Martin Banks, Personal Computer World 03/86 - checked

March 1986:

I had the temerity the other day (quite a few other days ago now, but only the one other day then, when I wrote this) to accept an invitation. Yes, I know that this isn't the type of thing one would expect an officer and gentleman to do, but then I'm hardly either, am I?

So, with no respect at all for the civilities of life, I girded my loins and set off to the wicked city to attend that great event in all journalists' poor and humble lives - the press conference. This one was being organised on behalf of a software company that many of you may have heard about. Novell Inc has made quite a nam e for itself producing networking software systems for personal computers, and various personages from that emporium were in town to tell us what they had done `now'.

This turned out to be, amongst other things, the final nail in the coffin of King Computasarfun. You remember him, of course, he was the one that had us all buying computers for splendidly hedonistic reasons, such as playing Space Invaders and Pac Man. Now the king is dead and there is no fun left in the computer industry at all.

Novell's nail came in the form of fault tolerant software for its networks. Whatever goes wrong, a simple disk crash in which the directory gets wiped or something more spectacular like a disk system grinding itself to dust, Novell has a product that will allow the user to recover lost data and look splendidly efficient.

This is, of course, the end so far as King Computasarfun is concerned. Not only does this reduce the `will it do anything at all that's understandable' fun of the early days of personal computing to little more than myth and saga, it even ruins the fun of owning a modern PC.

No longer will we sit at our desks, midnight oil near burned away, system memory full of fine prose or hot program, and hit the Return key with crossed fingers. From now on, we will have fully recoverable, don't-worry-about-a-thing-matey-I'll-find-it-for-you-it-must-be-in-here-somewhere fault tolerance systems that will, just to be on the safe side, actually do the work for us in the first place. Well, you know what humans are like, so damned.... human.

For those in business, or doing anything real with their computer systems, something like this is not only sensible but essential. But for all those funsters out there, hoping that this is just some passing aberration in the great adventure game, it is the long walk into the terminal void of boredom.

Yes, even the personal computer has now had to succumb and become just a boring old business tool, and the reason is....well, business. Last year was not just a turning point for many of the companies in the business, it was the year in which they drove straight into the brick wall. Why they did this is not that difficult to defin e, especially with the journalist's luxury of perfect hindsight. Where it now leads is also, to some extent at least, equally definable.

The brick wall, as I have writ before, is that the 'bought from' market got used up. This has meant the industry having to go out and `sell to'. In practice, they have not had too much to offer, especially when they have been trying to play the game of high volume, low cost merchandising. The machines have not been cheap enough to be mercha ndised, nor have they been, like cans of baked beans, consumable.

The market demand has therefore been easily met and, unlike the hi-fi business, there has been no standardisation of software formats like records and tapes to drive the market from outside. Acorn, Sinclair and, more recently, Apricot have found that low prices (even with good value for money) do not make for a continual market or a good enough profit margin.

The answer I feel, lies in a direction I first put into words many years ago; the prices of systems, I'm afraid, has to go up. What is more, that is exactly what it is now likely to do. The personal computer has at last been accepted in the bosom of the marketplace it has always fancied for itself - big corporations.

The effect on the PC industry could be quite marked, however. Big corporations will require different things from the PC industry than the small business user - fault tolerant networking software is but one of them. They are also going to require bigger systems generally. This does not necessarily mean bigger individual boxes. Rather, it means things like networking, multi-user systems, clustered work-groups and mainframe communications. These are the things which will make the integrated systems the corporate users need.

In practice of course, the individual boxes will also get bigger, and this threatens some additional changes. Bigger systems structures running bigger boxes will mean different operating system requirements. The 640K limit of MS-DOS is already a hindrance to many applications and its single-user, single-tasking orientation will contradict the aspirations of the corporate user. Presumably these are problems which Microsoft intends to face with Version 4 of the system, when it eventually appears.

It seems reasonable to suggest that, if it fails to face these questions, the marketplace will move on to other things, despite the dominance of the DOS applications software base. It is probably no accident that Intel, part-owned by IBM and supplier of Big Blue's processor chips, has chosen Unix as an operating system for the ne w 80386 32-bit processor which will certainly power a future generation of PCs. It is a system which has the power and facilities required, and its current user-unfriendliness can be overcome. I used to think Unix wouldn't make it in the commercial big league, but now I'm not so sure.

The integration of PCs into large corporate data processing structures will also lead to a growth in `old' products, such as applications written in Cobol. This is, after all, what the majority of corporations already use.

And the upshot of all this? The systems will get more expensive, and the sales volumes will drop a little. The margins for the PC makers will get better however, and they will be better able to genuinely service the needs of other, smaller users. These in turn will have to learn that `service' usually doesn't mean the cheapest box in the shop.

end