This article may contain URLs that were valid when originally published, but now link to sites or pages that no longer exist. To maintain the flow of the article, we've left these URLs in the text, but disabled the links.

flux@microsoft.com

Douglas Boling

The Book of the Future

he concept of an electronic book, a device that can download text and display it, has fascinated futurists for a generation. What's even more interesting to me is that this idea hasn't yet caught on. We can easily produce a tablet device that has sufficient storage and can display text. Why aren't people buying?       The reason may be that books—those antiquated things with hundreds of thin sheets of processed wood—are in fact excellent information display devices. Books are small, relatively lightweight, and cheap. The spatial relationship between the information and the page the information is displayed on also makes books useful—or so says a team of researchers at MIT. When you think about it, they're right. When I pick up a book to reread a section, I don't look at the index or the table of contents. Instead, I thumb through the book to find that item I remember being on a right-hand page about halfway through the book. It's faster that way.       The MIT researchers are attempting to create an electronic book that maintains the page feature of books by having hundreds of pages of text that can be changed electronically. They describe their project in "The Last Book" in IBM Systems Journal. This paper describes the team's quest to create an electronic book that has the form factor of a real book. The key to achieving this goal is in the display technology.       Current display technology uses a transistor array to display the matrix of pixels. These transistors are grown on a glass substrate, making the display heavy and inflexible. Another problem with current display technology is the high power requirement, approximately 2.5 watts. And because of the need for working transistors for each pixel, the production yield is low and therefore the cost of these displays is high.       Instead of evolving current display technology, the MIT team is developing something new and completely different called electronic ink. Electronic ink is a compound that is composed of bistable particles that can be manipulated so that they later look either black or white. Once you have such a compound, you can spray it on a piece of paper. You then use an electric field to flip selected particles, or what can now be referred to as pixels, so that the page displays text.       The team already has a compound composed of these bistable particles. The particles themselves are composed of an inner two-color particle that is electronically charged. This particle is then encapsulated in an outer clear shell. Changing the color of the particle is simply a matter of flipping the inner particle so that either the white or black side of the particle is showing. Since the particles are stable when showing either color, the ink only needs power when the display is changed. This dramatically reduces the electrical current required for the display.       The fact that this electronic ink can simply be sprayed on a surface opens up a number of new possibilities for displays. For example, you could put this type of a display on a curved or even a flexible surface. Try doing that with today's active TFT displays. Conceptually, you could even own a rather expensive T-shirt that could have a different saying on it each day.       Once you have a sheet of paper that has this ink on it, you still need to be able to address each of the particles. To do this, the paper is "printed" with a transparent array of microfine conductors that can be used to flip the particles to form the text.       The MIT team claims that the cost for this technology will be somewhere between one to 10 dollars for each 8.5 X 11-inch page. You have to figure that if these guys can pull this technology off, the cost will drop even more due to the anticipated production volume. Even at this price, the pages of a 300-page book might cost $300, but you'd only need one. You would download the text into the electronic book as necessary. You could also have lower cost books with fewer pages that only displayed a chapter or so at a time.       This, of course, brings us to the problem of storage. But with storage technology evolving at an exponential pace, that isn't an insurmountable problem. Storing a single book takes approximately 1MB. That's uncom-pressed, and text is, fortunately, very compressible. So you can see that a single 4MB flash chip embedded in the spine of an electronic book could easily store 10 books. Flash PCMCIA cards already have a capacity of over 300MB. So for a book with a PCMCIA slot, you're already talking about a good-sized personal library. By expanding the storage options beyond flash cards, you're easily within reach of tiny PCMCIA-sized storage capacities in the multigigabyte range. At that point, you've passed the number of books in many small-town libraries.       Once you have a book like this, you have to start dealing with the publishing infrastructure. Some bean counter is going to immediately try to make readers buy text on a pay-for-use basis instead of the buy once read forever method used today. I don't believe a pay-for-read process is going to fly when people are used to reading books bought by their grandparents 50 years ago. So it will be up to the publishers to determine if electronic books can be successful.

From the October 1998 issue of Microsoft Interactive Developer.