is a real bells and whistles version of Unix, and is already available on IBM's 6150 RT/PC. It was not made immediately apparent at the announcement by IBM, but this operating system is also to be made available on the top-end machine, the Model 80. If, over the years, you have got to understand the peculiar semaphore of nods, nudges and winks which constitutes real communication with IBM executives, this same operating system will appear on mainframe machines at some time in the future.

There are several things which should be understood about AIX. Firstly, it is available, which is more than can be said about OS/2. Secondly, it offers users more facilities than OS/2. In particular, it is said that the Model 80 variant will allow direct multi-tasking of existing MS-DOS applications. This will be done by running them as DOS 3.3 tasks, with AIX controlling their operation and providing all necessary inter-task communications.

This is what many users want, and is more than is possible with even the extended version of OS/2. It must be remembered that the first version of OS/2 does not have things like the Presentation Manager windowing system and can run only one MS-DOS task at a time, in `go-faster PC' mode. If users want multi-tasking, they must change their applications software to the specific OS/2 versions.

The last thing about AIX is that it is IBM's own system. It should take little guessing as to which operating system IBM will actually favour, long term.

And if this all happens, the PC market as we know and love it could start to disintegrate as each of the other main players tries to take over the leadership role vacated by IBM. They will all, inevitably, try different ideas to gain market leadership, and will, equally inevitably, divide and dismember the users' loyalties.

Ironically, that latest and most public clone maker, Amstrad, could easily end up with de facto leadership of the PC business, just by sitting tight and keeping faith with the standard as it currently stands. After all, most small business and individual users have barely reached PC level in the sophistication of their applications.

end