Menu
The Petreley theorem

The Petreley theorem

Last week, I painted a picture of the future based on the Petreley Theorem - to profit from progress, you have to develop a gestalt of the industry to determine the industry's direction, and then invest in companies and technologies moving in that direction. As promised, this week I'll demonstrate how I drew those conclusions by first looking at how we got where we are.

The economy of the PC era was built on three constants: 100 per cent IBM PC compatibility, compatibility with DOS and Windows, and increasing processor power. As CPUs got faster, opportunities increased to sell compatible software and peripherals to take advantage of that power. As a result, the single-vendor, vertical mainframe economy became, practically over-night, more of a commodity market and a horizontal bonanza for almost anyone who wanted to play.

Only now is it becoming obvious that the mood of the market is changing. When the Pentium was introduced, it was a breakthrough. Response to the Pentium II and MMX has been largely ho hum. Microsoft chief financial officer Mike Brown recently characterised fiscal 1998 as a product revenue valley, despite the release of Windows 98 and Windows NT 5.0 in that time frame. Windows 95, on the other hand, was billed as a life-changing experience.

But if a shift is taking place, where is the market focusing? David Moschella, of US Computerworld, puts his finger on it in his book Waves of Power (released this year by publisher AMACOM): "An easy way to understand the current shift is to recognise that, today, a doubling of modem speed would do more to move the industry forward than a doubling of microprocessor performance."

The new constants seem to be bandwidth and seamless connectivity. It is becoming less dangerous to deviate from 100 per cent IBM hardware or Windows compatibility and increasingly dangerous to deviate from emerging standard communications protocols such as TCP/IP and formats such as HTML.

The first observable result that tells you this shift is occurring should appear as increased profits for smart companies that supply the bandwidth. And that's exactly what seems to be happening. For example, Ciena, a company that produces high-capacity optical fibre for telephone companies, had a net loss of $US2.7 million in the second quarter last year. This year it had a gross profit of $US54.7 million in the second quarter, and its market value went up $US1.2 billion in just the past several months.

This new computer economy will eventually affect other markets, too. When you hold only bandwidth and seamless connectivity constant, the market becomes more horizontal, more open, and closer to a true commodity environment. There is the opportunity to build limitless types of devices that only conform to one principle - they seamlessly communicate with other devices.

Not that this will be easy. But the difficulty gives us one way to isolate the builders of the next empire. Look for software or hardware vendors that find ways to create mutually beneficial relationships among their products in real time. Their devices will discover one another.

Hewlett-Packard revealed itself as one such company when it recently announced its JetSend communications protocol. If adopted industry-wide, you will be able to do such things as connect any vendor's digital camera to any other vendor's printer and get a hard copy of your pictures.

The point is that the most successful companies in the future will build products that care about two constants: sufficient bandwidth and cooperation with other products over standard protocols.

The platform-neutral zone's dangers

Captain's log, star date 1-900-555-5555. Due to a transporter malfunction caused by incompatible data formats, science officer Ilia and I beamed aboard with a mission - to unify all of the known races in the universe.

"Set the coordinates for the planet XML in the Internet sector," I cried.

"But Captain, that's in the neutral zone. The Redomulands will consider it an act of war."

"I'm aware of the danger."

"Fascinating, Captain. You realise that you're risking your career by disobeying Star Fleet, and we don't even know if the mythical standardisation device really exists."

"Spock, you don't understand. This standardisation device. It must exist. It holds the key to the future of all living beings."

"I see, Captain. The needs of the many outweigh the needs of the few. Or the one."

Platform neutrality is the impetus for many a battle these days. But on the Internet, the needs of the many indeed outweigh the needs of the few. That's why I've always said that the Internet represents an external pressure to standardise despite the self-interest of individual vendors.

That is also why there's a movement to support XML, or Extensible Markup Language. XML is a platform-neutral, somewhat object-oriented approach to organising and exchanging complex information. Compare with HTML, which is (theoretically) a platform-neutral approach to presenting simple information. XML may eventually replace HTML, but for now it complements it well.

For example this week, Lynda Radosevich points out that XML could let an emergency room physician browse medical records from a foreign database and see that you are allergic to penicillin.

Sun Microsystems' Jon Bosak offers a more detailed explanation in an essay, available at sunsite.unc.edu/pub/sun-info/standards/xml/ why/xmlapps.htm. Bosak says that the goal is for XML to allow foreign data to be presented the way the emergency room doctor expects to see it. XML should ultimately enable the hospital physician to drop an icon representing your personal records into a folder representing the local application's database.

The data would transfer seamlessly between the two "incompatible" database formats.

Astute readers will note that this follows exactly the new market economy trend I have been describing in recent columns. The movement toward a communications-centric industry necessitates that any two objects needing to communicate will do so seamlessly, even if they were designed by companies that had no idea these two objects would ever be used together.

XML is one more step in that direction and therefore has a good chance of success. (XML is actually far more than just a data intermediator.)Ah, if only it were as simple as Shatnerian bravado. Unfortunately, the interests of the few often supersede the interests of the many, and standards can be subverted even through sheer sloppiness.


Follow Us

Join the newsletter!

Error: Please check your email address.
Show Comments