Modular Windows Debuts

Doing TV is Microsoft’s new challenge

Late in October, Microsoft publicly announced its plans to enter the world of consumer electronics by creating operating system software for digital consumer devices. Unlike Apple’s Newton strategy, which calls for an entirely new operating system for the new devices, Microsoft developed a stripped-down version of its immensely successful Windows 3.1 called Modular Windows.

Microsoft, which envisions an entire product family based on its Modular Windows technology, has mapped out three areas of focus: consumer information appliances, cable television and personal communications devices. It is now selling a software developers’ toolkit, with source code and sample applications such as an online TV guide, for $99. (For an indepth review of Microsoft’s consumer strategy, see Vol. 1, No. 10, p. 3.)

Modular Windows is a subset of the full Windows operating system. The so-called modular nature of the software allows Microsoft to add or remove functionality depending upon the application. It was designed to take up a very small amount of space, unlike the fully featured Windows software that requires up to 6 MB on a hard disk.

Today Modular Windows is focused on television applications. It is most dramatically different from Windows 3.1 in its elimination of system facilities (such as printer drivers, TrueType fonts and Dynamic Data Exchange) as well as adapting the graphical user interface to one more appropriate for television, which is viewed from a distance measured in feet, as opposed to the normal 18-inch distance to a computer monitor.

Buttons, cursor arrows and the standard display font were all enlarged or redesigned to stand out from the screen. In addition, since the input device is some form of infrared remote control and not a keyboard, IR drivers were added.


Of the three categories of appliance recognized by Microsoft, only one is on the market. Tandy’s Video Information System (VIS), a consumer-oriented multimedia player, was the first commercial device to use Modular Windows.

The operating system is contained on a 1-MB, “read-only” chip that functions as the software brains of a Modular Windows-powered consumer device. The hardware itself is essentially a low-end PC clone that uses a remote control instead of a keyboard for input.

Rob Glaser, vice president of multimedia and consumer systems at Microsoft, says VIS was the first “proof of concept” device for Modular Windows, proving at least to Microsoft that there could be a consumer-oriented, television-based information appliance derived directly from personal computing technology. (Our first impressions of VIS are on p. 24.)

Sanjay Parthasarathy, Modular Windows product manager at Microsoft, believes that interactive CD boxes will begin to show their market potential next year, after full-motion video adapters can be regularly included in such products. “Online” or network connectivity will require the addition of a modem (unless manufacturers wise up and spend the extra few dollars to build them in), and will take more time to catch on.


Windows as TV interface for programming and services has been quietly under development for more than a year. At the Western Cable Show last month in Anaheim, CA, Microsoft came out of the closet, demonstrating its view of how Modular Windows could add functionality to the cable system, and assist the industry in increasing customer response to advertising via the power of interactivity.

Modular Windows for TV applications will take much longer than CD-based consumer appliances to evolve. Microsoft itself predicts it will take about seven years before most TVs and cable decoder boxes include an operating system of some kind.

So Parthasarathy’s goal at the Western Show was “blow away” the cable industry with examples of how an operating system — Modular Windows, for example — can help the cable industry move into the digital world. As Microsoft continues to develop working relationships with the cable industry, it hopes it will be able to learn what cable viewers want in a TV interface.

On the trade show floor, Microsoft teamed up with Jerrold Communications, a division of General Instruments and makers of cable converter boxes, to demonstrate the potential of a Modular Windows converter box. Along with Insight, Prodigy and TV Guide, Microsoft demonstrated a prototype system that provided online programming services, including searching and sorting of listings by type of program, channel, etc.; supplemental information such as sports statistics or song lyrics; interactive services over Prodigy; and most importantly, transaction capabilities such as the ability to purchase record albums or event tickets.

Beginning in 1993, Microsoft will be conducting field tests of cable delivery systems based on Modular Windows. Working with a number of different cable operators and equipment manufacturers in testbed operations throughout the country, Microsoft has set a number of assumptions and goals to prove to itself and the cable industry.

With different programming options, such as enhanced MTV, ESPN or CNN, as well as interactive services and online transaction capabilities, Microsoft and its partners (whom Parthasarathy would not name for print, but which include some of the biggest operators in the cable industry) hope to measure people’s desires for new services and user interface preferences. (This explains why Microsoft, and many large cable operators, are not members of the Austin-based First Cities group discussed in last month’s Digital Media.)

Glaser hopes these trials will determine the economic viability of both Modular Windows and the products it would enable. Then it would be a matter of “rolling the truck down the neighborhood” signing people up. In the cable industry, says Glaser, “understanding how to get 2–4 percent of the market is extremely difficult, but how to get from 20 to 60 percent is well known. In the computer industry, the first 25 percent is understood; getting above that is difficult.” In other words, selling a new product or service to cable television customers is extremely difficult. Once it takes hold, however, it is relatively easy to sell to large audiences. This is in sharp contrast to the computer industry, which has yet to understand how to mass market its products.

No PDAs yet. Personal communications and organizational products, like those based on Apple’s Newton system technology or the upcoming releases from GeoWorks (with Tandy and Casio, see p. 18), PenPoint (with EO) and others, are still down the line for Microsoft. While these hardware and software companies are actively experimenting with new personal devices, executives at Microsoft have chosen to remain quiet about the PDA phenomenon, except for the expected occasional cheap shots at archrival Apple’s efforts with Newton.

While nobody would directly say so, Microsoft does not appear to have crystallized its strategy/rationale for Modular Windows-powered devices. For that matter, Glaser and others at Microsoft do not seem confident in what the new devices will look like, what they will do or who will buy them. “We’re not sure what to even call those products,” said Glaser.

The arguments remain compatibility of data shared between Big and Little Windows applications, which for some applications might be a valid concern, and consistent development environments between the two systems. This concern assumes that many applications will be shared between the PC and PDA environments, an assumption that is fundamentally flawed since there is virtually no reason to port a PC application to one’s television. The applications are completely different.


According to Parthasarathy, each application of Modular Windows can be developed independently, with the appropriate functionality, and yet each will still be able to utilize the tools and applications that have been developed for the personal computer and the full Windows system.

The Microsoft model follows that of traditional consumer electronics, in which people purchase peripheral devices depending upon their desires or needs. A home may have only a radio, or it could have a stereo receiver, with CD player, turntable, television set and VCR. Each new box can easily be plugged in to the other boxes and function properly. The cable television world has not been as successful at this, with incompatible “cable-ready” televisions, VCRs and converter boxes.

According to Glaser, Microsoft is “breaking down applications piecemeal and trying to find meaningful compatibility.” As an example, he imagines a cable decoder box which, by simply purchasing an additional peripheral, could be upgraded to an interactive player like the Tandy Video Information System.

The underlying theme for Microsoft is the easy migration of information and application development methodology from one device to another. “The approach is to not throw the baby out with the bath water,” according to Glaser. The content should be compatible, even if the user interface and form factors change from device to device. “The only thing that needs to change between television and portable [devices] is the user interface,” said Parthasarathy.

But this assumes that PC software developers and those creating consumer applications actually have something in common. For the most part, they do not. The two are dramatically different. While data compatibility would certainly ease the load in moving the same application across different devices, how many applications would you want to access from a personal computer, your cable system, an interactive media player and a personal communications device?

As we’ve witnessed in many of the “reference” works available for the VIS platform (or Philip’s CD-I for that matter), it is clear that the process of looking up information without a keyboard is extremely cumbersome, and would drive back to the bound pages of a book anyone who does not have hours of free time on their hands (and a very high tolerance for psychic pain). What good, for example, is an electronic cookbook that cannot be attached to a printer and is located in the living room, not the kitchen?

As for cable television, Microsoft is offering something that the cable industry does not have: an operating system and experience in user interface design. But the company certainly does not have the solo offering; Apple, GeoWorks and 3DO are just three companies that have publicly announced their interest in the “device” market, and there are sure to be many others in the coming months.

The questions for Microsoft are whether leveraging off of the personal computer development platforms can actually help create programming and services for the cable industry, and whether Microsoft’s offering will satisfy the average consumer.

David Baron