Microsoft 'extends' Windows: What does that mean?

This morning in What's Now | What's Next, we reported on the early word from a keynote address to the Computex trade show in Taiwan, from Microsoft Corporate Vice President Steven Guggenheimer. What might have been big news there was already leaked in advance: Windows 7 will be available to the public October 22. The #2 story was supposed to have been the company casting its net wider, making Windows available on a broader range of devices.

Yet in Taiwan, where IT device production is shifting away from PCs and toward smaller, more customized devices, the question is just how broad that new range will be. The industry there (which locals refer to as "ICT" for "information and communications technology") has drawn a borderline around a concept called smartbooks -- devices whose blueprints can be assembled using pre-existing intellectual property that's licensed to vendors, typically using ARM processors. Meanwhile, Microsoft has drawn some borderlines of its own -- again -- by way of announcing that Windows may be addressing new market segments in the near future, extending its reach to new platforms. But now, there's dispute and confusion over whether the ICT industry's boundaries and Microsoft's have any overlap.

This morning, Microsoft is saying a lot without actually answering that particular question directly. In a statement to Betanews, a Microsoft spokesperson said Guggenheimer's keynote centered around a concept called consumer Internet devices (CID), defined as components that are not necessarily portable, not necessarily PCs, but which are connected to the Internet and whose functionality depends on it. Hand-held GPS, PMPs, and set-top boxes all fall into this category. The spokesperson said that the Embedded Division's general manager, Kevin Dallas, joined Guggenheimer to discuss how Windows would "drive innovation" in this category.

Meanwhile, at the very same time, Microsoft's corporate press office issued a statement saying Guggenheimer's keynote centered around a concept called ultra-low-cost PCs (ULCPC), a class of non-portable device that's connected to the Internet, and which may include such things as networked electronic picture frames and stationary e-mail and IM receivers. That statement started by referring to this segment as nettops, before mentioning that the more common name is ULCPC (if true, perhaps the first case in history where a catchy title is tossed aside for a five-letter abbreviation without a vowel in the middle).

As it very likely turns out, Guggenheimer took time to reflect on both categories, and different departments of the company took home their slice of the pie that best suited their respective flavors. But are smartbooks anywhere in the vicinity of these categories that Microsoft will support? According to Reuters, in an exclusive interview, Guggenheimer directly answered no, saying, "For people who want a PC, albeit a different chipset, we don't think those will work very well." In other words, Guggenheimer repeated the Microsoft message that when folks want to do PC-style work, they prefer a PC-style computer, going on to suggest that with any other kind of platform, users wouldn't be guaranteed the use of their favorite software or their plug-in devices such as printers.

So the discussion of "no Windows for smartbooks," while dominating much of the online traffic this morning, may not take into account a very important point: Windows Embedded CE is already one of the two dominant operating systems preferred for use on ARM processor-based components by ARM itself. When ARM executives brought up the topic some months ago of whether Microsoft should extend its support to ARM devices, they were talking about Windows 7 -- whether Microsoft should make a version of its desktop class PC operating system for ARM-based smartbooks, in light of engineers who successfully made Windows XP work there. The "no" answer from Guggenheimer appears to be an answer to that question.

In the end, however, Windows 7 may not be best suited for such devices anyway, for reasons that Guggenheimer alluded to and which may be even more numerous. Windows Embedded CE is designed to be deployable on a small device such as an ARM-based component, in such a way that it receives only as many features and functions as is necessary to run the thing. Windows 7, meanwhile, could be transferred by way of hard drive from one PC to a completely different PC, and despite the product activation issues that would likely ensue, chances are that it would run. The desktop PC operating system contains so much more overhead than an ARM device would ever put to use.

But by that same token, you could substitute "networked picture frame" in place of "ARM device" in that last sentence. You're not going to want your printer or Outlook or SharePoint Server running from, say, your refrigerator. Why didn't Guggenheimer's logic apply the same way to "nettop" devices as it did to "smartbook" devices?

Does that mean there will be "no Windows for netbooks," as some press sources have extrapolated Guggenheimer's statement(s) to mean? Absolutely not. As the company made clear to Betanews last February, while there won't be something entitled "Windows 7 for Netbooks" or even "Windows 7 for ULCPCs," OEMs that produce what they call "netbooks" will be eligible to pre-install Windows 7 Starter Edition. That's the version whose three application maximum was axed last week; certainly there wouldn't be an unlimited number of apps running on Windows 7 on netbooks if Windows 7 couldn't run on netbooks.

In the end, you do have to wonder if Microsoft is willing to extend its marketing umbrella for Windows to encompass one class of devices that's actually serviced by Windows Embedded, why not extend it to encompass another very similar class that's serviced by Windows Embedded, and that probably runs on the same processors? Once more, a Microsoft keynote address succeeds in substituting one set of questions with another.