When it comes to Windows CE, the engineering community is involved in a classic love-hate relationship. Interest in the Microsoft operating system for handheld computers is insatiable, as anyone who's posted a CE article on the Web and watched the hits go through the roof can attest. But ask engineers about their development plans and you'll see them wince, because they're finding themselves between an operating system and a hard place.

Here's the problem: Microsoft recently retargeted CE from palmtops, which are selling poorly, to embedded applications such as antilock braking systems and factory-floor controllers, which constitute an immense market. Unfortunately, CE does not yet have the feature set required to handle heavy-duty embedded apps.

That's no problem, according to the folks in Redmond. In April, they announced plans to toughen up CE to make it a more potent embedded player. The goal was to equip the 3.0 release, due in 1999, with support for "hard" real-time applications. This means interrupt response times on the order of 50 microseconds, instead of several hundreds of microseconds today.

Adding hard real-time support to an existing OS-as opposed to building it in from the ground up-is a big challenge. Yet Microsoft hasn't provided technical specifics on how it will meet its objective. Still, company sources are rejiggering the software's innards and report the effort is on track.

Software considerations aside, an even bigger problem is that Microsoft can't seem to make up its mind whether it wants CE to be a "consumer" OS aimed at cable-TV set-top boxes or a true embedded contender.

Set-tops are a good short-term strategy, since announced design deals build up momentum. For example, Microsoft has inked a pact with Japan's Matsushita, which will use CE for digital TV reception and display. Microsoft also snared a contract with cable giant TCI, which is supposed to combine CE with Sun's PersonalJava for use in set-top boxes.

But licensing deals do not always result in real-world deployments. Indeed, I believe set-tops are a poor long-term bet for CE, since set-top software must do the bidding of the hardware, not the other way around. Embedded apps may be a tougher row to hoe in the near term, but they promise bountiful harvests well into the millennium. There's just one rub: Embedded developers live or die not on marketing, not on promises, but on performance. It's a lesson Microsoft must take to heart if it wishes to succeed.

	eeProductCenter Launches SpecSearch®, New Parametric Parts Search Engine In our continuing effort to enhance our site, eeProductCenter introduces SpecSearch® powered by GlobalSpec. Click here.
Free Subscription to EE Times First Name Last Name Company Name Title Email address   Click here for your Free Subscription to EETimes Europe