Guy Kewney's Weekend Diary

Written by Guy Kewney, Contributor

Reviewing loudspeakers was not something I expected to be doing when I got into the computer business but today I find myself not only reviewing speakers, but reviewing serial speakers. I also find myself staring at a slide talking about USBHUBICS.

I suppose you're going to ask me what on God's earth a serial loudspeaker might be. It's like this: there's a serial port on the PC architecture which doesn't work, and another on the Mac architecture which also doesn't work. So Intel on the one hand and Apple on the other have come up with new, improved versions that run a very great deal faster. Intel has evolved the Universal Serial Bus, and Apple has sponsored FireWire.

And guess what: they are incompatible.

So here I am looking at Philips chips. One chipset does FireWire, which is a much higher specification for a serial bus, and could easily cope with USB protocols - if only the USB had a model for sending messages from peer to peer. And the USB chipset does USB, and is going to ship in gazillions, and couldn't talk to FireWire if it wanted to, because it's too slow.

In this situation, a wise product manager generates smoke and snake oil. Philips has trundled out a serial hub which drives a serial loudspeaker. It contains the sound card technology. The speaker accepts the serial (digital) signal from the sound hub. It then turns that into audio. Yes, this does involve a loudspeaker with a complete high grade amplifier and a D/A converter and a USB controller inside it.

The payoff, says the Philips man, is that it doesn't really matter how long the wire is. The digital signal will be converted into the best possible sound. I'll admit it was awesome. They showed two purpose-built speakers, and they were simply amazing. So I offered to test them in the labs.

"Oh, no; these aren't going to be a product. They're proof-of-concept prototypes," said the Philips man, amazed that I should think otherwise. "If we launch a product, it won't be of this quality. These are £1,000 level - each - speakers."


The man from Reuters arrives to try and sell me a Business Briefing subscription. He has a CD, and as he installs the system, he explains how for £20 a day, I can search for all the relevant data from all publishers on any subject I choose.

We choose Reuters, and the data I search for is my column, now one month on the streets, about their market research. You may remember I reported a survey which disclosed the fact that people who had to work hard at interesting and creative jobs found it difficult.

Interestingly, the text from my magazine isn't online. Even more interestingly, it turns out we have a corporate feed from the Reuters satellite. Somehow, I'm starting to suspect that this demonstration isn't going to lead to a sale.


This was going to be a day in Manchester doing a seminar for my publisher. Instead, I find myself in a Sikorsky 76 helicopter heading West in a Force seven westerly gale.

I don't know if you're expert on winds and navigation, but this means that we are travelling to Bristol at about 35mph than we would in still air and bumping around the sky in visibility so poor that you can't see the ground.

All for Orange, which is starting to put together its plans for UMTS - Universal Mobile Telecom Service. You make your own explanations for these acronyms, I quickly discover; the planned standard for mobile comms for handicapped people is UMPTIDUMPTI, and the expansion is too far-fetched to embarrass you with.

UMTS is going to involve very, very much higher data rates over the air when it happens. And it looks like it will happen really quite soon, in the next five years certainly, we'll have mobile systems capable of doing PC comms at well over the speed of today's ISDN. And they'll also talk to internal systems like DECT in offices, say the Orange researchers.

On the way out, I notice a display referring to 'ASCI'. The mis-spelling is drawn to their attention. "Advanced Speech Call Items," they tell me, pityingly.


The reason I'm sitting on a tatty carpet with a Toshiba on my lap, is that I'm an old hand at TV.

The basic thing you need to understand about TV studios is this: they are single-tasking. So everything proceeds at the speed of the slowest process. And the Computer Channel wants me to arrive at 11 to talk about the merger between BT and MCI. Peter Bonfield does his bit about how he can "leverage" the size of the two companies. I sit in the make-up chair, chatting to Lucy about bags under the eyes. And the presenter, who got caught in the same thunderstorm, gets her hair done so that it hides her earphone.

It actually takes another two hours before I am summoned to the studio for my five minutes of wisdom about BT taking over MCI. "It probably doesn't mean much for the computer industry right now, because until liberalisation, there's no profit in it for BT to reduce prices or to increase data capacity." Oh, and I try to slide in my own prediction: that BT's next merger will be with a bank. "But what about merging with Microsoft, wouldn't that be more sensible?"

No, actually it probably wouldn't. But then again, I expect Microsoft to try merging with a bank too. Fortunately, as I said, I've done this before, and I've brought a notebook, GSM phone, and data adaptor.

By the time I emerge into the rain again, Fate has played an odd hand. Into the Greenroom (the waiting room for the studio's guests) strolls Simon Walker, wearing a Toshiba anorak. He ought to be 6,000 miles away in Rio, on the BT Global Challenge boat "Wave Rider" which Toshiba sponsors. Instead, he's back in London to see his brand-new daughter.

An opportunity to chat with one of the skippers of the round-the-world race doesn't come often; but very, very rarely would you be able to interview one of them in London when their boat is half way round the planet. Simon's a computer junkie so we talk about the Web site, Inmarsat comms, future satellite video, and the interesting fact that the next generation of BT boats will have glass-fibre reinforced decks. This is really important to the computer world, because it means they can mount full-scale satellite dishes below deck.


This is the day of a new buzz-word. You've heard of RAD (rapid application development)? Now is the time to learn RAC (rapid application change). Yes, it's guru interview day. I find myself in Battersea, admiring the ears of James Martin, and wondering if it's safe to tell him that he's the spitting image of Jack Tyrrell, the Formula 1 racing car constructor.

James Martin is a legend in computers. He's got a book called "CyberCorp -- the new business revolution." I have a review copy, and an "interview opportunity" at his house on the South end of Albert Bridge.

Discussing the future with Martin is not unmitigated cheerfulness. He foresees a time of very rapid change, corporate evolution at high rates, and rapid death for those corporations which, as he says, "are designed to resist change".

Naturally, the conversation moves rapidly onto IBM, concerning which Martin can be quite scathing. But his most cutting criticisms are aimed at Bill Gates. "The Road Ahead was quite a stupid book," he says matter-of-factly. "It's clear that Gates has no strategic vision, and not even a long-term tactical vision. For example, he could have bought Express, which is now part of Oracle, and which fills Oracle's previous weakness in data warehousing, for which SQL is unsuited. Gates now has that hole and no obvious way of plugging it."

I don't know that I'd necessarily recommend the book. If you're a Martin junkie - and many are - then this will be the 100th in your collection. It's pretty different from the rest, most of which are "pretty tutorial" in nature, says Martin, when I ask point blank whether this might be his best book.

What it does have is lessons. There are dozens and dozens and dozens of anecdotes and reports from how various large corporations cope with change. Each report is disguised as aphorism, but they are valuable for all that. ISBN 0-8144-0351-4 from American Management Association, 1601 Broadway, NY NY 10019.

At the end of the day, after several chats with several other industry figures - including Charlie Tierney (Sun, inventor of the JavaStation).

Editorial standards