Bill Gates talks Vista and Linux

Bill Gates is pretty confident that when he spots an emerging technology, it will emerge. Exactly when that happens, though, is sometimes a question mark.

Tablet computing and interactive TV are just two examples where Gates has pushed Microsoft to get involved, though the markets have taken a long time to develop.

"My instincts, if they're wrong, it's usually not about what's right," the Microsoft chairman said. "It's about timing."

In the second part of our two-part interview, Gates reveals his favourite feature in Windows Vista, why he'll keep pushing for a new file system, and the role of open source software.

Q: It's been five years since the last big Windows release. Windows Vista is coming out really soon for businesses, fairly soon for consumers. What's your favourite feature in Vista?
A: Well, it is such a broad release that it's hard to pick. [Desktop] search may be my favourite. You know, you immediately say to yourself, "How did I live without it? Well, I bought add-ons and things like that." But the integrated user interface makes it quite different from even the add-ons.

One project you have long championed is the idea of WinFS, an all-new file system for Windows. When you made the announcement that you'd be moving to part-time work eventually, you said that you were going to try and convince Ray Ozzie to keep championing the need for a new file system. How's that going?
Well, you definitely still want a structured look for certain kinds of rich query. And if we're going to bring all these things of email and files and photos, bring it together fully, we need more than just the search indexing. Search indexing takes you further than people expect, I would say. But eventually you'll need more of a database-type look to these things.

And, you know, we're taking our SQL technology and using it in broader ways — that's our fundamental code base for all the rich storage things. I still think that the question is: is it the next major release or the one after that? I think that'll be a big breakthrough when we do it.

Are Ray and Steve on board? Or is that still something you've got to do before 2008?
Well, Steve — Steve Sinofsky, you mean, or Steve Ballmer?

Steve Ballmer.
Well, Steve knows that my instincts, if they're wrong, it's usually not about what's right, it's about timing. So he knows it's "when", not "if." But he'll rely on the teams to say, in terms of memory size and performance, is it right?

We have really clear criteria, now, for how we would choose. Before you pick the key features for release. Is it ready to be picked for that? So, yeah, I feel quite confident Ray will pick the right time.

And, you know, it's not like I disappear. I'm around even after mid-2008. There are some things I'm closely associated with, like tablet [PCs] and WinFS, that people will expect me to shepherd. I want to do that.

Is that the big lesson coming out of Longhorn? To have a different process in place, to look at what's doable right now?
That's one thing. I mean, you learn so much when you do these releases. We had to do breakthroughs in terms of security. We had to do XP SP2, which was, from our point of view, a full release, even though it wasn't an end-user feature release.

But we feel that some of the things we've learned about layering and how you organise the feature teams will let us take this next five-year period and do a lot more than we did in the last five-year period. Do we wish we'd done more in some areas? Absolutely. I feel super-good that in these last eight months, even as we're getting the product done, we got organised for that next step forward.

Windows is nearly ubiquitous on computers, but computers aren't ubiquitous in the world. And you guys have been looking for a long time at different ways of bringing computing to more of the world's population. One of the approaches that has been talked about, inside Microsoft, is the idea of mobile phones and using them with nearby screens make them more of a PC-like device, as well as just bringing down the cost of PCs. Is one of those really more of the answer for more of the world?
Well, you always have a spectrum of capabilities. Today, there are people who buy $3,000 gaming PCs, and it's the most cool product they have in their life. And there are people who buy $200 PCs. So we have an amazing range.

By doing the right thing on the [mobile] phones, we can take that down. It's not equivalent to that $3,000 thing, or even the $200 thing. We can take that down, say in the $50, $60 range. It's not a broadband network, but it's a network.

And there are a lot of scenarios that that works for. We can project onto your TV, we can connect up a keyboard. You can have AC power, so that you can maybe clock [the processor] a little higher. We have an incubation team working on that as an entry point. It's not a replacement [for a PC], you know.

We also have a team that works on what we call "shared-use PC," where you go into a library or a cafe and you just bring, on your USB stick, your files. Or you have those in the [server] cloud — the cloud store that we'll make…