Chapters 1
-
1 open library vision 17:27
Acknowledgements. It is an honor to work with so many people in nonprofit, government, commercial, and educational institutions to pursue the ancient dream of an open library. The Internet Archive's research and development have been supported by a generous grant from the Alfred P. Sloan Foundation and the Cale Austin Foundation. PDF Books has been developed in association with HP Labs, Luritech, and Adobe. Deja Vu scans have been developed with Lizardtek. The scanners used have been designed and manufactured by the Internet Archive and Kirtis Technology. Librarianship and Scanning has been supported by Carol Moore of the University of Toronto and eight other Canadian libraries, the University of California Library System, and with the generous support of Yahoo!
Developers on the Internet Archive Scanner and associated software include Stuart Blair, Todd Cass, Molly Davis, John Devinesia, Tracy Jacketh, Mark Johnson, Gabe Jussel, Brewster Kale, Joanne Moth, Tom McCarty, Robert Miller, Brad Newburgh, Stephen Rosenberg, Ariana Roxborough, Earl Sterling, Ronna Tannenbaum, and Roxanne Williams. The Million Books Project has been supported by Carnegie Mellon University, the National Science Foundation, and the governments and libraries of India and China.
Vision of an Open Library Today we are making a step towards building the library we have all dreamed of. The library that makes all the published works of humankind available to everyone in the world. While large in scope and ambition, the goal is within our grasp. Achieving it requires librarians, authors, government officials, and technologists to build towards a shared vision. Fortunately, we are seeing this happen. And what a world it could be! Imagine a neighborhood library that has millions of books on its virtual bookshelf, So that the talented math whiz in a rural community can explore high math without restriction.
Imagine a recent immigrant from a distant country who can now share the children's books that were so important to them growing up, with their own children. Imagine an elderly person able to have a large print edition of any book ever published, or any book read aloud to them. In this future, a talented young scholar can publish a new book to this great library, in order to illuminate some subject that might not previously have made it through the difficult publication process. Haven't we already achieved this with the World Wide Web? Well, yes and no. There is more information available to those with computer access than was ever imaginable, but few writings on the Web are more than ten years old.
There is much worth reading from all the ages. Also, The book format is likely to be useful for many years to come, even if it is moved from place to place via computer networks. Won't some of the big commercial digitization projects deliver this future? They are a part of it, but if we go no further, we may have an expanded bookstore, or a single means of organizing the materials, but we may not be building on the open tradition of a library, a place where librarians help organize and present materials, where patrons can draw from and add to the holdings, where the right to private study is upheld, and where physical facilities offer a meeting place and environment for learning.
We have been asked if authors and publishers will play a role in this vision. The answer is a resounding yes. Publishers and authors create the works in our libraries. In the United States, we spend $12 billion a year on our library system. Of which three to four billion go to publishers. We are not suggesting we spend less, but rather that we spend it better. With such a public commitment to a free flow of information, we can build a great library leveraging these new technologies.
So why an open library? Because we can combine the best that the library system and publishing industry offer to build towards universal access to all knowledge. This book describes a vision of an open library and how we are building it. Many of us are coming together precisely because of the potential of such a project. Together, we can build a great library. From book to book. To make this concrete, let's step through a piece of the bigger task, how to make the books published through the ages, universally available in this open library. How can we take advantage of the new technologies and leverage the interest of authors, companies, and governments to build the future we want to live in? Let's outline how we can begin with the books on the shelves of our libraries, digitize them, catalog them, serve them on the internet, and view them on a screen, but do more.
Print and bind, creating a book that looks like the original. In other words, what is the process of going book to book? People are inventing interesting digital devices for reading and browsing books, but for many readers, a faithful facsimile of the original book may be the best way to enjoy the work. This also holds us to a high standard of digitization, if the digital book is to be judged by at least the standards of the original. Selection. To begin, librarians select works in their collections and oversee their digitization. Based on the economics of digitization, they can bring works online, subject by subject, or by physical location. The Open Library is built by librarians to create comprehensive collections with finding aids.
Since the Open Library isn't limited to a single library's or librarian's selections, it represents many points of view. We can join multiple collections to create new virtual collections that never existed in physical form. For example, An early open library collection is a cooperative collection of American literature developed from the holdings of the University of California libraries and beyond. Ultimately, we would like the open library to include all books. But for the moment, a selective collections-oriented approach will help us build a bibliographically coherent collection of immediate use to educators, researchers, and the public. Digitizing can now be economically done inside the library.
This avoids the uncertainties of shipping books and permits librarians to provide active feedback on how the digitization process should proceed. The skills of selection, prioritization, handling, cataloging, and quality control are all part of the training and passions of our librarians.
Recording images of the pages of a book without hurting the book or its bindings, quickly and inexpensively, is now possible. Fortunately, three recent improvements in technology have made high-quality scanning much more economical. The best digital cameras now have sufficient resolution to equal scanning using flatbed or sheet-fed scanners. Hard disks have the larger capacities needed to store and deliver these images. Lastly, clusters of computers can process these images into compressed and accessible formats. The Internet Archive integrated these into the Scribe system, but other scanning systems are also available. With the Scribe, the book is held facing up, open at only a 90-degree angle, as an operator manually turns the pages.
The pages are held flat by a glass platen that is raised and lowered to hold the page flat. The majority of books are small enough to be scanned at 500 color pixels per inch, but even books 16 inches high can be scanned at 300 pixels per inch, which is a scanning standard.
Current image quality. Taking a picture of the page with controlled lighting and fixed angles helps create a consistent image. Getting even lighting requires special lamps to illuminate an angled cradle. Similar care in lighting and compensation is necessary in order to capture true colors and make a facsimile that looks like the original. While the work done on the scribe does not compare to the fine art imaging of Hewlett-Packard Laboratories or the artisanal scanning of Octavo, we have worked with these groups to do what we can to maintain excellent image quality while maintaining a steady workflow.
Maintaining Quality. With the high-end digital cameras we are now using, we can record most books at 400 pixels per inch, and many at 600 pixels per inch, in full color. While this is insufficient to reproduce some treasured historical artifacts, many believe this is more than enough to read and enjoy books on computer screens, to print faithful facsimile editions, and do advanced data processing to recognize text and graphics. The result is a set of images and processed images that are searchable, readable, downloadable, and printable. Popular access formats are supported, and future formats can be created because the uncompressed original images are saved.
Cataloging. Cataloging and organizing these works is necessary so that patrons can find the books. Cataloging can also enable the coordination of digitization activities in different libraries and assure completeness of the digital record. We therefore need to match the digital catalog record to the digital version of the book. In most cases, this is easy, because we are fortunate that the library community has converted its card catalog data into digital form over the last 30 years. While not all books are digitally catalogued, the majority of the works in our research libraries are. Metadata is read and processed in the standard formats of MARC Record and Dublin Core, easing interoperability with other systems.
This is added to other metadata we collect from the book, such as reviews and descriptions.
Copyright. While many important books are still under copyright, There are many classic works that could find broader audiences if they were easily available. Determining the United States copyright status of a work is sometimes difficult, even for professionals, says Jessica Littman in her seminal book Digital Copyright, Protecting Intellectual Property on the Internet. But in many cases there are useful formulas, such as Stephen Fishman's The Public Domain, How to Find and Use Copyright-Free Writings, Music, Art and More. Evidence of copyright status can be determined by the publication date, whether a work was originally registered for copyright, and whether the copyright has been renewed.
The open library strives to make materials as openly available as is legally allowed. The initial collections will be built using materials that are now in the public domain. Digitized, out-of-copyright books can be offered for free access without restrictions, allowing patrons to read, print, share, study, and excerpt. The Open Library will rely on Creative Commons licenses to encourage the greatest possible degree of access to and reuse of the materials, consistent with respect for the rights of content owners and contributors. As the technology of digital books develops, we hope the market also makes distributing in-print materials profitable, so all books will be available online.
The uncompressed image of a page is 20 megabytes, which means that a 300-page book occupies 6 gigabytes. A million books, then, would consume 6 petabytes. The scale goes mega, giga, terra, peta, which is a challenge with today's technology. The Internet Archive designed a system specifically for preservation and access to petabyte-sized collections, called the Petabox. A commercial company, Capricorn Technologies, now manufactures these machines. Preservation of this data requires moving it onto new systems every three years and replicating in multiple locations. The Internet Archive has partnerships in Europe and Egypt to help ensure the long-term care of these digital artifacts. Book readers.
Despite all the work on e-books, Allan Kaye's Dynabook, and e-Ink, reading books on a screen is still frustrating. Hopefully, the availability of many desirable books in electronic format that are open for experimentation will encourage the development of new ideas and new designs. One version, built on the ideas of a British library reading kiosk, is shown here. Integrated with the browser, this animates page turning and highlights search matches with stickies and a virtual highlighter pen.
New readers possible. With an open library, there may be new experiments in on-screen books that combine the advantages of both books and screens. This interface by the University of Maryland Human Interface Group is a Java-based program designed for children. Please try www.icdlbooks.org to try their creative interface.
Print on Demand. For hundreds of years, most books have been made in editions of thousands. The setup costs of lead type or offset printing has meant that smaller runs are often quite expensive. Computer printers are more expensive per page, but do not have these setup expenses. This makes print-on-demand or editions of one or a few economical. One such company doing this is lulu.com. which prints a single book and ships it for the price often charged by a larger-scale publisher. This exciting development can mean that the books in the open library can be printed and sent to patrons for a nominal fee.
Internet bookmobiles. Printers and binding machines are now inexpensive enough That a small van can carry all the equipment necessary to make books from the Internet. Connecting to the Internet with a satellite connection or just by storing the books on hard drives, the librarian can help anyone download, print, and bind a book. If the book is black and white and 100 pages long, then the materials to make the book cost one dollar. This is less expensive than total costs of lending a book, as reported by some libraries. This technology is already in use in Egypt, India, the United States, and Uganda.
One hundred dollar laptop project to bring digital books to millions of children. An ambitious project to build an appropriate laptop for children all over the world will be offering public domain books to their users. The Chief Technology Officer, Mary Lou Jepson, said, Since the $100 laptop is designed to be an electronic book reader, as well as a general-purpose computer, We are very happy to offer the international book collection of the Open Content Alliance to our users. Audio versions. Audio versions of classic books can be shared as easily as the images of the pages. It may turn out that a major use of the books in the open library will be to create audiobooks. A new group, LibriVox, is coordinating the reading of classic literature by volunteers and offering the recordings in the public domain.
They have recorded an open library book, an international episode by Henry James, for the launch event, which has been integrated into the book display.
Books accessible to all. The visually impaired can now enjoy all the books in the library because the digital form can be transformed into the formats used by reading machines and special printers. Bookshare.org is a non-profit organization that brings in-copyright and out-of-copyright books to the visually impaired. Kristen, Kara, Brad, Gord, Sureka, Eddie, Alex, Hugh, Chris, and Sally.