DIGITAL VS. TRADITIONAL LIBRARIES

The shift from traditional libraries to the digital is not merely a technological evolution, but requires a change in the paradigm by which people access and interact with information.

A traditional library is characterized by the following:

• emphasis on storage and preservation of physical items, particularly books and periodicals
• cataloging at a high level rather than one of detail, e.g., author and subject indexes as opposed to full text
• browsing based on physical proximity of related materials, e.g., books on sociology are near one another on the shelves
• passivity; information is physically assembled in one place; users must travel to the library to learn what is there and make use of it

By contrast, a digital library differs from the above in the following ways:

• emphasis on access to digitized materials wherever they may be located, with digitization eliminating the need to own or store a physical item
• cataloging down to individual words or glyphs
• browsing based on hyperlinks, keyword, or any defined measure of relatedness; materials on the same subject do not need to be near one another in any physical sense
• broadcast technology; users need not visit a digital library except electronically; for them the library exists at any place they can access it, e.g., home, school, office, or in a car

Everything Can Be Stored

The total number of different books produced since printing began does not exceed one billion. (The number of books now published annually is less than one million.) If an average book occupies 500 pages at 2,000 characters per page, then even without compression it can be stored comfortably in one megabyte. Therefore, one billion megabytes are sufficient to store all books. This is 10¹⁵ bytes, or one petabyte. At commercial prices of $20 per gigabyte, this amount of disk storage capacity could be purchased for $20 million. So it is certainly feasible to consider storing all books digitally.

Very Large Databases

A database of a billion objects, each of which occupies one megabyte, is large but not inconceivable. Once one is comfortable with sizes of this kind, it is feasible to imagine a thousand such databases, or to envision them all as portions of the same global collection. This amount of storage is sufficient to house not only all books, but all of the following:

• photographs
• legislative material, court decisions
• museum objects
• recorded music
• theatrical performances, including opera and ballet
• speeches
• movies and videotape

Distributed Holdings

When information is digitized and accessible over a network, it makes little sense to speak of its "location," although it is technically resident on at least one storage device somewhere, and that device is connected to at least one computer. If the information is available at multiple mirror sites, it is even less meaningful to speak of it being in a "place." While traditional libraries measure their size by number of books, periodicals and other items held, the relevant statistic for a digital library is the size of the corpus its users may access. This means that digital libraries will want to expand their "holdings" by sharing digital links with other libraries. Unfortunately, there seems to be very little sharing of this sort taking place at present.

How can we understand the unwillingness of libraries to share content? The question goes back to the old measure of the size of a traditional library¾the number of books it holds. When a library expends funds to assemble digitized works, it loses a portion of its prestige (or thinks it does) by allowing other libraries to copy or access its data. Ultimately, however, all material should be accessible from every library.

Gore's Digital Earth

In 1998, Vice President Gore stated that "A new wave of technological innovation is allowing us to capture, store, process and display an unprecedented amount of information about our planet and a wide variety of environmental and cultural phenomena. ... I believe we need a 'Digital Earth.' A multi-resolution, three-dimensional representation of the planet, into which we can embed vast quantities of geo-referenced data" (Gore 1998). He then called on scientists to create a digital map of the earth at a resolution of one meter. Such a project will require technical innovation beyond that required even for a digital library containing every book ever written. The area of the earth in square meters is about 5 x 10¹⁴. Storing two megabytes of data per square meter (which would include terrain data, imaging, environmental and other pertinent information) will require 10¹⁸ bytes, an amount roughly equal to the amount of digital storage currently present on earth.