Our vision of the NII is based on a wide-area, high-performance, client/server model. This model includes large parallel database and compute servers, high-speed networks as well as alternative network solutions (e.g., ISDN), and distributed networks of personal computers in the classroom. How will the promise of the NII be delivered to the classroom? Technical issues to resolve include integration of system components (e.g., parallel databases, ATM networks, classroom interfaces) and scalability (e.g., storage and retrieval of large video archives). Classroom user requirements include ease of exploration, the opportunity to learn from many types of linked visual - language materials, and extensive resources that can be called upon to answer learner-generated questions and solve curiosity-driven problems along with problems posed by the teacher to ensure an integrated educational context. We propose the Living Textbook as a prototype Education Information Infrastructure of the future. Our software development and delivery model is based on the vision of multiple-use services. Our K-12 education applications take advantage of technologies previously developed for large scale scientific simulation and parallel database applications. Our delivery model takes advantage of bandwidth installed for other purposes. We are sharing software (e.g., NASA terrain rendering) and expertise with related applications (e.g., MPEG compression) and re-using a rich high-performance computing and communications (HPCC) base infrastructure. 4.1 Technology Vision Information Dissemination Model

We designed the Living Textbook to support information-on-demand for education, and follow certain guiding principles agreed upon by the national HPCC community. We use community standards whenever possible such as High Performance Fortran, Message Passing Interface based software development models, and MPEG viewers within standard web interfaces into multimedia libraries. We build applications on top of pervasive and commercially viable base technologies. HPCC technologies are intrinsically cost-effective as they are replications of mass produced components. We emphasize scalable technologies and strategies that work across a range of current and future systems and technologies such as scalable software standards, parallel architectures, and communication strategies from ISDN to ATM digital delivery. We developed a multimedia client-server system using a distributed computing model where nodes can be either sequential or parallel computers. Information access leverages World Wide Web technologies. We integrate users friendly applications on the client side (e.g., WorldView Inc. 3D Macintosh navigation environment) with special data production services (e.g., parallel terrain rendering modules on the server side). Our server model includes stand-alone Macintosh or PCs in the classroom accessing subsets of the Living Textbook project via CD-ROM storage. The full client-server system connects workstations in the classroom via an ATM Gigabit network to parallel multimedia servers with terabytes of hierarchially arranged storage. Several smaller clients connected by ISDN bridge these two solutions. NPAC is able to develop this model in a highly demanding and diverse environment with facilities that include Thinking Machines CM5, Intel iPSC 860, two DECmpp's, IBM SP2, and nCUBE 2, and clusters of high performance workstations (IBM RS6000's and DEC Alpha's) NPAC also has a number of high-end database products installed on our MPP platforms, including Oracle's Parallel Server running on the nCUBE and SP2, and Oracle's Parallel Text Server, running on the nCUBE2. We implemented our own video pump software on the nCUBE2 and use this platform as a Video-on-Demand server. Delivery Model We are building a scalable, balanced model for delivering interactive multimedia software to the classroom. We currently focus on ATM delivery, with future extensions including widely available network solutions such as twisted pair, ISDN, T1-T3, and possibly cable/fiber hybrids. Our testbed includes systems which differ in performance (computer power, network bandwidth, storage capacity) and cost by several orders of magnitude. Interface Our classroom interface is based on standard web browsing software (NCSA Mosaic, Netscape). Developments efforts underway at NPAC to extend the level and kinds of interactivity supported by the Web server and client will be incorporated into the Living Textbook as these tools become robust. We envision a future classroom interface built on top of web tools tailored for educational use, including tools such as a book reader, video clip editor, and tools for managing information collected from the World Wide Web. New Learning Model in K-12 The new national standards for math (NCTM) and science (NAS goals 2000) emphasize improved problem-solving abilities and cross-disciplinary studies that address basic science and mathematics in real world problem solving situations. The explosion of resources and users on the World Wide Web has created a vision of the information world of the future. This information infrastructure must effectively become integrated with content-rich interactive learning environments. These environments must (a) provide a basic framework of knowledge, (b) support and encourage interactive experimenting by the learner, typically including various processes that initiate simulations and other computations or processes that gather information or manipulate data, and (c) operate as windows into relevant information and databases on the global Internet. We will require multiple representations of data and information to enable knowing about something in multiple and complementary ways (eg, through sound, 2-D and 3-D imagery, text descriptions, statistics and graphics of properties, and both static and dynamic representations). The integrated learning/information environment of the future will support the ways that learners obtain real experience, do science, explore, create their own environment, ask questions, find resources, solve their own problems, and answer their own questions. One of the current issues for which existing technology supports exploration is linking related data, information, and knowledge in a more information driven fashion than the WWW supports. This issue is being addressed in a different context, for example, by the NASA EOSDIS group that is devising methods to maintain links between data and the metadata needed to make it meaningful in addition to the methods used for different types of analysis and manipulation of the data. The Living Textbook applications will become more than just visual or visual-audio elements. An upcoming workshop is designed to obtain teacher input on how to begin integrating Living Textbook applications and technologies into their classrooms. Our model includes introductory workshops to allow teacher exploration and feedback on the applications. We examine how the application is conceptually useful, practical issues, and what gets in the way. We will be in a position to evaluate whether the Living Textbook facilitates introduction of curriculum material in related areas and how it affects interest. Students will use Living Textbook tools to create their own multimedia resources. We can evaluate what they create, study how they use resources, and obtain feedback on additional support the tools might provide on effective location and integrations of resources. One measure of success will be to see computers in the classroom used NOT as electronic typewriters and spreadsheets, but as research tools. Another will be to see teachers integrate lectures with student interactive learning, and learn how to intermix them effectively.