November 10, 2003

Longhorn enters the "witching season"

Author: Joe Barr

In case you've been off planet -- or so otherwise occupied that you haven't noticed -- Microsoft is building a new operating system. It's called Longhorn, and it has now officially entered a new phase of the development process. It's not scheduled for launch until 2006. Longhorn's new phase of development does not show up in any of the literature I can find on systems life cycle. Of course, the reason may be that it belongs in marketing texts rather than technical. The new phase is known as "the witching season." Be warned, gentle readers: It's a time when the line between reality and imagination simply disappears.

Traditionally "the witching season" only lasts for 10 days, but Microsoft has embraced and extended it. Now it now runs from a public unveiling until it is publically released. Never mind that is three or maybe four years down the road. Time is no problem during "the witching season." It's an object-oriented phase, encapsulating as it does the entire lifecycle of vaporware, then being destructed by the real launch-event.

Promises, promises

Microsoft products never run as quickly, are more robust, or stay more secure than during this magical period. Every promised feature is present and works perfectly. Use of the product is life-altering. It is right on schedule. For good measure, Microsoft sometimes tosses in unfulfilled promises from earlier releases. Like an all-new (but fully backward compatible) API, or perhaps an all-new file system.

All English majors are familiar with the concept of a "willing suspension of disbelief." That's what allows us to enjoy theater, film, and election-year politics. But forget Broadway and the Beltway, during the witching season it's also a requirement for nearly all PHB reading material. If you check any mainstream trade press publication or news portal during the witching season, you might find some news mixed in with the hype about Microsoft's next OS.

You can't blame Microsoft for wanting to get a running start at ramping up their incredible hype-machine over Longhorn. Traditionally, that's what vaporware is all about. It freezes purchasing decisions as PHBs decide to take a "Let's wait to see what X does before we buy Y" attitude. Remember "Chicago"? Never mind that with at least three years to go (the launch date has already slipped one year), Longhorn cannot yet even be certified as ANSI-standard vaporware. It's not that far along.

Is it all a smokescreen?

Redmond has been given extra incentive to start banging the drums for Longhorn earlier than usual by other forces at work in the wilds of Realityville. Things like continued unhappiness over licensing and pricing, security breaches in Windows 2003, and the increasing popularity of the competition from Linux and Mac OS/X. Beating the drums for Longhorn gives Microsoft the opportunity to take the focus off these problems and to keep it fixed on the glittery and glamor of its promises for tomorrow.

How do you beat the drums for vaporware? The answer is as simple as rigging an election: You do it early and often. Day in, day out, years before it exists. Stooges in the trade press report the all the latest "leaks" and PR fluff about the vaporware just as if it were real news about a real product doing real work. If it happens that real news about the latest critical "security gaffe and patch" from Microsoft gets shoved off the front page, all the better. Hey, if people want reality these days, they can watch prime time TV.

So how successful have they been thus far? Believe it or not, a Google search on "+Longhorn +Microsoft" this morning returns more than 800,000 hits. That's 20,000 more than when I started writing this two days ago. This for an operating system that isn't going to be released for at least three years. Not bad for a product whose design lacks the surface tension of Jello. The "witching season" is magical, indeed. It's also an especially incredible display of Microsoft's influence on the trade press.

Renaming is name of the game

Don't count on everyone in that same MS press corps to remind you that Longhorn's WinFS sounds vaguely like the Cairo file system promised in 1995. Or that Longhorn's new API is pretty much the same as described for dot Net three years ago. Cairo, of course, has undergone a number of curative renamings in the years since then. So don't be surprised when Longhorn is renamed a time or two before it gets real. Renaming things has been a traditional cure at Microsoft since the GPF days.

Don't look for a makeover renaming anytime soon, though. That's because you won't be reading about a real OS for some time. Everything you read in the hundreds and thousands of stories destined to appear between now and a public release is straight from Microsoft's propaganda factory.

Analysts in the trade press will have to be very creative in trying to make the material Microsoft hands them sound new and unrehearsed. But since Microsoft is the only source of information about the vaporware, it's going to really take some doing to try to stamp the "news" with a unique voice. That may be a problem for egotistical writers, but it is exactly what Microsoft wants to read: their story in their words unfettered by harsh realities.

But the "witching season" is not all bad. It does provide one benefit for those who prefer to read honest journalism rather than warmed-over propaganda from One Microsoft Way. It helps you to easily identify dishonest scribes.

Take names and write them down

You can cast out the demons by doing your own "de-witching." On a blank pad of paper, start keeping a list of writers. Add to that list any author who:

  • Writes about Longhorn as if it existed today
  • Makes any claims about Longhorn's abilities to do anything at all, now or in the future, without pointing out that the claim is unsubstantiated in any way.
  • Uses Microsoft claims about Longhorn in comparing it to real products

Then start the "de-witching process" by taking names and ... OK. Wait. I blew it. I was wrong about something. You're going to need a whole stack of blank pads to do this, not just one.

Click Here!