This section includes a brief history of the Internet, the TCP/IP communications protocols, the World Wide Web (WWW or 'Web') and of Mosaic, a tool for accessing the Web.
The Internet is not a single computer network, such as one that might interconnect the computers in an office or building. Rather it is a network of networks - a network that spans much of the earth. All of these networks have in common their use of the Internet protocols which are described below. Computers, acting as smart switches, serve as gateways connecting the various networks together. They can also translate messages to and from computers that use different communications protocols. The gateways use the Internet addressing scheme to move packets of information (text, graphics, audio) back and forth. Packets are passed from one computer (or node) to another until the final destination is reached. If a pathway is unavailable, the packet is rerouted automatically by the switches. Upon arriving at their final destination, the packets are automatically reassembled (packets are often too small to contain entire messages or graphics) and provided to the reguestor. Typically, from only a few seconds will pass from the time the request is initiated until it completed.
The Internet provides a conduit for electronic mail, including mail sent to LISTSERVS of people with some shared interest. The file transfer protocols, or FTP, are used for sending files from one computer to another. Normally you must have a password for the computer you wish to exchange files with, but there are also many anonymous FTP sites which permit anyone to log on and retrieve specific files. Telnet is an internet protocol permitting a user to log on one computer by accessing it from another (for instance, if you are visiting a friend in Melbourne you might log on to his account and use it to sign on to your own back in Dallas). Usenet News, another feature of the Internet, are bulletin boards containing discussions on any conceivable topic - for instance, specialists in environmental accounting. A variety of other special services have been developed for the Internet. Several, including Archie and WAIS are search services, capable of locating files or articles from throughout the world. Gopher, developed at the University of Minnesota, is a relatively easy to use navigation tool for finding and retrieving information on the net. In 1994 the most exciting Internet capability was the World Wide Web, described below.
In July of 1994 there are an estimated 10 million users accessing the Internet from 20,000 unique networks. Growth continues to be tremendous, with a near doubling in subscribers reported each year. Although the Internet had long been in use on college campuses, much recent growth was in commercial accounts. However, many of those commercial subscribers, fearing security breaches, had chosen to either isolate their commercial networks or to establish very limited access between them and the Worldwide Web.
The Internet is not an organization with stockholders, president, and a board of directors. Instead it can be thought of as a set of communications protocols, as functions users can perform, as a great inventory of available information, or as a scheme for intelligently connecting the people of the world into one fast learning organism. The governance structure for the Internet, thus far, has been fairly loose. Net etiquette (or Netiquette) and a strong culture had provided some discipline in the past, but the massive growth and increased commercial exploitation of the web were beginning to create clashes of the old and new cultures. Among the most notorious such clash in 1994 was the transmission of a law firm's advertisement to several thousand listserv participants.
At its simplest the Internet was little more than a set of communications protocols that many people had agreed to use in connecting computers to a network and in connecting networks to each other. The Transmission Control Protocol/Internet Protocol (TCP/IP) had come out of work started by the U.S. Department of Defense in 1969. The intention was to design a communications network that could survive nuclear war. TCP/IP was designed to support a distributed network architecture. That is, there were multiple pathways between computers, no central point of vulnerability, and intelligence in the network that could make corrections as necessary.
TCP/IP is in marked contrast to, for instance, the Systems Network Architecture of IBM which assumed that a mainframe computer would control communications routings. Furthermore TCP/IP was a non-proprietary standard that provided for open competition across the hardware platforms of various computer and communications equipment providers. The decreasing costs of both workstations and communications bandwidth had, in the early 1990's, spurred intense customer interest in the so called client-server computer and communications architecture that TCP/IP represented. In such an environment the workstation (or client) was in control while larger computers acted as servers for those clients. In the past information was pushed at users; now it could be pulled.
The World Wide Web (WWW) began as an attempt by high-energy physicists at the CERN research center in Switzerland to share files with colleagues around the world. Its subsequent use has probably far exceeded their expectations. Essentially, the WWW or Web provides easy to use access to an ever expanding cornucopia of information sources located throughout the world. Files to be shared can be in a variety of formats (text, audio, video) and from many sources. The intelligence of the user's desk top workstation (or client) coupled with the intelligence of the computers that provide the information (servers ensure that incompatibility problems are automatically resolved.
Unique suffixes attached to file names helped communicate to client
and server alike the nature of a particular file. For instance, a picture
of the
Edwin L. Cox School of Business might be
stored on the school's server under the name of:
The first part of the name was constructed by the creator of the file and was intended to be shorthand for COX School at SMU: PICture of. The gif suffix informs both client and server that this is a graphic image (rather than, for instance, text (html) or audio (au). Assuming that someone in Singapore wished to use that picture in a document of their own they could access it using a uniform resource locator or URL. The URL is like an electronic mail address, but for accessing a document rather than a person. The address for this particular document is:
http://www.cox.smu.edu//mis/coxsmupic.gif.The last 13 characters we have already discussed. The //mis/ characters point at a specific directory within that server - one serving the management information Sciences department. The www.cox.smu.edu is the Internet address for the server computer on which that picture resides. The last three characters of that address - edu - signifies that this server is connected to the internet from an educational institution. Finally the first 7 characters, http:// inform client and server that it is looking at an address coded using the hyper text transmission protocol.
Although the web had been popular with some physicists since its introduction in 1992, it really began to take off in 1993. That new found popularity largely came from the development of Mosaic by the National Center for Supercomputer Applications (NCSA). Mosaic is a world wide web reader that typically resides within a user's desk top machine, sometimes referred to as a client. Mosaic, and servers like it, rely on the TCP/IP communications protocols to retrieve files of information from server computers. If the user's employer permits web requests to go through their security firewall), Mosaic can retrieve data from computers located throughout the world. It was also relatively easy to set a desk top computer up as a "server" capable of providing information to other users.
Mosaic's graphical user interface (GUI) is easy to learn, powerful, and relatively consistent across three desktop architectures (PC, Macintosh, and workstations ). The primary feature of web browsers like Mosaic is their ability to link from one piece of text to another using a publishing technology called Hyper Text. Creators of hyper text documents can designate that particular words are 'hot'. When the user later uses their computer mouse to select one of those specially designated words Mosaic (using the TCP/IP tools) will go out on the web and retrieve the desired file and display it on the user's screen (or play it through the sound card on her workstation). The retrieved file can be either text, video, or sound. By following linkages from subsequent documents users might soon find themselves looking at Old Master's in Le Louvre in Paris, boning up on Le Cordon Bleu recipes, or, from a server on the other side of the world, being spurred to action by The Information Technology Vision of the 'Intelligent Island' of Singapore. Mosaic's interface lets users easily accumulate interesting destinations in so called, 'hot lists' so that they could be easily found again or shared with others.
Some have called Mosaic the Web's 'killer application', because it has made the Internet so accessible to so many. Less optimistic observers feel that the popularity Mosaic has brought to the web could kill the Internet because of the heavy load it is placing on communications trunks and server computers. Partly this load is created by the thousands of new users drawn to the Internet each week and the increased availability of information and services, but partly it is the ease with which Mosaic users could handle large graphic, audio, and visual files. A Mosaic option permits users to turn off the automatic downloading of graphics to speed up processing. This is particularly important for users connected over modems who for the most part in 1994 are using line speeds of 14,200 bits per second or less. At this speed, graphics take several seconds to download. Besides Mosaic, other reader/browsers for the Web had been developed, including Lynx which could read the textual parts of the html but would largely ignore any graphical or audio links. HTML developers could, however, put a textual alternative to the unavailable grapic as in, 'picture of Cox School'. Increasingly, information and service providers on the Web, are giving users the choice of either graphic or text-based versions of their documents. Estimates in 1994 indicated that about two-thirds of web users are using just text.
Mosaic and the TCP/IP communications software that all web readers require is available at very small cost from several sources including both the Macintosh and PC versions of Adam Engst's Internet Starter Kit (Engst also publishes TidBITS an electronic newsletter focused on new developments in computer technology which was available, free of charge, over the WWW). A pull down menu on the the Macintosh version of Mosaic interface permits feedback from the users of Mosaic to the systems developers including reports of programming bugs and suggested enhancements. New releases of the software can be downloaded directly from the NCSA servers in Illinois or various NCSA mirror installation's server computers located throughout the world. These mirrors reduce the transoceanic communications burden though there are still some problems ensuring that they truly mirror images. Even with the mirrors, however, users can wait for hours or even days when new releases of Mosaic are instantaneously announced all over the world. Although connected by very high speed communications lines, the NCSA computers have become a bottleneck to efficient electronic distribution. Any user of the Internet could someday find themselves constrained by similar bottlenecks. Such bottlenecks can occur anywhere along the line between a particular user and the server that providing information to him or her. To reduce the distribution bottleneck for Mosaic itself, the University of Illinois announced in August of 1994 an agreement with Spyglass, Inc. whereby Spyglass would license copies of Mosaic to vendors in units of at least 10,000 copies.