% fortune -ae paul murphy

What happened in IT in 2006?

Overall 2006 can easily be seen as discouraging, as proving the notion that the normal curve bulges badly on the left: Google was briefly the world's highest valued public company; Vista was widely hailed as breakthrough, must-have, technology; the mainstream media easily beat the bloggers - making the average American see the least corrupt party as the most corrupt; YouTube demonstrated how boringly repetitive mass creativity can get; and, Apple's sepuku strategy continued to win it plaudits on Wall Street; but the overall reality, was, I think, far more positive.

2006 was the year some long term research projects at Sun, IBM, Freescale, and others produced their first marketable products. IBM got its grid on a chip into the market, Sun got it's SMP on a chip into the market - and software development on both sides is proceeding rapidly. Second generation products are on their way from both - with Toshiba getting ready to launch a cell/Linux based Asian PC, Sony's PlayStation already in the market, IBM's cell blades in test use, and Sun selling everything it can make on increasing order backlogs.

On the Wintel side, Intel more or less caught up with AMD on delivered performance, but did so mainly on the strength of its manufacturing investment rather than its design prowess - taking advantage of its lead in getting to 65nm to compete with AMD's more efficient 90nm products, but sacrificing profitability and writing off a five year engineering direction to do it. IBM, meanwhile, got its high gigahertz technology out of the lab and will continue Intel's temporarily discarded high cycle rate strategy with a 6Ghz PPC sometime next year - leaving Intel stuck between the lesson in Microsoft's demonstrated failure to fully support it's now discarded hyperthreading technology on the one hand and AMD's patent sharing agreement with IBM on the other.

Probably the most important thing about the industry this year, however, was the extent to which the mainstream technology press got things wrong. If it were possible to measure the messages the technology press delivered in some kind of "weighted by volume and authority" format, I imagine that the companies offering the biggest breakthroughs - Freescale, Texas Instruments, IBM, Sun - would score lowest while the companies that got forced to shuffle their executive suite, cannibalise their own advantages, and re-introduce previously discarded engineering strategies would score highest.

Vista, for example, is important in the sense that it will cost lots of people lots of money but the action in software for 2006 took place mainly with respect to handhelds, in IBM's work on getting Linux on Cell to work well, at Sun in getting Solaris ready for Niagara and Rock, and in dozens of labs and graduate student carrels where people invented new ways of working with digital media products. But, if you read about this stuff at all, you read about it on specialised sites, in small distribution newsletters, or in emails from friends - and not in your newspapers, on the popular news sites, or in "best selling" blogs.

It's a good bet, for example, that your daily newspaper will have some kind of year end technology review - and that most will mention both playphones and big screen TV's as this year's hottest selling consumer products while unanimously failing to tell the reader that not a single major player in any market -from TVs to airplanes - outside the PC world embeds either x86 or Windows.

A big part of what happened seems to have been a simple consequence of noise: we can only take in so much, and when the same people who benefited from their professed disdain for MacOS X mount a tsunami of Wintel adoration celebrating Microsoft's commitment to layering an Apple like GUI on top of NT, it's hard to see that, in reality, the tide on this stuff is steadily going out - and even harder to notice that after five years of effort Vista meets essentially none of Microsoft's announced "Longhorn" expectations.

Basically, there's so much noise supporting nonsense that important things become invisible. Did you notice, for example, that Microsoft didn't produce an integrated, ready to run, small to medium business application suite? Instead it choose to support its supporters: the consultants and PC businesses offering to perform the "some assembly required" services needed to merge Microsoft's small to medium business servers with Microsoft's Great Plains and Navision derived small to medium business applications.

But, bottom line, what happened in 2006? it was the year new foundations, some of them lain years ago, became the visible determinants of new directions -in hardware, in software, in legislation, and in corporate strategies.

At the hardware level we saw SPARC and PPC sweep the field for everything except the PC, we saw the cell and CoolThreads ideas emerge in marketable products, and we saw software supported media convergence in consumer products from 60 inch TV screens to handhelds. In software we saw new emphasis on compiler technology, particularly for Linux on cell; the full Solaris 10/ZFS release, new emphasis on developer protection in patent umbrellas, evolution in open source licensing ideas, and the appearance of brand new directions in hardware supported multi-media digital search and retrieval. In business change we saw a new emphasis on energy use, the coming into force of ediscovery legislation, and exploding public consciousness of the identity information risks associated with particular corporate strategies and technologies.

And, all of these should have interesting consequences next year - on which more tomorrow.


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.