Dying Dreams, Rising Hopes

In 1997 I was having to think about saying goodbye to OS/2. But the good news was that my other favorite operating system, NeXTStep -- long thought to be dying -- had just been bought by Apple, although it still wasn't clear what they intended to do with it...

[Future 64: Prelude to Rhapsody][PCPlus][4 Oct 97][chb]

This column is about the future of computing. But from time to time I find it necessary to indulge in a little breast-beating, as the present catches up with the future, and predictions made here fail to come true. Two such failures that have become horribly evident are my confident assertions in these pages about the the OS/2 and Taligent operating systems. I did not say that either of them would take over the world, but I predicted that as they came to be adopted by users we would find ourselves entering into a new era of computing.

Well, OS/2 was taken up by some users, but the earth did not move for them, and many are back on the Microsoft trail. Even IBM seems to be edging its way out of commitment to its operating system, and is currently caught up by a sudden enthusiasm for Windows NT. As for Taligent -- that costly development that was nurtured for five years by Apple under the name of Pink, and ripened for another three years in a cooperative venture between Apple and IBM -- it simply never happened. IBM will tell you that its valuable technologies continue under the covers in such products as VisualAge for Java, but if pressed even they will admit that Taligent was a costly failure.

IBM's message when I attended the company's Technical Interchange at St Louis this year was that the magic no longer lies in the operating system. Java and the kind of thin-client we were talking about here last month is IBM's preferred solution for the desktop. But what you run there really doesn't matter, says IBM, because at the other end of the network is the real magic -- a pot-pourri of old so-called "legacy" applications, new applications, and new "glue" software that encapsulates them into an organic whole and delivers that across the network. This is the model IBM is unfolding for its corporate customers. Ultimately I think Big Blue would like to see all of us sitting at thin "Network Computers", with IBM in charge of the real computing system at the other end of the wire.

But the "PC" in the title of this magazine will continue to stand for "Personal Computer" for a good few years yet. That should mean that we all get to choose what we run on our own machines, but lately the options seem to have been closing down around whatever Microsoft cares to offer us. For me the positive part of the IBM scenario is the message that "what you run on the desktop doesn't matter". It promises an escape from the tyranny of the "standard desktop operating system".

As the millions of Linux users world-wide know, there is at least one very viable alternative to Microsoft Windows. Linus Torvelds' "home-made" operating system has grown up over the seven years of this decade to become a plug'n'play modern UNIX that can be easily installed on any of the machines you're likely to be running today or in the future. At the moment I'm sitting at the Seimens-Nixdorf Scenic Mobile laptop I mentioned last month, writing this on Caldera's version of Linux. The machine is set up to dual-boot between Windows 95 and Linux, mostly to demonstrate to friends that you can have your cake and eat it. Ah, but what about the applications, they ask. A couple of years ago that used to fox me. Although Linux did everything I needed, there were no equivalents to Microsoft Office. Today there are at least two, and one of these, StarOffice, is freely downloadable for private use. And whereas two years ago Linux had to develop an emulator environment to be able to run the many applications written for SCO Unix, today the boot is on the other foot -- SCO has just come out with a Linux emulator!

There's one other desktop operating system contender in the pipeline that, for the last seven years at least, I never thought I would be mentioning in the context of "future computing". It's a bit like saying that airships will make a comeback (by no means an impossibility, by the way, but that's another story). I'm referring to the extraordinary renaissance of NeXTStep, the operating system devised by Steve Jobs ten years ago, adopted by IBM in 1990 (and dropped again immediately afterwards), and purchased by Apple towards the end of last year.

NeXT was founded by Steve Jobs following his acrimonious departure from Apple. Four years ago it attempted to overcome market indifference by dispensing with hardware manufacture and porting its operating system onto more popular platforms, notably Intel. But there was no rush of customers to take up the product -- in part because the complete development system cost around $5000, and in part because all eyes were focussed on the upcoming wonder from Microsoft, Windows NT. And the Intel hardware needed to run the NeXTStep operating system at the time was far from standard -- 32 Mb of RAM, a 500 Mb hard disk and a 20 inch monitor to make the most of the high definition Display PostScript graphical user interface.

The move out of hardware didn't help much. This time last year Steve Jobs's company, NeXT, was in the doldrums. As an addicted NeXTStep user myself I might have been forgiven for feeling depressed too. But far from it -- there was still plenty of NeXTStep freeware and shareware in circulation (mostly because the environment is so delightful and so easy to develop applications for), and as an Internet machine my NeXTStep Canon object.station couldn't be bettered (the World Wide Web was originally developed under NeXTStep). If, heaven forfend, Microsoft died tomorrow, we'd all be lost without the next upgrade, and the fixes to that upgrade (and the fixes to the fixes). Life as a NeXTStep user is much less exciting. I would be sorry to see NeXT go, but everything else, including my own NeXTStep system, was ready to carry on happily for at least the next five years.

And then, astonishingly, Apple stepped in and bought the company. Suddenly I was faced with the prospect of the orphan operating system that was working so well for me, actually coming to the rescue of millions of other users world wide (to say nothing of rescuing Apple from its own impending fate). For months afterwards my thoughts about this were confused and rambling -- and so evidently were Apple's, because not long afterwards the chief technical officer who recommended buying NeXT quit Apple, and the CEO who made the decision was sacked shortly afterwards. Perhaps this is why the test version of NeXTStep released to developers was given a comic opera name, "Prelude to Rhapsody".

It seems to me that the Apple-NeXTStep saga is still a laughable mess today. But Microsoft is not laughing. It took the opportunity of buying into Apple, I believe because Bill Gates sees in NeXTStep much the same huge potential that I do. Next month I'll be spelling out in detail exactly what I think that potential is.