InformationWeek

October 19, 1998

TechView:
Microsoft's PromiseWare Strategy

By Sean Gallagher

o spice up his keynote at the Microsoft Professional Developers Conference last week in Denver, Bill Gates became a TV star. Attendees were treated to a high production value video that included a spoof of Volkswagen's famous "Da da da" commercial--two guys aimlessly driving around what looks like a Silicon Valley neighborhood, picking a chair out of someone's trash and then abandoning it because it smells funny.

In the Microsoft production--which painstakingly reproduces the Volkswagen ad, right down to the last detail--Steve Ballmer is driving, and Gates is riding shotgun. And sitting on top of the chair is a Sun Microsystems workstation--which Gates and Ballmer determine to be the source of the offensive smell. The faux ad ends with the Internet Explorer logo and the Volkswagen advertising slogan, "Drivers Wanted."

In every joke, there is a bit of truth. The gag was a demonstration of the mutual disrespect between Microsoft and Sun--and the feelings surrounding the lawsuit filed by Sun over Microsoft's alleged misuse of Java technology. I'm sure that the image of Gates and Ballmer out

driving around and not getting anywhere could be used as a metaphor by Sun and other Microsoft competitors. You could also compare Windows NT to the Volkswagen Golf they were in--user friendly, but with a limited payload. But I'll leave such deconstruction to all those former film students out there now writing Visual Basic code.

The tension between Sun and Microsoft over Java is probably part of the reason why Microsoft has taken direct control of another cross-platform product--the implementation of the Common Object Model for the Solaris platform. Delivering COM on Unix is a key part of Microsoft's enterprise strategy to turn the plumbing of Windows into the plumbing for the entire enterprise.

I've visited the subject of Microsoft's Distributed interNetworked Applications architecture and the COM-everywhere strategy before in this column ("Microsoft's COM-unist Manifesto," June 8, p. 92). The first problem with the stra-

tegy today is that it's based so heavily on products you can't buy yet. Until Windows NT 5.0 and COM+ ship next year, a lot of Microsoft's vision is just that--a vision.

But what makes that vision so attractive to the hordes of Windows developers--like the 6,000 or so that gathered in Denver--is that it is at least rooted in something they can put their hands on today, and they can be relatively certain of the future of the Windows platform.

That's more than can be said confidently about Enterprise JavaBeans and Corba--and they're already available. It's no wonder that Microsoft can beat products with promises.