Ballmer on Whistler, UIs, services and Windows.NET
More clues as to what the next generation will be about
Microsoft's big .NET presentation this week didn't get too specific about next year's first Windows implementation, confining itself to describing it as "the next generation of Windows" while implying that it would be built on Whistler. But company president Steve Ballmer, who's been doing interviews again, was maybe a little bit more forthcoming later.
Speaking to an AP reporter Ballmer didn't quite get around to saying Windows.NET version 1.0 (AP's designation - we think 1.0 is verboten usage in Redmond) was the same as Whistler, but he provided a clearer picture of what the user's eye view of .NET will be. There might be "evolution in the user interface", he said, and you might also see "more opportunities for services".
Why, you may well ask, does The Register class vagueness of this order as a clarification? Well, it suggests that the various UI projects kicking around in Microsoft at the moment are still pretty much up in the air, that these are going to continue to be viewed separately from base OS development, and that although Microsoft has an idea of what it wants to do, it's entirely unsure of how it's going to do it. If you think about it there's an obvious imperative to get a more fluid, user- and ISV-customisable UI together, because .NET is supposed to provide information and services to clients of all shapes and sizes, and the old hard-wired Windows UI obviously doesn't play. Which must be a considerable psychological hill to climb, if you've spent several years defending the integrity of the "Windows Experience" (which, of course, Ballmer has).
His not confirming that Windows.NET is Whistler is also possibly significant. It's really an add-on, and as .NET is supposed to be available to all classes of device, Windows.NET could, next year, be hooking in mobile phones, X-Boxes and set-top boxes. In the case of X-Box, it surely has to be doing this, as Microsoft has identified online gaming as one of the key categories that will power it into a services-based business model.
But back to Steve holding forth: "Windows.NET is not a service, so it has to come with a service." Rickety logic here, we know, it's a bit like saying Windows 98 is not a cowpat, so it has to come with one, but you kind of know what he means. "You would see certain kinds of services that would be an integrated aspect of that user interface."
So Windows.NET, aside from not being an operating system, isn't a service either, but a mechanism for delivering services which are bound up within a customised UI. To some extent, you could say Microsoft is doing this already via the MSN empire, and Steve uses MSN as an example. He says that "no matter what Web site you're looking at, you have access to those kinds of services... MSN would certainly take advantage of it. There would be a version of MSN that would substantiate the .NET platform through Windows.NET on consumer PCs".
If you think of MSN client software so far as a user interface rather than an application, then you can spot some of Windows.NET's ancestry, and maybe consider the Mars MSN UI/client a sibling. The UI is designed to give you access to specific sets of services in a simplified way, and by its very nature this approach is restrictive. Look at it one way and you might criticise MSN (and AOL) for trying to lock users out of the general Web and into their proprietary subsets, but you have to set against this the fact that people are going to want to deal with subsets, and are going to want a restrictive approach in at least some circumstances.
The difference between the MSN-style tied client approach and the Windows.NET one that Ballmer is currently evangelising is of course that Windows.NET is intended (so Microsoft says, and we should take it on trust till we find the gotchas) to be an approach available to everybody, while at the same time not breaking existing systems. Says Steve: "If the Web site doesn't take advantage of them [advanced features] they're still there for you in a simpler way." So long as Microsoft's playing this straight, the rationale must be that the range of services available and the excellence of their delivery will allow Microsoft to succeed on a level playing field. But you don't have to believe that; nor do you have to entirely believe that the playing field is level, as Ballmer did touch on Microsoft-only aspects to its XML implementation in his presentation.
Will Microsoft come unstuck when it comes to actually selling these services? The presentation itself was pretty vague as to what they might be, but during interviews Steve tends to firm up quite a bit, and this time is no exception. "We're used to getting, say, $200 for an Office upgrade, and instead you get 20 bucks a month or 25 or whatever the price ends up being." Once the shift from packages to rented services is made, you can see how nicely the numbers stack up annually. And of course as Microsoft has never quite managed to shift Office upgrades annually, and certainly never managed to get everybody to upgrade as a matter of course, the numbers will ramp considerably, if the plan works.
But he sees this as evolving, rather than it being a case of one type of revenue stream ceases and another suddenly takes over.
Finally, there's the matter of Plan B to contend with. Actually, as the judge has stayed everything pending appeal, Plan B isn't anything like as important as it was a couple of weeks ago. The gist of the matter is that, as .NET is all about integrating everything, and all of the company's divisions working together, it would be illegal if Judge Jackson's ruling was applied. So Microsoft ought to have a Plan B to cater for the eventuality of its appeals failing, and the company being split.
On previous occasions Ballmer has specifically denied that Microsoft has a Plan B, but this time around... AP asked: "But if the US Court of Appeals or the Supreme Court uphold even parts of the remedies against Microsoft, will Microsoft still be able to execute this plan?"
And here's what Steve said: "This is the right thing, and we're going to execute on this one way or another. I know which way is infinitely better for users... [but] we'll take care of this either way. We'll get this programming model and user experience to you either way, and by the way, we believe we'll be able to get this to you the right way. We remain confident. We'll give it to you no matter what. Can we really give the level of integration and simplicity in this thing that we know we can? Well, we think so."
So there you go, there is a Plan B, so if Microsoft wounds up split into an apps and an OS company, then the apps (it's got to be, hasn't it?) company is going to be able to do .NET by alternative means. Tune in next week, and no doubt we'll be reporting Ballmer saying there isn't a Plan B again... ®
And the greased pig link
We were remiss in not reporting a Jim Lehrer interview with Steve Ballmer on PBS a couple of weeks back. Lehrer gave Steve a hard time on Plan B, court delaying tactics and the PR/lobbying campaign, and it's well worth checking out. As a contributor to the Silicon Valley Linux User Group mailing list (which drew the interview to our attention) memorably commented:
"I know who I'd bet on in a wrestling match between him [Lehrer] and a greased pig, and it wouldn't be the pig." Couldn't have put it better ourselves. You can find the interview text here.
Sponsored: Hyper-scale data management