Before I elaborate on what this means for service-oriented architecture (SOA), let me get this confession out of the way: I like Steve Ballmer, as a man (that's his best bet, anyway). One reason I like him is that he'd probably laugh at that joke. At least he has laughed when I've poked fun at him and at Microsoft in the past. More important, the laughs were genuine, not a pretense of peace with a columnist who has been openly critical of Microsoft business practices. (Say what you will about Ballmer, but he does not disguise his feelings well. He's an easy read; even an amateur can tell when he's faking something.) That's why Ballmer also strikes me as a guy you could enjoy a few beers with. He's genuine enough in person that it would be fun to get to know him even if we disagreed on everything. Perhaps he even has enough of a life that the topic of Microsoft wouldn't even come up in the conversation.

I also respect Ballmer. He has an honest bone in his body; maybe even more than one. I'll never forget the day Steve Ballmer visited me at InfoWorld to convince me that Windows 95 would be the wave of the future. In our conversation, and in front of a room full of editors and skilled technicians, he unapologetically admitted that IBM's OS/2 was superior to Windows 95. He added that Windows NT was superior to OS/2, but that was certainly true even then. I can think of one or two other times when Ballmer spoke the truth even when it was uncomfortable to do so. If you have spent as much time around marketing people and executives as I have, you can appreciate the significance.

Steve Ballmer is also realistic enough to know when it is time to break the mold. He may talk like a rabid dog when he is wearing his marketing hat, but he is not above making an unpopular decision (unpopular at Microsoft, that is) if he feels it is in the company's best interest. In fact, I predict that Microsoft will give more than lip service to open-source and open standards sometime in the future, and Ballmer will be at least partly responsible for this shift.

No, I'm not predicting that Microsoft will ever release Microsoft Office as open source. But don't be too quick to single out Microsoft as the pinnacle of evil if Microsoft Office remains proprietary. IBM hasn't exactly released its most prized software as open source, either, and IBM is often regarded as one of the most beneficent forces behind corporate acceptance of open source and Linux.

So how does this all fit in with SOA? Microsoft is most threatened by the presence of Unix and Linux on the back-end. The increasing popularity of SOA could be one of Microsoft's best tickets into the server room. Microsoft already has a solid software foundation for good SOA with .Net. All Microsoft has to do is polish up its offerings and push them as part of its new mission to interface cleanly with open standards.

Ballmer is just the guy to make it a reality. That's not to say Microsoft couldn't have been a successful player in SOA under the Gates dynasty, but Microsoft could not be as successful as it should be with Ballmer in charge. With all due respect to Bill Gates, he was blinded by his single vision of crushing all competition and locking customers into a Microsoft world. Ballmer shares that vision to a point, but unlike Gates, Ballmer is pragmatic enough to think outside the box labeled One Microsoft Way. Ballmer is the kind of man who would accept the possibility that Microsoft does not need to be the only game in town if it means making Microsoft more successful.

None of what I have said is meant to offset or excuse the harm Microsoft has done in terms of shipping insecure software, influencing the OOXML standardization process, or [name your own pet peeve]. But do not count out Microsoft in the SOA game. Microsoft is a powerful company with a lot of talent, and it has the capacity to do good as well as evil.

Despite outward appearances and some of the outrageous things Steve Ballmer has said and done in the past to toe the party line, I have a great deal of confidence in his ability to adapt to challenges in a way we have not seen Microsoft adapt before. I'll be watching, and I hope you will be, too.