WEB HISTORY DAY:
PIONEERING SOFTWARE AND SITES
Program * Abstracts * Exhibit
Andy van Dam
Theodor Holm Nelson
Professor of Environmental Information,
Keio University SFC Campus
Wake-Up Call From Ted Nelson
Ted Nelson coined the term "hypertext," and along with Doug Engelbart is one of two independent inventors of hypertext on computers.
In 1960 he began to design electronic documents and literatures for digital delivery, having decided simply as a leap of faith that all text would soon migrate to the digital screen. He made this leap of faith on the basis of one computer course, his readings in science fiction, his desire for a very different world, and his background in theater and writing - in particular from grappling, as a writer, with problems of version management and self-publishing.
Ted's vision of 1960 was a giant network of hyper-linked documents of a certain structure distributed on servers throughout the world, which he was later to call Xanadu. (Ted did not hear of Doug Engelbart's parallel work until he had already published his first design in 1965.)
In 1968, Ted collaborated with his former frisby-mate at Swarthmore, Andy van Dam, to build HES, the Hypertext Editing System, at Brown University. Like Doug Engelbart, Ted and Andy had to invent what we now know as real-time interaction with computers, with no support and some hostility from the computer science community of the time.
HES was originally going to be built around Ted's hypertext ideas, but against his wishes the project turned to an emphasis on printout and formatting, becoming instead the prototype of the word processing systems of today. While most people would count this as an achievement to be proud of, Ted considers it a career disaster. He believes that word processing is based on a fundamental misunderstanding of the nature of text and the nature of ideas, and believes that the HES project is one of the key junctures where computerdom went wrong. (All that remained in HES of his hypertext ideas were the link and the return stack, now immortalized as the "back" button on Web browsers.)
For years Ted continued improving his Xanadu designs, seeking support from people who thought very differently. Because his ideas were so different, many people who talked to Ted thought he was technically ignorant when he was in fact trying to steer them in other directions, which in those days he tried to do politely.
But the politeness began to wear thin. Growing sick of the computer establishment, he wrote up many of his ideas impolitely in Computer Lib (1974), which became a bible for a number of software visionaries, and Literary Machines (1981), which detailed the Xanadu project of that time.
Ted insisted from the beginning that Xanadu have a specific document structure which is still not widely understood. While Xanadu's design had basic hypertext features somewhat like those Doug Engelbart built into NLS, Ted foresaw and dealt with problems that have only just begun to confront the Web today. Xanadu's highly-generalized link structure would allow many links to and from the same materials - links that could be followed in either direction, and would survive most editing changes. This was because the links were not embedded in the data, but attached sideways to characters and other elements from a different zone of the document, and themselves individually addressable. All documents were virtual, structures of addresses pointing to frozen base files.
But the cornerstone of the Xanadu design was transclusion. Transclusion is the logical center of Nelson's document universe, and he believes it must be fully implemented at whatever cost. Transclusion provides methods for deep intercomparison, version management, re-use and republication, ownership and fine-grained micropayment. Carried even further, transclusion provides an alternative to hierarchical directories in new kinds of operating systems.
Transclusion has been widely misunderstood. A logical structure rather than an implementation method, transclusion is a special kind of inclusion by reference, which people tend to mistake either for a kind of link, or for "only one copy of anything." It means neither of these. It means tracking all the manifestations of any data, and keeping them connected to their virtual cosmic original, throughout the docuverse. (Ted strongly opposes trying to state this in more conventional terms, which he believes confuse difficult issues to the point of hopelessness.)
Ted firmly believes that this is the structure everyone is groping toward from many different directions and for many different reasons, and that when even a few people understand, it will abruptly and universally prevail.
Most people are unaware of the breadth of Ted's paradigm or its theoretical foundations. However, these are widely appreciated in Japan, where he is finally implementing this alternative virtuality with the help of his close collaborator, Marlene Mallicoat.
Ted will discuss the history of his vision, emphasizing features which need to be brought to the Web today, and his efforts to do so through the ongoing Xanadu project.
Founder, Bootstrap Alliance
Lunch With Doug Engelbart
Douglas Engelbart is the inventor of the mouse, and with Ted Nelson one of two independent inventors of computer-based hypertext. In the 1950s, Doug started thinking about ways of augmenting human intellect with computers. At the start of the 60s he began building what became NLS (oNLine System), one of the first two computer-based hypertext systems. Along the way he had to promote the concepts of word processing and real-time interaction with computers, ideas that seem obvious now but which were very radical in those batch-processing days. NLS grew to include all the features we now associate with hypertext, including many which have yet to be implemented on the Web; typed links, links to links, automatic multiple views of the same information for different levels of users, and more. Two fundamentals of NLS remain blue sky concepts in the contemporary software industry; the idea that any document, note, link etc. can be addressed as an independent object, and the "document centric" model of documents being editable by multiple applications rather than belonging to a single one. In the late 60s, Doug's SRI lab become one of the first sites on the infant ARPAnet (direct ancestor of the Internet). Doug's team developed one of the best early email systems, and founded the Network Information Center (NIC), the first attempt to organize a comprehensive "card catalog" for the information available over the net. Both the email system and NIC were direct outgrowths of the work on NLS. Doug will discuss his original NLS system, demo AUGMENT, its successor, and discuss the Journal System, an important and unique component of AUGMENT. He will describe the features which he believes most urgently need to be introduced to the Web today, and his current efforts to do so through the Bootstrap Alliance.
Andries van Dam
Founder/Professor, Brown University Department of Computer Science
Thomas J. Watson Jr. University Professor of Technology and Education
Executive Vice President and CTO, Macromedia
The Web Before the Web: HES, FRESS and Intermedia
Andy van Dam and his colleagues at Brown University have produced four seminal hypertext systems. Andy will discuss HES, the mid 60s system he developed with his college friend Ted Nelson, and demomstrate FRESS, the late 60s system he developed after meeting up with Douglas Engelbart and seeing NLS. FRESS made use not only of alphanumeric display terminals but also of a then cutting-edge IMLAC graphics minicomputer 16 bit PDP-8, attached to mainframe via communications line and used as an intelligent terminal, which let the system incorporate windows and vector graphics. FRESS could represent any character (non-Western alphabets, math symbols, etc.) with a system of vector strokes. FRESS was the first hypertext system to see systematic use in for-credit courses, including a poetry course in the mid-70s sponsored by a National Endowment for the Humanities. FRESS was used for over two decades at Brown for personal hypertext libraries and courses and had several commercial spinoffs. Andy will also discuss his late 70s Electronic Document System. This graphics-based hypertext system presaged Web image maps by featuring various types of links between graphic objects. It was used for research into online maintenance and repair manuals, where users could follow links between pictures, text, and thumbnail views of pages or diagrams.
Thumbnails were created on the fly to present graphical views of the chronological path, and had three uses: for a time-based/historical view, for the ability to show a "you are here" map for any page showing previous and next links, and for boolean retrieval results showing pages in each page and chapter that satisfied the query. Like NLS, all three systems had integrated browsing and editing environments.
Andy is the founder of Brown University's Computer Sciences department and was its first chairman; he is also a pioneer in computer graphics. He co-founded ACM SIGGRAPH, and co-authored, with Jim Foley, the standard reference works in computer graphics, Fundamentals of Interactive Computer Graphics, and with Foley, Feiner, and Hughes, wrote Computer Graphics: Principles and Practice.
Norm Meyrowitz, the "father" of Shockwave, will demonstrate Intermedia, the hypermedia system he designed after working with Andy Van Dam. Mr. Meyrowitz helped found IRIS (Institute for Information and Scholarship), at which he designed Intermedia. Founded by huge grants from IBM and the Annenburg Corporation, IRIS ran under Mr. Meyrowitz's direction from the early to late 1980's, at which point Mathan Mhyrvold of Microsoft tried but failed to buy the Intermedia team. Intermedia was perhaps the best-realized and most sophisticated of the mid-eighties hypertext systems, and was used extensively in Brown University classes on a variety of topics. Less sophisticated hypertext systems began to be marketed commercially in that era, including Apple's HyperCard and OWL, and were the ancestors of the proprietary navigation systems used on many CD-ROMs today.
Director and Founder, World Wide Web Consortium
The Web's Inventor
Tim Berners-Lee is the inventor of the World Wide Web. In 1980 he wrote Enquire Within, an information organizing program which foreshadowed the Web in significant ways, including an open-ended set of relationships between documents and typed links. He will discuss and possibly demo Enquire. Tim proposed the World Wide Web in 1989, then again in 1990 in collaboration with Robert Cailliau. Tim wrote the code for the Web much as we know it today in the fall of 1990. All of this occurred at CERN, the giant European physics research laboratory based in Geneva, Switzerland, where Tim was a programmer. Tim will demo the first browser in the world, his original GUI NeXT browser-editor from the fall of 1990, on his original NeXT cube. This browser seamlessly integrated browsing and editing, and resembled a simple word processor plus Web capabilities. Tim later studied the works of Ted Nelson in detail, but invented most of the core concepts of the Web independently. He will discuss his vision of the Web at that time, which included many features since reinvented by the Web community; multimedia capabilities, server APIs, intranets, fuller implementation of SGML standards, and virtual pages. He will also demo features proposed in the pre-Mosaic era which remain "lost"; typed links, multiple options for how to display linked objects (embed, jump to, show in separate window, etc.), scripts to automatically generate documents from Web pages, and full integration of browsing and editing.
Samba, the First Mac Browser
Robert Cailliau was proposing his own hypertext system for documents at CERN when he discovered the Web project. He joined Tim Berners-Lee in 1990 and adopted the much broader Web philosophy. Robert evangelized the Web to CERN management, the physics community, and government bodies. He implemented the first Macintosh browser, Samba, and the current Web conference series. Robert will discuss early history, particularly the difficulty of getting the Web concept across, and demonstrate Samba.
Director and Founder, Infodesign, Geneva, Switzerland
Building the Web
Just after Tim Berners-Lee created the first version of the Web in late 1990, Jean-Francois Groff joined him as the first technical student on the project. Together, they worked on designing and implementing significant portions of the Web's architecture and protocols. Jean-Francois made significant contributions to the original WWW code library at CERN, and helped worked out many of the features of the Web as we know it as well as many yet to be implemented. Jean-Francois started www-talk in 1991, the online mailing list which served as the central "forum" for the early years of the Web's development. Jean-Francois was a founder of the Virtual Library, a prototype "card catalog" for the Web, and ported a version of the Web to DECNet. Other tasks included an API-based HTTP server, and the login trap which helped early adopters learn about the Web. After leaving CERN in 1992, Jean-Francois founded InfoDesign, the world's first Web services firm. Jean-Francois will assist Tim in demoing the NeXT browser-editor, and discuss in detail many of the now "lost" features he helped develop in the very early 1990s. Jean-Francois will also talk about design decisions and the process of innovation.
Mosaic: The Browser That Popularized The Web
In Fall of 1992, the Software Development Group at the National Center for Supercomputer Applications (NCSA) was producing Collage, a program to let researchers collaborate over networks. As the project neared completion, programmers at the Group noticed an ambitious project on the Internet called the World Wide Web. They quickly realized that Web compatibility could turn the Collage project into something much broader than a collaboration tool, and the rest is history. Marc Andreessen joined www-talk, the mailing list forum for early Web developers, and received encouragement from Tim Berners Lee, the Web's inventor. Tim had been actively soliciting interested people over the net to develop UNIX and other versions of Web browser-editors. Marc wrote a crude web client interface over Christmas in 1992, and soon enlisted the services of his senior colleague and veteran UNIX programmer Eric Bina. The manager of the Software Development Group, Joseph Hardin, backed the idea. When the SDG demoed an early version to Larry Smarr, NCSA's founder, he was thrilled, and saw Mosaic as a possible successor to popular NCSA Internet applications such as NCSA Telnet in the '80s. Eric and Marc released the first version of Mosaic for X-Windows in early 1993. Backed by NCSA's credibility and resources, Mosaic provided would-be Web users with exactly what they needed to get started. Like many other browsers of the period, Mosaic was essentially an interface on top of the basic WWW code libraries provided by CERN. But unlike earlier and technically more advanced browsers, Mosaic was stable, very simple to install, and included simple bitmapped graphics. It was the first browser backed by sufficient resources to provide technical support and versions on different platforms, which the programming team moved quickly to produce for Mac and Windows. Key managers and programmers from NCSA will demo early versions of Collage, Mosaic, and the NCSA Web site, and discuss Mosaic's development history.
Founder, Internet Archive
The Web Before the Web: WAIS and Indexed Information
In the late 80s Brewster Kahle, a developer at Thinking Machines, began developing an advanced system for sharing information over networks including the Internet. Though his ultimate vision was strikingly similar, in many ways Brewster's approach was the exact opposite of that taken by the Web and many hypertext-based systems. Rather than encouraging user-directed "browsing" through relatively raw source material, WAIS (Wide Area Information Servers) was designed to give users targeted answers to sophisticated queries, from databases of carefully indexed information. Brewster saw the development of WAIS clients as a simple and inevitable consequence, and concentrated on the server side. WAIS achieved considerable commercial success with major publishing houses, large companies, and government bodies, who had both the need and resources to index their source materials for WAIS. But it failed to catch on with more casual users. In 1989 Brewster moved to the San Francisco Bay area and founded WAIS incorporated, one of the first successful companies based primarily on the Internet-related technology. He later sold WAIS to America Online, and is the founder and director of the Internet Archive. Brewster will demo historic versions of WAIS and discuss its evolution. He will also briefly discuss the Internet Archive and its relation to his original vision for WAIS.
Director, Multimedia Research Group
The Web Before the Web: Microcosm
In 1989, Wendy Hall and her team began developing another proto-Web, the open hypermedia system Microcosm. Like most hypertext systems, and unlike the Web, Microcosm stores links in a database rather than "hard-wiring" them into documents. This model can potentially solve many problems of version control and links to other systems. Microcosm developers will demo early versions and also Webcosm, the recent Distributed Link Service (DLS) which brings Microcosm linking principles to the Web. One of Professor Hall's assistants will be demonstrating Microcosm.
The Web Before the Web: System 33
In the late 80s Mark Weiser, Steve Putz, and others at Xerox PARC developed System 33, which foreshadowed some of the Web's multiple document format capabilities. This document sharing system let users interactively exchange documents of different sorts over a network, with format conversion on the fly. Tim Berners-Lee visited PARC in 1992, and incorporated some of System 33's ideas into later Web specifications. A 23-minute videotape about System 33 and its format capabilities will be shown.
IICM, Graz University of Technology, Austria
Frank M. Kappe
IICM, Graz University of Technology, Austria
HyperWave Research & Development
From Hyper-G to HyperWave: The Evolution of the Hyper-G Project
Hyper-G is another independent invention of something very much like the Web. At around the same time Tim began the Web at CERN, Hermann Maurer and his group at Graz University of Technology in Austria began developing Hyper-G. They produced what was in many ways a far more mature networked hypertext system than the Web, using a link database to manage hyperlinks, providing constructs for structuring information, definining rich, searchable metadata for objects, and integrating a full text inverted index. The original Hyper-G server was simultaneously a WWW and a Gopher server, speaking the corresponding protocol to clients.
With the popularity of the Web, emphasis shifted to enhancing Hyper-G's WWW facilities. Today, Hyper-G is a sophisticated Web document management system, marketed commercially under the name Hyperwave. Frank Kappe and Keith Andrews are two of the project's original developers, and they will demonstrate software clients from the text-only 1992 version to current commercial ones.
The First Applet Browser: Viola
In 1989, Pei Wei began creating an information sharing system he called Viola on his Amiga personal computer. He developed a vision of clients and servers using executable "applets" as a primary means for parsing and displaying information. The applet model has recently achieved fame through Java, independently reinvented at Sun Microsystems. When Pei heard of the World Wide Web project via the Internet in late 1991, he incorporated Web browsing capabilities into Viola, and wrote a UNIX version. ViolaWWW grew into the most sophisticated of the early Web browsers, and was the browser of choice for pre-Mosaic demos by the CERN team and most others. It was the first browser with support for stylesheets, tables, and nestable HTML elements, features which took years to resurface in later browsers. Viola was one of the primary examples of browsers used by NCSA in developing Mosaic, though the Mosaic team chose to leave out most of Viola's more sophisticated features to get a simple browser out quickly. Pei was hired by O'Reilly and Associates, a key early Web publisher, and developed later versions of Viola for O'Reilly. Pei played a key role in O'Reilly's launch of GNN (Global Network Navigator). He stayed with GNN after its purchase by AOL, and now works for the Internet Archive. Pei will demo and discuss the development of Viola, with a focus on early Viola applet technology.
Manager, Distributed Computing Support
Academic Computing Services at the University of Kansas
An Early History of Lynx
Lynx started as an innovative new way of organizing the CWIS (Campus-Wide Information System) at the University of Kansas in the early 90s. It soon became yet another proto-Web, based on HyTelnet, with documents transferred between servers and browsers via an integrated interface. Early Lynx visions included many features we now associate with the Web (though minus full hypertext linking), including easy remote access to "pages" with text and images. The Lynx team in that era included Computing Support manager Michael Grobe and their programming student Lou Montulli. When the Lynx team discovered the World Wide Web project over the Internet, they quickly adapted Lynx to incorporate Web browser capability. Programmed mostly by Lou, the Web-enabled Lynx became the premier text-only Web browser, far easier to use than CERN's reference text browser, and an important access point to the early Web for people using terminal connections to mainframes. Lynx was a major contemporary of the early Mosaic. The name Lynx was a word play on the Internet navigation protocol Gopher, and the final choice in a series of names of animals which hunt gophers. Lou Montulli was later recruited by Netscape Communications as a founding member, and works there today on core features within the Netscape client. Michael Grobe and possibly Lou Montulli will demo historic versions of Lynx and the University of Kansas site, and discuss its development.
Law Department, Cornell University
Cello, the First PC Browser
In 1992, Tom Bruce of the Cornell School of Law began developing Cello, the first Web browser designed for the IBM PC and compatibles. His primary motive was to improve access to legal information, which had long been computerized but on a variety incompatible, mainframe-based systems. He knew his legal users were mostly PC-based, and had little access to contemporary UNIX-based browsers Unlike other browser writers of that early era, including the Mosaic team, Pei Wei's Viola, Tony Johnson's Midas, and Ari Lemmke's Erwise, Tom couldn't simply copy the CERN code libraries developed by Tim Berners-Lee for the "guts" of his browser. The PC of 1992 was simply not powerful enough to run that UNIX-oriented code. So unlike the others he essentially rewrote the functionality of the CERN library for his own purposes. Cello achieved a fair amount of use and recognition within the legal community, and some amongst PC users - on the order of 150 to 200,000 of them. But like other "one-man show" browser writers, Tom simply couldn't keep up with the support requests from users and eventually let the project fade away. Tom and/or assistants will demo and discuss Cello and the various stages of its development.
Hypermedia Engineer, VeriFone Internet Commerce
Pioneering Sites: The Hawaii Site
In May 1993, when Kevin Hughes set up a Web site at Honolulu Community College, it became the first educational institution to integrate sounds, images, and movies over the Web. This site pioneered the use of many design techniques common on the Web today, as well as the use of interactive images and narrated media. The "Hawaii site" became one of the handful of sites most commonly used for early Web demos, from audiences ranging from the top members of the European Commission to the U.S. White House. Kevin went on to work for Marty Tenenbaum of EIT and designed the early CommerceNet site as well as the Internet Shopping Network, sites which along with GNN established many conventions for doing Web-based advertising and shopping. Kevin was entered into the World Wide Web Hall of Fame at the First International Web Conference in 1994. He will demo and discuss the early Hawaii site and the evolution of Web-based design.
Welcome to the White House: An Interactive Citizens' Handbook
Vice President Al Gore is well known for his interest in networked information, and proposed an "information superhighway" as a plank in the '92 campaign platform. But his interest and involvement were both longer and deeper. In 1991 he had Brewster Kahle demonstrate WAIS to him, and with his and the President's support the Democrats pioneered the use of email for communication during the '92 campaign.
In the spring of 1994, Tom Kalil, David Lytel, Jock Gill and many others helped produce the first White House Site, and spent many months facilitating the creation of sites by most of the largest government agencies. They did this with crucial backing from the President and Vice President, and technical assistance from NASA's Ames Research Center, the AI Lab at the Massachusetts Institute of chnology, Tony Rutkowski and the Internet Society, and the University of North Carolina at Chapel Hill's SUNSite.
Pioneering Sites: CommerceNet
In 1993 Marty Tenenbaum, the founder of EIT, launched CommerceNet, the first consortium for companies interested in doing business over the Web. CommerceNet played a seminal role in transforming the net into a marketplace, pioneering core technologies for security and payment, and literally introducing Corporate America to the net's extraordinary commercial potential. The CommerceNet web site was designed by Kevin Hughes, whom Marty had hired based on his work on the Honolulu Community College site. Marty later hired Marc Andreessen, who worked on the first commercial web server, prior to co-founding Netscape. Applauded by Vice President Al Gore as being an influential early player in the Internet commerce arena, CommerceNet celebrates its 3rd birthday in 1997. Marty Tenenbaum will give a demonstration of its early online evolution and present a video of the original launch event.
Pioneering Sites: Global Network Navigator
In 1993 the publisher O'Reilly and Associates launched Global Network Navigator, one of the first sites to offer "one-stop shopping" for news and information. Conceived of by Dale Dougherty and implemented partly by Pei Wei, GNN set the standard for many similar sites to come, with regular articles and snazzy graphics. GNN pioneered many practices for advertising and doing business over the Web. In 1996, O'Reilly sold GNN to AOL, where it remained a hugely successful site for nearly two years. AOL closed down GNN in late 1996, which in many ways was a sign of its success; AOL wanted to bring the kinds of services offered by GNN into its core business. Dale Dougherty, GNNís president, will demonstrate the site and discuss its development.
GVU Center, Georgia Tech, and Xerox PARC
Tracking the Web: The History of Web Demographics and Usage Collection
James Pitkow established the first major demographic studies of who uses the Web and how. He also developed analytical and visual techniques for tracking groups of users in log files, which remain the basis of much Web usage tracking today. This talk will cover major demographic trends on the Web from 1994 to 1997 and the evolution of usage tracking technology. For each time period, Jim will review major demographic findings and events along with the controversy behind the numbers. The emerging demographic and usage trends will also be discussed.
From Scrolling to Strolling: The Birth of VRML
In 1994, Mark Pesce brought something unexpected to the first International Web conference, held at the Web's birthplace in Geneva, Switzerland. It was a full proposal for Virtual Reality Modeling Language, or VRML, which he had developed with his partner Tony Parisi. Mark and Tony hoped it would become a general specification for sending virtual reality worlds over the Web, much as HTML defines a standard format for documents. Their ideas was that one could experience a Web site as a virtual room or world, rather than a static page. The first demo was hardly impressive; a banana and a few cubes. But many of the other conference-goers were ignited by the idea, and a host of interested programmers and companies have been steadily developing VRML specifications and browsers to this day. Now, Tony Parisi's startup Intervista, Silicon Graphics, and many others are producing commercial VRML browsers and environments. If the time for virtual reality has indeed come, the near future of the Web may be much more interactive than many users realize. Mark and Tony will demo and discuss early VRML specifications and browsers, as well as recent work.
Interface Director, HotWired
Pioneering Sites: The History of HotWired's Interface
Wired Magazine's Web publication HotWired was the first widely known online periodical, and pioneered the use of many interface techniques now used today on the Web, as well as the heavy use of dynamically generated pages and per-user subscription and preference technology. Jeffrey Veen, HotWired's Interface Director, will demo various versions of the interface over its three year history, and discuss how it evolved along with Web publishing technology.
Netscape Navigator: The Web Goes Commercial
In 1994 Jim Clark, Marc Andreessen, and Eric Bina hastily recruited most of the original Mosaic development team at NCSA, and founded Mosaic Communications. Their goal was to seize Mosaic's role as leader of the emerging Web market with a "Mosaic killer". The team settled down to the chore of writing completely new, cleaner versions of the Mosaic browser and server, partly because NCSA still owned the code to Mosaic and partly because they knew a rewrite would clean up their own rushed original efforts. The results became the first Netscape Navigator and Netscape Server. NCSA did indeed sue the new company over the use of the Mosaic name and source code, and after settling out of court Mosaic Communications changed to Netscape Communications. Netscape Navigator quickly eclipsed Mosaic as the leading browser, and dominated the market until Microsoft launched Explorer, igniting the current "browser wars". Original Netscape developers will demo historic early versions of Navigator, and discuss its development. They will also show the hugely successful Netscape Web site at different stages of its evolution.
Internet Explorer: The Browser Wars Begin
In the late summer of 1995 Microsoft released Windows 95, perhaps the most publicized software release of all time. Included with Windows 95 was built-in connectivity to the Microsoft Network (MSN), a proprietary network which Microsoft hoped would make the World Wide Web redundant. Just after releasing Windows 95, Microsoft performed one of the fastest turnarounds of any multi-billion dollar corporation ever; they announced that they were making MSN a site within the Web, and would from then on make the Web the key focus of their entire corporate strategy. They pulled the copies of Windows 95 which included MSN connectivity, and began work on Internet Explorer, which became a successful competitor to the then-dominant Netscape Navigator. The race is still on. Key developers of Internet Explorer will demo early versions of the software and the Microsoft site at the time. They will discuss the origins and development of the browser, including which models were used in its architecture. They will also discuss the development of the Microsoft site, and the corporate history of the shift from MSN to a Web-based strategy.